University of Southern California Dissertations and Theses

Leading change initiatives: communication and bounded agency in a health care organization

(USC Thesis Other)

Leading change initiatives: communication and bounded agency in a health care organization

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content LEADING CHANGE INITIATIVES: COMMUNICATION
AND BOUNDED AGENCY IN A HEALTH CARE ORGANIZATION
Copyright 2002
by
Jordana Kanee Signer
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(COMMUNICATION)
December 2002
Jordana Kanee Signer
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UNIVERSITY OF SOUTHERN CALIFORNIA
THE GRADUATE SCHOOL
UNIVERSITY PARK
LOS ANGELES. CALIFORNIA 90007
This dissertation, written by
under the direction of h . ^ c...... Dissertation
Committee, and approved by all its members,
has been presented to and accepted by The
Graduate School in partial fulfillment of re
quirements for the degree of
DOCTOR OF PHILOSOPHY
Dean of Graduate Studies
Dafe December 18 , 2002
DISSERTATION COMMITTEE
Chairperson
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
DEDICATION
This dissertation is dedicated to the memory of my grandfather, Dr. Ben Kanee, MD,
and to my loving and devoted husband, Daniel G. J. Signer. You both continue to
inspire me to grow from strength to strength.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
iii
ACKNOWLEDGEMENTS
It makes no difference whether the project is a novel, a script, a published article, or
a dissertation—professional writing is always a collaborative effort. In producing
this dissertation, I was extremely fortunate to have the help of an exceptionally
wonderful crew.
I am very grateful to Paul Adler and my advisor, Patti Riley, for inviting me
to participate in their Diffusion Capability Project, and to Ben Lee, who knew of my
interest in health care and suggested that I get involved in that project. “Hospital for
Children,” where I conducted my research, is a very special place. I was truly blessed
to have had the opportunity to spend time with the people working there. My thanks
also go to the Packard Foundation and the Institute for Knowledge Management,
which generously funded our research project. Norma Panico and her employees at
ATM Inc. were much appreciated for their help with the extensive transcribing.
My committee members, Patti Riley, Bill Dutton, and LaVonna Blair Lewis
have my heartfelt gratitude for their thoughtful feedback, advice, and support. Patti,
in particular, sustained me with her unfailing encouragement
Without my Mends and family, this dissertation would not exist. I treasure
my Annenberg Mends, Mary Wilson, Traci Hong, and Elisia Cohen, who have been
such good companions on this journey. Mary, thanks for always checking in on me.
My parents, Judy and Robert Kanee, have supported me in every way imaginable
through this process. Thank you both for being so generous and inspiring, and for
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
being such good listeners. The love, laughter, and patience of my husband, Dan
Signer, were indispensable throughout my years in the doctoral program. Dan, thank
you for embarking on this adventure with me. We have given each other the courage
to pursue our dreams.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
V
TABLE OF CONTENTS
DEDICATION........................................................................................................... ii
ACKNOWLEDGEMENTS...................................................................................... iii
LIST OF TABLES................................................................................................... vii
LIST OF FIGURES AND EXHIBITS................................................................... viii
ABSTRACT.............................................................................................................. ix
CHAPTER 1: INTRODUCTION...............................................................................1
Leadership and Sensemaking................................................................................2
Health Care and Sensemaking..............................................................................3
Quality in Health Care..........................................................................................4
Reasons for Poor Quality......................................................................................7
Leadership and Communication in Quality Improvement...................................11
Outline of This Dissertation................................................................................ 16
CHAPTER 2: QUALITY IMPROVEMENT IN HEALTH CARE
ORGANIZATIONS.................................................................................................. 19
History of the Quality Movement.......................................................................21
Research on Quality Improvement in Health Care.............................................24
QI & Leadership..................................................................................................28
Commitment....................................................................................................31
Challenges for Leaders Implementing QI...........................................................38
Managing Resources........................................................................................39
Developing Culture......................................................................................... 43
Addressing Resistance.....................................................................................46
Managing Projects........................................................................................... 50
Coordinating Change Initiatives........................................................................ 53
Organizational Communication & QI.................................................................56
Intended Functions of Communication in QI......................................................56
Communication Media Used in QI....................................................................64
Emergent Communication in QI....................................................................... 67
CHAPTER 3: METHODS........................................................................................72
Sources of Data.................................................................................................. 73
The Research Site............................................................................................ 73
The Three Quality Improvement Projects...........................................................77
Rationale for Study Design.................................................................................83
Data Collection................................................................................................... 86
Interviews....................................................................................................... 87
Observation.................................................................................................... 89
Documents..................................................................................................... 90
Survey........................................................................................................... 90
Data Analysis..................................................................................................... 91
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
vi
CHAPTER 4: A SYNCHRONIC ANALYSIS OF LEADERSHIP AND
COMMUNICATION RESOURCES FOR CHANGE............................................. 93
Leadership: A Definition.................................................................................... 94
Resources for Leading Change...........................................................................99
Leaders Enact Resources through Systems and Structures................................106
Formal Organizational Structure & Leadership.................................................109
Networks & Leadership................................................................................. 123
Communication Media & Leadership.............................................................. 130
Information Systems & Leadership..................................................................138
Summary.......................................................................................................... 150
CHAPTER 5: A DIACHRONIC ANALYSIS OF LEADERSHIP AND
COMMUNICATION RESOURCES FOR CHANGE........................................... 152
Leadership across the Project Timeline............................................................156
Project Phases and Resource Enactment...........................................................159
Job Attributes of Change Leaders....................................................................175
Project Teams and the Dynamics of Representation...........................................178
Leaders Respond to Project Challenges............................................................ 188
Conflict.........................................................................................................189
Project Leadership Transitions........................................................................ 194
Bureaucratic Obstacles................................................................................... 195
Summary.......................................................................................................... 197
CHAPTER 6: DISCUSSION & CONCLUSION................................................... 199
Summary of Findings.......................................................................................200
Contributions to Theory................................................................................... 205
Limitations.......................................................................................................212
Implications for Practitioners........................................................................... 214
Directions for Future Research........................................................................ 218
Conclusion.......................................................................................................221
REFERENCES.......................................................................................................223
APPENDIX............................................................................................................240
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
LIST OF TABLES
Table 2.1 Leadership Loci and Variables in Health Care QI Research....................29
Table 2.2 Indicators of Management’s Commitment to Quality Improvement 33
Table 3.1 Summary of Proj ect Goals, Processes, and Outcomes..............................81
Table 3.2 The Project Leaders.................................................................................82
Table 4.1 Top Managers’ Enactment and Non-Enactment
of Resources through Formal Organizational Structures.......................113
Table 4.2 Project Leaders’ Enactment and Non-Enactment
of Resources through Formal Organizational Structures.......................118
Table 4.3 Project Leaders’ Enactment and Non-Enactment
of Resources through the Organization’s Communication Media.........132
Table 5.1 Leadership, Communication, and Sensemaking
in Expected Project Sequences.............................................................. 158
Table 5.2 Proj ect Leader Attributes....................................................................... 178
Table 5.3 Unplanned Challenges in Quality Improvement Projects.......................189
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
LIST OF FIGURES AND EXHIBITS
Figure 3.1 Hospital for Children Organization Chart.............................................75
Exhibit 4.1 Patient Access Quality Monitoring and Evaluation Plan,
presented July, 1999 ...........................................................................148
Exhibit 4.2 Patient Access Quality Monitoring and Evaluation Plan,
presented September, 2000.................................................................. 150
Figure 5.1 Factors Affecting Meeting Attendance and Implications
for Information Flow...........................................................................183
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ix
ABSTRACT
Change is an increasingly important part of what leaders do, and
communication is essential to both leadership and organizational change. However,
few studies have examined the communication processes involved in leading change
projects, especially at multiple organizational levels. A comparative case study was
conducted in an American pediatric hospital to investigate how top managers and
project leaders enacted resources for three change initiatives.
The analyses examined how four types of resources (information, knowledge,
authority, and allocative resources) were enacted through four organizational
systems or structures (formal organizational structure, informal networks,
communication media, and information systems) throughout the projects’ life cycles.
Top managers mainly enacted authority and allocative resources through formal
structures; project managers were much more concerned with collecting and
disseminating information. Top managers tended to be more involved in the first few
phases of the project, after which they generally “managed by exception” (Bass,
1990). Both groups of leaders often underestimated the need to communicate with
various stakeholders, to gain or maintain support for the innovations.
All three projects introduced changes demanding new behaviors of
communities of practice. Successful implementation involved establishing action
cycles and accountability structures that institutionalized the new practices; episodic
mechanisms alone could not maintain the changes.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
X
The project leaders faced multiple constraints. These included organization-
wide fiscal austerity, poor information systems, cumbersome bureaucracy, and
resistance from various stakeholders. Project leadership turnover and meeting
attendance problems created communication difficulties. Also, some project leaders
lacked authority (if not appointed by top management), or knowledge (e.g., project
management skills), or both. The leaders’ responses to these challenges ranged from
skillful problem solving to damaging neglect
Scholars have argued that resources are created, transformed, fixed and
circulated through human interaction (Clegg, 1989; Giddens, 1984), but few have
empirically investigated this phenomenon. The findings here suggest that resourceful
leaders are entrepreneurial problem-solvers; they creatively enact resources that help
them overcome constraints that would have impeded other leaders. The results also
indicate the importance of understanding the role of meeting dynamics and conflict
management in planned change. Directions for future research, and lessons for
organizations are discussed.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1
CHAPTER 1: INTRODUCTION
[C]hange leadership in pluralistic organizations could be viewed as a cascading
process that involves chains of interlocking role constellations at different levels.
There is a need for much more attention to the flow of leadership and change through
organizations.
Denis, Lamothe, & Langley (1999, p. 46)
[Leadership in the future will be distributed among diverse individuals and teams
who share responsibility for creating the organization’s future....In knowledge-
creating organizations, [formal and informal leaders at different levels] absolutely
rely on one another. None alone can create an environment that ensures continual
innovation and diffusion of knowledge.
Senge (1997, p. 32)
Global competition, increasingly informed consumers, and, in some sectors (notably
health care), increasing regulation are pressing organizations to improve their
performance. As they straggle to do things better and cheaper, many organizations
have turned toward quality improvement (QI) to strengthen their competitiveness.
Managers are beginning to acknowledging that leadership is needed at all levels in
order to achieve organizational proficiency in these processes. Change management
has become practically synonymous with leadership.
Scholars have argued that communication may be “the most important
mechanism” of leadership (Cyert, 1990, p. 35), and even have stated that leadership
cannot occur without it (Harris, 1993). Most of the literature on organizational
change, however, does not explicitly address communication (Eisenberg, Andrews,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2
Murphy, & Laine-Timmerman, 1999). I agree with Eisenberg and his colleagues’
claim that “communication forms the implied but neglected foundation of the change
literature” (p. 135). Recently, Lewis and Seibold (1998) have called on
communication researchers to pay more attention to the implementation of
organizational change. The study reported here represents an attempt to do so, and
focuses on both “interaction during implementation and communication-related
structures regarding implementation” (p. 93).
Leadership and Sensemaking
A large part of what leaders do involves perceiving and interpreting
ambiguous environmental and organizational information for their followers (Weick,
1978). As organizations strive to become more flexible, adaptive, and innovative,
ambiguity and uncertainty have become more pronounced in organizational life.
Over the last 25 years, there has developed a substantial literature that sees leaders as
“managers of meaning.” Many scholars have emphasized how leaders’
communication practices affect their followers’ cognitions, and thereby influence
organizational behavior (e.g., Conger, 1991; Fairhurst & Sarr, 1996; Pondy, 1978;
Thayer, 1988; Weick, 1980; also Smircicb & Morgan, 1982 in Bryman, 1996; and
Pfeffer, 1981 in Bryman, 1996). However, little empirical research has addressed
what resources leaders use when making sense of their experiences, particularly
when undertaking organizational changes.
This dissertation proposes to address that question: How do leaders enact
communication and sensemaking resources in managing organizational change? In
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
addressing this issue, the proposed study extends organizational and communication
theory in a way that will have clear practical implications for organizations and
employees attempting to plan and implement change. Three main theories inform the
research: Giddens’s theory of structuration, Weick’s theory of sensemaking, and
Pearce and Cronen’s Coordinated Management of Meaning (CMM) theory. These
theories share a symbolic interactionist perspective on social phenomena. Blumer
expressed the core of this perspective most succinctly in his classic premises of
symbolic interactionism: that people’s actions are based on the meanings they
ascribe to things, that those meanings come out of social interaction, and that
meanings change through individuals’ interpretative processes (Blumer, 1969). For
the purposes of studying change management, symbolic interactionist theories focus
attention on the communication processes of leadership, and on the social
construction of new and existing organizational realities.
Health Care and Sensemaking
“Information is the principal input processed by professional organizations”
(Weick & McDaniel, 1989, p. 336), and it follows that the role of communication in
healthcare is “pervasive, ubiquitous, and equivocal” (Fisher, 1978, p. 91 cited in
Kreps, 1988). On top of the ambiguity related to the political dynamics of health care
organizations, and the inherent uncertainty of health care information, a great deal of
ambiguity has arisen out of the tremendous institutional changes have shaken the
U.S. health care industry for the past 20 years. Managed care has forever altered the
face of health care delivery in the United States, and never before has the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4
government been so involved in regulating the industry. Health care organizations,
especially hospitals, have gone through wave after wave of organizational change in
response to these pressures, from the adoption of Diagnosis Related Groups (DRGs),
to learning Total Quality Management (TQM, which is generally known in health
care as “Continuous Quality Improvement” or “CQF), to reengineering, to mergers
and acquisitions. Since ambiguity, negotiation, and change are integral to health care,
it is an ideal setting for understanding the communication and sensemaking resources
that leaders use in managing organizational change.
The research reported here is an embedded case study of leadership in three
quality improvement projects at an American children’s hospital. As such, the
findings have direct implications for health care workers who are struggling to
improve the performance of their organizations. Most hospitals in the United States
(and many in other countries, including Canada), adopted quality improvement
programs in the early 1990s, when CQI was mandated by the national regulatory
body, the Joint Commission for Accreditation of Health Care Organizations
(JCAHO). The history of the diffusion of CQI is described in more detail in the
second chapter of this dissertation. In 1993, 69% of American community hospitals
reported that they had adopted some sort of QI program (Bareness et al., 1993a), and
by 1998 their ranks had swelled to 93% (Arthur Andersen, 1999).
Quality in Health Care
The results of the quality movement to date have, unfortunately, been
disappointing (Blumenthal & Kilo, 1998; Shortell et al., 1995b). According to the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
teachings of TQM, quality improvement efforts are supposed to focus on the needs
and desires of customers within and outside of the organization, and in health care
the primary customer is the patient. Despite the widespread adoption of quality
management programs in health care, however, the industry has yet to show
improvement in the area that patients and clinicians care about most: health
outcomes.
In 1998, three major reports were issued that addressed the scope of the
quality problems in health care. RAND Corporation published a literature review on
the quality of health care in the United States covering peer-reviewed journals
between 1993 and mid-1997; the Advisory Commission on Consumer Protection and
Quality in the Health Care Industry concluded that the health care industry was
“plagued with overutilization of services, underutilization of services and errors in
health care practice” (1998, quoted in Institute of Medicine, 2001, p. 24); and a
report by the IOM’s National Roundtable on Health Care Quality drew the same
conclusions. That year, the IOM formed the Committee on the Quality of Health
Care in America to advance the cause of improving the quality of health services in
the U.S.
The Committee’s first report, To Err Is Human: Building a Safer Health
System, released in 2000, shocked the American public with its conclusion that tens
of thousands of American patients were dying each year from medical errors, and
hundreds of thousands suffer from or barely escape preventable injuries. In their
most recently released report, which calls for a quality revolution in health care
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6
delivery, The Institute of Medicine (2001) noted that between 1991 and 2001, more
than 70 articles were published in peer-reviewed journals documenting “serious
quality shortcomings” in medical care. The sobering statistics suggest that:
one-fourth of hospital deaths may be preventable. Each year 180,000
people may die, partly as a result of iatrogenic injury: one-third of
some hospital procedures may expose patients to risk without
improving their health; one-third of drugs prescribed may not be
indicated; and one-third of lab tests showing abnormal results may not
be followed up by physicians (Dubois and Brook 1988; Leape 1994;
Brook et al. 1990). (Shortell, Bennett, & Byck, 1998, p. 593)
Less research has been conducted on the quality of pediatric care (Ferris,
Dougherty, Blumenthal, & Perrin, 2001) but existing evidence suggests that it suffers
from similar deficiencies. A review of the literature on quality of care for children
concluded that “care for children is variable, at times inappropriate, and that at times
it is not meeting quality of care criteria; and.. .that parents report opportunities for
improvement in coordination and involvement” (Homer, Kleinman, & Goldman,
1998, p. 1096). For example, the authors noted that some studies have shown that
“one-fifth to one-quarter of pediatric hospital days could be justified neither on
severity of illness, intensity of service, nor social risk criteria (Kemper, 1988; Kreger
and Restuccia 1989)” (p. 1095). Furthermore, a recently published study suggested
that medication errors and resulting adverse drug events are commonplace among
pediatric inpatients. Kaushal and his colleagues (2001) reviewed 10,778 medication
orders in two teaching hospitals over a six-week period in 1999 and found
medication errors in 5.7% of orders, potential adverse drug events in 1.1%, and
actual adverse drug events in 0.24%. They claimed that 19% of the actual adverse
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
drug events were preventable, and noted that the rate of potential adverse drug events
was three times higher among the children in their study than the rate found in a
1992 study of medication errors among adults.
Reasons for Poor Quality
How is it possible that, despite quality improvement programs apparently
having diffused throughout essentially the entire population of health care
organizations, such disturbing problems with quality of care have been allowed to
continue? There appear to be three reasons related to the implementation of quality
improvement: (1) QI is not usually implemented across health care organizations, (2)
most QI projects have not focused on improving quality of care, and (3) QI projects
that are conducted seem to have a veiy poor success rate.
The research on health care organizations suggests that commitment to CQI
implementation may have been lacking throughout the first decade of “CQI
adoption.” For example, a national survey of adult acute care hospitals in 1998 found
that, on average, only 22% of the hospitals’ physicians had been trained in
CQI/TQM, and only 35% of their staff had received such training (Arthur Andersen,
1999). These hospitals reported having achieved an average of only 3 significant
improvements in patient outcomes out of a list of nine different areas (up from 1 in
1993), and only 22% of them had integrated clinical databases, which are crucial for
undertaking large scale, ongoing clinical quality improvement efforts. A case study
of 10 hospitals found that some hospitals were implementing CQI very slowly
(Boerstler et al., 1996; Carman et al., 1996). Similarly, in a survey of 12 long-term
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
care facilities, 2 facilities reported 10-year implementation schedules, one said their
timeline for implementation was “open-ended with no end in sight,” and 2 did not
mention a timeframe at all (Graves & MacDowell, 1994). According to quality
experts, it takes five to ten years to see organization-level results from a shift to
quality management in organizations that have seriously committed to adopting it
throughout their operations (Deming, 1982, 1986; Juran, 1989).1 Since industry-wide
changes typically lag years behind organizational changes, and since most health
care organizations do not appear to be strongly committed to implementing CQI, it is
not surprising that CQI has yet to deliver on its promise.
The majority of CQI projects, especially in the early stages of adoption, have
been focused on administrative process improvements, particularly with a view to
controlling costs (Blumenthal & Kilo, 1998). For instance, a 1993 survey found that
69% of non-govemmental acute care hospitals were actively involved in CQI
activities, but more than 70% had not yet devoted significant effort to examining
clinical processes or conditions (Shortell et al., 1995c). Shortell and his associates
noted in their (1998) review of the clinical CQI literature that physician researchers
did not begin to apply CQI methodologies to improving clinical processes until the
mid-1990s. In a recent review of quality improvement in pediatrics, the authors
found that half of the studies that met their inclusion criteria had been published
since 1994 (Ferris et al., 2001). These two reviews found that
1 Results are not guaranteed even over the long term, since many organizations “have spent
the several years but failed to reach the goal” (Juran, 1989, p. 71).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
many published reports described clinical CQI interventions aimed at reducing
lengths of stay for various conditions or medical interventions, which is a means of
controlling costs and is not directly related to actual health outcomes.
There are macro-level structural reasons for these organizations’ lack of
commitment to implementing quality improvement. The IOM acknowledges many
of these in its 2001 report: declining third-party payments, nurse staffing shortages,
and increasing numbers of uninsured patients receiving uncompensated care (p. 3),
as well as the challenges of organizational restructuring, “misaligned incentives,
professional entrenchment, competing priorities, organizational inertia, and lack of
adequate information systems (Shortell et al., 1998)” (p. 128). Pediatrics faces
additional challenges to CQI implementation, including a dearth of outcomes
research in children, compounded by the fact that many conditions or treatments in
children are of such low volume that it is difficult to collect the data necessary for
rapid cycle change. In addition, most pediatric care is delivered in office settings,
which have less developed information systems and less resources to devote to CQI
(Ferris et al., 2001). Pediatric hospitals have also been experiencing declining
revenues and competition with adult facilities, due to managed care (Thompson,
Chesney, Stocks, Shmerling, & Herron, 1999). They have lost income due to the
Balanced Budget Act of 1997, which cut funding to Medicaid (a major source of
revenue for some of these hospitals) and graduate medical education (which many
pediatric hospitals provide through affiliation with medical schools).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
10
These pressures have led to a conundrum for health care organizations: on the
one hand, they are under increasing pressure to improve their performance, while at
the same time they have decreasing resources to devote to improvement. In any case,
CQI is still widely seen as the best hope for hospitals to improve their performance,
both by physicians concerned with improving health outcomes (e.g., Blumenthal &
Kilo, 1998; Greenfield, 1995; Institute of Medicine, 2001; Mulley, 1995), and by
administrators struggling to keep their organizations afloat. A survey of
administrators of children’s hospitals conducted in 1996 found that the top two
strategic responses that administrators planned to use to deal with declining revenues
were quality improvement (94%) and cost containment (82%) (Thompson et al.,
1999). (All of them also planned to downsize: 59% by reducing the number of beds,
88% by reducing the number of nurses, and 70% planned by reducing the number of
nonmedical support. There is evidence that turnover is a threat to QI success (Graves
& MacDowell, 1994; Parker, Wubbenhorst, Young, Desai, & Chams, 1999;
Zbaracki, 1998).)
While on the one hand, CQI clearly has not been fully or deeply implemented
in most health care facilities, at the same time most of these organizations appear to
have adopted some abbreviated form of it, and most have applied its methods in
quality improvement projects. The track record of these projects is difficult to gauge,
but reports suggest that it is not impressive. There have not been any empirical
studies of the failure or success rates of these interventions—in fact, the average
number of CQI projects undertaken by hospitals, or other kinds of health care
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
organizations, remains unknown. Reviews of the literature on CQI interventions
have consistently pointed out that failures are underreported and that study designs
are weak (Bigelow & Amdt, 1995b; Ferris et al., 2001; Shortell et al., 1998). In
general, studies with stronger designs (controls and random assignment) have been
more likely to report failures (Fenis et al., 2001; Shortell et al., 1998), or effects of
lesser magnitude (Ferris et al., 2001). Ferris and his colleagues interviewed nine
experts on quality improvement in child health care, and were told that failures were
“relatively common” (p. 152). Informants in a case study comparing TQM in five
different organizations (including one hospital) reported that typically only one in six
quality improvement projects actually succeeded (Zbaracki, 1998).
Leadership and Communication in Quality Improvement
Given the fact that CQI is still believed to be the best way to improve
performance in organizations, it seems critical to ask how health care organizations
might improve the success rate of these projects. To date, research on the facilitators
and impediments of CQI has focused on organization- and market-level factors. This
research has considered how variables such as the size of the organization and
managed care penetration affect CQI implementation and performance. QI program
implementation and QI projects are both complex types of organizational change,
and they are effected through communication. Yet very few studies have looked at
quality improvement from a communication perspective. Gail Fairhurst’s (1993;
Fairhurst & Wendt, 1993) research on the implementation of a total quality program
in five plants of a manufacturing company, and Laurie Lewis’s study of quality
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
12
improvement programs in four organizations (Lewis, 2000b) represent the main
contributions in this area.
Communication is how QI gets done: it is how organization members decide
what projects to pursue, how team members coordinate their activities, how teams
and organization leaders monitor team progress, and how specific process
improvements are ultimately implemented. Often those process improvements are
designed to improve communication among co-workers, between departments, or
with others (such as patients, families, or referring physicians) in order to improve
outcomes that may range from patient satisfaction with wait times to prevention of
medication errors. In looking at QI from a communication perspective, this study
sheds light on how individuals’ actions and interactions shape QI processes.
Several studies have looked at how the activities of key organizational actors
or groups affect CQI implementation and outcomes. These studies have tended to
focus on the activities of top management (Arthur Andersen, 1999; Weiner,
Alexander, & Shortell, 1996; Weiner, Shortell, & Alexander, 1997), the board of
trustees (Arthur Andersen, 1999; Barsness et al., 1993b; Weech-Maldonado, Zinn, &
Brannon, 1999; Weiner et al., 1996), physicians (Barber, 1963; Barsness et al., 1994;
Blumenthal & Edwards, 1995; Boerstler et al., 1996; Carman et al., 1996; Shortell et
al., 1995c), and the quality improvement council (Boerstler et al., 1996; Lammers,
Cretin, Gilman, & Calingo, 1996).
Shockingly, very little comparative research has looked at the activities of
quality improvement teams—in fact, I was only able to locate a handful of studies
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
13
that specifically surveyed or interviewed team members. One of these was a quasi-
experimental study (Steel, Mento, Dilla, Ovalle, & Lloyd, 1985), two were cross-
sectional surveys (Lammers, 1992; Solberg, Brekke, Kottke, & Steel, 1998) and five
were case studies (Blumenthal & Edwards, 1995; Lewis, 2000b; Savitz & Kaluzny,
2000; Waldman et al., 1998; Zinn, Weech, & Brannon, 1998). This is particularly
remarkable given that TQM and CQI methodology is entirely built on doing quality
improvement work in project teams. Quality gum Joseph M. Juran (1989) stated this
plainly: “[a]ll quality improvement takes place project by project and in no other
way” (p. 35). By examining the activities of several quality improvement projects
that spanned years, this dissertation will help researchers and practitioners to better
understand the actions and structures that facilitate and impede team progress—
which is the foundation of organizational performance improvement.
Practitioners have complained that the advice they are given about quality
improvement is too general, especially with regard to the social aspects of behavior
change and communication (Bigelow & Amdt, 1995b). The TQM literature assigns
the responsibility for ensuring organizational commitment to quality to top
administrators, so most of this advice has been directed at upper management.
Weiner, Alexander and Shortell (1996) argued that the concept of “leadership” for
quality improvement should be expanded beyond the team of upper administrators to
include the roles of boards of trustees and physician leaders. Another body of
literature has addressed physicians, particularly the problem of how to win them over
to CQI. Lower level leaders (i.e., managers and supervisors) are rarely discussed
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
14
except insofar as they are identified as a locus of resistance to QI implementation,
and nursing is almost completely absent from the empirical literature on QI
(Lammers et al., 1996 and Calomeni, Solberg, & Conn, 1999 being rare exceptions).
The neglect of lower level leaders and nurses is particularly striking given
that managers are disproportionately involved in QI compared to other employees
(e.g., Counte, Glandon, Oleske, & Hill, 1992) and that nurses form the “backbone”
of QI (Savitz & Kaluzny, 2000). Nurses, middle managers, and ancillary staff (e.g.,
respiratory therapists, physical therapists, pharmacists, laboratory workers, social
workers, and nurses’ assistants) are the ones who actually do the work of QI:
tracking performance, determining root causes of performance deficits,
brainstorming solutions, and implementing them. They organize the meetings,
collect and process the data, and design the policies and procedures for improving
processes. By focusing on such 01 activities, and by looking at how team leaders
work together with other organizational leaders to achieve their goals, this study
aimed to find specific communicative and social mechanisms by which leaders help
their teams make progress and achieve success. It also looked at missed opportunities
or critical moments when leaders at different levels could have potentially avoided
pitfalls or obstacles.
Several authors have linked CQI with organizational learning. According to
Blumenthal and Kilo (1998), the management philosophy underlying CQI includes a
“vision of leadership that encourages the creation of what Peter Senge has called ‘the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
15
learning organization’ (Senge 1990; Chawala and Renesch 1995)” (p. 627). Savitz &
Kaluzny (2000) claim that successful clinical process innovations require
the development of what Peter Senge (1990) has described as the ‘learning
organization,’ which Tichy (1998) has more recently termed the ‘teaching
organization,’ that is, an organization where people continually pass on
learning and energize others to be teachers (p. 376).
Senge (1996a; 1996b; 1990) has argued forcefully that organizational learning
cannot take place without fostering leadership throughout the organization, at all
levels.
Scholars have recently recognized that if the current environment demands
organizational learning, and if organizational learning demands leadership from all
levels, then new theories of leadership will be needed. Leadership theory will have to
move beyond telling managers and supervisors how to better motivate their
subordinates, and instead will have to look at how different kinds of employees
might best interact in order to coordinate and maximize the performance of
leadership networks. Communication under this model of leadership will likely be
exponentially more complex than in the past, and will probably require the
development of more sophisticated, systems-based theories. The research reported
here, which looks at the interactions and relationships among different types of
organizational leaders over time, will contribute to an understanding of how
distributed leadership functions in organizational change efforts. By looking at how
team leaders interact with various organizational constituencies, and how they deal
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
16
with obstacles to achieving their goals, the study builds on the tradition of leadership
theory that is concerned less with what leaders are and more with what they do.
In sum, little is known about how communication operates in QI processes,
and to date research has focused on organization-wide QI programs to the exclusion
of understanding QI projects and the role of QI project leadership in project team
dynamics. This study specifically tries to understand how leaders at different levels
communicate in and about QI projects, with a focus on project team processes.
Studying this important phenomenon helps fill important gaps in theories of
organizational communication, change management, leadership, and quality
improvement.
Outline of This Dissertation
This study aims to address the following research questions about
communication and the leadership of organizational change. The term “change
leaders” includes both project leaders and executive leaders.
Research Question 1: How do the communication practices o f change leaders
enact (a) allocative resources (i.e., control over things), (b) authoritative
resources (i.e., control over people), (c) information resources, and (d)
knowledge resources, in organizational change projects?
Research Question 2: How do change leaders enact resources through their
reproduction and transformation o f organizational systems and structures,
including (a) formal reporting structures (as documented in job descriptions and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
17
organization charts) , (b) informal communication networks, (c) communication
media systems, and (d) information and communication technology?
Research Question3: What resources do change leaders need fo r the different
project phases? How does their enactment o f resources vary across those
phases?
Research Question 4: How do resources needs and patterns o f enactment
across the project phases vary fo r leaders (a) at different organizational levels,
and (b) across occupations?
Research Question 5: How do change leaders manage emergent social
conditions (e.g., transitions in leadership) that are not addressed in formal
models o f the change process? What role does communication play in the
management o f unexpected developments and unintended consequences?
The next chapter gives a brief account of the history of the quality movement,
and reviews the literature on quality improvement in health care. The review focuses
on empirical research concerning the role of organizational leadership in
implementing QI, challenges for QI leaders, and the role of communication in QI.
Chapter Three describes the methods used in this study, which is a
comparative case study. Data were collected using ethnographic techniques
including interviews, observation, and some limited participant observation. A
variety of organizational and project-specific documents were also obtained. Data
were analyzed using largely qualitative techniques, including coding practices
derived from grounded theory (Strauss & Corbin, 1998), and comparative cross-case
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
18
displays (Miles & Huberman, 1994). This chapter also includes a description of the
pediatric hospital that was the site of this research, and a description of the three
quality improvement projects that were the focus of this study.
Chapters Four and Five develop the study’s analytical framework, and apply
it to the data in reporting the findings. The analytical framework draws f om
structuration, sensemaking, and CMM theories in answering the research questions,
but is also influenced by emergent patterns in the data. Chapter Four takes a
synchronic approach in examining how leaders enacted different types of change
resources for the three change initiatives. It considers how leaders’ enactment of
resources was related to their organizational systems and structures. Chapter Five
takes a diachronic perspective, focusing on change processes and their evolution
over time. It considers leaders’ resource enactment within the expected the
progression of project phases, the differences among change leaders with differing
job attributes, and the leaders’ responses to unexpected challenges.
Chapter Six summarizes the findings and discusses implications for
researchers and practitioners. It considers the theoretical contributions of this
research, addresses limitations of the study, suggests applications for organizations,
and recommends areas for future research.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
19
CHAPTER 2: QUALITY IMPROVEMENT IN
HEALTH CARE ORGANIZATIONS
This study is investigates communication and leadership in organizational change
projects by considering three quality improvement (QI) efforts in a pediatric hospital.
QI efforts are defined as any attempt to change an organizational process in order to
improve outcomes such that they will measurably attain a standard that represents an
improvement over the baseline level of performance. For the past decade,
organizations throughout the health care industry have been pursuing quality
improvement through the introduction of Total Quality Management (TQM)
methodologies, which the health care industry prefers to call “Continuous Quality
Improvement” (CQI). TQM/CQI is usually implemented as an ongoing organization-
wide program, which operates through specific, local, and finite quality improvement
projects.
This chapter reviews the research on quality improvement that pertains to
leadership and communication. It begins with a brief history of TQM in business and
its diffusion to the health care industry, followed by a short overview of the
published research on QI in health care. The discussion then turns to leaders’ roles in
implementing organization-wide QI programs, including how they demonstrate
commitment to QI. The next section discusses several challenges that leaders
commonly face when implementing QI programs. These challenges mean that QI
leaders may need to ensure that resources are appropriately managed, supportive
organizational cultures are developed, resistance is addressed, projects are managed,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
20
and that multiple ongoing change initiatives are coordinated. Finally, the chapter
concludes with a discussion of the role of organizational communication in QI
implementation, including intended functions of communication, communication
media usage, and emergent communication processes.
TQM is usually described as a holistic management approach or paradigm
that encompasses both a management philosophy and a set of practices geared
toward delivering customer-oriented goods and services. Shortell and his colleagues
(1995a) identified four organizational dimensions of TQM in their review of the
quality literature: the cultural, the technical, the structural, and the strategic. The
practitioner literature emphasizes that top management is responsible for ensuring
the development of these dimensions throughout the organization. They are expected
to lead continuous improvement in core processes by utilizing statistical
methodologies to track performance, by building organizational structures and
systems that empower employees to make changes, and by ensuring that everyone is
trained in the requisite skills to pursue and measure improvements. Implementing an
organization-wide QI program requires the participation of all employees, and is
meant to involve leaders at all levels: upper managers (e.g., departmental nursing and
medical directors), middle managers, and supervisors. In any given QI project, the
project leader must coordinate with managers and other employees at various levels
in order to investigate problems and to design and implement solutions.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
21
History of the Quality Movement
Much of the earliest theoretical work on quality control in manufacturing was
done in the 1920s by the quality-assurance department of the Bell Telephone
Laboratories, and much of the pioneering application took place that same decade
within the Hawthorne Works of the Western Electric Company (Juran, 1989). This
work had little impact on industry at that time, even within those companies. It was
not until the early 1940s that American business and the U.S. military adopted Dr.
W. Edwards Deming’s statistical tools and Joseph M. Juran’s management practices
for quality control and quality engineering. After World War H, these innovations
spread to Japan, where the Deming Prize for quality was established in 1962 to
recognize companies that excelled in Total Quality Management (TQM).
By the 1980s, Japanese companies had significantly eroded corporate
America’s market share. This situation prompted a quality revolution in American
companies, with many U.S. corporations implementing TQM programs throughout
the 1980s and into the 1990s. The U.S. government established the Malcolm
Baldrige National Quality Award in 1987 to recognize American organizations that
excel in TQM.
By the 1990s, TQM had begun to diffuse within the healthcare industry. A
few hospitals adopted TQM in the late 1980s, “largely in response to pressure from
employers, purchasers, and payers to provide more cost-effective health care
(Berwick, Godfrey, and Roessner, 1989, 1990; Laffel and Blumenthal, 1989)”
(Westphal, Gulati, & Shortell, 1997, p. 379). Editorials appeared in prominent
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
22
medical journals advocating the use of TQM in health care organizations (Berwick,
1989; Ellwood, 1988; Laffel & Blumenthal, 1989). Then Berwick, Godfrey and
Roessner published the results of a large-scale national demonstration program of
TQM in hospitals (Berwick, Godfrey, & Roessner, 1990). A year later, the Joint
Commission on Accreditation of Healthcare Organizations (JCAHO) announced that
it would be significantly revising its standards in the 1992 Accreditation Manual fo r
Hospitals. The name of the “Quality Assurance” chapter from the 1991 Manual was
changed to “Quality Assessment and Improvement,” reflecting a shift in emphasis
from “quality control” (monitoring a limited number of quality indicators with a goal
of maintaining minimum standards) to continuous quality improvement (following
the principles of TQM: patient-outcome focused, multiple feedback sources, process-
oriented, and emphasizing ongoing reassessment and improvement).
The Joint Commission (JCAHO) is a major regulatory and institutionalizing
force in the health care industry. Accreditation by JCAHO is required for all
hospitals participating in Medicare and many state Medicaid programs. Under
JCAHO’s influence, hospitals across the United States quickly adopted TQM/CQI.
In 1990, less than 400 general medical surgical hospitals had adopted TQM, but by
1991 the number had jumped dramatically to more than 1,200 of these hospitals, and
by 1992, more than 1,800 hospitals had adopted TQM, with all such hospitals (about
2,000 total) having adopted by 1993 (Westphal et al., 1997). Other surveys
documented similarly steep diffusion rates of TQM/CQI programs in acute care
hospitals (Barsness et al., 1993a; Eubanks, 1992).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
23
Other health care organizations followed suit. In the late 1980s, nursing
homes experienced increased competition and heightened regulation, and many
nursing homes responded to these pressures by adopting TQM/CQI. A survey of 615
nursing homes in Pennsylvania during 1994-1995 found that 76% of the 241
responding facilities were practicing quality improvement activities (Zinn et al.,
1998). QI methods have been used in primary care clinics (e.g., Solberg et al., 1998),
with controlled, randomized studies of QI interventions demonstrating both success
(Palmer et al., 1996b) and failure (Solberg et al., 2000). Pediatric hospitals have also
been turning to QI as a strategic response to competitive pressures and declining
revenue (Thompson et al., 1999), and some organizations have pursued collaborative
initiatives to apply CQI to community health issues (Kaluzny, 1995).
Although several physicians (e.g., Berwick and Blumenthal) were among the
earliest proponents of applying TQM to health care, administrators were the driving
force behind TQM adoption in most hospitals and nursing homes, and in most health
care organizations physicians lagged behind administrative and nursing departments
in undertaking CQI projects. At the same time, in the 1990s physicians saw the
growth of the evidence-based medicine movement. Research on medical outcomes
proliferated, and clinical guidelines, pathways, and algorithms were developed and
promoted by a range of organizations, including professional medical associations
and the U.S. Agency for Healthcare Research and Quality (AHRQ, formerly the
Agency for Health Care Policy Research). Some physicians argued strongly in favor
of combining CQI methods with outcomes research (e.g., Greenfield, 1995; Mulley,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
24
1995), and recent literature reviews have shown that CQI is increasingly being used
in the development and implementation of clinical guidelines (Ferris et al., 2001;
Shortell et al., 1998).
Research on Quality Improvement in Health Care
The American private sector adopted TQM throughout the 1980s, yet
American “researchers did not begin to seriously study the phenomenon until 1990”
(Shortell et al., 1995a, p. 7). Several studies conducted in 1990s drew conflicting
conclusions about the value of continuous quality improvement programs in
corporations. A study of Fortune 1000 corporations found that an overall TQM
approach was associated with aspects of organizational effectiveness (Lawler,
Mohrman, & Ledford 1992, cited in Shortell et al., 1995a). However, only one-third
of the 500 organizations that participated in a survey by Arthur D. Little (1992, cited
in Shortell et al., 1995a) felt that their CQI efforts had significantly improved their
competitiveness. Similarly, Powell’s (1995) research concluded that most of the
best-known elements of TQM do not generally produce competitive advantage.
A number of reviews of the burgeoning literature on TQM in health care
were published in the mid-1990s (Bigelow & Arndt, 1995b; Gann & Restuccia,
1994; Motwani, Sower, & Brashier, 1996; Shortell et al., 1995a). Most empirical
reports that had been published up to that point were anecdotal, with the exception of
a few surveys on TQM/CQI diffusion in hospitals. Gann and Restuccia (1994) noted
an “interesting paradox—that we do not use continuous improvement techniques in
the core business of health care delivery” (p. 494). High-profile physicians called for
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
25
the application of TQM to improve clinical outcomes (e.g., Greenfield, 1995;
Mulley, 1995), and health services professors (e.g., Bigelow & Amdt, 1995a;
Bigelow & Amdt, 1995b) called for research into why TQM was not being used to
improve clinical processes. Not until the mid-1990s did significant numbers of
physicians begin to apply CQI methodologies to improving clinical processes
(Shortell et al., 1998).
Two later articles systematically reviewed the CQI literature with a focus on
efforts to improve clinical processes. Shortell, Bennett and Byck, (1998) located 55
articles published between 1991 and 1997 that reported on specific clinical
applications of CQI addressing problems of overuse, underuse, and misuse of health
services. Similarly, Ferris et al. (2001) found 68 articles published between 1985 and
1998 (though half had been published since 1994) that described QI interventions
aimed at improving overuse, underuse, and misuse of health services for children,
and that included a quantitative assessment of the primary outcome. Both reviews
found that the vast majority of the articles they reviewed (93% in Shortell et al.,
1998; 68% in Ferris et al., 2001) had weak single-group pre- and post-treatment
observation study designs, and most of the studies reported successful results.2 Both
reviews also noted that the few published articles on controlled or randomized
clinical trials of QI interventions were more likely to report negative results.
2 This has also been an issue in the general management literature on TQM. A review o f 99
articles about TQM effects (published between 1989 and 1993) found that more than 80%
only reported on program results in a single organization (Hackman & Wageman, 1995).
Only 4% o f TQM effects studies actually assessed the extent o f TQM implementation.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
26
A notable theme in the literature on TQM in health care is the belief— which
appears to be fairly widespread among healthcare workers—that hospitals and other
health care organizations are so special or different that the lessons of other sectors
are generally untranslatable to their facilities. Ironically, similar reactions have been
observed in other sectors, including higher education (Koch & Fisher, 1998). Joanne
Martin has dubbed the shared tendency of institutions to tell stories about their
exceptional status the “uniqueness paradox” (Martin, Felman, Hirsch, & Sitkin,
1983). Smith (1995) cites Martin and the uniqueness paradox in his critique of Ardnt
and Bigelow’s (1995) argument that TQM is probably not well-suited to the health
care industry. In their review of the research on TQM in health care, Gann and
Restuccia (1994) discuss several typologies of service organizations and ultimately
conclude that health care fails to fit neatly into any of the reviewed categories. For
them, this is evidence of how distinct the industry is: “The inability of health care to
fit neatly into the current service typologies emphasizes its unique nature. It adds an
additional confounding layer to the study of TQM and reminds us not to assume
automatically that findings from other types of services apply to health care.” (p.
492)
This perspective of uniqueness carries through to discussions of how to talk
about quality. Motwani, Sower and Brashier (1996) warn in their review that when
hospital administrators are implementing TQM and are naming their organization’s
quality improvement process, “every effort should be made to disassociate TQM in
the health care industry with TQM in other industries, since medical employees view
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
27
themselves as different from manufacturing employees” (p. 75). Similarly,
Blumenthal and Kilo (1998) write that one of the lessons of the first decade of CQI is
“Avoid jargon and management language; emphasize the scientific approach to
problem solving that lies at the heart of CQF (p. 644). Such recommendations
appear to be based on empirical research showing that physicians are resistant to the
language and tone of CQI management discourse (Blumenthal & Edwards, 1995).
Blumenthal and Kilo predict that in the future, the CQI movement will have
abandoned its jargon, including the “controversial label” of CQI.
Considering the energy that has been spent promoting quality improvement
in the health care industry, relatively little empirical research has looked at
organization-level quality improvement programs across health care facilities. None
of the previously mentioned reviews has attempted to synthesize this literature. A
literature search located 35 separate references where empirical research
methodologies were used to compare the adoption, implementation, or performance
of quality improvement practices across two or more organizations (including at least
one health care organization). The vast majority (21) of these studies were cross-
sectional surveys. Eight were comparative case studies that looked at quality
improvement across organizations. Five of the case studies compared the
implementation of QI programs in several health care organizations (Blumenthal &
Edwards, 1995; Boerstler et al., 1996; Caiman et al., 1996; Savitz & Kaluzny, 2000;
Shortell et al., 1995c); the other three included a hospital among the organizations
they studied (Lewis, 2000b; Waldman et al., 1998; Zbaracki, 1998).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
QI & Leadership
“The concept of leadership is central to all of the writing on TQM” (Amdt &
Bigelow, 1995, p. 8), and the QI literature generally equates leadership with top
management. Top managers (i.e., the administrators, the executive team) are
responsible for the implementation of their organization’s QI program, and they are
supposed to be actively involved in that change process. Shortell and his colleagues
(Shortell et al., 1998; Shortell et al., 1995a; Shortell et al., 1995c) say that
organizations must develop their QI capabilities along four key dimensions in order
to successfully achieve continuously improving performance: the cultural dimension,
the structural dimension, the technical dimension, and the strategic dimension.
Executive leaders are the only ones who are authorized to direct the necessary
resources for developing these areas across the organization.
Although the QI literature generally assumes that “leaders” are
administrators, health care organizations are pluralistic arenas, where “strategic
change.. .requires collaborative leadership involving constellations of actors playing
distinct but tightly-knit roles” (Denis et al., 1996, p. 673). Therefore, some
researchers have argued that the concept of “leadership from the top” in health care
organizations should be broadened beyond administrators, to include the board and
physician leaders (Weiner et al., 1997). As can be seen in Table 2.1, four leadership
loci have been studied in the literature on QI in health care. The roles of physicians
and of top management in leading QI have received the most attention, with 10
studies having addressed each one; five studies have looked at the role of boards of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
29
Table 2.1 Leadership Loci and Variables in Health Care QI Research
Article
(by Date)
Leaders tip Loci
Top Management Board of
Trustees
QI Council Physicians
Eubanks (1992),
Grayson
(1992) &
Koska
(1992)
enthusiasm for
TQM
time spent directing
QI
enthusiasm for
TQM
enthusiasm for
TQM
timing of
involvement
inQI
Barsness, et al.
(1993a,
1993b.
1994)
barriers to QI
implementation
involvement in
QI
clinical QI
applications
Blumenthal &
Edwards
(1995)
physicians’
perceptions of
their
commitment to
TQM
extent of and
approach to
involvement
in TQM
specific roles in
TQM
Shortell et al.
(1995c),
Carman et
al. (1996)
involvement in
QI
Boerstler et al.
(1996)
strong senior
leadership
involvement in
QI
if it exists, with
senior
management
membership
involvement in
and
leadership of
clinical QI
Lammers et al.
(1996)
commitment to
TQM
commitment to
TQM
team supervision
meeting
frequency
commitment to
TQM
Weiner,
Alexander &
Shortell
(1996)
involvement in
governance
involvement in
governance
Li (1997) leadership for
quality
Weiner, Shortell
& Alexander
(1997)
leadership for
quality
involvement in
governance
involvement in
TQM/CQI
Solberg et al.
(1998)
involvement in QI involvement in
QI teams
Arthur Andersen
(1999)
involvement in QI monitoring of
QI
% trained in QI
% involved in QI
teams
Parker et al.
(1999)
commitment to QI
Weech-
Maldonado
etal. (1999)
leadership for
quality
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
30
trustees, and only two have studied the role of QI councils (which are considered an
essential part of the QI structure).
In order to implement QI throughout their organizations, upper-level leaders
must ultimately instill QI practices, attitudes, and beliefs at all organizational levels.
Thus, for instance, Crosby’s (1984) model of a quality-oriented organization
specifies the requirements for the chief executive officer (CEO), the chief operating
officer (COO), the senior executives, managers, professional employees, and all
employees (p. 8). Although middle managers, supervisors, and line staff are often
involved in QI teams and although their areas are directly affected by the
“improvements” that QI teams recommend, very few studies have looked at the role
of lower level leaders in QI, let alone the employees who report to them. The
research on lower-level leaders in health care QI is discussed below, in relation to
their commitment to QI, and their role in achieving strategic coordination of QI
efforts between organizational levels. The empirical research on the role of project
leadership in QI is even more scant It appears to consist entirely of two surveys of
primary care clinicians (physicians in one, nurses in the other) who were assigned
responsibility for implementing or facilitating the implementation of improvements
in controlled, randomized QI intervention studies (Calomeni et al., 1999; Palmer,
Hargraves, Orav, Wright, & Louis, 1996a). This lack of attention to project
leadership is particularly shocking given the fact that, as Juran (1989) has written,
“[a]ll quality improvement takes place project by project and in no other way”
(p. 35).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
31
Commitment
The proscriptive QI literature emphasizes that commitment, and especially
the commitment of top leadership, is a necessary condition for its success. There is
growing empirical evidence that supports this claim. Very few studies have asked
respondents directly about commitment to QI. A survey of 44 primary care clinics in
the metropolitan Minneapolis/St. Paul area asked clinical personnel to indicate their
agreement with the statement “Leadership at this clinic is firmly committed to using
CQF’ on a 5-point scale (Solberg et al., 1998). Since 35.7% of respondents were
neutral and 7.0% either disagreed or strongly disagreed with this item, the authors
concluded that this indicated that there was “some ambivalence about the perceived
commitment of clinic leaders to CQF (p. 268).
Lammers et al. (1996) asked 36 organization-wide quality coordinators and
168 QI team leaders in 36 Veteran’s Administration hospitals to rate various
employees (their director, associate director, chief of staff, chief of staff for
ambulatory care, quality coordinator, all service chiefs, all medical staff, all nursing
staff, and all other workers) on how committed they were “to newer methods
advocated by Deming, Juran and others, which we know of as CQI or TQI.” They
then aggregated these scores into indices of organization-wide commitment,
physician leadership commitment, physician commitment, and employee
commitment. Organization-wide commitment and physician commitment scores
significantly correlated with the amount of training given to top management, and
they significantly predicted the hospital’s total perceived improvement scores in each
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
32
facility. The regression model with the physician commitment scale predictor
explained more of the variance in total perceived improvement than the model where
the predictor was the facility-wide commitment scale (41% vs. 32%). There is ample
evidence that active and committed physician leadership—by formal physician
leaders (Boerstler et al., 1996) or by opinion leaders who are “early converts to
TQM” (Blumenthal & Edwards, 1995)—is necessary for successful implementation
of clinical QI programs and of specific QI innovations like clinical guidelines
(Hernandez, 1998; Palmer et al., 1996a).
Most researchers and organization members infer managers’ commitment
levels indirectly, either from the actions of individual leaders (especially their
personal involvement in QI activities) or from the organization’s allocation of
resources. Some surveys have measured CEO involvement and board involvement in
QI with sets of items concerning their participation in QI activities, such as their
monitoring reports, planning and setting QI goals, and initiating QI studies (Arthur
Andersen, 1999; Li, 1997; Parker et al., 1999; Weech-Maldonado et al., 1999;
Weiner et al., 1996; Weiner et al., 1997). A survey of physicians who were appointed
to be QI leaders used similar behavioral measures (Palmer et al., 1996a). Case
studies have found that organization members also view upper managers’ personal
involvement in QI activities as indicators of their commitment (Waldman et al.,
1998, p. 190; Zbaracki, 1998, pp. 619,621). Qualitative research has pointed to
additional indicators of managers’ commitment, including leaders’ language and
discourse (Waldman et al., 1998; Lewis, 2000b), consistency of leaders’ speech with
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
their actions (“walking the talk;” Zbaracki, 1998), depth of leaders’ QI knowledge
(Blumenthal & Edwards, 1995; Zbaracki, 1998), leaders’ feedback to and follow-up
with employees about QI efforts (Waldman et al., 1998; Lewis, 2000b), leaders’
implementation of QI team recommendations (Steel, 1985; Blumenthal & Edwards,
1995), leaders’ substantive recognition of QI accomplishments (Savitz & Kaluzny,
2000), and persistence of leaders’ personal involvement (Waldman et al., 1998).
Table 2.2 gives specific examples of several of these indicators.
Table 2.2 Indicators of Management’s Commitment to Quality Improvement
Indicator Examples
Language “Another person noted: ‘Then we (top management) went on to
set up three pilot projects. So just even calling it a “pilot
project” made you realize that we don’t have total commitment
to it’” (Waldman et al., 1998, p. 192).
Consistency of
Speech & Actions
(“Walking the
Talk”)
“[A] hospital director checked the rhetoric of the CEO against
his activities” (Zbaracki, 1998, p. 621).
“Employees wondered why they had to use the TQM process
while their managers apparently didn’t.” (Zbaracki, 1998,
p. 619).
Knowledge of QI “In some cases, chief executive officers were described by
subordinates as supportive of a general quality thrust but not
fully familiar with the content and methods of TQM.”
(Blumenthal & Edwards, 1995, p. 250)
Resource
Allocation for
Implementation
“As one hospital middle manager described it, ‘You need a lot
of support from upper management to continue QI. And not
only a pat on the back, but the acknowledgment that this is a lot
of work.. .a lot of extra work.. .We didn’t quite get it. So we felt
we were let down...’” (Waldman et al, 1998, p. 188)
Feedback &
Follow-Up
“In the Manufacturing case, top management foiled to do any
follow-up.” (Waldman et al., 1998, p. 190)
Implementation of
QITeam
Recommendations
“Particularly discouraging for participants is the failure of
leadership to implement recommendations of quality
improvement groups” (Blumenthal & Edwards, 1995, p. 250).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
34
Table 2.2 Indicators of Management’s Commitment to Quality Improvement
(continued)
Indicator Examples
Recognition of
Accomplishments
“Innovating teams are often left stunned when there is no
recognition of ‘system’ savings that results from their
implemented CPI and they are asked to cut departmental
budgets along with other departments that may have benefited
from these efforts. ” (Savitz & Kaluzny, 2000, p. 373).
Persistence of
Personal
Involvement over
Time
“|T]n the Hospital and Manufacturing plant, top management
were originally involved in the QI efforts, but were then
distracted with other business matters.” (Waldman et al., 1998,
p. 189)
“At the Hospital, top management actually reduced its
involvement in the QI effort.” (Waldman et al., 1998, p. 190)
Involvement of top managers (Arthur Andersen, 1999; Parker et al., 1999)
and boards (Arthur Andersen, 1999; Weech-Maldonado et al., 1999) in QI activities
have each been found to significantly predict the extent of QI implementation, and
the success of QI performance outcomes. Li’s (1997) survey found that top
management’s QI leadership activities positively affected their hospitals’ perceived
competitiveness in health service quality through influencing organizational
cooperation, technology leadership, workforce development and information/process
analysis. Staff at five high-QI implementation hospitals viewed top management as
strong QI advocates with a clear and consistent vision for QI and a clear
understanding of QI principles (Parker et al., 1999). Palmer et al. (1996a) found that
physician leaders who modeled process improvements by adhering to guidelines and
by participating in other QI activities were able to influence their peers to adopt the
guidelines’ standards.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
35
Two case studies (Lewis, 2000b; Waldman et al., 1998) found that in the
most successful QI organizations, managers demonstrated persistent commitment
together with flexibility. Adaptability may be particularly relevant in the early stages
of QI adoption: studies have found that high performing QI hospitals and their
managers tend to adapt or customize their QI programs to their own needs, rather
than simply adopting an “off-the-shelf’ QI program (Parker et al., 1999; Westphal et
al., 1997). Later in the process, evidence suggests that implementation of QI team
recommendations is a predictor of successful QI program implementation and
performance (Arthur Andersen, 1999; Steel et al., 1985).
Several comparative case studies have discussed the negative consequences
of management behavior that conveys uncertain or hedging commitment to QI. In his
comparison of two quality circle programs, Steel (1985) argued that the program was
much less successful in the military hospital organization than it was in the military
maintenance facility the because the managers in the former organization were
considerably less committed than the managers at the medical facility. Similarly,
Waldman et al. (1998) contrasted the “wavering commitment” of management in a
hospital and a manufacturing company where the QI programs were neglected and
poorly implemented with the more persistent long-term commitment of management
in a Canadian national police force that succeeded in “going beyond the project
mentality” and institutionalizing QI as part of the culture. QI programs languished
and stalled in all four of the organizations that Lewis (2000b) studied, a phenomenon
that she called a “crisis of direction and commitment”
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
36
Executives signaled their lack of commitment in different ways. Parker et
al.’s (1999) interviews at five low-QI hospitals found inconsistency among top
managers regarding QI vision and philosophy, employees who questioned their
directors’ motives in pursuing QI (they saw it as a way for the director to get
noticed), and top managers who were inattentive to cultural factors associated with
QI implementation. Even in health care organizations identified by experts as being
at the forefront of QI, commitment was not uniform among upper managers
(Blumenthal & Edwards, 1995). Chief financial officers were particularly likely to be
skeptics, and in some decentralized institutions there were managers of major
operating units who were “openly hostile to TQM.” The authors cautioned that such
“[divisions among leaders in their attitudes toward or support of TQM undoubtedly
are recognized by physician staff and can contribute to physician passivity or
outright resistance” (p. 250).
Savitz and Kaluzny (2000) define commitment as referring to “resource
allocations associated with providing the necessary training, staff, and equipment to
facilitate the innovation and dissemination process” (p. 368). Several other case
studies have made the same inference (Blumenthal & Edwards, 1995; Waldman et
al., 1998; Zbaracki, 1998). Two case studies (Blumenthal & Edwards, 1995; Savitz
& Kaluzny, 2000) found that managers in health care organizations were often
reluctant to devote the necessary resources (especially labor and information
systems) to initial clinical QI projects until after they had proven their effectiveness
in pilot studies, and the pilot projects frequently received inadequate support. This
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
37
tentative attitude toward QI was also recognized by an employee at a defense
contractor, who commented that “when the managers implement TQM, they are
‘dating’ it, trying it out to see if they want a long-term relationship with TQM”
(Zbaracki’s 1998, p. 620). Steel (1985) also took resource allocation to be an
indicator of management’s commitment. In his study of two organizations, the
hospital with the failed quality circle program had hired a program facilitator in a
temporary, lower level position, who left partway through the study and was not
replaced, so their duties were assumed by hospital staff as additional work. In
contrast, the maintenance organization with the successful program hired a facilitator
in a higher level position, retained them through the program, and this person
attended 90% of their quality circle meetings.
At the end of the QI project cycle, researchers report that failure to
implement QI team recommendations is discouraging for participants (Blumenthal &
Edwards, 1995, p. 250; Steel, 1985), who are in effect disempowered by their
inability to effect desired improvements (Lewis, 2000b). Some research suggests that
when management fails to implement QI team recommendations, the viability of the
entire QI program is threatened (Lewis, 2000b; Steel et al., 1985). Kotter (1995;
1996) calls this “not removing obstacles” to change and identifies it as one of the top
eight reasons that transformation efforts fail in organizations. Indeed, Lawler and
Mohrman (1985) noted middle managers’ failure to implement the suggestions of
quality circles resulted in the demise of “a significant percentage of QC programs” in
American corporations during the 1980s (p. 68).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
38
It is not uncommon for health care organizations to plan extremely slow,
drawn-out implementation timetables (Carman et al., 1996; Boerstler et al., 1996;
Graves & MacDowell, 1994). There is evidence that early adopters (Westphal et al.,
1997), faster and deeper implementers (Blumenthal & Edwards, 1995), and
organizations that undertake more radical forms of QI, such reengineering (Waldman
et al., 1998) have more success with QI. Managers that adopt vague, extremely long
term (or even unending!) implementation schedules for QI programs may not be as
committed to QI as those who keep their organizations on a more timely
implementation schedule. Drawn out implementation schedules may not convey the
sense of urgency that is needed for organization-wide QI implementation. This sense
of urgency is essential to the success of most complex organizational changes
(Kotter, 1995; Kotter, 1996), even if the changes only affect one department
(Hernandez, 1998).
Challenges for Leaders Implementing QI
Why is it so hard to obtain and sustain commitment to QI? Leaders may face
a number of obstacles in implementing and institutionalizing QI. Research on these
challenges has focused on identifying and describing factors that facilitate or impede
QI implementation—leaders’ responses to these challenges have rarely been
investigated. The following discussion of impediments to and facilitators of QI
implementation also considers implications for QI leadership. QI leaders must
manage resource issues, including the expense of QI, resource scarcity due to
environmental pressures, employee turnover, and competing demands on employee
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
39
time. Leaders also need to recognize and address inhospitable organizational cultures
and recalcitrant units or groups of employees. QI projects that are grass-roots
initiatives often struggle without adequate support, and improvement projects with
innovations that are not a good fit with the organization are not always identified.
Executives must work with lower level QI project leaders to manage these projects
(individually and as a set), and to coordinate their implementation with other
organizational changes.
Managing Resources
Managers frequently balk at the investment represented by high
implementation costs for QI training, additional support staff for data collection and
analysis, upgrading information systems, and the labor expenses of releasing staff to
work on QI projects (e.g., Blumenthal & Edwards, 1995). In health care, the high
costs of QI implementation are compounded by an increasingly resource-poor
operating environment, which tends to starve QI programs both directly (e.g.,
through shrinking operating budgets and resulting layoffs) and indirectly (e.g.,
through strategic responses like mergers and acquisitions, which shift focus away
from QI efforts)/ A national survey of non-governmental acute care hospitals
J Conventional capitalist wisdom holds that increasing market pressures will ultimately lead
organizations to deepen their commitment to QI (e.g., Blumenthal & Kilo, 1998), but the
empirical evidence for this relationship is mixed. While one study found that hospitals in
markets with high managed care penetration were more likely to be involved in formal
CQI/TQM, and had adopted CQI/TQM earlier (Arthur Andersen, 1999), a different study
found that objective measures of market competition did not predict adoption of TQM in
nursing homes—only perceived levels of competition did (Zinn et al, 1998), and in a third
study, actual market competition negatively predicted the length of time a hospital had been
involved in CQI/TQM (Weiner et al., 1996). These studies report similarly mixed results for
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
40
(Arthur Andersen, 1999) found that “too many other changes going on in the
organization” was the number one organizational barrier to QI. Waldman et al.’s
(1998) case study described how a Canadian hospital let its QI program founder due
to environmental pressures: “[B]udget crisis, and interference from government
became more important. ‘We were supposed to set up a quality council and have at
least one of our senior management meetings each month only to talk about quality
issues. That could never happen. Finance always overshadowed everything.’” (p.
190). These challenges are not unique to health care. For example, the QI program in
the University Services unit of a public American university ultimately failed in the
face of serious budget cuts (Lewis, 2000b), which had caused two rounds of layoffs
and the elimination of an entire department.
Zbaracki (1998) found that turnover in management can threaten an
organization’s commitment to QI, especially when QI champions are replaced with
managers that are less enthusiastic about quality. In Parker et al.’s (1999) case-study
interviews, high-QI hospitals reported long tenure (10+ years) of key QI champions
in top management, whereas low-QI hospitals reported significant turnover in key
top management positions, such as the chief of medicine and the chief of staff (the
tenure of hospital directors in this group ranged from 18 months to 3 years). High
the effects o f market competition on various measures o f QI implementation and QI
outcomes. Weiner and his colleagues suggested that low competitive pressure “may permit a
hospital to commit greater time, attention, and resources to a multiyear CQI/TQM
campaign.” (p. 412) Shortell and his associates (1998) have also noted that “competition that
is primarily based on price or cost only works against CQI implementation” (p. 611).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
41
turnover was also a barrier to QI implementation in nursing homes (Graves &
MacDowell, 1994). This research implies that leaders who are serious about QI will
invest in their human resources to minimize turnover, and will screen candidates for
management positions to ensure their commitment to QI.
Time-related problems are frequently mentioned in the practitioner literature
on QI, and time is also a major theme in research on barriers to the implementation
of quality improvements (e.g., guidelines) in health care organizations. Both the
1993 and the 1998 surveys of non-governmental acute care hospitals listed “lack of
time” among the top three barriers to QI program implementation (Arthur Andersen,
1999). Lack of resources and time was also the most frequently mentioned barrier to
implementing QI programs in a survey of long-term care facilities (Graves &
MacDowell, 1994). Lack of time has also been identified as a major barrier to
pediatricians’ adoption of clinical practice guidelines (Cabana et al., 2000), and to
nurses’ adoption of care standards in long-term care facilities (Lekan-Routledge,
Plamer, & Belyea, 1998).
When health care organizations were first adopting QI, some health care
executives appear to have had difficulty judging how to best manage the time they
devote to participating in QI. This may have been in part due to the advice of health
care QI consultants. For example, Kaluzny and his colleagues (1993) seemed to
think that the Hospital Corporation of America was effectively “signaling
management’s commitment” by requiring their hospital CEOs to “study TQM
intensively and then teach the fundamentals to the hospital staff themselves” (p. 82),
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
42
in addition to chairing the QI oversight council. However, the SSM Health Care
System reported that the time required of senior management to teach QI courses in
their nursing homes was a barrier to QI implementation (“Thinking Barriers,” 1994).
A 1992 survey of 781 U.S. hospital CEOs found that in 39.5% of the facilities, the
CEO was primarily responsible for guiding the QI program (Grayson, 1992).
Another 21.2% said they had given this responsibility to their COO or a senior vice
president (only 22.5% had assigned it to the director of quality assurance).
Executives at 12 east-coast long-term care facilities reported that half had assigned
the responsibility for coordinating implementation to various members of the top
management team—and in four of the facilities the program coordinator was the
president or CEO (Graves & MacDowell, 1994). The authors noted that while this
arrangement has certain advantages (“it shows their obvious commitment, and
employees will see the importance and significance of the program”), the down side
is that “the CEO may not have the time it takes to coordinate such an extensive
educational effort” (p. 13).
Clinicians also face challenges in finding the time to implement QI,
especially when faced with insufficient resources for their QI projects. In one case
study of four health care systems, inadequate support meant that employees working
on improvement projects were not paid for training or education, they were not
compensated or offered relief substitution for committee services, and there were no
funds for new positions or acquiring enabling information systems (Savitz &
Kaluzny, 2000). In such cases, physicians and senior managers often spend large
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
43
amounts of their own time on tasks such as data collection and data analysis (e.g.,
Blumenthal & Edwards, 1995; Hernandez, 1998). Organizations that are very
committed to implementing QI have found that involving physicians in QI can be
very costly and strains system resources (Blumenthal & Edwards, 1995). Blumenthal
and Edwards found that in these hospitals, other physicians were having to cover for
their colleagues while they participated in QI team activities. In some organizations,
physicians participating in QI activities had to choose between seeing more patients
in less time or doing their QI work on the side, which meant sacrificing their
personal lives for QI activities.
Fortunately, the leaders in some health care organizations have realized that
making these demands on physicians “might not be a sustainable approach” to QI.
They have been exploring different solutions, such as hiring staff to fill in and paying
departments for their losses, or designing compensation schemes to alleviate the
burden on physicians’ colleagues while they cover for QI team members, cutting
back doctors’ time on teams by using them as consultants rather than as full team
members (Blumenthal & Edwards, 1995, p. 254), using physicians on “guidance
teams” (which support QI teams in areas of their specialized expertise), and using
physicians only in the initial phase of the project to help frame problems and issues,
with solutions executed by others (Shortell et al., 1995c, p. 220).
Developing Culture
Threats posed by resource scarcity interact with other factors, including
organizational culture and politics. An organization’s culture consists of a relatively
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
44
stable set of values, beliefs, norms, and practices, and these may either support or
undermine an organization’s adoption of QI (Shortell et al., 1995a). QI leaders need
to be aware of whether or not their organization’s culture is likely to embrace and
sustain the principles and practices of QI. Research has shown that hospitals with
participative, flexible, and risk-taking cultures are more successful at implementing
QI programs (Carman et al., 1996; Levin & Shortell, 1996; Parker et al., 1999;
Shortell et al., 1995b), and are able to demonstrate improvements more quickly
(Boerstler et al., 1996). Incompatible cultures need to be redirected if QI programs
are to succeed. Thus, leaders in organizations with cultures that are incompatible
with QI have a lot more work to do in implementing their QI program than those in
organizations with supportive cultures.
Larger-sized hospitals tend to have more bureaucratic cultures, which in turn
pose difficulties for implementing QI (Carman et al., 1996; Shortell et al., 1995b).
Similarly, a study of primary clinics found that larger clinics had lower QI
implementation scores than smaller ones (Solberg et al., 1998). Levin and Shortell
(1996) found that size moderates the effect of culture on QI implementation, such
that group/developmental cultures were most beneficial in large facilities (>400
beds), whereas in small hospitals (<100 beds) group/developmental cultures had such
a weak effect on QI implementation that it was not statistically significant.
Other organizational attributes are closely related to culture. For example,
teaching hospitals have different organizational cultures from those without
residency programs. The presence of attending physicians (teaching faculty),
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
45
residents, and medical students creates different dynamics with the hospital staff.
Teaching hospitals generally adopted CQI/TQM earlier than other hospitals (Arthur
Andersen, 1999; Barsness et al., 1993a), and achieved more improved patient
outcomes than non-teaching hospitals (Arthur Andersen, 1999). However, teaching
hospitals have reported more barriers to QI from both organizational infrastructure
and senior management (Barsness et al., 1993a). An early survey found that a
smaller proportion of their physicians had QI team experience than non-teaching
hospitals (Barsness et al., 1994), and other research suggests that the CEOs and
boards of these hospitals are less involved in CQI/TQM activities (Weiner et al.,
1996).
Teamwork and cooperation are also important cultural elements for QI. For
example, Blumenthal and Edwards (1995) found that physicians in a health care
system where problems were often solved in committees found it “quite consistent
and comfortable” to participate in QI teams (p. 251). Conversely, in an organization
where physicians were historically quite separate from all other staff, QI teams were
prone to difficulties such as blaming individuals and discounting nonphysicians’
contributions. Good relations between physicians and administrators tend to foster
physician involvement in QI and the advancement of clinical QI initiatives (Boerstler
et al., 1996; Shortell et al., 1995c). Parker et al. (1999) also found that
interdisciplinary collaboration and good union-management relations helped QI
implementation, and Li (1997) found that organizational coordination significantly
predicted perceived organizational competitiveness in quality. On the other hand,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
46
poor labor-management relations can be a barrier to QI implementation (Parker et
al., 1999). Cooperation and coordination may be encouraged and reinforced by
appropriate organizational structures. A national survey of acute care hospitals found
that hospitals in the top quartile of overall QI performance tended to have
organizational structures that were more aligned with core processes than with
traditional functional departments (Arthur Andersen, 1999).
Addressing Resistance
Commitment to QI not only varies depending on an organization’s culture, it
also tends to vary to across organizational subgroups. Not surprisingly, Lewis’s
(2000a) study found that change implementers who reported cooperation problems
(including low participation and lack of cooperation from managers and front-line
staff) also tended to perceive that their organizational change programs were less
successful. Parker et al. (1999) found that in low-QI implementation hospitals,
administrators overestimated their managers’ and employees’ commitment to QI. To
successfully introduce and maintain QI changes, leaders need to accurately anticipate
which individuals, groups, or units will oppose their implementation. These
considerations affect many decisions about QI, including leaders’ choice of projects.
“Projects that will meet a favorable reception take precedence over projects that may
encounter strong resistance, for example, from the labor union or from a manager set
in his ways” (Juran, 1989, p. 53). hi QI projects where employees resist the changes,
leaders must respond and address their concerns.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
47
There is evidence that commitment levels are frequently lower at lower
organizational levels, that middle management is often resistant to QI, and that
physicians are generally unenthusiastic about QI. In an early survey, hospital CEOs
reported varying degrees of enthusiasm for QI among different members of the
organization ^Hospitals!Service Master Survey,” 1992). On a 5-point scale, where
“5” was “most enthusiastic,” the CEOs gave the enthusiasm of senior management
an average rating of 4.08 and their trustees an average rating of 3.71. Nursing’s
enthusiasm rated 3.69, middle management rated 3.63, employees were given 3.59,
and the physicians came in last at 2.97. Lammers et al. (1996) found a similar pattern
of commitment among Veterans Administration hospitals: the mean commitment
score for management was 8.2 out of 10 (range: 6-9), 7.0 for physician leadership
(range: 2.5-9), 5.2 for physicians (range: 2-8), and 4.7 for employees (range: 2.7-7).
Their analyses further showed a nested structure of commitment levels across
organizational levels: “a pattern of commitment in which individuals are most
similar in their commitment to those closest to them in the medical center hierarchy”
(p. 469).
Research suggests that middle managers often resist QI implementation.
Studies have shown that middle management resistance impeded QI implementation
in nursing homes (Graves & MacDowell, 1994), and in hospitals (Boerstler et al.,
1996). A case study of QI in three organizations found that in a hospital and a
manufacturing plant that each had weak QI programs, middle management had failed
to promote the QI philosophy (Waldman et al., 1998, p. 188). Anecdotal practitioner
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
48
sources often suggest that middle managers feel threatened by QI, but the reasons for
middle management resistance do not appear to have been systematically
investigated. It is possible that many managers simply do not view QI as a priority.
Much of the practitioner-oriented literature on QI warns that physicians are
generally resistant to QI. However, research paints a more complex picture. Most
physicians are interested only in personally relevant applications of QI. A case study
of QI in 10 hospitals found that “physicians were almost universally more
comfortable with data-driven clinical studies—particularly with those focused within
their own specialty—than with the notion of hospitalwide or organizationwide QI or
of developing a continuous improvement culture” (Shortell et al., 1995c, p. 216).
Shortell and his colleagues believe that most physicians are neutral in their attitudes
toward QI, which may be related to the fact that many of the doctors they
interviewed did not see the relevance of QI to their work: “as one interviewee
expressed, ‘It’s okay if the hospital wants to do it, but it doesn’t affect me’” (p. 217).
Another reason that physicians may be perceived as resistant is that they have
been trained in the tradition of scientific reasoning, which demands a skeptical stance
toward innovation. Most physicians require evidence of efficacy before adopting
new practices, and often their standards for evidence are beyond the capabilities of
the QI teams that are attempting to promulgate change (Blumenthal & Edwards,
1995; Savitz & Kaluzny, 2000). Interestingly, Savitz and Kaluzny note that actual
standards for evidence tend to vary quite widely among health care organizations.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
49
Attitudes toward QI also vary substantially among subgroups of physicians.
Shortell and his colleagues (1995c) recommended that health care leaders take a
“market segmentation” approach toward involving physicians in QI. They identify
four different groups that require distinct strategies for involvement in QI: (1)
salaried hospital-based or group-practice based physicians, (2) selected physicians
with special interests in quality (these are the potential “champions” or change
project leaders) (3) high-admitting physicians of high-cost, high-volume conditions
(which have strategic importance for the hospital), and (4) the “neutral majority” of
indifferent physicians. Empirical evidence supports taking a segmentation approach
to communicating with physicians about QI. One group of researchers found that
whereas active-staff physician involvement in hospital governance was positively
related to CQI/TQM adoption and board activity in QI (Weiner et al., 1996), as well
as clinical involvement in QI (Weiner et al., 1997), physician-at-large involvement
in governance was negatively related to both CQI/TQM adoption (Weiner et al.,
1996) and clinical involvement in QI (Weiner et al., 1997). Similarly, Blumenthal
and Edwards (1995) found that doctors who were employees of the health care
organizations “seemed to accept TQM more readily and were more likely to lead its
implementation than were those physicians who worked in affiliated private
practices and who had a heightened sense of autonomy” (pp. 242-243). A study of a
clinical guideline implementation effort found that ancillary staff were less likely
than staff radiologists and residents to support the need for cost constraint in local
practice (Hernandez, 1998).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
50
Researchers have looked at the roles of middle managers and physicians in
QI, but they have conspicuously neglected a major occupational group. The nursing
profession is almost completely absent from the majority of health services research
on QI, and yet nurses are integral to QI in health care. In a study of six health care
organizations nationally recognized to be advanced in QI, nurses were consistently
identified as “the backbone of CPI [clinical process improvement] implementation”
(Savitz & Kaluzny, 2000, p. 372). Amazingly, this observation was made
parenthetically, as a passing comment—no explanation or further details were given.
In the Hospital magazine survey (‘ ‘ Hospitals/Service Master Survey,”1992), CEOs
reported that the nurses were the most enthusiastic group of employees regarding
QI—more enthusiastic than middle management, other employees, and the
physicians (only senior management and the board of trustees were thought to be
more enthusiastic than the nurses). This may explain why, in the 34.5% of
responding hospitals that were implementing CQI one department at a time, the
department that was most commonly mentioned as the first to adopt CQI was nursing
services. A randomized control trial of CQI implementation to improve the delivery
of preventive services in primary care clinics (Calomeni et al., 1999) is apparently
the only multi-site study that has reported on the QI experiences of nurses. A survey
of and interviews with 13 nurses who were QI team leaders or facilitators found high
levels of satisfaction with the experience of process improvement, self-reported
perceptions of increased knowledge and skills, and the shared belief that their
involvement in QI had a positive impact on their nursing role.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
51
Managing Projects
According to the QI literature (e.g., Juran, 1989), administrators are supposed
to participate in a QI council that oversees and coordinates the ongoing QI projects.
This council decides which projects will be initiated, receives progress reports and
recommendations from the project teams, and helps institutionalize the project
changes. Empirical studies suggest that certain project and innovation characteristics
have significant implications for project leaders and QI council members. Research
has found that managerial commitment varies according to the origins of a given QI
project, and that the perceived characteristics of an innovation will affect its adoption
by organization members.
One of the main goals of QI is to foster an innovative organizational culture
where employees are both motivated and empowered to initiate improvements on
their own, without direction from above. However, some empirical research of QI in
practice has found that top-down projects have an advantage over bottom-up
initiatives because “they tend to be better supported by organizational resources and
therefore experience fewer delays” (Blumenthal & Edwards, 1995, p. 257). In her
study of QI programs in four different organizations, Lewis (2000b) found that the
one case of a bottom-up QI initiative faced particular problems with regard to
legitimacy, as compared to the QI programs in the three other organizations, which
were top-down initiatives. This may explain why the bottom-up initiative involved
the lowest amount of employee training of the four programs, the least creation of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
52
reward structures (i.e., none were created), and the fewest changes in work
procedures (i.e., none).
QI project leadership is especially affected by the characteristics of the
proposed improvements. Research on the diffusion of innovations has found that five
characteristics affect the adoption of an innovation: relative advantage, compatibility,
trialability, observability, and complexity (Rogers, 1995). In general, the more that
potential adopters view an innovation as compatible with their values, relatively
advantageous (compared to existing options), trialable (i.e., amenable to small-scale
testing), and observable, the more likely they are to adopt the innovation.
Conversely, the more complex and unfathomable the innovation is, the less likely
people are to adopt it
Hernandez’s (1998) multi-hospital experimental study of a program to
implement a clinical guideline illustrates the effects of some of these innovation
characteristics. The purpose of the guideline was to reduce costs by restricting the
use of a more expensive variety of radiographic contrast agent to patients who were
identified as at risk for adverse side effects from the less expensive variety. Some
physicians were reluctant to follow the guideline because they did not want cost
concerns to influence their clinical decisions. This is an example of incompatibility
with professional values. Ultimately, the guideline was only successfully
implemented in hospitals that were in financial distress, where such costs were
generally expensed to departmental budgets (as opposed to the hospital’s pharmacy
budget). In those hospitals, the relative advantage of adhering to the guideline was
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
53
much clearer to department heads, whom hospital administrators were pressuring to
lower costs. In those hospitals, the department heads were much more likely to give
feedback to staff and to enforce the guideline with their medical and ancillary staff,
which made the innovation more observable. Hernandez suggests that one reason
that the intervention was unsuccessful overall (across all participating hospitals) is
that right before the study commenced, there was a drop in the price of the expensive
contrast agent whose use was being restricted to reduce costs, and several department
heads expected further declines in price. This obviously affected their perceptions of
the relative advantage of implementing the time-consuming cost control measures.
Health care organizations have many different stakeholders, and different
groups of adopters often have different views of an innovation’s characteristics.
These views are shaped by their values, interests, practices, and communication
patterns. Thus, for instance, administrators tend to emphasize and initiate
administrative, cost-saving improvement projects, whereas most physicians are only
interested in clinical (as opposed to administrative) QI innovations in their own area
of specialty that focus on improving patient outcomes (as opposed to lowering costs)
(Blumenthal & Edwards, 1995; Shortell et al., 1995c). These different perceptions of
innovation characteristics may help explain the differing levels of commitment to QI
programs that were discussed previously. QI leaders at different levels need
understand how the professional cultures and agendas of different groups influence
perceptions of and commitment to QL Often, local leaders are better positioned to
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
54
obtain and interpret such information (Fairhurst, 1993), which should be shared with
project leaders and, if appropriate, with the administration.
Coordinating Change Initiatives
Top managers are responsible for the strategic alignment of QI activities
across their organization. This kind of coordination is can only happen when there
are robust communication networks across organizational levels and between
departments. These networks are underdeveloped in most health care organizations
because of the traditional functional or “silo” structure that persists within the
industry.
Savitz & Kaluzny’s (2000) case study of clinical process innovations (CPI) in
four integrated health care delivery systems “revealed a uniform lack of awareness of
the overall strategic initiative” involving the diffusion of a CPI (p. 374). This is a an
example of the “visibility barrier,” which is “the gap between the strategic initiative
launched within the organization to achieve some designated goal and the emergent
learning that occurs as a result of the ongoing process of the organization” (pp. 375-
376). According to the authors, the visibility barrier is largely attributable to poor
communication flow between upper and lower organizational levels, especially
through formal communication channels with middle management. The authors state
that middle managers “serve as a conduit to senior management” and warn that the
trend toward downsizing middle managers in healthcare has increased this problem,
especially with respect to upward communication of grassroots learning.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
55
Another factor may contribute to the visibility barrier is the lack of
communication about change with and through line supervisors.
Several authors have noted the potential power of using line
supervisors to communicate information to employees (Baronas &
Louis, 1988; Larkin & Larkin, 1994; Argote, Goodman, & Schkade,
1983). These authors believe that line supervisors play a significant
role in affecting the attitudes of employees toward change initiatives.
Supervisors influence the opinions of their directly reporting
subordinates and are best positioned to translate major change
initiatives into everyday realities for employees. (Lewis, 2000b,
p. 152)
Through their day-to-day discourse, middle- and lower-level leaders implement,
embed, and sustain QI change visions within the culture by transforming tools and
concepts into institutionalized practices (Fairhurst, 1993, p. 335). However, Lewis’s
(1999) survey of 89 change implementers in a wide variety of organizations found
that line supervisors were rarely used to disseminate information about large-scale
organizational changes such as QI implementation. Her case study of four different
QI programs (Lewis, 2000b) found the same thing: supervisors did not play a
significant role in the implementation of these programs. Ignoring middle- or lower-
level managers during implementation not only threatens the depth of change, it
probably also hinders alignment of bottom-up and top-down change initiatives.
Another problem that probably contributes to the visibility barrier is the
legitimacy hurdle faced by emergent, bottom-up initiatives. If managers don’t
adequately support grass-roots initiatives that are aligned with their strategic
objectives, employees will become discouraged and leaders will fail to foster the
innovative culture that is needed to sustain continuous performance improvement.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
56
The visibility barrier can also be compounded by resistance among managers
or physicians, which can have negative effects on communication at all
organizational levels. Members’ attitudes may range from the indifference of the
“neutral majority” of physicians (Shortell et al., 1995c), to the “wait and see” attitude
of lower level employees (Lewis, 2000b), to the skepticism commonly found among
CFOs and other executives who may be unconvinced but willing to “go along”
(Blumenthal & Edwards, 1995), to the outright hostility of particular division heads
(Blumenthal & Edwards, 1995). When groups or individuals are indifferent,
resistant, or hostile toward QI, it may affect information flow among these groups in
ways that contribute to the visibility barrier.
Organizational Communication & QI
The empirical literature on QI discusses organizational leaders’ roles and
their commitment to QI, and it has investigated the conditions that pose challenges
for QI leaders, but it offers little insight into the actual process of implementing—
and especially doing—QI. Processes are certainly alluded to in articles that discuss
leaders’ and physicians’ participation and involvement in QI, and processes are
clearly implicated in discussions of problems with time (e.g., complaints that
practitioners lack time for doing QI). However, aside from looking at leaders’
actions as symbolic demonstrations of commitment, few studies have looked closely
at how the social interactions and communicative practices of organization members
affect QI processes and outcomes. This section reviews the literature on the role of
communication in QI, with a focus on the intended functions of communication in
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
57
QI; types of communication media used in QI implementation; and the research on
emergent communication in implementation processes.
Intended Functions o f Communication in QI
Many (if not most) of the functions of communication in QI are in fact
broadly characteristic of most organizational change implementation efforts. For
example, as Lewis (1999) has observed, “implementers must execute basic
communication tasks concerned with educating, informing, persuading, and
overcoming resistance to changes they seek to introduce” (p. 45). These themes are
very prevalent in practitioner-oriented publications on how to implement QI. Some
functions characterize particular phases of the implementation process. For instance,
education and training is usually most relevant near the beginning of the undertaking,
while feedback and follow-up are particularly important at later stages of the process.
The popular literature on quality improvement unanimously emphasizes the
importance of training and education in implementing QI (e.g., Crosby, 1996;
Deming, 1982, 1986; Juran, 1989). Implementing an organization-wide QI program
requires a major investment in training, because QI processes and structures are
complex and require a range of skill sets that most employees in traditional,
bureaucratic organizations have not developed. Various types of knowledge and
skills that are important for performing QL including the QI philosophy, basic
problem-solving skills, statistical thinking, basic statistical process control, employee
involvement, team building, leadership/facilitation, outcomes measurement,
competitive benchmarking, robust design, teaching/training, quality policy
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
58
deployment, and experimental design (Huge, 1990, cited in Simpson, McLaughlin, &
Kaluzny, 1994). Other skills may also need development, such as meeting
management (Lewis, 2000b), and project management (Hutchins, 1999). Different
groups of employees typically need training in specific combinations of these skills,
and the education plan must be designed accordingly.
On any given change project, customized and targeted training is usually
necessary as QI teams work to measure, design, and implement their changes.
Medical, nursing and/or ancillary personnel often must be trained to consistently
document process and outcome measures so that the team will have reliable and
meaningful data. Training and education is also usually required when the
interventions involve instituting changes in clinical practices, which is usually the
case for new guidelines, pathways, and related policies. All of this training suggests
that implementing organization-level QI programs and project-level QI changes is
extremely complex. Complex innovations diffuse with more difficulty (Rogers,
1995), which helps explain why health care organizations have been so slow to fully
implement QI.
Most health care organizations do not provide adequate QI training for
employees, and therefore most health care workers remain untrained in QI. A 1993
survey found that 43% of non-governmental acute care hospitals in the U.S. had
conducted at least some physician training in QL but on average only 14% of the
active staff physicians in these hospitals had been trained (Shortell et al., 1995c).
When the survey was repeated in 1998, it found that on average only 35% of the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
59
hospitals’ employees, and only 22% of their physicians had been trained in QI
(Arthur Andersen, 1999). Many hospitals chose to train their staff and physicians on
a targeted or just-in-time basis, in order to save money (e.g., Boerstler et al., 1996),
even in health care organizations with national reputations as QI leaders (Blumenthal
& Edwards, 1995). Even active QI team members often remain untrained in QI
techniques (Blumenthal & Edwards, 1995; Solberg et al., 1998; Steel et al., 1985).
There is evidence that clinical QI project teams that introduce new guidelines
or process innovations may not take a rigorous approach to training employees in
how to alter their documentation practices and other clinical behaviors demanded by
the new standards. Researchers have noted that hospital staff training in these areas
is often brief, informal, and haphazard (Hernandez, 1998; Savitz & Kaluzny, 2000).
Education serves several functions beyond transferring knowledge of new
practice standards. It informs employees about what is going on in the organization,
which helps reduce anxiety that can arise from uncertainty. Lewis’s (1999) review of
the literature on information dissemination during organizational change gives
examples that emphasize the importance of reducing members’ uncertainty by
addressing their information needs during organizational change. Education (together
with other forms of communication) also persuades employees that the changes are
needed and desirable. Blumenthal and Edwards (1995) found that in organizations
where many doctors had been trained in QL physicians tended to be much more
understanding of their colleagues who had to take time away from their practice in
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
60
order to work on QI projects. Thus, training is an important part of the “internal
campaign” that serves to legitimize and thereby institutionalize QI (Fairhurst, 1993).
Legitimacy is a major theme in the QI literature and legitimation processes
are closely bound to communication. Lewis (2000b) concluded that one manager’s
clumsy attempt to promote her bottom-up QI program through a newsletter actually
had the opposite effect of undermining the program’s legitimacy (pp. 147-149).
Zbaracki (1998) found that the hotel and the defense contractor in his case study
misrepresented QI progress by only reporting successes, by including successful
change stories that had nothing to do with their formal TQM programs, and by
reporting out-of-date success stories as though they were current (pp. 623,628).
These messages convinced upper management that the program was effective, while
lower-level employees were left frustrated and disappointed by the credibility gap
between these glowing reports and their personal experiences with QI.
Some authors have considered strategies for legitimizing QI in the eyes of
physicians. Near the beginning of this chapter, several authors were cited as having
noting that physicians are often put off by the management-oriented language of QI.
QI experts have advocated “selling” QI to physicians by emphasizing its
compatibility with their professional training to respect the scientific method and
rational decision making (e.g., Blumenthal & Kilo, 1998). However, translating QI
discourse out of “management-speak” and into the language of medical science is
not enough: in practice, QI practices must be seen as meeting physicians’ standards
for the scientific method, which is often difficult. Clinical data must meet high
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
standards in order to establish the legitimacy of QI in the eyes of physicians
(Blumenthal & Edwards, 1995), though there is some evidence that physicians’
informal “observed evidence bar” does vary across organizations (Savitz & Kaluzny,
2000). Many organization members also judge the legitimacy of a QI program by
gauging whether top management is demonstrating “true commitment,” or whether
they are doing QI “just for show” (Waldman et al., 1998, p. 192).
Legitimation strategies are also implicit in the practitioner literature that
suggests ways of pre-empting middle management resistance. For example,
implementers in the SSM Health Care System included middle managers in training,
so that every employee on a QI team would be trained at the same time as their
supervisor and manager; the implementers asked the teams to keep middle managers
informed of their progress; and they encouraged middle managers to become TQM
facilitators (“Thinking Barriers,” 1994; Ryan, Jean, & Capozzalo, 1992). However,
the common assumption that resistance is always harmful is a symptom of “pro
innovation bias” (Rogers, 1995). “Resistance can be beneficial to organizations ‘by
preventing the installation of systems whose use might have on-going negative
consequences’ (e.g., stress, turnover, reduced performance; Markus, 1983, p. 431)”
(Lewis, 2000b, p. 146). Lewis recommends that change leaders establish robust two-
way communication channels for listening to and dialoguing with unconvinced or
disenchanted employees. These two-way channels may transmit valuable feedback to
implementers that can enable them to adapt and improve their innovations or their
implementation approaches. By listening to and addressing employees’ concerns,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
62
they may also avoid rounds of political conflict over the changes. Because lower-
level disenchantment with QI often develops over long periods of time (Lewis,
2000b; Steel et al., 1985; Waldman et al., 1998; Zbaracki, 1998), two-way feedback
and evaluation systems should be maintained beyond the implementation period. The
institutionalization of such communication channels may be a key factor in ensuring
that the organization’s QI program is not allowed to “fizzle out.”
Feedback and follow-up are critically important aspects of any change effort,
serving many vital functions for individuals and organizations, “including improving
performance, reducing uncertainty, enhancing self-image, and managing self
presentation goals (Ashford & Cummings 1983)” (Lewis, 2000b, p. 146). Feedback
from employees to managers allows organization members to monitor the progress
of a QI project: to understand if the improvements are being achieved, and why.
Feedback to managers can also be used to channel employees’ resistance to change
efforts (Lewis, 2000b), while feedback to employees helps keep them up-to-date on
changes, and can be used to recognize the achievements of QI team members.
Feedback in either direction (up to managers or down to employees) also
serves an important legitimating function, since it demonstrates management’s
ongoing commitment to change and it makes the innovations more observable. In a
study of leaders’ self-improvement efforts, managers who gave employees updates
tended to be seen as more effective by their employees, especially if they followed
up consistently at regular intervals (Goldsmith, 1996). The reverse is also true:
people who feel cynical about organizational changes are more likely to feel
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
63
uninformed about what is going on in their workplace (Reichers, Wanous, & Austin,
1997). Research on clinical guideline implementation has repeatedly shown that
education alone is usually insufficient to change the behavior of clinical
professionals and their patients—initial education must be combined with feedback
in order to be effective (Ferris et al., 2001; Hernandez, 1998). Unfortunately,
empirical studies of QI in practice shows that just as education is often handled
poorly, feedback is frequently inadequate (Hernandez, 1998; Lewis, 2000b; Savitz &
Kaluzny, 2000). It is often conducted informally and sporadically, if at all, leaving
out significant groups (notably nursing staff), failing to address important issues
(such as skills acquisition), and containing questionable data.
When feedback suggests that a project is encountering difficulties,
management must respond in a timely manner, lest they give the impression that they
are uncommitted to the QI effort and thereby contribute to its delegitimation. For
example, Lewis (2000b) described how, at one of the organizations she studied, large
portions of the monthly staff meetings were devoted to QI progress reports. One
committee gave repeated reports, “for a few months in a row”, on their inability to
get a sign put on the building. Lewis suggested that “such publicity may have added
to the frustration experienced at this organization” (p. 149).
Aside from implementation functions (i.e., education, legitimation, and
feedback), communication may be integral to the particular process improvement or
innovation that is being implemented. QI in hospitals often focuses on improving
communication among medical, clinical, and support staff in order to improve a
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
64
process and its outcomes, be they administrative (e.g., emergency room wait times)
or clinical (e.g., medication errors). Sometimes the changes that are implemented
take the form of communication technology. For example, reminder systems that
alert physicians or patients to take certain actions during the care process have been
an effective mechanism for implementing clinical practice guidelines (Alemi et al.,
1996; Ferris et al., 2001). Reminder systems may be as simple as a standardized
order form or as technical as a computerized telephone system that calls parents to
remind them to immunize their children. Unfortunately, these simple decision aids
are not yet widespread in health care organizations (Institute of Medicine, 2001,
p. 176).
Communication Media Used in QI
Implementers use a variety of media to inform, persuade, and train employees
and to monitor the progress of their efforts to implement QI programs and project-
based improvements. Young and Post’s (1993) study of communication in
exemplary companies found that these organizations tended to use multiple channels
for communicating during crises and major organizational changes. Some experts
emphasize the need for upper management to communicate change messages,
whereas others argue that supervisors are the preferred and most important channel
of communication for line staff (see Lewis, 1999 for a review), but the multi-channel
best practice suggests that all levels of the organization should be used together to
promote, seek feedback on, and make adjustments to organizational change efforts.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
65
The organizational change literature recommends that implementers use
interpersonal channels to communicate changes, and empirical research suggests that
implementers tend to favor face-to-face channels (Lewis, 1999). Lewis’s survey of
89 implementers from a range of sectors, states, and countries found that the most
common channels for disseminating information regarding change were meetings
and small informal discussions. Much of the work of QI takes place in meetings.
Some quantitative studies have implicitly treated the frequency of QI team meetings
(Solberg et al., 1998) or QI council meetings (Lammers et al., 1996) as indicators of
commitment to QI. Indeed, as a hospital employee in Waldman et al.’s (1998) case
study observed about their QI council, QI work cannot be accomplished if people are
never able to get together and hold the necessary meetings. However, the quality of
the meetings is as important as their quantity. If project meetings are not well
managed and do not result in progress toward the team’s goals, they will be seen as a
waste of time. According to an employee told in Lewis’s (2000b) study, many
employees in their organization found CQI to be “too time consuming. A lot of time
spent in meetings and not enough work getting done” (p. 138). Dissatisfaction with
meetings also affected QI project team dynamics in this study, as discussed in
Chapter Five.
Printed messages are common in organizational change programs. Written
communication often includes meeting minutes, articles in the organization’s
newsletter, project reports, new policies related to process improvements, and other
similar documents. Lewis’s (2000b) case study of four QI programs found that
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
66
documents were created in all cases, but they were used differently in each
organization. Organizations varied in the extent to which they circulated and re
visited such documents throughout the implementation (p. 151), and the content also
differed in significant ways. The newsletter in one of these organizations ended up
contributing to the delegitimation of its QI program, as has been mentioned. Lewis
also contrasted the focused and well-written mission and vision statements of a
technology company with the uninspiring and vague mission statement of a
university’s continuing education unit.
The effectiveness of written communication depends not only on content, but
also on how documents are used. The technology manufacturer in Lewis’s (2000b)
study prominently displayed their “Quality Policy,” Mission Statement, Vision
Statement, and Values Statement throughout meeting rooms and other places in the
company. The management group consistently referred to them and used them as a
basis for decision-making, which reinforced the messages, demonstrated their
significance, and helped employees internalize them. However, when implementers’
messages in different media are inconsistent, the messages in the official documents
may be invalidated. A middle manager in another case study reported that “there was
a lot of written communication. I know from my viewpoint here in finance, my
senior management never really believed in it” (Waldman et al., 1998, p. 192).
Lewis has noted that “little empirical work has addressed precisely how
implementers tailor their use of communication channels with different internal
audiences” (Lewis, 1999, p. 53), and that “systematic research about the relative
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
67
effectiveness of communication strategies about change is scant” (Lewis, 2000b,
p. 153). My study contributes to a better understanding of how change leaders use
various media to communicate with different stakeholders. Implementers have to
communicate about change with their project teams, their superiors, and with the
people who will be affected by the changes they implement This research addresses
differences in leaders’ patterns of media use, and suggests important differences in
effectiveness of the channels in conveying different kinds of messages to these
audiences. Media use also has implications for understanding implementers’
strategies for involving employees, which is another area in need of research (Lewis
& Seibold, 1998).
Emergent Communication in 01
The above discussion of communication goals and communication media in
organizational change suggests that communicative acts—or their absence—often
have unintended consequences during actual implementation processes. This may be
one reason that managers need to demonstrate flexibility as well as commitment in
order to both implement and maintain a successful QI program (Waldman et al.,
1998; Lewis, 2000b).
Two studies have proposed process models that forefront the role of
communication and emergent organizational phenomena in strategic initiatives like
QI (Zbaracki, 1998; Savitz & Kaluzny, 2000). Savitz and Kaluzny adapted a model
by Mintzberg to describe how the visibility barrier represents a gap between
emergent learning and strategically directed learning goals. Zbaracki applied
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
68
Weick’s evolutionary model of sensemaking to describe how formal and informal
organizational discourse interacts with organization members’ experiences with QI.
Both models also emphasize how different experiences at different organizational
levels shape communication and knowledge about QI among different groups.
Zbaracki details how social norms and political pressures that encourage face-saving
and impression management lead to distorted upward information flow that filters
out stories about QI failures. Information about grassroots change efforts also tends
to get left out of upward communication (Savitz & Kaluzny, 2000). Such
communication patterns pose barriers for organizational (system) learning and for
strategic alignment of problem-solving initiatives across the organization.
Interorganizational learning is also affected, since QI success stories tend to be
spread beyond the organization, whereas QI failures are retained inside (Zbaracki,
1998).
Disenchantment may be a common unintended consequence of many QI
efforts, especially among lower-levels employees (Lewis, 2000b; Waldman et al.,
1998; Zbaracki, 1998). Zbaracki (1998) and Lewis (2000b) note that disenchantment
happens through sensemaking processes, as employees interpret their negative
experiences with QI and decide for themselves the meaning of QL regardless of any
pro-QI messages from above. Disappointment is a by-product of managerial
communication that raises expectations for change during the implementation of the
QI program (Steel et al., 1985; Waldman et al., 1998; Zbaracki, 1998); Zbaracki
(1998) notes that this pattern “follows the ‘paradox of positive value’ that seems to
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
69
occur in many implementation efforts (Sproull and Hofmeister, 1986: 57)” (p. 628).
One lesson for change leaders is that poorly executed organizational changes can
have harmful results (Steel et al., 1985). Indeed, employees’ negative feelings can
extend beyond the QI project or even the QI program in their organization—resulting
in lower organizational commitment, satisfaction, and motivation to work (Reichers
et al., 1997), negative perceptions about their organization (Steel et al., 1985) or
about a whole class of improvements, such as clinical guidelines (Hernandez, 1998),
or pessimism about future change initiatives (Reichers et al., 1997).
The literature suggests that several steps may be taken to pre-empt and deal
with employee cynicism about QI. Steel and his associates (1985) write that
organizational changes are most effective when they are “prudently initiated and
carefully nurtured, and when realistic expectations are created...(Beer, 1980)” (p.
117). This suggests that not all programs should be initiated in all organizations, but
that when managers do decide to implement them they will need ongoing attention.
Shortell et al. (1998) make a similar point, using the same organic metaphor
The weaknesses of CQI do not lie so much in the approach itself but,
rather, in the infrastructure required for its success. CQI’s major
weakness, if you will, is that it is very demanding of individuals and
organizations along multiple dimensions: cognitively, emotionally,
physically, and, some might say, spiritually. For the CQI rose to
flourish, it must be carefully cultivated in a rich soil bed (e.g., a
receptive organization), given constant attention (e.g., sustained
leadership), assured of appropriate amounts of light (e.g., training and
support), and water (e.g., measurement and data systems), and
protected from damaging pests (e.g., overly burdensome regulation
and parochial views), (p. 605)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
70
One way to address the cognitive, emotional, and perhaps even spiritual
demands of QI is to invest in robust multi-directional feedback mechanisms (Lewis,
2000b) that help people gain knowledge from QI failures so that they can succeed in
future efforts, rather than become discouraged. Reichers et al. (1997) list several
communication tactics and strategies that can help prevent and mitigate employee
cynicism. However, communication is not a panacea—it can only be effective if
employees also witness and experience some successes. Research suggests that in
order to build people’s perceptions of self-efficacy, they must personally achieve
success in challenging and meaningful endeavors; the same is true for social groups
and perceptions of collective efficacy (Bandura, 1997). That is why change experts
advise implementers to start with projects that are more likely to be successful, in
order to build support for the initiative (e.g., Kotter, 1995; Kotter, 1996). For
example, Juran (1989) says “the fust project should be a winner,” and should meet
several specific criteria that will help assure its success: it should deal with a chronic
problem that has been awaiting a solution for a long time, it should be feasible, it
should be significant, and it should be measurable (p. 52). (Note that these are
closely related to several of Roger’s [1995] characteristics: they help ensure that the
improvements are observable, not too complex, compatible with adopters’ values.)
To date, comparative QI studies have taken either an entire QI program or a
specific innovation (e.g., a clinical guideline for asthma) as their unit of analysis, and
have compared these phenomena across organizations. There is almost no empirical
research on QI project teams or leadership, in spite of experts’ comments that,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
71
* ‘[s]ince much improvement occurs at the group/team level, hospitals must...focus on
increasing group and team effectiveness, particularly with regard to communication
and conflict management” (Arthur Andersen, 1999, p. 5). In order to begin to
understand how QI projects are actually conducted, my study takes the QI project as
the unit of analysis, comparing organizational communication in three projects
within the same organization. By taking a longitudinal, qualitative case study
approach, my study could examine the actions leaders took in implementing change,
and how these actions both affected and were affected by their contexts (i.e., the
project, the organization).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
72
CHAPTER 3: METHODS
This dissertation is a case study focusing on the leadership of change initiatives in a
pediatric hospital in the United States. The research followed a study design that is
known as “collective case stud/’ (Stake, 1994) or “multiple case stud/’ (Yin, 1994).
Three change projects were examined; all three projects conducted in the same
pediatric hospital, and all three were quality improvement initiatives. The research is
an “embedded” case study (Yin, 1994), involving three units of analysis at three
different levels. The first level and unit is the organization (i.e., the hospital), the
second is the project, and the individual leaders are the lowest unit of analysis. This
study takes a qualitative approach to the case analysis, though the qualitative
analyses are supplemented with some quantitative information from a survey
conducted as part of the CHAI research project, as well as quantitative archival data
about the projects and the hospital.
In trying to understand how configurations of leadership interacted in the
three quality improvement projects, I examined the roles of various individuals who
were directly involved in projects. The analyses reported in this dissertation focus on
the people who led the three QI projects, and.how they interacted with their project
teams, as well as with the broader hospital structure. The QI activities of various
levels of hospital management, ranging from the top executives down to the front
line supervisors, were examined insofar as they pertained to the three QI projects.
Because organizational change projects are inevitably enmeshed within a
broader context, this research also considered other QI activities in the hospital and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
73
the system to which it belonged. QI activities took place at multiple levels:
departmental (e.g., departmental QI committees and employees who handle
departmental QI), hospital-wide (e.g., the hospital-wide QI council, the Medical QI
Committee) and the system (e.g., the system-wide QI council, and QI training
through the system’s Organizational Development Department). I looked at how
people “did QI” in the hospital’s QI department, in die Nursing Department, and
within several other functional areas of the hospital that were relevant to the change
projects (e.g., the laboratory, the pharmacy, die emergency department, the intensive
care unit). Among the physicians, I looked at how the formal leaders (e.g., the chief
of staff, the chief resident, the administrators of the medical school and its pediatrics
department) and physicians who were active committee members and committee
chairs each related to QI within the hospital, with an emphasis on their experiences
concerning the three QI projects of interest
Sources of Data
The Research Site
The data for this dissertation were collected as while working on the USC
Diffusion Capability Study,4 which was funded through grants from die Institute of
4 The USC Diffusion Capability Study was conducted from 1999 to 2000, and included case
studies of four pediatric hospitals and two surveys in six hospitals. All participating hospitals
were members of the Child Health Corporation of America (CHCA), a “business alliance of
38 non-competing children’s hospitals” (www.chca.com). The participating hospitals also
belonged to a subgroup of CHCA hospitals called the “Child Health Accountability
Initiative” (CHAI), which was formed in 1998. To date, the CHAI continues to operate as “a
multi-site collaborative of children’s hospitals formed to design, evaluate and implement
national measures of quality and health outcomes for children, using twelve hospitals that
care for children as learning laboratories” (www.chaioutcomes.com).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
74
Knowledge Management and the Packard Foundation. The case study research was
conducted in a hospital that was participating in the USC Diffusion Capability Study.
That hospital will be referred to here as “Hospital for Children” or “HFC.” HFC is a
regional center for tertiary pediatric care. At the time of this study, it operated 175
beds, employed 1,250 employees, had a medical staff o f485 physicians, and saw an
average of 14,000 admissions and 100,000 outpatient visits each year. HFC was
affiliated with a university medical school, and the offices of the school’s department
of pediatrics were located on site. In 1995, the hospital merged into one of the
region’s major non-profit health care systems (“the Network”). The organization
chart for Hospital for Children is given in Figure 3.1. The hospital’s top management
team consisted of the President, the COO, the VPs, the Medical Director, and two
administrative directors (the Administrative Director of Performance Assessment
and Improvement, and the Administrative Director of Diagnostic/Outpatient
Services). The directors of the various departments constituted the next level down in
the hierarchy: upper management.
HFC was located in a poor, predominantly African American neighborhood
of a mid-size American city. Many (about 60%) of the hospital’s patients were on
Medicaid. At both the state and federal level, tight fiscal policies toward health care
expenditures had directly and indirectly affected the hospital’s income. The state had
subcontracted the management of its Medicaid patients’ insurance and related
utilization issues to private managed care companies. Furthermore, the federal
Balanced Budget Act (BBA) of 1997 had negatively affected the income of most
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHIEF EXECUTIVE OFFICER
Network Healthcare System
PRESIDENT
Hospital for Children
CHIEF OPERATING OFFICER
Hospital for Children
VICE PRESIDEN"
Operations
VICE PRESIDENT
Operations
DIRECTOR
Perioperative Services
DIRECTOR
Security
DIRECTOR
Engineering Services
DIRECTOR
Environmental Services
ADMINISTRATIVE DIRECTOR
Diagnostic/Outpatient Services
DIRECTOR DIRECTOR j DIRECTOR
Rehabilitation Services CVS, Neuro Diagnostics, ! | SCU & PST
Urotiynamics
DIRECTOR DIRECTOR
Food & Nutrition i - U Radiology
MANAGER DIRECTOR
Laboratory ■ — I Emergency Department
DIRECTOR
Education Services
DIRECTOR
Respiratory Care Services
DIRECTOR
Environmental Health
ADMINISTRATIVE DIRECTOR
Nursing Administration
DIRECTOR
ICU&TCU
DIRECTOR
DIRECTOR
7 West & 4 Central
DIRECTOR
7 Central & 6 West
DIRECTOR
Behavioral Health ■ — • 6 Central & 5 West
DIRECTOR
Metabolic Unit
Figure 3,1 Hospital for Children Organization Chart
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHIEF EXECUTIVE OFFICER
Network Healthcare System
PRESIDENT
Hospital for Children
CHIEF OPERATING OFFICER
Hospital for Children
: MEDICAL DIRECTOR
_L
VICE PRESIDENT
Operations
DIRECTOR
Education Services
DIRECTOR
Respiratory Care Services
DIRECTOR
Environmental Health
ADMINISTRATIVE DIRECTOR
Nursing Administration
DIRECTOR
Health Information
Management
IECTOR
J&TCU j -
i
DIRECTOR
- j 7 West & 4 Central
i
SECTOR ;
J S PST L
DIRECTOR
- 7 Central& 6 West
: ector 1
ioral Health j -
j
DIRECTOR
- ! 6 Centrals 5 West
ECTOR
boIicUnit -
ADMINISTRATIVE DIRECTOR
Performance Assessments Improvement
DIRECTOR
Pharmaceutical Services
DIRECTOR
Psychosocial Services
SUPERVISOR
Utilization Management
CLINICAL RESOURCE MANAGER
Quality Improvement
DIRECTOR
Decision Support
DIRECTOR
Patient Access Services
STAFF ASSISTANT
Medical Staff Services
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
76
U.S. hospitals, especially teaching hospitals like HFC. Financial resources had been
tight for the Network (HFC’s health care system) since the BBA, which affected its
hospitals7 budgets. HFC and the Network attempted to keep their compensation and
benefits packages competitive, but the entire U.S. health care industry has long been
experiencing a major manpower shortage in many areas and there were always many
positions waiting to be filled in the hospital and in the Network system.
At the time of this study, QI was somewhat decentralized at HFC. Each unit
monitored its own quality measures through their respective Quality Improvement
(QI) Committees. Within many departments, there were one or two employees who
had more QI responsibilities than their peers. The amount of time that these
employees devoted to QI activities (collecting data, data entry, preparing reports, and
attending related meetings) varied considerably between departments. The hospital
had a Quality Improvement Department, which was staffed by four people: one
Clinical Resource Manager (who monitored interdepartmental quality improvement
projects and clinical pathways, and managed the department), two QI Coordinators,
and one Clinical Data Coordinator. Two hospital-wide standing committees were
dedicated to clinical quality issues: the Nursing QI Council, and the Medical QI
Committee. The Medical QI Committee included administrators, ancillary
professionals (e.g., pharmacists), nurses, and QI Dept employees, in addition to staff
physicians.
After the merger, the hospital adopted the Network’s quality improvement
methodology. The Network practice was called QIP (Quality Improvement Process)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
77
and was developed in the early 1990s by the Network’s Organization Development
Department. The QIP was a project life-cycle model, with nine steps that the QIP
project teams were supposed to complete:
1. List and prioritize improvement opportunities.
2. Define opportunity and set objectives.
3. Define customer requirements.
4. Collect and organize data.
5. Analyze data and select root cause.
6. Generate potential solutions.
7. Select best solution.
8. Implement solution and evaluate results.
9. Track effectiveness.
QIP project teams were generally formed when issues crossed departmental lines,
and they involved a certain amount of paperwork, both in initiating the project, and
in submitting reports for tracking.
The Three Quality Improvement Projects
Three QI projects were studied at the Hospital for Children: Patient Access
(PA), Mislabeled Specimens (MS), and the Asthma Clinical Pathway (ACP).
Although the data in this study were collected over the span of one year, each of the
three projects took place over several years, and implementation-related activities
continued past die point at which this study ended. This section will briefly introduce
the goals and outcomes of the three projects, and will mention commonalties
observed among them.
The Patient Access project sought “to improve the registration process so that
it [would afford their] customers quick efficient access to patient care and [would
obtain] the information needed for payment and continuity of care.” Following the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
78
project team’s recommendations, the hospital hired a new director for the Patient
Access Department. This director implemented the project team’s other
recommendations: she centralized certain pre-registration and pre-certification
functions, and standardized the registration process. A new dedicated center was
created for scheduling, pre-registration, and pre-certification. Jobs were redesigned;
employees were retrained and crosstrained. Gatekeeper policies were introduced so
that no outpatients were seen in specialty clinics without referrals, and planned
inpatient visits required advance pre-certification to be admitted. The results were
dramatic: the average proportion of planned, inpatient admissions that were pre
registered jumped from 24% to 89%, the average proportion of pre-cerdfied
outpatients went from 83% to 100%, the number of insurance claims declined due to
missing information decreased by 75% between December, 1999 and July, 2000.
The Mislabeled Specimens project attempted to decrease the incidence of
errors in labeling specimens by developing and implementing an organization-wide
policy for collecting and labeling laboratory specimens. This implementation faced
many difficulties, and the new policy was never institutionalized because most
employees, including some managers, did not appear to learn or adhere to it A year
after the new specimen labeling policy was introduced, no changes had been affected
in the rate of mislabeled specimens.
The ACP project sought to standardize the care process for the majority of
patients admitted for asthma so that physicians’ practices would be consistent with
the national guidelines. The pathway was also designed to meet educational goals
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
79
(e.g., teaching the residents how to classify their patients’ asthma), and financial
goals (e.g., ordering home nebulizers upon admission so patients would not have to
extend their length of stay because they were waiting for the equipment). The
pathway was considered to have been implemented when a team instituted
standardized clinical forms for the physicians (especially the residents in the
Emergency Department) to use with asthmatic patients. Many residents did
apparently use the asthma forms in the first six months after they were introduced.
However, by the end of this study, there were efforts underway to recruit new and
broader membership in the asthm a pathway team, because several parts of the
pathway (e.g., social workers ordering nebulizers for home use upon the patient’s
admission) still were not being followed consistently.
I deliberately chose to study three interdepartmental projects because the
literature has shown that these are generally more challenging (e.g., Boerstler et al,
1996; Jennings & Westfall, 1994) and therefore would be a richer source of data
about how leaders negotiate obstacles and enact resources while pursuing
organizational change. My choice of projects was also guided by considerations of
variational sampling (Strauss & Corbin, 1998), which is a theoretical sampling
strategy wherein the researcher looks for “incidents that demonstrate the dimensional
range or variation of a concept” (p. 210). The literature distinguishes between
administrative and clinical QI projects, which involve different kinds of units and
actors (importantly, physicians are much more likely to be involved in clinical QI
projects), thus it was useful to have one of each type in my study. The Asthma
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
80
Pathway was the clinical QI project, and the Patient Access project was the
administrative one. The Mislabeled Specimens project represented a hybrid type of
project, in that it included both clinical and administrative elements. This third,
mixed type is not discussed in the literature; because of this, I believed it could
potentially reveal some new insights into QI processes in health care organizations.
The three projects represent three different scenarios, in terms of goals,
processes and outcomes, which are summarized in Table 3.1. The Patient Access
project was initiated by the hospital’s top management, as was Mislabeled
Specimens. The Asthma Clinical Pathway was originally a bottom-up initiative, but
the administration stepped in to reinvigorate the project after it had stalled in the
implementation phase. On PA and the ACP, top managers selected the subsequent
leader, whereas the subsequent leader on the MS project was chosen by the project
team.
The MS encountered considerable difficulty in institutionalizing the practices
described in their new hospital-wide specimen-labeling policy. The ACP project was
somewhat more successful at introducing new work practices, at least over the short
term, in that the residents appeared to use the new asthma pathway forms during the
first asthma season that they were available. Usage of the forms was monitored, and
when patterns of error appeared (e.g., use of the forms with patients who were not
known asthmatics), they were corrected. Other desired outcomes (e.g., to eliminate
delays in ordering home equipment) were not achieved. The PA project, which
required the most resources and involved changes to organizational structures, also
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
81
appeared to be the most successful at achieving improvements. However, later
developments arose that potentially threatened the hospital’s ability to sustain the
changes.
Table 3.1 Summary of Project Goals, Processes, and Outcomes
Mislabeled
Specimens
Asthma Clinical
Pathway
Patient Access
Goal To eliminate the
incidence of
specimen labeling
errors
To increase adherence to
national guidelines,
eliminate delays in
ordering home equipment
To save revenue lost
due to insurance
denials by
improving the
registration process
Origins Initiated by top
management
Initiated by upper
managers (non
physicians)
Initiated by top
management, with
deadline for
recommendations
Leadership
Transition
Project team
selected subsequent
leader after initial
leader left the
organization
Top management
appointed the subsequent
leader to implement the
pathway that was
developed by the initial
leader’s team
Top management
hired a Director of
Patient Access to
implement
recommendations of
initial leader’s team
Innovation New hospital-wide
policy for labeling
specimens
Clinical pathway with
standards for treating
asthma patients.
Subsequent leader
introduced standardized
forms for physician
orders, histories and
physicals.
Reorganization of
the Patient Access
Department,
including job
redesign, a new unit
for pre-registration
and pre-certification,
new employees.
Outcomes Not Successful
(no change in the
rate of mislabeling)
Partially Successful
(residents used the forms,
and more patient
education was
documented, but home
equipment delays
persisted)
Successful
(number of claim
denials declined due
to missing
information
decreased by 75%)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
82
All three projects had experienced transitions in project leadership by the
time of this study. Two different teams worked on the Asthma Pathway at two
different points in time, each with a different leader. The PA and the MS projects
each had only one project team, but two leaders. The Patient Access project team
ceased meeting after the new leader was hired. On Mislabeled Specimens, the team
continued meeting and working under their new leader (who they had chosen after
the old leader left). Table 3.2 “names” the six project leaders5 , and identifies their job
titles, project, and order of leadership (i.e., initial or subsequent). In order to help
prevent confusion, the text will identify these leaders by their names together with
initials for their projects and a number for their order. Thus, the initial leader of the
Mislabeled Specimens project will be referred to as “Jane (MSI).”
Table 3.2 The Project Leaders
Project Leader Order & Name Title
Mislabeled
Specimens
(MS)
Initial: Jane Director of Decision Support
Subsequent: Karen Quality Improvement
Coordinator
Asthma
Clinical
Pathway
(ACP)
Initial: Dr. Song Attending Physician
(Professor of Pediatrics)
Subsequent: Dr. Newman Attending Physician
(Associate Professor of
Pediatrics)
Patient
Access
(PA)
Initial: Madeleine Director of Rehabilitation
Services
Subsequent Alexandra Director of Patient Access,
Decision Support
5 The true names o f all individuals in this study are withheld to preserve confidentiality.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
83
Rationale for Study Design
Leadership research has suffered from an over-reliance on survey and
experimental research (Bass, 1990) which has limited the kind of discoveries that can
be made about leadership (McCall & Bobko, 1990). Qualitative, longitudinal
methods are more sensitive to the ecology o f leadership (Bass, 1990), which is a
central concern of this study. This is because qualitative research designs are holistic
(looking at “the larger picture, the whole picture, and [beginning] with a search for
understanding of the whole,” [Janesick, 1994, p. 212]), and because qualitative
approaches to data analysis, such as grounded theory, aim to preserve the complexity
of social phenomena in their explanations (Strauss, 1987).
A related aspect of qualitative research is that it is committed to
understanding the context of that which is being studied. An example of this is
Strauss and Corbin’s grounded theory paradigm, which is an analytic perspective
toward qualitative data that integrates structure and process. Two of the main
components of the paradigm are a social phenomenon (such as a repeated pattern of
interactions) and its conditions (the events creating the situations, issues, and
problems that pertain to a phenomenon) (Strauss & Corbin, 1998). “Conditions
might arise out of time, place, culture, rules, regulations, beliefs, economics, power,
or gender factors as well as the social worlds, organizations, and institutions in which
we find ourselves” (p. 130). Thus, conditions are very closely connected to the
context of a phenomenon.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
84
A case study (which may be quantitative or qualitative or a mix of both) “is
an empirical inquiry that investigates a contemporary phenomenon within its real-life
context, especially when the boundaries between the phenomenon and context are
not clearly evident” (Yin, 1994, p. 13). Indeed, the present study of leadership and
organizational change does display these characteristics. Because the leaders in this
study are both acting upon their organization and are acted upon by that
organization, the boundaries of context can at times be ambiguous. Furthermore, the
boundaries of each project tend to expand and contract over time, particularly as they
move from the problem definition and analysis phase, into solution design and then
implementation. Thus, context can at times be somewhat fluid.
Case study inquiry is also particularly suited to studying complex phenomena
in that it “copes with the technically distinctive situation in which there will be many
more variables of interest than data points” (Yin, 1994, p. 13). Such conditions
existed in this study of the three change projects, as is evidenced in the reported
findings that follow.
Qualitative approaches enable researchers to probe deeply into the
relationships between social structures and social processes (Strauss & Corbin, 1998;
Strauss, 1987), and such relationships are the focus of this study. All three theories
that inform this study—sensemaking, structuration, and the coordinated management
of meaning—are concerned with such relationships. The case study approach, which
involves a variety of data sources (documentary, interview, and observational),
represented the best method for providing a detailed and reliable picture of how
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
85
leaders and other organizational actors involved in the three projects interacted, how
these interactions were affected by the local contexts of their projects, and how their
actions affected those contexts. The rich variety of data sources allowed for cross
verification of information within and between projects; such triangulation increases
the reliability of findings in case study research (Yin, 1994).
Scholars who study health care organizations have called for “[f]ine-grained
studies of the learning behaviors and practices of health care organizations as they
implement QI work” (Shortell et al., 1995a, p. 20), and for research “that provides
insight into how organizational processes and power affect the steps in the TQM
process in order to help managers better understand and manage it” (Arndt &
Bigelow, 1995, p. 12). This dissertation represents one of the few extant attempts at
such an investigation. Levin and Shortell stated that “measuring implementation and
change over time, using a longitudinal design, would be helpful” (1996, p. 16), and
several experts have explicitly recognized the need for thorough, rigorous case
studies of QI implementation processes (Bigelow & Arndt, 1995b; Li, 1997; Weiner
et al., 1997). The multiple case approach, which was taken here, increases the
strength of the study design by allowing for comparison of multiple projects within
the same study, thus offering further opportunities for discerning patterns and
relationships, and for validating or invalidating emerging theories during data
analysis (Strauss & Corbin, 1998).
The field research for this study began shortly after the pediatric hospital
agreed to participate, in late November 1999. After a few initial visits, three quality
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
86
improvement projects were identified for in-depth, longitudinal analysis. Data
collection methods included observation, participant observation, semi-structured
interviews, and organizational documents. The interviews and observations
examined how the leaders within the hospital related to quality improvement (QI) in
general, and also addressed the progress and development of the three QI projects,
with a focus on the role of the project leaders, team members, and hospital
management. All interviews and any observation notes containing any reference to
leadership or quality improvement were typed up and imported into Atlas.ti for
analysis. Data analysis involved coding these interviews and observations, reviewing
field reports that summarized the histories of the QI projects (Signer, 2001), and
analyzing findings in network and tabular displays.
Data Collection
Eight visits, each lasting 7 to 10 days, were made to the site, beginning in
November, 1999, and ending in November, 2000. Data collection included
observation of performance improvement activities and daily operations in the
pediatric hospital, participant observation in training workshops at the hospital, and
in-depth semi-structured interviews. Archival documents, including written policies,
reports, slide presentations, newsletters, and meeting minutes were consulted in order
to obtain qualitative and quantitative data that supplemented and cross-validated
information obtained in the interviews. Some of these materials were obtained from
the system’s intranet, which I was able to access remotely. Data were also gathered
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
87
from the hospital and Network system websites, the CHCA website, and the
restricted CHAI workgroup website.
Qualitative and quantitative data from documents (e.g., annual reports,
patient satisfaction surveys), interviews, observations, and a survey of hospital
managers also provided valuable insight into the context of the projects, which
included both the organizational and the institutional environment. The
organizational environment comprised the hospital, the larger nonprofit multi
hospital health care system to which it belonged, and the medical school affiliated
with the hospital. The institutional environment consisted of the local, state, and
federal health care system.
Interviews
In qualitative research, the primary criterion in sampling cases or informants
is whether the relevant choices represent significant opportunities to leam about the
phenomenon of interest (Stake, 1994). According to Patton (1990, cited in Morse,
1994), “the logic and power behind purposeful selection of informants is that the
sample should be information rich” (p. 229, italics in original). More specifically,
informants are systematically chosen in a way that will assure, “variety but not
necessarily representativeness, without strong argument for typicality,.. .weighted
by considerations of access” (Stake, 1994, p. 244). In this study, semi-structured and
unstructured interviews were conducted with four different groups of leaders in the
hospital: (1) the top management team, (2) physician leaders, (3) project leaders, and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8 8
(4) departmental QI leaders. Other employees, such as project team members, line
staff, and upper and middle managers were also interviewed.
In a longitudinal qualitative study, the sampling strategy is expected to
evolve through progressive iterations of data collection (Huberman & Miles, 1994),
until theoretical saturation is reached. Straus and Corbin (1998) discuss theoretical
saturation in terms of theoretical categories that emerge in grounded theory building.
In terms of sampling interviewees in a case study context, theoretical saturation has
occurred when the following criteria have been met: “(a) no new or relevant data
seem to emerge regarding a [case], (b) the [case] is well developed in terms of its
properties and dimensions demonstrating variation, and (c) the relationships among
[constructs of interest in the case] are well established and validated” (Strauss &
Corbin, 1998, p. 212).
With these principles in mind, this study began with interviews with three of
the hospital’s top executives and key members of the Quality Improvement (QI)
Department in order to identify appropriate projects for investigation. Next, lists of
project team members and other project participants were compiled for each of the
three projects, and 7 to 8 individuals were interviewed from each team. In order to
gain a better understanding of the hospital QI context, semi-structured interviews
were also conducted with 9 physicians in leadership positions (not including the
Medical Director, who was a member of the hospital’s top management team), all 4
employees in the QI Department, and 6 people in other managerial positions or
management-related departments at the “corporate” or system level (e.g., the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
89
Network’s Information Systems Department). Several key informants (project
leaders, hospital administrators, and QI Department employees) were interviewed
more than once. Numerous unstructured interviews also took place, several during
observation throughout the hospital.
A total of 64 semi-structured interviews were conducted, of which 59 were
taped. Extensive notes were taken at all interviews, and notes regarding
conversations were made within a 24-hour period. Handwritten interview notes were
typed up, and all tape recorded interviews were transcribed. Twenty-seven of the
taped interviews were transcribed by a paid transcriptionist; these transcripts were
reviewed against the original tape for accuracy and were corrected accordingly. All
transcripts and conversation notes were imported into the Atlas.ti software program
for qualitative data analysis.
Observation
Direct observation took place in four settings: hospital committee meetings
and presentations, training sessions for hospital employees, formal observation in the
departments most affected by the QI projects, and informal observation throughout
the hospital (e.g., in offices, the cafeteria, the hallways). In the employee training
sessions, I participated fully as a trainee in group exercises, taking quizzes, and other
assigned tasks. I took extensive notes at each event, except for certain committee
meetings where the hospital’s policy was that no-one in attendance except the person
taking minutes could record the proceedings. In those situations, my observations
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
90
were recorded promptly upon conclusion of the meeting. The handwritten notes were
typed up and imported into Atlas.ti for analysis.
Documents
Documents and related archival records (in both paper and electronic form)
were used to provide information directly about the three cases, and, more indirectly,
about their context. These documents included hospital publications for employees
(e.g., newsletters), for managers (e.g., consultants’ reports), for patients’ families
(e.g., educational pamphlets), and for other stakeholders (e.g., annual reports). Some
of these documents were stored electronically on the system’s intranet. While I was
on site, I reviewed the archived records (usually stored in files or binders) for each of
the three QI projects, and obtained copies of all available documents. These records
included meeting minutes, email messages, written policies, forms, progress reports,
and slide presentations.
Survey
The Learning, Leading & Improving (LLI) Survey was administered at the
hospital in March, 2000 as part of the USC Diffusion Capability Study, in order to
leam about cultural and structural variables related to performance improvement and
leadership in the hospital. The questionnaire was administered to HFC’s
departmental directors and the medical committee chairs, who form the hospital’s
upper management (i.e., the next full level down from the top leadership team).
Respondents were asked to rate their agreement with descriptive statements about
their hospital, their unit, and the hospital’s top management on 5-point scales that
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
91
ranged from “Strongly Agree” to “Strongly Disagree.” Six hospitals participated in
the survey, with a total o f436 respondents; 39 of these respondents were from HFC,
representing an approximate response rate of 78%. (The questionnaires were
administered by the hospitals, which did not keep track of exactly how many surveys
were distributed, so the response rate is estimated.)
The survey was analyzed using statistical techniques, including descriptive
statistics and correlations, focusing on leadership and diffusion constructs. These
analyses were reported to the participating hospitals at a presentation in June, 2000.
With respect to the comparative case study of the three QI projects, the LLI provided
valuable contextual information about the organization, including upper
management’s perceptions of the organization’s support for QI, its attitudes toward
and opinions about the hospital’s top administration, and an overview of the
organizational communication culture (Bantz, 1993) with respect to QI.
Data Analysis
Data analysis for this study involved coding the interviews and observations
in the Atlas.ti software program order to determine patterns in the leaders’ enactment
of resources through organizational systems and structures. The coding process was
neither as rigidly structured as the a priori approach preferred by Miles and
Huberman (1994), nor was it as purely inductive as classic grounded theory (e.g.,
Strauss & Corbin, 1998). Instead, it used a coding method somewhere between these
two approaches, “that of creating a general accounting scheme for codes that is not
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
92
context specific, but points to the general domains in which codes can be developed
inductively” (Miles & Huberman, 1994, p. 61).
After six intensive visits to the site over a six-month period, with data
collection focusing on QI and leadership, it became clear that a central theme of the
interviews and observations was resources. The theoretical literature on resources,
communication, and leadership was consulted in order to refine the research
questions, which suggested four categories of resources (allocative resources,
authority, information, and knowledge). Likewise, both the data and the literature (on
QI, and on communication and leadership) were consulted to identify organizational
systems and structures that were relevant for resource enactment. A field report
detailing the histories of the QI three projects (Signer, 2001) was used analyzed to
determine the sequence of phases that was common to the three projects, and to
identify the unexpected challenges that the leaders faced as the projects progressed.
These categories of systems, structures, resources, and phases provided the starting
point for coding the transcripts. Some of the categories became “families” of codes,
which is a grouping function in the Atlas.ti program (e.g., “Comm Media: Email”
was in the family “Communication Media”). In vivo codes (Strauss & Corbin, 1998)
were also used to capture relevant concepts that emerged from the data (e.g.,
“fragmentation,” “empowerment,” and “accountability”).
The codes were used to identify quotations that were summarized in matrices,
where they exemplified and highlighted relationships among the various constructs
(see Chapter Four). In looking for patterns in leaders’ enactment of resources as the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
93
change projects unfolded over time, two approaches were taken. Working from the
data coded in Atlas.ti, the program was used to create network diagrams that mapped
the relationships among various phenomena. Also, a matrix comparing the activities
and roles of various kinds of leaders in the three QI projects were constructed based
on project histories contained the field report (Signer, 2001). The data were
displayed and summarized in a time-ordered, conceptually-ordered matrix, and
cross-project time-ordered matrices (Miles & Huberman, 1994).
The data analyses focused on two main areas: (1) understanding how leaders
enacted the communication and sensemaking resources in working toward project-
related goals, and (2) understanding the dynamics of leaders’ project-related
communication with others, as they unfolded over time. These two investigations are
related, the first, being the syntagmatic structure of the communication resources
that were used and enacted by these leaders, and the second being the diachronic
evolution of the process in which they were enacted—or not. Chapter Four focuses
on the structure of these resources, whereas Chapter Five analyzes the processes of
enactment, structuration, and the coordinated management of meaning that took
place surrounding the leadership of the three change projects.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
94
CHAPTER 4: A SYNCHRONIC ANALYSIS OF
LEADERSHIP, COMMUNICATION, AND RESOURCES FOR CHANGE
How do leaders enact communication and sensemaking resources in
managing and promoting organizational change? This study addressed that question
by investigating the actions of leaders and other actors involved in three quality
improvement projects at a pediatric hospital. It focused on understanding how
leaders at different levels communicated in and about QI projects in order to
accomplish their objectives with respect to the change initiatives.
This chapter examines how leaders enacted resources for change through
organizational systems and structures. The first part of the chapter introduces several
key concepts, explaining what is meant by leadership and defining the four kinds of
social resources for leading change (information, knowledge, authority, and
allocative resources). The next section looks at examples of how change leaders in
the hospital enacted these resources in the course of negotiating four types of
organizational systems and structures (formal organizational structure, informal
networks, information systems, and communication media); it also considers
situations in which leaders did not enact resources that could have aided their change
efforts. Finally, the chapter concludes with a brief summary of patterns in resource
enactment across the three projects.
Leadership: A Definition
Northouse (1997) has defined leadership as a process involving influence
over a group to achieve a common goal (p. 3). This definition is in keeping with the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
95
interactionist premises that underlie this study. Leadership is social and involves
mobilizing others toward an objective. This implies why communication is essential
for effective leadership. Influence is inherently symbolic—it cannot take place
without communication. In organizational change efforts, leadership involves
influencing (coordinating, mobilizing) project team members, but it also involves
influencing other stakeholders (individuals and groups) who are affected by the
changes. This study looks at how organizational leaders at several levels were
involved in the QI projects, including top management (the hospital administrators),
upper management (the departmental directors), and physician leaders (e.g.,
committee chairs, the chief resident, the chief of staff). However, the analysis
focuses on the projects’ change leaders, those most directly and consistently
involved in directing and enabling the three projects. These change leaders belonged
to one of two groups: top management, or the project leaders.
Another important implication of Northouse’s definition is its emphasis on
agency. Leaders are fundamentally concerned with achieving goals—their actions
are intended to “make things happen” and to “get things done.” However, several
authors (e.g., Miner, 1982; Pfeffer, 1978) have argued that leaders are constrained by
the socio-political structures that exist within and outside their organizations, and
leadership researchers have found various individual-, group-, and organization-level
variables that moderate the effects of leadership behavior on performance (Howell,
Dorfman, & Kerr, 1986). While some have questioned the efficacy that has
traditionally been attributed to leaders (e.g., Pfeffer, 1978), it is clear that sometimes
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
96
individual leaders are able to make a tremendous difference. One clear example of
this is the case of an entrepreneur who founds and builds an organization where none
existed before.
The theoretical perspective taken here suggests a an even-handed view of
what leaders can accomplish, acknowledging that leaders are constrained, while
insisting that individuals have the potential to influence the structures within which
they operate. Structuration theory says that individuals create and recreate social
systems and structures through their actions, actions which are both enabled and
constrained by those systems and structures. For Giddens (1979), “the notion of
human action logically implies that of power, understood as transformative capacity:
‘action’ only exists when an agent has the capability of intervening, or refraining
from intervening, in a series of events so as to be able to influence their course” (p.
256). Giddens (1984, p. 176) recognizes three types of constraint on human actions:
material constraint (deriving from the material world and the physical limits of the
body), sanction (deriving from discipline and punishment), and structural constraint
(deriving from properties of the social context).
Coordinated Management of Meaning theory also implies that agency is
limited. For example, CMM looks at the ratio between the prefigurative force and the
practical force that actors believe to be operating in a given communicative situation.
Prefigurative force refers to the actors’ perceptions of the degree to which their
actions are determined by the context of their interaction; practical force refers to
their view of their own actions as more or less motivated by their own goals and thus
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
97
under their own control (Cronen, Pearce, & Harris, 1982). Like CMM, sensemaking
theory understands action as enmeshed within dynamic cycles of interaction. Weick
(1979, 1995) notes that individuals enact their own environments. However, action is
also embedded within and constrained by causal systems that may be loosely or
tightly coupled, and where relationships among operating variables are not
necessarily known or even predictable (Weick, 1979; Weick & McDaniel, 1989).
Taken together, the three theories (structuration, CMM, and sensemaking) adopt a
perspective on human action that might be called—with inspiration from Herbert
Simon—“bounded agency.”
The leaders in this study faced both material and structural constraints. The
daunting environment in which children’s hospitals have been operating for some
time was described extensively in Chapter One. Revenues have declined, financial
resources are scarce and tightly controlled by managed care companies, and
regulation makes increasing demands for quality and compliance. Although the
hospital in this study was very efficiently run compared to other pediatric hospitals,
it was still feeling fiscal pressures, as could be seen in budget cuts and widespread
shortstaffing.
On top of the material constraints was the relatively new reality of belonging
to a larger health care network. (This structural reality had come about in response to
the fiscal environment: HFC had merged with the Network for financial protection.)
Many decision making processes took longer than they had before the merger. Also,
the new, bigger Network departments (e.g., Human Resources, Information Systems)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
98
tended to be less responsive than what the HFC employees were used to.
Furthermore, a number of HFC employees with QI experience felt that the Network
generally had a less open and less innovative culture regarding problems and how to
solve them. These structural constraints are consistent with what is known about
organizational culture and operational processes in large bureaucracies.
With these contextual constraints in mind, we can turn our attention to the
different kinds of resources were enacted by the change leaders involved in the three
projects. The term “change leaders” refers to both the project leaders and the
hospital’s administrators who sanctioned and supported their efforts. Research has
shown that two types of leaders are critical for implementing innovations in
organizations: product (or process) champions, and management (or executive)
champions (Frost & Egri, 1991). Product champions may come from any level or
unit of an organization and rely heavily on forming coalitions through their
communication networks. In contrast, management champions are top-level
administrators who “serve to buffer the innovation process and the activities of the
product champion from outside interference as well as procuring the needed time and
resources” for the change effort (Frost & Egri, 1991, pp. 266-267). Top managers are
often “controllers,” actors with the power to permit or stop a change effort (DeLuca,
1984). The research on QI in health care organizations, as reviewed in Chapter Two,
further underscores the importance of active support from top management in
achieving performance improvements. Thus, although some people equate change
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
99
leaders with project leaders, this analysis also considers the activities of
administrators in enacting resources for change.
Chapter Three described how each of the three QI projects at the hospital had
been led by two project leaders, an initial leader and a subsequent leader. The data
analyses that follow, in Chapters Four and Five, tend to report less on the initial
project leaders than on the subsequent leaders. This is largely because of the timing
of the fieldwork, which was conducted from December 1999 to December 2000. At
that time, the subsequent leaders had already taken over each of the three projects;
they had started leading their projects around mid-1999. The initial leaders of the
Mislabeled Specimens and Patient Access projects had worked on these projects in
1998 and 1999; the initial team worked on the first version of the Asthma Clinical
Pathway from 1994 to 1998. Interviews were conducted with people who had
worked with all three initial leaders, and with Madeleine (PA1). The initial leaders of
the other two projects were unavailable for interviews. Because members’
recollections about the projects under the subsequent leaders were more current than
their memories about the earlier efforts, and because the subsequent leaders gave
multiple interviews, the data about these later leaders is more abundant and detailed.
This difference in data richness is evident in the analyses, which give more
information about the subsequent leaders.
Resources for Leading Change
Scholars have long recognized the importance of resources in understanding
social dynamics, especially power dynamics. An entire body of literature on resource
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
100
dependency between and within organizations grew out of the work of Peter M. Blau
and Richard M. Emerson on power in social structures. Several theorists have noted
that actors derive certain forms of power from their control over resources (e.g.,
Clegg, 1989; Giddens, 1984; Pfeffer, 1977). However, most of the literature on
power and resources in organizations ignores the role of communication in these
phenomena, and those writings that do address communication tend to view it as
peripheral and subordinate to the deterministic conditions of resource relationships
(Mumby, 2000). Rather than viewing communication merely as a means of
representing resources, this study considers communication to be constitutive—that
is, the main process of enacting and instantiating resources.
Theorists have traditionally conceived of resources as tangible, material
entities, but some writers have departed from this view, adopting instead a relational,
circulatory conception of resources. Giddens (1984) emphasized that phenomena
only become resources when actors put them to use in social interaction, and claimed
that material resources could only be developed through the transmutation of
authoritative resources (p. 260). Clegg (1989) asserted that resources are contextual
phenomena whose natures are determined by social relations (p. 217), and that
resources are transformed through social interaction (p. 209). He saw resources as an
elemental aspect of power, and argued that power moves “through circuits in which
rules, relations and resources that are constitutive of power are translated, fixed and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
101
reproduced/transformed” (p. 211).6 If resources are realized, transformed and fixed
through social interaction, then it follows that communication is of central
importance to these processes.
In her classic study of innovation and change leadership in six companies,
Rosabeth Moss Kanter (1983) identified three “organizational power tools,” which
she described as three ‘“basic commodities’ that can be invested in action:
information (data, technical knowledge, political intelligence, expertise); resources
(funds, materials, space, time); and support (endorsement, backing, approval,
legitimacy)” (p. 159). Her list is similar to the set of resources that were important
for the change leaders in the Hospital for Children. These four types of resources
were allocative resources, authority, information, and knowledge.
Allocative resources involve control over material phenomena, such as goods
or objects (Giddens, 1979; 1984), as well as control over monetary wealth and
related derivatives. Giddens notes that allocative control over materials derives at its
base from human domination over nature, and that it forms the basis of economic
institutions. In the context of an organization, allocative resources are often
controlled by mechanisms such as budgets, which are subject to hierarchical
approval processes. Because their salaries or wages are budgeted, human resources
constitute a form of allocative resource in organizations. In this hospital (where
6 This view was influenced by Foucault, who wrote that power is primarily a relation o f
force, one that exists only in action, and that must be analyzed as something that circulates
(1976/1980, p. 89).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
102
budgets were tight), managers considered FTEs (full-time employees) to be precious
assets.
Authority refers to control over persons or actors (Giddens [1979; 1984] uses
the term “authorisation resources” or “authoritative resources”). Giddens notes that
authority relations form the basis of political institutions in societies. In
organizations, authority is closely linked to the command and control structure of the
formal hierarchy. In an early article, Bennis and his associates (Bennis, Berkowitz,
Affinito, & Malone, 1958) defined authority as the legitimized right to reward and
punish other organizational members, a right that is “legitimized” in that it is derived
from a formal position (p. 144). These authors noted that influence or power is not
always coterminous with the formal authority structure, and they claimed that when
leaders do not have adequate control over the punishment and reward system, then
the stability of the organization may be compromised.
Change initiatives can involve a significant investment of allocative resources
(i.e., employee work hours) in project participation, and they often involve
influencing others to adopt new behaviors and work practices. These dimensions of
organizational change are among the many factors that contribute to political
conflicts surrounding innovations in organizations (DeLuca, 1984; Frost & Egri,
1990a; Frost & Egri, 1990b; Frost & Egri, 1991). Conflicts are more likely in
organizations where resources are scarce (Pfeffer, 1977), as in the case of this
hospital. Such political conflicts may include challenges to actors’ legitimacy and
authority. In this study, leaders exercised authority through formal empowerment of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
103
individual employees (e.g., hiring, promoting, appointing them; assigning duties to
them), through giving a mandate to a team or committee, and through activities
concerning formal policies and protocols (e.g., developing, approving, enforcing).
The discussion of the literature on QI and leadership in Chapter Two showed that the
legitimacy of a change initiative can also be inferred from other actions of
organizational members who are in high positions of authority (e.g., attending
meetings). This implies that indirect or informal symbolic acts should also be
considered in order to understand the enactment of authoritative resources.
Of the four types of resources, information is the most essentially symbolic.
This symbolic dimension can be seen in Davenport and Prusak’s (1998) distinction
between data and information. Data are “a set of discrete, objective facts about
events” (p.2), whereas information is a message, which is communicated between a
sender and a receiver. Thus information is “data that makes (sic) a difference” (p. 3);
it is meaningful, hence it has social implications. In organizations, information flows
to and from internal and external actors, through many different media.
Some organizational theorists have treated information as subordinate to
allocative resources. For example, Pfeffer (1977) acknowledges that control over
information can be used to influence decisions regarding financial resource
allocation in an organization, but he fails to recognize how central information is to
every aspect of organizational operations. Information is critical to organizational
stability and change; it is just as necessary for organizational functions as allocative
resources and authority are. Information processing is necessary for interpreting the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
104
organizational environment (e.g., Daft & Weick, 1984), and for internal decision
making (e.g., Connolly, 1977). Without information resources, organizations could
not get core tasks done, monitor performance, or develop and implement
innovations. Information is especially critical in the everyday operations of health
care organizations (Kreps, 1988), where clinical decisions are heavily dependent on a
complex and continuous stream of information.
Organizational change creates information needs above and beyond those
related to everyday operations. On QI projects, information gathering and
interpretation are particularly intense surrounding problem analysis and during the
post-intervention monitoring phases. Information dissemination is greatest during the
implementation phase. A primary task for change leaders is to manage the flow of
information over time.
Chapter One noted that health care organizations deal with large amounts of
ambiguous information. Weick (1995) has explained that uncertainty in
organizations can be resolved by obtaining more data about the object of uncertainty,
but confusion can only be resolved by choosing an appropriate framework for
interpreting ambiguous information. When confusion is the problem, individuals
need knowledge resources to make sense of the information.
Just as information derives from data, so knowledge, in turn, derives from
information. More specifically.
Knowledge is a fluid mix of framed experience, values, contextual
information, and expert insight that provides a framework for
evaluating and incorporating new experiences and information. It
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
105
originates and is applied in the minds of knowers. In organizations, it
often becomes embedded not only in documents or repositories but
also in organizational routines, processes, practices, and norms.
(Davenport & Prusak, 1998, p. 5)
Weick (1995) enumerates a variety of knowledge resources for sensemaking,
including ideologies, premises (i.e., unconscious assumptions), paradigms, theories
of action (cognitive maps, mental models, and knowledge structures), traditions,
stories, and vocabularies. People draw on these frameworks when they interpret
ambiguous or equivocal information in their environment. Following the principle of
requisite variety, Weick has argued repeatedly (1978,1979, 1980,1995) that leaders
need a repertoire of knowledge that is as rich and varied as the information in their
environment. In the present study, several different kinds of knowledge were found
to be relevant to change leadership: knowledge about QI, knowledge about the
specific problem area, knowledge about the innovation, and managerial knowledge
(e.g., project management knowledge, meeting management knowledge).
Regarding these four types of resources, this study considered the following
question:
Research Question 1: How do the communication practices o f change
leaders enact (a) allocationresources, (b) authoritative resources, (c)
information resources, and (d) knowledge resources in organizational
change projects?
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
106
Leaders Enact Resources through Systems and Structures
Social systems and structures are sustained and altered through the actions of
individuals, including leaders (Giddens, 1979; 1984).7 For example, the Medical
Director of the Hospital for Children had an extremely rich and wide-reaching
communication network among the hospital’s physicians. He consistently attended
all medical committee meetings, ate lunch in the physicians’ dining room, carried on
many informal conversations with physicians and hospital staff (both in person and
by telephone), and was also very responsive via email. In maintaining his informal
network of relationships with the physicians, the Medical Director was able to enact
various resources that were useful in his job. His activities reinforced his positional
authority and legitimacy in the formal organizational structure by showing that he
was a responsive and accessible leader. Networking also gave him access to a
tremendous amount of information about the events in the hospital, as well as the
perceptions and beliefs of the physicians and other groups.
When individuals enact resources, their actions circulate these resources
within social systems. An example from the leadership research can be seen in the
“Pelz effect.” Pelz (1952, cited in Lee, 1997) found that subordinates were more
satisfied with their supervisors’ supportive behavior when the supervisors had more
' Structuration theory conceives o f systems as “reproduced relations between actors or
collectivities, organized as regular social practices” (Giddens, 1979, p. 66), and structures as
“rules and resources, or sets o f transformation relations, organized as properties o f social
systems” (Giddens, 1984, p. 25). Giddens’s technical usage will not be used in this study.
Instead, the common usage o f terms such as “organizational structure” and “information
system” will be employed.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
107
influence over their superiors. Lee has suggested that resources may help explain the
dynamics of the Pelz effect. Lee’s (1997) study was based on Leader-Member
Exchange (LMX) theory, which posits that workers have differing relationships with
their bosses, and these relationships may be ranked, according to quality and
closeness, from high (in-group) to low (out-group). Lee found that the quality of
superiors’ LMX with their bosses predicted the cooperative communication behavior
of the superiors’ subordinates. He suggested that “those subordinates whose
superiors had the higher qualities of LMX with their bosses are more likely to come
to possess greater amounts of ‘trickle down’ resources than their counterparts whose
superiors had the lower qualities of LMX with their bosses” (p. 279).
According to Giddens (1984), resources not only generate power over people
and things, they also possess “transformative capacity” with respect to structures and
systems. The transformational character of resources is bound up with other social
institutions, including social norms and symbolic orders (e.g., language). With
respect to leadership, such a transformation can be seen when an individuals’
demonstrated ability to access resources results in their emergence as a leader. This
could be seen in some of the reasons that members of the Mislabeled Specimens
(MS) team selected their subsequent leader (Karen) to head the team after their initial
leader (Jane) left the hospital. An MS team member said that the Karen (MS2) was
motivated, vocal, and committed, so the team believed that she would move the
project forward. That team member said the group believed that their new leader
“could get the resources we needed, like meeting space from Facilities” (i.e.,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
108
allocative resources). Other assets that were mentioned included Karen (MS2)’s
network of relationships that she had built while working in the Emergency
Department, her history of involvement in QI (i.e., QI knowledge), and her respected
reputation in the hospital (which conferred legitimacy, or informal authority).
The relationship between resources and organizational structures was also
illustrated in a quote from an upper-level manager regarding some implications of
the hospital’s fiscal situation:
They keep cutting the budget, so you’re not going to get caught up.
I’ve always been taught that good managers make sure they provide
their resources that the staff needs to accomplish their goals. If you
don’t have the money and the resources to give them the equipment
and the supplies and the manpower they need to accomplish those
goals, that is a frustration. It leads to low morale, and low
productivity.
This quote shows that this manager saw the acquisition of allocative resources as an
important aspect of her job. Her words, and those of the MS team member, suggest
that resources are intimately connected to efficacy of both formal and emergent
leaders.
Resources are enacted and circulate in social systems through
communication. Change leadership involves many communication events (meetings,
presentations, telephone calls, and emails), as well as various communication
strategies (e.g., bargaining, negotiation, and coalition building; Kanter, 1983) in
order to enact and access organizational resources. For change leaders to be
effective, it is important that they understand how their organizational systems and
structures generate, transform, and distribute information, knowledge, authority and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
109
allocative resources. Leaders may need to alter or extend organizational structures
and systems in order to access resources for improvement efforts. Furthermore,
structures and systems may need to be reorganized in order to change the work
processes that need improvement. These issues suggests the following research
question:
Research Question 2: How do change leaders enact resources through their
reproduction and transformation o f organizational systems and structures,
including (a) formal reporting structures (as documented in job descriptions and
organization charts), (b) informal communication networks, (c) communication
media systems, and (d) information and communication technology?
Formal Organizational Structure & Leadership
An organization’s formal structure consists of a relatively stable “pattern of
relations among positions” in the organization (Monge & Eisenberg, 1987, p. 305).
Monge and Eisenberg note that each position comes with a set of roles, and a set of
attendant expectations regarding behaviors. In medium- and large-sized
organizations, structure also encompasses the division of labor among employees,
the functional differentiation of subunits, and “the hierarchical embedding” of these
subunits and the employees in them (McPhee & Poole, 2000). The classical, or
“positional,” view of communication structure in organizations took a deterministic
view of the relationship between formal structure and communication: that is, that
structure determines who communicates, about what, and how in organizations
(Monge & Eisenberg, 1987). More recently, some theorists have argued the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
110
opposite—that communication constitutes formal structure—though others have
questioned the viability of this perspective (McPhee & Poole, 2000). The data in this
study suggest that organizational structures constrain and enable communication, but
that they also need communication in order to be actualized, to be produced and
reproduced.
The leadership literature also takes two different approaches toward
understanding organizational structure. The dominant perspective in the literature
takes structure for granted, assuming that all leaders are managers, particularly
executives or adm inistrators. Another area of research looks at leadership in small
groups, and tends to see structure as emerging out of the group dynamics that lead to
role differentiation. Very little has been written about com m unication patterns
amongst project leaders (emergent or appointed) and formal organizational leaders
(at any level of the managerial hierarchy).
Organizational QI structures are the committees, councils, teams, and formal
relations of accountability that relate to QI in the organization. Most experts consider
these structures to be necessary for implementing QI programs (e.g., Shortell et al.,
1995a). The research on these structures has largely been in the form of descriptive,
diffusion studies in hospitals (Arthur Andersen, 1999; Barsness et al., 1993a),
nursing homes (e.g., Graves & MacDowell, 1994), or medical school departments
(e.g., Graz, Vader, Bumand, & Paccaud, 1996). Very little research has looked at
how QI structures operate, or how they affect QI performance. A survey of Veterans
Administration hospitals found that the QI council’s involvement in QI activities was
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
I ll
related to how often they met, and was a significant, positive predictor of the
hospital’s number of QI teams (Lammers, 1992). A comparative case study found
that hospitals that worked on QI without establishing a quality council that included
senior management had more difficulty making progress on interdepartmental
quality issues (Boerstler et al., 1996).
QI projects, like all organizational change initiatives, inevitably involve
multiple structures. Two kinds of structure that are involved in most change projects
are the project team (usually a temporary structure), and the more enduring
organizational structure that exists outside of the project team (i.e., the departments,
divisions, facilities and other units that comprise the organization). On the three QI
projects at HFC, four of the six project leaders were appointed by the
Administration; two were emergent leaders, asked to lead by others involved in the
change project. The directors of departments that were most affected by the change
chose representatives to serve on each of the teams. Often they chose employees who
were responsible for QI activities in their departments. Individuals from other units
were also involved with one or more of the projects, including top managers from
Administration, and employees from the QI Department and from the Risk
Management Department. As stated earlier, the following analyses focus on the
actions of the administrators and the project leaders.
According to Hackman (in press), effective team leaders “tweak the
organizational structures and systems so they provide teams with ample support and
resources” (chap. 7). Administrators also engage these systems and structures to
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
112
enact resources in ways that can either support or impede a change project. Table 4.1
gives examples of how the hospital’s top managers enacted resources for the change
projects through their interactions with the organization’s formal structure, to
varying degrees of success. Overall, top managers mostly enacted allocative
resources, authority, and knowledge resources through the formal organizational
structure for the three QI projects. They did not actively enact information resources
for these projects through the formal hierarchy.
Top managers are able to shape organizational structures through hiring,
promoting, appointing, and reorganizing departments and their relationships to each
other. The Hospital for Children and the Network had both experienced many
changes to their organizational structures by the time of this study. One employee
said “I think we get a new organization chart every two or three weeks.” The
restructuring of the QI Department in 1998 freed up allocative resources for the ACP
and MS projects. It enabled the Clinical Resource Manager to assume responsibility
for the clinical pathways and to devote some of her own time to supporting Dr.
Newman (ACP2), the subsequent leader of the Asthma Clinical Pathway project. It
also enabled Karen (MS2), a nurse working as a QI Coordinator in the QI
Department, to devote time to the Mislabeled Specimens project.
Due to the difficult financial situation facing the Hospital for Children and its
Network, the administrators were very conservative about expending allocative
resources. Thus, when the QI Department was restructured, they were careful to do it
in such a way that only one FTE (full-time employee) would be added to the staff.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
113
Table 4.1 Top Managers’ Enactment and Non-Enactment of Resources through
Formal Organizational Structures
How Administrators Enacted Change
Resources
How Administrators Did Not Enact or
Unsuccessfully Enacted Change Resources
ALLOCATION:
A committee of the hosnital’s administrators
and directors restructured several departments
and jobs to assemble a QI Department that
would support the hospital’s QI efforts,
especially clinical initiatives. Only one
additional FTE (full-time employee) was hired.
The Administration save Dr. Newman
(ACPI) administrative and data collection
support from the QI Department. The Clinical
Resource Coordinator was her “lieutenant,”
and another QI employee collected data.
AUTHORITY:
The COO’s attendance at earlv meetinss o f the
PA project showed some team members that
Administration was committed to the Droiect
The PA team recommended centralization of
the registration processes under one director for
the whole hospital. When the Administration
hired this new Director of Patient Access, she
became the project’s subsequent leader.
KNOWLEDGE:
Administrators selected Madeleine fP A ll and
Jane (MSI) to lead their projects because both
departmental directors had extensive
experience with QI initiatives. They hired
Alexandra (PA2) to be Director of Patient
Access because she had implemented similar
changes at another pediatric hospital.
Jane (MSI) did not have a clinical
background, so the administrators thought that
she would be able to approach the problem
“from a fresh perspective.” The Medical
Director asked a general pediatrician to be the
Dr. Newman (ACP2), and to implement the
pathway. He later thought her expertise as a
generalist may have given her a “big picture”
perspective on the pathway, compared to Dr.
Song (AC Pl)’s specialist viewpoint
ALLOCATION:
After they implemented the ACP forms, the
COO suggested that the ACP team merge
with the Asthma Group to reduce the number
of committee meetings. When the ACP team
realized that the group was only meeting
about once every six months, they requested
permission to split off again.
The Administration did not give
administrative help to the Dr. Song (ACPI)
or any leaders on MS or P A , so team
members shared data collection and
administrative tasks. Karen (MS2) and
Alexandra (PA2) handled the bulk o f their
own data analysis and reporting.
AUTHORITY:
Alexandra (PA2) only remained Director o f
Patient Access for a year before the
Administration promoted her to
Administrative Director of the Performance
Assessment and Improvement Division. This
left the Patient Access Dept without a
director, against the PA team’s original
recommendations.
INFORMATION:
At the start o f this studv. the Administration
had decided to establish an oversight
committee (of administrators and directors'!
that would ensure that the directors received
adequate information about the QI projects.
Also, a cadre o f advisors (“facilitators”) were
trained to help the leaders on the QI teams.
Neither the committee nor the facilitators
appeared to be active during this study.
Team members on two projects reported
problems conveying or obtaining information
because they could not attend upper and top
management meetings.
Key; underline = top management leader; bold = project leader; MS = Mislabeled Specimens project;
PA = Patient Access project; ACP = Asthma Clinical Pathway project; 1 = initial project leader, 2 =
subsequent project leader
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
114
Also, because clinicians’ time is a valuable allocative resource, administrators made
conscious efforts to maximize it by reducing the number of active committees.
Shortly before this study began, the Medical Director had consolidated several
medical staff and related clinical committees in order to reduce the number of
meetings that doctors and others had to attend. Thus, it is understandable that the
COO would try to save meeting time by having the Asthma Pathway project team
merge into the Asthma Group. It appears that the COO did not anticipate that the
Asthma Group would meet only once over a six-month period, stalling the progress
of the Pathway initiative. The Pathway team eventually requested to split off and
begin meeting again on their own, which the COO approved. In this case, it appears
that the COO was attempting to institutionalize the Pathway team by merging them
into an existing organizational body, but it also appears that she did not have
sufficient information in making her decision (i.e., information about the ongoing or
expected needs of the Pathway team, and about the activities of the Asthma Group).
Administrators enacted knowledge and authority through the appointment or
hiring of project leaders. Using their own knowledge of the various leaders’ skills
(e.g., QI expertise, knowledge from experience implementing similar organizational
changes), the top managers accessed those strengths through assigning individuals to
project leadership roles, in one case as a department’s director. A couple of
administrators used the word “empowerment” to describe this process of assigning
roles or responsibilities to employees. The choice of project leader also had
allocative implications. When project leaders were appointed from within then
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
115
project activities took time away from other duties; when the leader was hired on
from outside to lead the project (e.g., Alexandra [PA2], who became the Director of
Patient Access) then the there were other expenses (e.g., salary, benefits).
Ironically, the administrators who enacted resources for the Patient Access
initiative by hiring Alexandra (PA2) to be the Director of the Patient Access
Department were the same administrators who withdrew those resources a year later
by promoting her to Administrative Director over the Performance Assessment and
Improvement Division. After the promotion, they did not hire anyone to be the
Director of Patient Access, instead leaving the position vacant. This directly
contradicted the recommendations of the PA team, which had called for hiring a
director who would be “accountable” or responsible solely for the registration
functions throughout the hospital.8 Alexandra worried that without a Director over
Patient Access, the hospital might not be able to maintain the performance
improvements that were made when she held that position.
Meeting attendance was a major theme in the discourse of this organization,
one which is further explored elsewhere in this study. Through their attendance at
project team meetings, administrators conferred authority and legitimacy on the
change initiatives. One employee noted that the COO had attended several meetings
of the Patient Access QI project, and said that by “making her presence known,” the
8 In fact, even when she was the Director o f Patient Access, Alexandra was not exclusively
dedicated to that function, since she also held the position o f Director o f Decision Support
As Director o f Decision Support, she was the only employee in that department, and
therefore handled all analyses on her own.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
116
COO showed “that she cared, and that this was a decision that she wanted to help us
to make.” Another employee observed that the participation of top managers was
extremely helpful for change initiatives at the hospital: “On any committee that I’ve
been in that’s really set up to try to do something, it seems to make a huge difference
if you have somebody from Administration that has some authority being involved in
the committee.”
Administrators’ attendance at QI project meetings was important, but so was
team member access to standing hospital committee meetings. Operational meetings
were an important venue for communicating about the QI projects, especially when
implemented changes were being monitored and revised. Team members on two
different projects expressed frustration at their lack of access to meetings of top and
upper management. One of these member was concerned that important information
was not reaching her, whereas the other wanted to attend so that she could convey QI
information to the directors. In both cases, rank affected access to meetings—and
therefore determined access to communication channels.
Meetings were also essential for enacting organizational structures. The
Oversight Committee that was reportedly established at the start of this study was
only mentioned in one interview, early on in the research. It does not appear that this
body met at any point over the year of data collection. The purpose of the Oversight
Committee was to ensure that information about the projects got to the departmental
directors so that they would not be surprised by events such as the introduction of
new policies. Surprises had occurred because representatives on the project teams
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
117
did not always function as reliable communication links to their departments. The
Oversight Committee was top management’s attempt to establish a structure that
would facilitate communication between the QI project teams and upper
management. But because the Committee did not meet, it only existed in name, not
in practice. This supports Giddens’s argument that structure can only exist insofar as
it is reproduced through human action. McPhee and Poole (2000) have recently
argued that “there may be a tendency in organizational communication research to
overemphasize socially created or symbolic features at the expense of material
aspects of structure” (p. 538), but the case of the Oversight Committee shows how
social practices (e.g., meetings) are inextricable from the very existence of structure.
Table 4.2 lists various examples of how the project leaders enacted, and did
not enact resources for their change projects through the formal structure of the
hospital. If top managers were not concerned with enacting information, the project
leaders were much more so. The project leaders reported on their progress to the
Administration (either in written form, orally, or at the QI Council meetings); some
trusted that information flow to the affected departments would be handled by the
administrators. During periods when the teams were meeting, the leaders assumed
that team members would report on the projects to their respective departments. The
team members were also a source of knowledge for the projects. The project leaders
and department directors selected participants from a range of job levels so that the
team would benefit from different perspectives on the issues. A majority of team
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
118
Table 4.2 Project Leaders’ Enactment and Non-Enactment of Resources through
Formal Organizational Structures
How Project Leaders Enacted Change
Resources
How Project Leaders Did Not Enact or
Unsuccessfully Enacted Change Resources
INFORMATION:
All project leaders reported on their progress
(especially regarding implementation and
monitoring results') to the administrators.
Updates were given in written reports and in
person at meetings o f the Pathway Steering
Committee, the Medical QI Committee, and/or
the Operations Management Council.
When the Chief Resident stopped attending the
Asthma Pathway team meetings, Dr. Newman
(ACP2) stepped in to act as the liaison with the
attending physicians and residents.
When the physician representative to the
Mislabeled Specimens team stopped attending
their meetings, the Karen (MS2) made sure
that the hospital’s Medical Director, the Lab’s
Medical Director, and the Chief Resident all
received updates about the project and had
input into the specimen labeling policy.
KNOWLEDGE:
All leaders invited directors from affected
departments to send representatives to
participate on their teams. Teams consisted o f a
mix of clinical and non-clinical staff, at many
levels (from management to line staff).
INFORMATION:
None o f the subsequent leaders gave timely
or regular updates to team members from
earlier project stages on the status of their
projects during implementation. Alexandra
(PA2) did not update the Clinical Resource
Manager (the hospital’s QI Coordinator)
either.
KNOWLEDGE:
When Dr. Newman (ACP2) introduced the
new Asthma Pathway forms, the Education
Department was not involved, so non
physicians were not informed about it in any
systematic way. Later the team decided to
enlist the help o f the Education Department
in educating non-physicians about the asthma
pathway.
AUTHORITY:
On all three projects, some employees’ poor
attendance at team meetings resulted in
certain departments not receiving timely
information, and not having input into
decisions. When only one or two people were
involved, the Dr. Newman (ACP2) and
Karen (MS2) did not follow up. When
several individuals stopped coming, the
Madeleine (PA1) and the Karen (MS2) each
followed up with written messages addressed
to their groups.
Karen (MS2) (a non-manager) felt that her
low formal position hindered her. The initial
leader, Jane (M SI) had been a departmental
director. Karen felt it would have helped to
have had a formal sponsor fa VP-level
“owner”! who was both responsible for and
actively involved in the project
Key: underline = top management leader, bold = project leader; MS = Mislabeled Specimens project;
PA = Patient Access project; ACP = Asthma Clinical Pathway project; 1 = initial project leader; 2 =
subsequent project leader
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
119
members were clinicians on both the initial and subsequent Asthma Pathway teams,
and on the Mislabeled Specimens team.
In general, while the project leaders took care to maintain the upward flow of
information to their superiors, they were less concerned about keeping team
members informed about the status of the projects. This was observed repeatedly
among the subsequent leaders of the projects, though there were also examples under
the initial leaders. Concerns about the sensitivity of the project information
prevented Karen (MS2) from sending updates to team members between team
meetings. When the subsequent leaders arrived on the Patient Access and the Asthma
Clinical Pathway projects, the team members who had been involved under the
initial leaders were no longer active. The subsequent leaders of these projects were
apparently unaware that the members of the original teams wanted updates about
these projects. Likewise, Alexandra (PA2) did not seem to know that she was
supposed to update the hospital’s QI Coordinator.
Communication between project team leaders and their team members was
also affected by member attendance. Hackman (in press) writes that effective leaders
“make sure that they have created a real work team that will have some stability over
time” (chap. 7). On these three projects, some members’ nonattendance undermined
the stability and organizational structure of the project team, which was intended to
ensure departmental representation. On the Patient Access team, representatives
often sent substitutes whose lack of familiarity with the project impeded its progress.
The project leaders responded to this problem in one of four ways: (1) by ignoring it,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
120
(2) by acting as representative to the relevant constituency in place of the missing
team member, (3) by sending written messages to the team insisting on their
attendance, (4) by eventually recruiting a replacement from the lapsed member’s
department.
Dr. Newman (ACP2) favored the first two strategies, which avoided
confrontation with team members regarding attendance. She disregarded a
respiratory therapist’s lack of participation, and as a result the Respiratory Care
Department was surprised by the implementation of the Asthma Pathway forms, so
the therapists had to be educated after the forms were already on the charts. It also
meant that Respiratory Care gave belated input on these forms (though this was not
disruptive, because revisions were also suggested at that time by parties who had
been actively involved in earlier stages). However, Dr. Newman (ACP2) was able to
make up for the Chief Resident’s absence by personally communicating with the
residents. Similarly, Karen (MS2) was able to compensate for the absence of the
physician representative on her team by ensuring that information flowed to key
physician leaders in areas affected by the project. At the end of this study, Karen
(MS2)and Dr. Newman (ACP2) were both recruiting new team members to replace
lost members.
Reacting to a drop in attendance from 16 and 17 people at their February
1999 meetings to only 9 in March, Madeleine (PA1) sent out an email in late March
saying, “It is imperative that each of you attend these meetings and provide input.”
Although it appears that the message was received (e.g., a couple of team members
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
121
mentioned the email in interviews), team attendance after the email improved only
modestly (at the next eight meetings, the average attendance was 11). In November
2000, Karen (MS2, who had also been a member of the Patient Access team) asked
the Medical Director to send a similar message to the Mislabeled Specimens team
members, urging them to attend the project meetings. Because the message was sent
at the very end of this study, no data were collected on whether the message had any
impact on attendance.
The fact that Karen (MS2) asked the Medical Director to send the message to
her team on her behalf, rather than simply sending it herself, seemed to be related to
her feeling that she lacked authority in leading the Mislabeled Specimens project. Of
the six leaders on the three projects studied, two were physicians and three were
departmental directors. Karen (MS2) was the only one who was neither a physician
nor a director, and she felt that this left her in a weak position with regard to
persuading the departmental directors to enforce the new specimen labeling policy
(which their team had devised, and which the A dministration had approved and
adopted). Another employee explained how frustrating it could be to try to
communicate upwards with the directors at the hospital: “You have to be a director
in order for the directors to listen to you sometimes. I feel like I’m handicapped in
that I sometimes have a hard time getting my point across. It just seems as if they
listen to other directors sometimes before they listen to us managers.”
It is therefore understandable that Karen (MS2) believed that the active
involvement of a VP-level administrator would have been helpful: the vice
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
presidents each had authority over numerous directors. Throughout 2000, Karen
reported on the Mislabeled Specimen project monthly or bi-weekly to the
administrators at Operations Meetings. One of the VPs was assigned responsibility
for the MS project, but he did not attend meetings or otherwise actively participate.
In October 2000, a year after the adoption of the policy, the COO met with the MS
team to try to troubleshoot, because the policy was still being ignored. At the end of
this study (in December, 2000), the future of the policy enforcement issue and the
outlook for the project’s goals were still unknown (though by then the Network had
decided to implement the HFC specimen labeling policy across its other facilities).
Karen (MS2)’s desire for more support from above may also be related to the
way she came to assume her leadership position. After Jane (MSI) left her job at the
hospital, the Mislabeled Specimens team was without a leader for several months.
Eventually, the MS team selected Karen to lead them, and although this was never
formally announced in any official way, the Administration tacitly accepted her as
the team leader. This informal, emergent process may not have maximized Karen’s
authority. It could be that when an administrator formally appoints a leader, the act
not only symbolically grants authority or legitimacy to the project, it may also help
establish a more personal communication link between the project leader and the
Administration.
Another major theme in this study was the problem of knowledge transfer
during the implementation of organizational change. It was mentioned most
frequently with regard to the Mislabeled Specimens project and the Asthma Clinical
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
123
Pathway, because these projects sought to change people’s work behaviors, and to do
so without the organizational restructuring that was integral to the Patient Access
changes. Within the formal structure of the hospital, the Education Department was
responsible for training new nurses and other clinical personnel (e.g., therapists), and
also for education about new policies and procedures. However, neither the original
nor the subsequent Asthma Pathway team approached the Education Department for
help with implementing their pathway. When Dr. Newman’s subsequent team
implemented their standardized asthma pathway fonns, they relied on team members
to arrange education for nurses and respiratory therapists. Several months later, they
realized that a team working on a pathway for bronchiolitis had achieved better
results by enlisting the aid of the Education Department’s educators in training these
groups. When this study ended in November o f2000, the ACP team was arranging
to recruit several educators.
Networks & Leadership
In recent years, sociologists and organizational theorists have come to
recognize the value of informal communication networks in providing access to
resources in communities and in organizations. The term “social capital” has become
a popular way of referring to this relationship (see Adler & Kwon, 2002). For
leaders, networking is an essential skill for sensemaking (Hosking & Morley, 1988),
for acquiring influence (Brass, 1984; Krackhardt, 1990), and for implementing
change (Kotter, 1982).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
124
Research has shown that employees who hold central positions in
organizational networks also tend to have more perceived influence in their
organizations (Brass, 1984; Krackhardt, 1990). Brass (1984) found that employees
who held central positions in their organization’s communication network were more
likely to be promoted. In Kotter’s (1982) study of 15 general managers, the managers
who were better performers also networked more aggressively and skillfully, and
their networks were stronger and more robust.
Krackhardt (1990) found that part of the positive effect of network centrality
on employees’ power was attributable to the employees’ cognitive map of the
organizational network. Employees construct and enhance these mental maps
through active participation in their networks. Other research suggests that
networkers not only use their social relationships to gain information about their
environments, they also use their relationships to shape the perceptions of others in
the network (Brown & Hosking, 1986). The general managers in Kotter’s (1982)
study mobilized practically their entire network of relationships when implementing
their agendas in their organizations, using both indirect (symbolic, sensemaking) and
direct (asking, demanding, negotiating, intimidating) methods to influence the people
in their networks.
At the Hospital for Children, networks appeared to be strongest within
departments, and within disciplines. Employees’ professional communication
networks with colleagues outside the hospital were often more significant sources of
QI information and knowledge than their organizational networks within the hospital
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
125
or the larger Network health care system. This pattern of relationships is consistent
with the social scientific literature on professionals, which observed years ago that
members of a given profession often identify more strongly with their profession
than they do with their organization (Gouldner, 1957). Professionals’ extra-
organizational communication networks are a critical source of information and
knowledge (Coleman, Katz, & Menzel, 1966; Crane, 1972), which they use to make
sense of their environment. Such interorganizational sensemaking activities allow
professionals to operate strategically as they attempt to access allocative resources
such as research grants (Lievrouw, 1986).
Studies have shown that employees who are very involved in communication
networks are also more likely to talk about innovations (Monge & Eisenberg, 1987),
and those with diverse extra-organizational networks are more likely to import
innovations into their own organization (Monge & Contractor, 2000). The origins of
the asthma pathway project demonstrate how change initiatives can migrate into
organizations and then gain momentum through communication networks. In the
early 1990s, the Assistant Director of Pharmacy attended a seminar on disease
management, and started talking with his friend in Respiratory Care about starting an
initiative for asthma case management. They then connected with a manager in
Nursing Administration who was interested in working on clinical pathways, and
with an employee in the Education Department. All four individuals shared a desire
to achieve more consistency in the hospital’s delivery of care. The group approached
the Administration, who approved their initiative to work on clinical pathways.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
126
Several physicians joined them to form the Pathway Steering Committee. After
prioritizing the possible pathway projects, ten pathways were selected for
development and implementation, including the asthma pathway.
Several leaders in the hospital were particularly active networkers. Two
administrators made exceptional efforts to consciously cultivate their relationships
with a variety of employees and physicians. The Medical Director’s networking
activities were mentioned earlier. He was the central node of the hospital for QI
communication with the physicians, and with many other clinical personnel. In order
to monitor and communicate information with these people, he attended all medical
committee meetings and ate lunch in the physicians’ dining room. The hospital’s
President also made a point of eating in the physicians’ dining room, and during the
course of this study he began holding Employee Forums with different individuals
each month in order to foster communication with lower level employees throughout
the hospital.
On the MS project, Karen’s network played a role in her emergence as the
subsequent team leader. The Director of the Emergency Department said she had
selected Karen, one of her Assistant Directors, to join the Mislabeled Specimens
team because of her good relationships with the medical residents and with the Lab.
Karen also recognized the importance of these connections, saying, “I had a very,
very good, close relationship with the lab supervisors as an ED supervisor, and I
dealt with all the Lab-related issues directly. Even other issues that other managers
brought to me, I took them to the Lab so we could be consistent.” Later, when Jane
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
127
(MSI) left the hospital, the Mislabeled Specimens team selected Karen to be their
new leader, partly because of her network.
The story of how Karen became a project leader, and the origins of the
clinical pathway initiative, are both examples of how networks played a key role in
emergent organizational processes that affected the QI projects. However, the project
leaders rarely seemed to actively use their networks in order to advance their change-
related goals. An unusual example of such enactment occurred on the Patient Access
project. Early in 2000, Alexandra (PA2), was trying to launch a new unit for
scheduling, pre-registering, and pre-certifying patients, called the “Patient Access
Center.” She encountered multiple delays, but she finally thought she could open the
unit in the first half of March. Then, two weeks before their “go-live” date,
Alexandra (PA2) realized that they needed to have some programming done on the
new computers, and she was told that that the Network’s IS department would not
have the work done before the first week in April. Alexandra (PA2) spoke with two
of the IS employees who were doing other work on computers in her department.
These employees said that they could do the necessary programming themselves and
it would not take very long to do. Alexandra’s use of her network to access these
allocative resources was very effective, because the center opened on March 15,
three weeks ahead of the projected delay that would have resulted from waiting on
the IS Department’s service list.
One of the reasons that most of the project leaders did not make extensive use
of their networks in promoting change may have been that they did not have strong
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
128
enough relationships with the right people. Clearly, links to the Administration were
important. For example, in interviews, two separate employees mentioned the
Clinical Resource Manager’s close connection with the hospital’s Medical Director;
a member of the second Asthma Pathway team said that the CRM’s involvement
reassured her that the team would be heard and attended to because of this
relationship. On the Mislabeled Specimens project, Karen (the subsequent leader)
was not in upper management, so she lacked ties to the administrators and directors
who might have helped her more effectively implement the new specimen labeling
policy. A stronger network with top and upper manager might have compensated
(even if only in part) for her weak formal authority, and might have substituted for
more formal involvement by an administrator.
Leaders’ networks within the project team also appeared to be a factor. A
member of the Mislabeled Specimens team felt that Jane (MSI) had faced a
disadvantage in that she hadn’t known anyone on the team prior to the start of the
project. Other leaders may have faced obstacles because of weak relationships with
team members and other affected actors. On the Asthma Pathway, Dr. Song (the
initial leader) did not appear to have had strong ties with the other specialists who
were involved in designing the pathway, nor with the Emergency Department
physicians. One physician described Dr. Song (ACPI) as having “tunnel vision”
when it came to things that were not part in her interests or routine. The specialists
worked on the asthma path serially, passing it off from one to the other, voicing their
disagreements through their revisions of the path. The process took a long time, with
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
129
many drafts and changes (one person estimated that the path was changed 15 times
by the initial team). Stronger relationships between the doctors probably would have
facilitated communication among them, which could have increased the efficiency of
the design process.
Research has shown that intraorganizational networks can work to either
encourage or hinder the adoption of innovations (Monge & Contractor, 2000; Monge
& Eisenberg, 1987). In her previous job at a different children’s hospital, Alexandra
(PA2) had observed that social networks hindered the implementation of the
restructuring changes that were to centralize and standardize that hospital’s
registration and precertification processes:
We had major problems trying to implement the recommendations
because a lot of people didn’t agree with it, and there were some
physicians that were strongly opposed to it. They could not separate a
position from a person. They saw that person, that position with a
name, and a face and a family, so they could not see how it would
better the organization. And it did require reducing some FTEs, and
shifting some FTEs to another director. People had been there for
years, and they liked these people.
The in-house resistance was so strong that the hospital ultimately brought in outside
consultants to implement the changes. At Hospital for Children, when Madeleine
(the initial leader) was in charge of the Patient Access project, the team had also
encountered resistance to the prospect of centralization from directors who feared
losing FTEs. However, at HFC the loss of FTEs was framed as a human resource
problem, not as an interpersonal issue. One director explained “my secretary
registers, charges, discharges, codes, answers the telephone, does my typing. I said,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
130
‘If you take my secretary, the FTE part of my secretary, I’m stuck.’ So it was not,
like, real clear cut. We’ve got multi-talented people that do just a myriad of jobs.”9
Communication Media & Leadership
Communication media systems are integral to the operation of modem
organizations. A diverse array of media were used at the Hospital for Children. At
HRC, person-to-person messages were sent via email, telephone, voicemail, and
pagers, or if printed (e.g., clinical forms), then by vacuum tubes, fax, or mail. More
public messages were broadcast in newsletters, published on web pages accessed via
the intranet, posted on the wall in work areas, or sent over email. The clinical areas
had additional communication media related to the delivery of care, such as special
lights, alarm bells, and whiteboards that alerted clinicians to take various actions. Of
course, face-to-face interactions were the most widespread form of communication,
whether in meetings, in work areas, or in other parts of the hospital (e.g., the halls,
the cafeteria).
In such media rich environments, where the volume of information flow is so
great, change leaders have to be particularly skilled at selecting appropriate media
for sending and receiving messages. Daft, Lengel and Trevino’s (1987) survey
research found that high performing managers were more likely to use richer media
9 Ironically, in one department, the centralization o f the registration function seemed to
actually lighten workloads o f non-Patient Access employees. After the Patient Access
register in the Ambulatory Clinic was reassigned to die newly centralized Patient Access
department, the two unit secretaries in the Ambulatory Clinic communicated less with her.
The registration work that had been shared among the three employees was now done
exclusively by the PA register, leaving her with a disproportionately heavy workload
compared to the other two clerks.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
131
(i.e., media that conveys more informational cues) to send ambiguous messages than
low performers. Lengel and Daft (1988) argued that leaders’ media choices affect
how those leaders are perceived. “It is important for senior executives to realize that
in a very basic sense, a medium is an extension of self. A medium is an extension of
one’s personality and the basic senses. The selection of a medium signals the senior
executive’s management style, which may be cool and detached or warm and
personal” (p. 230).
When a leader chooses to communicate via a given communication medium,
that leader must consider more than just the message and the available media.
Ngwenyama and Lee (1997) have argued that, regardless of the communication
medium, the meaning of a given message can only be understood once the
organizational context and the actors’ actions are taken into account. Leaders need to
know “not only how to use the available technology but also how employees may
interpret different media choices” (Reardon & de Pillis, 1996, p. 405). They need to
be aware of organizational and subgroup norms that affect those interpretations, and
they need to be mindful of employees’ patterns of media use. For key actors, it may
also be important to consider individual differences in media habits.
Table 4.3 suggests that, in the three QI projects, the project leaders tended to
be more effective at enacting resources within the project team than they were at
transmitting information or transferring knowledge to the groups or departments
where the innovations were to be implemented. The main exception to within-team
communication effectiveness was that substantive information about the status or
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 4.3 Project Leaders’ Enactment and Non-Enactment of Resources through the
Organization’s Communication Media
How Project Leaders Enacted Change
Resources
How Project Leaders Did Not Enact, or
Unsuccessfully Enacted Change Resources
INFORMATION:
Meetings were the preferred mode of
communicating within the team. All leaders
on the three projects used email to schedule
meetings (on ACP, the Clinical Resource
Manager scheduled team meetings for the
leader).
An employee on the MS team reported that
Jane (M Sl)’s use o f multiple media (phone
& hard copy o f emails) to schedule meetings
was particularly helpful.
Alexandra (PA2) met one-on-one and in
small groups with managers in the various
points o f service to plan implementation of
improvements that affected their areas.
KNOWLEDGE:
A team member remarked on the strong
meeting management skills of Jane (MSI).
She was able to get all team members to
contribute to the conversation, and she kept
the meetings “on track.”
Team members noted that the Dr. Newman
(ACP2) conducted particularly focused,
task-oriented, and effective meetings. This
physician credited her meeting management
skills to her experience with project
management (having worked as an analyst
at a bank).
INFORMATION:
All leaders distributed minutes only at the
meetings. Employees who missed meetings did
not get the minutes.
Alexandra (PA2) tried to give advance notice
to employees and physicians about the planned
relocation o f Admissions. It was announced at
the physician staff meeting, in the hospital
newsletter, a page on the Network intranet,
email, and an auto-fax. Despite these efforts,
university-based doctors were surprised by the
move.
KNOWLEDGE:
Dr. Newman (ACP2) sent information about
the Asthma Pathway to the residents over email
(in addition to presentations). She later learned
that email was not always read or noted.
Karen (MS2) introduced the new specimen
labeling policy to affected areas through
presentations at meetings. The residents seemed
receptive, but it later appeared that the nursing
directors o f the two departments with the
highest volume of specimens (ED & ICU) had
not accepted the new policy as it was intended.
Karen (MS2) implemented the new specimen
labeling policy through the hospital’s
decentralized education structure. Several units
relied on written media to communicate the new
policy. Most people’s behavior did not change
after the policy was introduced. The rate of
mislabeling did not improve.
AUTHORITY:
When attendance at project meetings waned,
Madeleine (PA1) sent out an email to the team
telling them to attend.
Later, when attendance at MS project team
meetings waned, the Karen (MS2) asked the
Medical Director to send out an email to the
team to urge them to attend.
Key: underline = top management leader, bold = project leader, MS = Mislabeled Specimens project;
PA = Patient Access project; ACP = Asthma Clinical Pathway project; 1 = initial project leader; 2 =
subsequent project leader
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
progress of the project was not conveyed to team members outside of meetings.
Thus, the members of the initial Asthma Pathway team were not updated about the
changes being implemented or the indicators being tracked by the subsequent leader.
The members of the Patient Access team stopped meeting after they presented their
recommendations to the Administration, and they were not updated about the status
of those recommendations during the months that implementation and monitoring
took place. And on the Mislabeled Specimens project, the subsequent leader did not
update the team about the outcomes she was tracking (i.e., number of mislabeled
specimens) until she convened a meeting three months after implementation of the
specimen labeling policy. The hospital’s Risk Management Department was
concerned about the sensitivity of the Mislabeled Specimens information and they
tightly controlled its distribution (e.g., at Directors Council meetings, the reports on
MS were distributed to each attendee and then recollected at the end of the
meetings). Karen (MS2) therefore refrained from sending updates to team members
via email.
The project leaders relied quite heavily on meetings as the main mode for
communicating with their project teams. Generally, they did not update team
members between meetings, and those who missed meetings were not sent the
minutes. Since meetings were the dominant means of communication within the
teams, meeting management skills were very important. These skills were also
important because meeting dynamics and meeting content affected team members’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
134
perceptions of the project and their desire to attend meetings, as is discussed in
Chapter Five.
The two leaders who each sent email messages to their team members about
their poor attendance at project meetings attempted to enact authority by
admonishing the members to attend more regularly. A couple of employees referred
to these as “nasty letters” or “nasty notes.” The choice of a broadcast message
instead of a more personalized email message, and the choice of an electronic
medium instead of a medium like the telephone or face-to-face, which are also more
conducive to immediate feedback (Daft et al., 1987; Lengel & Daft, 1988), likely
conveyed an impersonal tone. This was reinforced by the content of the messages.
On the Patient Access team, the initial leader’s email emphasized the importance of
attendance and firmly told team members they should start coming to the meetings.
The Medical Director’s message, which he wrote to the Mislabeled Specimens team
on behalf of their subsequent leader, essentially said “You’re on this committee, and
you either come, or if you can’t, send a representative.” Neither message sought
information about why people weren’t coming or solicited any response other than
compliance. In other words, the message and the medium framed the communication
as a directive.
Redundancy of communication through various media can increase
effectiveness, but sending messages through numerous media may also be
insufficient. Jane (MSI), who used email and telephone to contact MS team
members about upcoming meetings, and then to remind them, was seen as effective.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
135
However, Alexandra (PA2) used five different media to announce the Admissions
area’s upcoming relocation (the physician staff meeting, the hospital newsletter, the
Network intranet, email, and an auto-fax), and yet did not reach an important
constituency: the university-based physicians. These physicians were on the
university’s email system, not on the hospital’s (where the email was broadcast).
They did not read the hospital’s newsletter regularly, and they had not been included
on the auto-fax list (which had been supplied by the Network’s Managed Care
office). Also, attendance at the quarterly physician staff meetings had declined
dramatically, apparently due to recent changes in attendance requirements. At least
one physician felt that the physicians should have been consulted and involved in the
decision to move Admissions, rather than merely being notified that it was going to
happen. This example shows that leaders need to consider the reach of the media
they employ, as well as stakeholders’ expectations about communication and
involvement in the changes to be implemented.
On change projects that aim to change employees’ work practices, the project
leader must coordinate team members and other organizational units to raise
awareness and transfer knowledge of the new standards. The Mislabeled Specimens
and Asthma Pathway projects both struggled with this during implementation. Dr.
Newman (ACP2) found that email was not a very effective medium for telling the
residents about the Asthma Pathway because many of them did not read their email.
In addition to the email messages, she presented information about the pathway at
several meetings with the residents (the 94 residents could not meet all together), and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
136
at a House Staff meeting. Many residents found the new Asthma Pathway forms to
be convenient and began to use them. However, team members later noted that not
all of the residents were consistently using the forms. One reason was the high
turnover rate among the residents. Also, the pathway was not a mandated hospital
policy; its status was “optional yet recommended.” The attending physicians did not
view the pathway and its forms as a protocol or standard of care: According to one
team member, “In most people’s minds, there’s not an awareness that the lack of a
pathway form is a problem. Some university faculty have given me compliments on
the form. They think it’s a good idea, but they wouldn’t know to care if it wasn’t
being used.”
Similar issues arose on the Mislabeled Specimens project. Karen (MS2)
introduced the new specimen labeling policy to affected areas through presentations
at meetings of clinical and administrative staff (e.g., division meetings, the residents
meeting). Then, the Education Department’s educators (each one assigned to a
different department) each arranged for training in units throughout the hospital.
Since the educators generally tailored their training to their particular departments,
different groups received the information in different ways. Several units relied on
written media to communicate the new policy. In the ICU, education involved
putting a copy of the policy in employees’ paychecks and then having them sign and
return a piece of paper saying they had read it. On other nursing units, the director
announced the new policy at a staff meeting, and then training involved reading the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
137
written policy, followed by a paper and pencil test. Reminder articles later appeared
in two employee newsletters.
Certain groups and subgroups of employees did change their behavior,
including many residents and some especially diligent nurses. However, most
people’s behavior did not change after the policy was introduced. The rate of
mislabeling did not improve. There were at least two reasons for this. First, the
written medium of communication for teaching the policy was probably inadequate
on its own. A face-to-face training session would have been richer, more interactive,
and therefore better for helping employees understand the nature of the specimen
labeling problem, or the expected changes in their everyday tasks. But even more
problematic was management’s attitude towards the new behaviors. One team
member observed that, “the education of the new policy wasn’t even all that
emphasized.” A couple of MS team members tried to remind the people working in
their departments to observe the policy, but they appear to have been the only ones to
do so. It eventually became clear that the nursing directors of the two departments
with the highest volume of specimens (ED & ICU) did not agree with the new
specimen labeling policy, as it was intended. If Jane (MSI) had involved these upper
managers more directly in the investigation of the problem and in the design of the
policy, then they might have ultimately been more supportive of the implementation
effort.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
138
Information Systems & Leadership
Leaders need ongoing information about their environments and feedback
regarding their own actions so that they can make sense of their unfolding situation
and make decisions accordingly. The need for process and outcome information is
particularly critical for leading organizational change. Interpersonal communication
networks can provide information about the socio-political environment (DeLuca,
1984), and team members can provide anecdotal or observational data about
practices and outcomes, but most organizations prefer more quantitative progress
measures (especially of outcomes). The philosophy and methodology of quality
improvement is built on the quantitative measurement of processes and outcomes,
which is why information systems and information technology are so important for
QI projects. Savitz and Kaluzny’s (2000) case study of clinical process
improvements (CPI) in four integrated delivery systems found that strong
organizational commitment to information technology was an important prerequisite
for successful dissemination and use of CPI. The use of clinical data for QI is closely
related to the successful implementation of QI programs, especially among
physicians (Arthur Andersen, 1999; Blumenthal & Edwards, 1995; Savitz &
Kaluzny, 2000; Shortell et al., 1998), so clinical information systems are particularly
essential.
Unfortunately, existing information systems (IS) in most health care
organizations are inadequate for quality improvement purposes. The Institute of
Medicine’s recent report states that although information technology (IT) has
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
139
tremendous potential to help improve the quality of patient care, especially through
automated patient-specific clinical information, and although efforts to automate
such information date back several decades, overall “progress has been slow” (pp.
16,171). The report notes that “IT has barely touched patient care. The vast majority
of clinical information is still stored in paper form” (p. 176). As late as 1998, only
22% of U.S., non-governmental acute care hospitals possessed the integrated
databases that are needed for clinical applications of QI (Arthur Andersen, 1999). In
both the 1993 and the 1998 national surveys of these hospitals, inadequate
information systems were identified as one of the top three barriers to QI (Arthur
Andersen, 1999).
Overall, HFC’s information systems (IS) were no exception to this trend. In
order to understand why they were deficient for QI, one must consider the hospital’s
information systems. In the mid-1990s, HFC was using a user-friendly and relatively
modem Windows-based admission-discharge-transfer (ADT) system called SMS.
But in 1998 they reluctantly converted over to the Network’s ADT system, a
proprietary legacy system that ran on a mainframe and was very cumbersome to use
(e.g., it was difficult to customize reports, and customization requests went to the
Network’s overburdened IS department). The Network resisted replacing their ADT
system because of the cost involved (they estimated that it would take at least S30
million). HFC’s local systems for laboratory, pharmacy, and radiology interfaced
with the Network’s ADT system through HBOC, an order-entry system. The
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
140
Operating Room’s system, ORSOS, and the SoftMed utilization management
program also interfaced with the ADT system.
Some clinical data were captured and retained in SoftMed. This included
anything that was dictated (e.g., histories and physicals, operation notes, operation
summaries, radiology notes, and discharge summaries). However, at the time of this
study (1999-2000) most of the hospital’s clinical information was still handled
manually. For example, the Emergency Department’s record-keeping was entirely
paper-based (all except basic registration and ancillary services order entry). Like
most hospitals, HFC did not have an electronic medical record (EMR), so all notes
(nursing, social work, respiratory therapy, etc.) and many physician orders existed
only on paper. Then, although the orders and results for ancillary services were
captured electronically, little of this data was electronically retrievable after patients
left the hospital. The heavy dependency on paper-based records made data
collection, retrieval and processing for quality improvement purposes very labor
intensive, especially for QI projects with a heavy clinical component.
Of course, physical IS technology is not sufficient for enabling QI; it must be
accompanied by adequate staffing with employees (including physicians) who are
capable of using the systems to collect and analyze the relevant performance data.
The Network operated its IS on a “super user” model, relying on a decentralized
network of employees in various departments who had expertise with different
programs. This model was necessary because the Application Support unit of the IS
Department was responsible for 400 applications—and it only employed five people.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
141
Needless to say, hospital employees with IS expertise were very busy. During this
study, the hospital invested in software training for a few employees in the QI
Department and in Decision Support. This did not much expand their pool of experts,
however, since two out of the three employees were IS experts before the training.
The Network contracted with a consultant to perform an IS needs assessment
in 2000. According their report in September o f2000, the consultant found
numerous lagging or obsolete systems that were not maintained by vendors or no
longer met business needs, highly customized systems that limited flexibility and
business agility, a “best of breed” vendor portfolio that limited ability to provide
integrated solutions, relatively low annual IT expenditures, a very high proportion of
IS staff devoted to maintaining current systems (with too few people available to
work on new systems), and a need for the Network to develop processes and skills to
manage technology risks and enhance opportunity for project success. At the time of
this study, the hospital was looking into purchasing new information systems for the
Emergency Department (a tracking and charting system), Radiology (a digital
imaging system), and the Quality Improvement Department (the QI module of
SoftMed). Two major obstacles in purchasing these systems were cost and
compatibility with existing systems.
Given this situation, it is not surprising that on the Diffusion Study’s LLI
survey, upper managers disagreed that the information technology system gave them
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
142
easy access to the information they needed to do their jobs (M= 2.62 out of 5).1 0
They also disagreed that the information technology system allowed them to easily
contribute to the databases (M - 2.56), and claimed that the information stored in the
databases was difficult to search and retrieve (M = 2.35). Finally, like most hospitals,
HFC was not found to be using their information systems for effective knowledge
management. The managers gave the hospital’s information technology system an
average rating of 2.76 on its ability to allow them to share knowledge about
problems and ideas about possible solutions.
The poor IS infrastructure limited the kinds of innovations and improvements
that could be made on the Patient Access and Mislabeled Specimens projects. The
subsequent leaders of both projects wished for more automation, including bar-
coding and SmartCard technology, but the ADT system was incompatible with these
technologies.
Madeleine (PA1), Alexandra (PA2) and Dr. Newman (ACP2) were able to
use existing hospital ISs to collect some of the data for investigating problems and
for monitoring the changes they implemented. All three projects involved manual
data collection (from incident reports, patient charts, or daily productivity reports
filled out by hand). The Mislabeled Specimens project was unable to rely on hospital
IS for almost any data; the only information either leader obtained electronically was
a tally of the number of specimens that the Lab processed each month. Both MS
leaders manually collected data from Incident Reports that were submitted to the
1 0 All survey items on the LLI survey were scored out o f 5, with 5 being best.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
143
Risk Management Department in order to track the monthly totals of mislabeled
specimens, their proximate causes, and the occupations of the responsible
individuals. Jane (MSI) had the team conduct a survey of employees and physicians
to determine their awareness of the problem and their knowledge of their
department’s policies for labeling specimens. She found that both knowledge and
awareness were low. After the policy was implemented, Karen (MS2) collected data
manually to determine whether it was being followed, keeping track of whether the
transporters had signed the paperwork to say they had checked that the paperwork
matched the label, and whether Incident Reports for mislabeled specimens were
being completed by both the Lab and the person who had made the error. Later,
Karen (MS2) had the Lab start keeping track of the names of managers who were
called about mislabeled specimens, so that she could see if they were following up
with their employees.
A number of studies have found that a lack of organizational support for QI
projects with data and human resources often limits the progress of these initiatives
(Blumenthal & Edwards, 1995; Hernandez, 1998; Savitz & Kaluzny, 2000).
Blumenthal and Edwards’s (1995) case study of clinical QI in six U.S. health care
organizations that are considered to be quality leaders found that even in the most
advanced o f these leading organizations, people involved in QI efforts expressed
frustration over limited information and the lack of personnel to support their data
needs. Lack of allocative resource support was an issue during the initial efforts to
implement the asthma pathway. The hospital’s Administrative Director of Nursing
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
144
Administration, who was originally in charge of all the clinical pathways, said that
when she and the initial ACP team had worked on the asthma pathway, they had
lacked the human resources to collect the data they needed to monitor the pathway.
At one point during that time, nurses and respiratory therapists manually collected
data from the charts of asthma patients to track asthma education, but the project did
not have enough dedicated resources to sustain or extend their work.
It has been mentioned that the QI Department collected data for the
subsequent leader of the Asthma Pathway. The Clinical Resource Manager (manager
of the QI Department and member of the subsequent ACP team) identified
appropriate asthma patients in SoftMed, and then the Clinical Data Coordinator had
Medical Records retrieve the charts so that she could manually collect data about
usage of the pathway. These data were collected on sheets of paper and then entered
into a computer spreadsheet. Totals and percentages for some measures (e.g., % of
patients with a diagnosis of status asthmaticus who were placed on the pathway, %
of asthma patient’s charts with documented education; % of PulmoAide home
nebulizing equipment ordered on admission as opposed to later) were compared with
data from earlier dates (e.g., Fall 1999 was compared with Fall 1998, and Peak
Season [January - March] 2000 was compared with Peak Season [January - March]
1999).
Case studies of QI in hospitals have found that it is not uncom m on for these
initiatives to suffer from seriously deficient data (Blumenthal & Edwards, 1995;
Hernandez, 1998; Savitz & Kaluzny, 2000). Several leaders encountered problems
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
145
with data on the QI projects at HFC. Sometimes these had to do with reports
generated by the Network’s Financial Services Department. A Patient Access team
member reported that for months the Network had not responded to HFC directors’
requests to have their bad accounts reported in a readable format. Only after the
hospital had lost enough money to spur the Administration to initiate the Patient
Access project did Financial Services finally modify these reports. Some team
members then met repeatedly with personnel from Financial Services to better
understand and modify their weekly reports and to establish relationships. This was
one of the first improvements to be implemented on the PA project; it was
undertaken while Madeleine (PA1) was still the leader.
Alexandra (PA2) encountered problems with Financial Services reports on
claims denied due to missing pre-certifications. Patients would routinely appear on
those reports labeled as missing pre-certifications, whereas the HFC system listed
them as having been pre-certified. Sometimes these were due to exceptional cases
(e.g., an infant who did not have her own insurance number was given a pre-
certification number under her mother’s insurance, which was later rejected). The
problems made it difficult to accurately monitor the new pre-certification processes.
The Network’s work productivity reports were also a continuing problem for
Alexandra because the measures had not been revised to account for the new job
duties of the Patient Access employees (i.e., scheduling, pre-registration, pre
certification). This problem also continued for months without resolution.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
146
Sometimes leaders encountered data problems that originated with employees
who were responsible for data collection. Karen (MS2) often had trouble getting
Incident Reports from the nurses caring for the patients affected by specimen
labeling errors. The specimen labeling policy required two Incident Reports for each
error, one from the Laboratory and one from the nurse in the department of origin.
However, throughout the year of this study, only the Lab employees consistently
completed their Reports—the nurses’ reports were often not submitted. Similarly, the
Patient Access staff tended to fill out their “daily” productivity reports once a month
instead of at the end of each work day (as they had been instructed), which led to
missing data for their director, Alexandra (PA2).
These documentation issues were not exclusively found on the QI projects.
Concerns about documentation were discussed in the Nursing QI Council (e.g.,
increasing the accuracy of nursing assistants’ documentation of liquid intake and
output), and in the hospital’s QI Council (e.g., improving the speed of physicians’
completion of medical records). Managed care companies had increased the amount
of documentation they required of attending physicians, and these companies were
making new and increasing demands regarding the usage of very specific language
on the charts in order for the hospital to be paid. The hospital had difficulty getting
nurses and residents to meet those demands. This context implies two constraints for
the QI projects. Operational documentation was an ongoing challenge for many
departments and disciplines (especially in light of increasing external demands),
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
147
therefore employees and physicians could be expected to resist additional demands
for better and more consistent documentation for QI purposes.
Occasionally, the project leaders seemed unclear about the data they were
collecting. Karen (MS2) changed the operational definition of her outcome measure
(the percentage of specimens that were mislabeled) shortly after the project team was
formed, so that data reported before August, 1998 were not comparable with data
reported after that date. Before August, 1998, the hospital used to count each
incident of a mislabeled specimen that involved a mix-up with two patients as two
cases of mislabeling. Shortly after the team was formed, Jane (MSI) changed their
measurement practices to conform to the reporting practices of most hospitals
participating in their benchmarking database. Because there were also cases of
mislabeling that did not involve two patients (e.g., missing labels), there was no way
to easily convert the aggregated data points from the pre-August, 1998 quarterly
reports. To achieve comparable numbers, each individual case from each quarter
would have to have been reexamined and retabulated.
Karen (MS2) did not recognize this problem with her MS trending data until
an administrative director pointed it out near the end of the study. Indeed, the data
were presented as comparable at presentations to the Administration and to the
Network’s QI Awards Competition. Karen (MS2) also tended to discuss the progress
measures in terms of raw totals of mislabeled specimens, even though the hospital
required formal reporting in the form of a percentage (percentage of specimens that
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
148
were mislabeled). Others in the hospital also emphasized the numbers more than the
percentages.
Another example of measurement confusion could be seen in the Quality
Monitoring and Evaluation Plan for the Patient Access project that was presented to
the Administration in July of 1999, as shown in Exhibit 4.1. At that time, the team
Exhibit 4.1 Patient Access Quality Monitoring and Evaluation Plan, presented July,
1999
Indicator Threshold Data Source Collection
Method
Time
Frame
Evaluation
Period
Preregistration Schedule Develop
process for
scheduling
visits vs.
prereg &
trending
overtime
Ongoing Monthly
Improved
internal
communication
Minutes Meetings Ongoing Monthly
Improved
external
communication
Minutes Meetings Ongoing Monthly
Decrease wait
time
Evaluate
BENCH-
marking
Patient
Satisfaction
Surveys (from
Company &
Patient Affairs
Dept.),
BENCHmark
data
Data Sheets Ongoing Quarterly
Reduce billing
errors
IS reports IS reports Ongoing Weekly
Increase staff
competencies
List o f cross-
trained staff
T est&
certification
Ongoing Orientation,
Annually,
Updates
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
149
was still lead by the Madeleine (PA1). Contrary to the teachings of QI, the
performance indicators were not quantified and numerical thresholds were not given.
Quantitative thresholds could have been given for certain indicators (e.g., decrease
billing errors, increase staff competencies), while others were quite qualitative (e.g.,
improved internal and external communication, as judged from meeting minutes).
In contrast, Alexandra (PA2; the Director of Patient Access) demonstrated
more sophisticated performance measurement skills, probably acquired through her
higher education (a bachelor’s degree in accounting, and an MBA). In her Patient
Access Service implementation proposal, submitted to the Administration in
September 1999, she quantified her objectives (e.g., obtain pre-authorization or pre
certification number on all patients requiring pre-certification before visit; prompt
patient registration, with a standard wait time of less than 5 minutes). The Quality
Monitoring and Evaluation plan she presented at the QI Awards Competition in
September 2000 included quantitative thresholds with each measure (see Exhibit
4.2). The Director tracked these measures (monthly, and some weekly) and regularly
reported this information to the QI Council and to the A dministration.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
150
Exhibit 4.2 Patient Access Quality Monitoring and Evaluation Plan, presented
September, 2000
Indicator Threshold Data Source Collection
Method
Time
Frame
Evaluation
Period
Number o f
Claim
Denials
0 Exception
Code Report
ADT
Information
System
Monthly Quarterly
Number o f
Claim
Denials
$0 Exception
Code Report
ADT
Information
System
Monthly Quarterly
Inpatient Pre-
registrations
100% Demographic
Sheets
Manual Monthly Quarterly
Outpatient
Pre
registrations
100% Demographic
Sheets
Manual Monthly Quarterly
Outpatient
Pre
certifications
100% Daily Activity
Report
Manual Monthly Quarterly
Patient
Satisfaction:
Courtesy o f
Personnel
90 Patient
Satisfaction
Survey
Reports
Opinion
Surveys
Monthly Quarterly
Patient
Satisfaction:
Speed o f
Admission
90 Patient
Satisfaction
Survey
Reports
Opinion
Surveys
Monthly Quarterly
Summary
Taking an approach that views communication as central to both leadership
and resources, the analyses in this chapter revealed several patterns of enactment
across structures, systems, and types of leaders. Top managers enacted knowledge,
authority, and allocative resources, primarily through formal structures. Information
was a key responsibility for project leaders, who demonstrated differing skill levels
with performance data. Data collection was often not automated and was therefore
labor intensive; also, the hospital’s legacy information system limited the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
151
innovations that the project leaders could implement. The project leaders were
conscientious about reporting information upward, but team members and other
stakeholders often did not receive updates. One project leader felt disadvantaged by
her lack of formal authority, and sometimes appealed to administrators for help.
Social networks operated in spontaneous processes (project initiation, leadership
emergence) but project leaders rarely leveraged them for achieving project
objectives. This may have been because their networks were inadequate.
This chapter considered examples from the three projects in order to
understand different ways in which the change leaders enacted resources through
various organizational structures and systems. Chapter Five goes beyond this
diachronic approach, to a synchronic perspective on the major dynamic processes
that were enacted through the three projects. It explores the roles of leaders in
managing expected project phases, and also how they coped with unexpected
challenges. Two issues that were raised in this chapter—poor meeting attendance,
and difficulties with knowledge transfer—will be explored in Chapter Five.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
152
CHAPTER 5: A DIACHRONIC ANALYSIS OF
LEADERSHIP, COMMUNICATION, AND RESOURCES FOR CHANGE
This chapter addresses the temporal dimension of change leadership and resource
enactment. It is concerned with dynamic phenomena that affect change processes in
organizations, including cyclical phases, recurring events, and unpredictable one
time events. Cyclical communication activities (e.g., project team meetings, data
collection and reporting) recur within the context of the change projects, which in
turn are embedded within the larger organization, which operates within the context
of its environment.
Activities and activity cycles can become enmeshed with other social
processes, affecting their pace, rhythms, or oscillations. This phenomenon is known
as “entrainment” (McGrath et al., 1984; McGrath & Kelly, 1986—both sources cited
in Anacona, Okhuysen, & Perlow, 2001; Anacona & Chong, 1996, cited in Mitchell
& James, 2001). An example was the way that certain disease cycles (the asthma
season, the bronchiolitis season) affected the timing of the clinical pathway projects.
The pathways for these conditions, and any associated innovations, had to be ready
and implemented before the start of the season, because the clinicians would not be
able to adopt any new work processes while working under the peak volume of
patients. Another cycle that affected the pathways was the academic year: new
residents started working at the hospital every July, and each new group had to be
educated on the pathways.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
153
Weick (1979) argues that people in organizations derive rules by engaging in
repeated interaction cycles. With subsequent iterations, the actors reduce the
equivocality of the cycles. Rules, once derived, offer a more efficient means of
understanding events, saving actors from discussions and debates that would
otherwise be needed to interpret the information. The QI philosophy is based on the
concept of learning cycles. The famous “PDCA” (Plan-Do-Check-Act) cycle is a
model of the ideal process for quality improvement efforts.1 1 The “do” step is a
small-scale, exploratory change, which is conducted as a pilot test of an
improvement that is “checked” (i.e., monitored and analyzed). If and when the pilot
is successful, then managers can “act” to implement the innovation throughout the
organization. Senge (1992) noted that when many American companies implemented
their total quality programs in the 1980s, they jumped from “plan” to “act”—leaving
out the “do” and “check” steps of the cycle. They did this because their managers
were impatient for quick and inexpensive results, but in their rush to act they
undermined their change efforts and cheated their organizations out of opportunities
for learning.
A very different approach was taken in Japan:
By contrast [to Americans,] the Japanese are masters of organizational
experimentation. They meticulously design and study pilot tests, often with
many corporations participating cooperatively.
Through repeated cycles, new knowledge gradually accumulates.
[When] organization-wide changes [are introduced], people adopt new
practices more rapidly because so many more have been involved in the
learning. (Senge, 1992, p. 33)
1 1 Deming attributes the model to Walter Shewhart o f Bell Labs.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
154
In the three QI projects that were studied, Hospital for Children tended to
skip the pilot stage and go straight for organization-wide implementation of
innovations. One employee described this approach, saying, “I think we try to
develop this thing whole cloth and then apply it, and then hope that it works.”
Granted, some innovations are not amenable to pilot testing, such as the
centralization of the Patient Access function. Yet the Asthma Pathway and specimen
labeling policy both could have been piloted, and they were not. Interestingly, the
Network’s managers treated two of the HFC improvement projects as pilots, and
decided to spread them to other hospitals in their system. In the case of Patient
Access, this was a logical choice, since the project had demonstrated significant
results. However, in the case of the specimen labeling policy, the Network’s
administrators apparently forgot to “check” for evidence of efficacy. If they had, it is
doubtful whether they would have bothered to diffuse the policy to their other
facilities.
HFC’s approach to learning from experiences with QI projects was much
more informal than the PDCA model. Employees did seem to have faith that such
learning would occur. One of the members of the initial Asthma Pathway team felt
that the group had struggled with what to include on the path, considering a wide
array of clinical decisions as candidates for standardization. He felt that the hospital
would improve and get faster at designing pathways as they gained experience with
developing more of them. When another employee was asked if she had observed
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
155
any trends in what constituted an effective QI project leader, she replied that there
had not yet been enough projects to make such an assessment.
Knowledge management regarding the QI projects was extremely haphazard.
There were no standards or rules about archiving the records from the projects, and
projects were not subject to reviews or assessment to try to glean lessons for the
future. Some employees saw individuals’ tacit knowledge about QI as valuable. For
example, the Clinical Resource Manager asked two members of the subsequent
Asthma Pathway team to help a new pathway team so that the new team could
benefit from their experience.
The rest of this chapter examines how the change leaders—the project
leaders, and the administrators—managed the process of their change initiatives, and
how they responded to unanticipated challenges over the course of these projects. It
focuses on the communication strategies and resources that these leaders enacted in
managing their projects and in meeting these challenges. The first part of the chapter
focuses on the leaders’ enactment of resources during the expected phases of the
change projects. First it discusses the roles of different leaders in performing the
sensemaking activities that define the different project phases. This is followed by a
discussion of how leaders’ job attributes affected their roles and abilities on the QI
projects. The next section analyzes the dynamics surrounding meeting attendance on
the projects, and suggests some implications for projects leadership. The final section
of the chapter looks at how change leaders enacted sensemaking resources when
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
156
faced with three different kinds of unexpected challenges: conflict, bureaucratic
obstacles, and leadership transitions.
Leadership across the Project Timeline
The academic and practitioner literatures contain numerous models of
organizational change processes. Change models are important for leaders because
they constitute an important knowledge resource for sensemaking. Leaders refer to
such knowledge in deciding what actions to take and in what order; they help guide
their planning and execution of change initiatives.
Comparative analysis of the three QI projects revealed that they all went
through seven major phases in essentially the same order. These seven phases were:
(1) problem awareness and definition, (2) project initiation and problem investigation
and analysis, (3) team recommendations, (4) implementation, (5) monitoring, (6)
revision, (7) institutionalization. The phases were not always distinct and separate.
For example, Monitoring almost always bleeds into and overlaps with the Revisions
stage. On the Patient Access project, some changes were implemented before the
formal recommendations were presented to the Administration. And on the Asthma
Pathway, Implementation was attempted twice, by two different teams, so the second
team ended up redoing some of the investigation phase. Also, the recommendations
on the Patient Access project were implemented in stages, so the Monitoring phase
overlapped with the Implementation phase. Keeping this looseness of the model in
mind, the seven project phases are still useful for understanding the sequence of
events: for the most part the projects experienced the phases in that order.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
157
This study is concerned with leadership and resource enactment, so the QI
project phases raise this research question:
Research Question 3: What resources do change leaders need for the different
project phases? How does their enactment of resources vary across those
phases?
Table 5.1 lists the sensemaking issues, leadership tasks, and communication
modalities for each project phase. These distinguishing characteristics of the
different phases were observed for all the three projects. The relationship between
these elements and the four change resources (information, knowledge, authority,
and allocative resources) is discussed shortly.
Chapter Four indicated that different kinds of leaders (project leaders vs. top
managers, upper-management project leaders vs. non-management project leaders)
enact different kinds of resources through different organizational systems and
structures. For example, top managers were most active in enacting authority and
allocative resources. They enacted knowledge through appointing project leaders, but
they did not enact information for the QI projects. Project leaders, however, were
very concerned with enacting information. In Table 5.1 the “Leadership” column
identifies certain tasks with top managers, other tasks with project leaders or other
kinds of employees (e.g., project team members, departmental directors). These tasks
also vary across the project phases. This suggests the following research question:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
158
Table 5.1 Leadership, Communication, and Sensemaking in Expected Project
Sequences
Project Phase Sensemaking
Focus
Leadership Communication
1. Problem
Awareness &
Definition
Is there a problem? Awareness originates at
any level, eventually
tunneling information to
top management
Informal (one-on-one)
Organizational
committee meetings
2. Project Initiation
& Problem
Investigation
What is the
problem?
What is/are the
solution/s?
Project leader is
appointed. Project leader
forms project team in
consultation with
directors, top
management
Project Meetings
(and meeting minutes)
Data Collection
3. Team
Recommenda
tions
Here is/are our
recommended
solution/s.
Project leader & team
prepares & presents
recommendations to top
management
Presentation(s) to
Administrators
Final Report and/or
Slides
4. Implementation What is the new
practice?
What does the new
practice mean for
each o f us, and all
o f us?
Project leader coordinates
implementation efforts.
Directors or team
members assume
responsibility for ensuring
local implementation.
Employee and/or
physician education
(variation in media and
formats among units)
Sanction o f formal
documents (e.g., forms,
policies)
5. Monitoring Is the innovation
working?
Project leader tracks
results & disseminates
them to top management
Directors are also
updated.
Data collection and
analysis
Ongoing reports
6. Revision What changes need
to happen for the
innovation to be
more effective?
Project leader works with
stakeholders to adjust the
innovation.
Meetings
Revised documentation
Data collection adjusted
accordingly
7. Institutional
ization
Is the innovation
integrated into
everyday work
routines?
Project leader ensures that
organizational practices
are aligned and consistent
with the innovation, and
that long-term monitoring
& re-evaluation systems
are in place.
Formal communication
routines:
■ Documentation
practices
■ Staff meetings
■ Employee
education
Informal interactions
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
159
Research Question 4: How do resources needs and patterns o f enactment
across the project phases vary fo r leaders (a) at different organizational levels,
and (b) across occupations?
The table included in the Appendix describes the major actions taken by
those involved in each of the three QI projects, at each phase of the project. It also
identifies the hierarchical level of the actors who were organizational leaders (upper
and top managers) or project leaders. The following discussion of the seven phases
refers to information summarized in Tables 5.1 and the Appendix.
Project Phases and Resource Enactment
Project management experts recommend that project managers ensure
“administrative closure” by “generating, gathering, and disseminating information to
formalize phase or project completion” (Project Management Institute, 1996, p. 103).
This sort of demarcation occurred naturally after the investigation phase, with the
presentation of the teams’ recommendations. In team meetings, the Jane (MSI) and
Madeleine (PA2) identified their project’s current stage of the nine-step Network QI
process. After implementation, however, the subsequent project leaders were less
likely to manage or indicate transitions between project phases, and people often
expressed uncertainty about the status of the various projects.
Problem Awareness and Definition. The first phase of the process addresses the issue
“Is there a problem?” It can be described with reference to Weick’s (1979) notion of
“bracketing,” which is a sensemaking step that people take in response to
environmental change. “When differences occur in the stream of experience, the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
160
actor may take some action to isolate those changes for closer attention” (p. 130),
and that action constitutes bracketing. In the QI language of the Network, this phase
involves the identification and specification of an “opportunity for improvement.”
Interestingly, the data from the three projects (summarized in the Appendix) suggest
that while two of the projects began in reaction to newly negative performance data,
the Asthma Clinical Pathway project was motivated by information about
improvement approaches that could address a chronic situation (i.e., variability in
care, including deviation from national care guidelines). The problem data were
generated by HFC or the Network; the innovation information came from
professional sources (e.g., a seminar) outside the organization.
Awareness began with employees in a range of departments (e.g., Risk
Management, Quality Improvement, Nursing Administration, Pharmacy, Respiratory
Therapy, Network Administration) and positions (e.g., manager, assistant director,
director, administrator). On the MS and ACP projects, employees enacted
information by discussing the issue with peers and then bringing it to the attention of
the hospital’s top managers. The administrators were approached because the
projects were interdepartmental in scope, and therefore needed to be authorized and
sponsored (with allocative resources) by top administration. A member of the Patient
Access team noted that HFC employees had been complaining for some time to
people at the Network about the PA problems that were resulting in lost revenue, and
they had met with the Network to discuss these problems sometime after the merger,
but no action was taken. The urgency of the Patient Access problem was finally
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
161
recognized when the Network’s top management reacted directly to information they
received through other channels.
Project Initiation and Investigation. The line between problem definition and
project initiation is actually somewhat indefinite, as can be seen in this quote from a
PA team member:
It took some planning to decide what the team was going to be. I think
that took maybe a month and a half or a little longer. That was
decided on the senior management level, as to what the problem was,
and who would be the ideal people to help supply that information to
really define what the problem was, and try and get some resolution to
it.
This quote highlights how iterative the sensemaking cycles were on these projects:
senior management “ decided...vshat the problem was,” so that they could then decide
“who would be the ideal people to help supply that information to really define what
the problem was.”
One of the first tasks in initiating the projects was to select a project leader.
HFC’s top managers appointed the initial leaders of PA and MS; the Clinical
Pathway Steering Committee recruited Dr. Song to lead the Asthma Clinical
Pathway. (Although there were administrators on the Steering Committee, they do
not appear to have been very involved in recruiting leaders for the pathways.) The
rest of the project team was then selected. Project leaders worked with others (e.g.,
top management, the QI Dept.) to identify units to participate on the team. The teams
consisted of employees from a range of organizational levels (from directors down to
nursing assistants) and a wide variety of disciplines. Administrators chose which
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
162
directors to involve in the projects; directors chose which employees to send as
representatives from their departments.
The formation of the team is very important because, as noted in Chapter
Four, leaders and team members enact knowledge and allocative resources for the
projects (and for their own departments). When top managers appoint project
leaders, they also enact authority. Once these resources are mobilized, the team can
turn its attention to enacting information. After the teams were formed, the leaders
began holding regular meetings in order to conduct and coordinate their investigation
into the central problems of their projects. At this point, the team meetings concerned
data collection tasks, and sharing information learned as a result of data analysis.
Toward the end of this phase, team meetings turned to discussion of
recommendations; on the MS and ACP projects, the teams also came up with more
detailed designs for solutions (e.g., a pathway, a policy, an education plan).
Disagreements were mentioned by team members on all three teams, but the Asthma
Pathway experienced the most conflict. Of the three teams, only Patient Access was
given a formal deadline for completing their investigation and presenting their
recommendations to the Administration.
Team Recommendations. In Phase 3, recommendations were presented to decision
making bodies (e.g., the QI Council, the Administration, the Clinical Pathway
Steering Committee) for approval. In presenting their recommendations, the project
leaders and their teams enacted information about the problems and also conveyed
knowledge, in that they were attempting to persuade the decision makers to adopt
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
163
their respective frameworks for solutions. On the Mislabeled Specimens project, the
decision makers enacted their own knowledge in requesting some minor changes to
the recommended policy. But the main resource that the decision makers enacted in
this phase was, of course, their authority—by approving the recommendations.
Implementation. The implementation phase is widely regarded as the most
treacherous time for organizational change efforts:
Perhaps the most vulnerable time of the innovation process is during
the implementation stage when the dysfunctional nature of
organizational politics is most often highlighted. It is responsible for,
among other things, unnecessary delays, excessive conflict,
compromised outcomes, and sometimes, ultimate failure (Corwin,
1972; Debecq & Mills, 1985; Guth & Macmillan, 1986; Nelkin, 1984;
Pelz & Munson, 1982; Yin, 1977). (Frost & Egri, 1991, p. 243)
Because this phase introduced changes into the larger organization, it involved
communication with a much larger number of employees than in earlier stages of the
three QI projects in this study. Direct and indirect communication from the project
teams to the affected departments increased dramatically during the implementation
effort.
The project leaders conducted and coordinated this phase, working with local
management to ensure that implementation happened in all affected departments. All
three projects in this study experienced changes in leadership around the
implementation phase. For the PA and MS projects, the transition occurred right
before implementation began. The first Asthma Pathway team attempted to
implement their clinical pathway under Dr. Song’s leadership, but were
unsuccessful, leading to a second implementation attempt that was led by Dr.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
164
Newman. HFC administrators actively appointed or hired new leaders for the Patient
Access and the Asthma Pathway projects; on Mislabeled Specimens, the subsequent
leader was chosen by her fellow team members.
Implementation approaches have been found to vary in their overall
effectiveness (e.g., Nutt, 1986). Some research has suggested that the effectiveness
of different implementation strategies varies depending on the attributes of the
innovation (Leonard-Barton, 1988b), and several “ideal types” of change processes
have been identified that pair particular implementation approaches with specific
kinds of innovations (e.g., Huy, 2001b). The task of implementing an innovation that
affects many members of an organization has been likened to “delivering a new
product to market,” requiring that change leaders act like marketers, and demanding
organizing skills as well as interpersonal ones (Leonard-Barton, 1990). Similarly,
Fairhurst (1993) has argued that organization-wide change visions are implemented
via internal campaigns. These internal campaigns are “like their external counterparts
in that they are planned, organized efforts to mold corporate images, manage issues,
and articulate values” (p. 334).
Education was the primary implementation mechanism for two of the
projects. The MS team had a detailed education plan for ensuring that the directors of
the different divisions (e.g., Nursing, Ancillary Services), plus the residents, were all
oriented on the new specimen labeling policy. Karen (MS2) coordinated that plan by
ensuring that all of the relevant groups and individuals received the relevant
information and training materials before the policy was in effect. Dr. Newman
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
165
(ACP2) personally educated physicians about the asthma pathway and the new
pathway forms (physicians were to use the forms for taking asthma patients’
histories and physicals, and for ordering asthma-related tests and treatments).
As mentioned previously, Dr. Newman (ACP2) used lecture-style
presentations and broadcast email messages to educate the residents on the asthma
pathway, whereas a couple of departments relied more heavily on written documents
(the policy and a post-test) in their specimen labeling training. The MS team also
used internal newsletters to remind employees about the policy, and they had the
Education Department create a computer-based training module on the policy that
each employee was required to complete once a year with their annual safety
training. The goal of these educational activities was to not only to transmit
information about the new policies but also to transfer knowledge about how to do
things differently in their everyday work activities.
The implementation approach was quite different on Patient Access because
the changes involved radically restructuring a department to support new work
processes. These changes included designing new jobs and a new career ladder,
cross-training employees, opening a new unit (the Patient Access Center) to
centralize the pre-registration and pre-certification functions, and enforcing a new
“gatekeeper” policy requiring insurance pre-authorizations for scheduled procedures.
Alexandra (PA2), implemented these changes over a longer period of time than the
other projects (six months, as opposed to one or two). As with the other two projects,
implementation required the project leader to enact information (e.g., publicizing
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
166
changes to hospital employees and physicians), and knowledge (e.g., cross-training
Patient Access employees). Alexandra (PA2) also enacted authority when she pushed
the Network to make allocative resources available to support her implementation
efforts (e.g., raises for employees whose jobs had been upgraded).
Monitoring. Information is the focal resource of the monitoring phase. After
implementation, the project leaders observe whether or not their innovations are
being used, how, by whom, and to what effect On these projects, the project leaders
ensured that data were collected and analyzed, and that information was transmitted
to top management. As mentioned in the previous chapter, team members did not
receive updates unless project meetings were held and they attended. Team meetings
were held after implementation of the specimen labeling policy and the asthma
pathway forms. However, neither the Patient Access team nor the initial asthma
pathway team met during die monitoring phase, so none of their members received
status reports on their projects. Therefore, these team members were not able to
update their home departments.
Information flow to the departments was also poor from other directions.
Alexandra (PA2) did not update employees in departments affected by the Patient
Access changes except on issues that were directly relevant to their operations, nor
did the Administration report on the project’s status to these departments. It was
unclear who received Asthma Pathway updates aside from the QI Department (which
collected the data) and the Medical QI Committee. Directors received regular
updates from Karen (MS2) with the total number of mislabeled specimens in the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
167
hospital and in their department, but it was almost a year after the policy had been
implemented before the team reconsidered what kind of information the directors
would find most useful for following up in their departments. Around that time there
was also talk of the administrators monitoring the directors to ensure their follow-up.
Monitoring was halted on the ACP after the COO asked the team to merge
with the Asthma Group. The group met so infrequently (once in six months) that the
ACP team stopped collecting data for a while, until the ACP team split off (with the
permission of the COO) and started meeting regularly again. This example illustrates
how senior managers can dramatically affect information resources through enacting
their authority and influencing the distribution of allocative resources (i.e., employee
time).
Revision. Monitoring often reveals aspects of the innovation that need tweaking.
Innovations affect organizational practices, which is why implementing them can
require a heavy investment in knowledge resources. Then, project leaders use their
information resources to decide how to improve their innovations. The Appendix
notes that Alexandra (MS2) modified the application of the “gatekeeper policy” in
order to account for the needs of a particular group of patients. She was able to
identify this issue because of careful monitoring: she tracked and reviewed rejections
for every day since the policy came into effect. Dr. Newman (ACP2)’s revisions to
the asthma pathway forms were based on more informal information-gathering from
the residents who were using them.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
168
Feedback and revision processes are not always apolitical. On the Mislabeled
Specimens project, Karen (MS2) faced considerable conflict with the Emergency
Department (ED) over the language and intent of the policy. Since Karen (MS2) felt
that she lacked authority, so she appealed to a decision making body in the hospital.
She brought the conflict to the Medical QI Committee, which seemed sympathetic to
her framing the problem as a noncompliance issue. The Committee asked the
Laboratory’s Medical Director to talk with the ED. Eventually, a compromise was
reached and a revised policy was implemented in February, 2000.
Project leaders were not the only ones who initiated revisions. Other team
members took steps to adapt organizational practices to support the innovations. The
Clinical Resource Manager did this when she reassigned the responsibility for
restocking the asthma pathway forms to the QI Department. She did so because the
other departments were not reordering these forms, a problem that arose after the
hospital began charging departments for their printing expenses (instead of allocating
those expenses to the Print Shop’s budget). Some ICU nursing assistants also
adjusted their practices to support the new specimen labeling policy. They began
taping the labels to the patients’ beds in order to remind the nurses and other clinical
staff to attach the labels to the specimen at the bedside (as stated in the policy). This
practice did not seem to have a great impact on clinicians’ behavior. Nurses
continued to look for their labels at the nursing station, and would take the labels and
walk away from the bedside.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
169
Institutionalization. The institutionalization phase is concerned with the stable
integration of project innovations into the ongoing work practices of the
organization. The three projects were all grappling with institutionalization—with
varying degrees of success—when this study ended. One reason that the change
leaders had difficulty with the institutionalization phase may have been that they had
not thought far enough ahead in earlier phases about the requirements for
maintaining these changes. Frost and Egri (1991) found that successful
administrative innovations were championed by change leaders who were aware of
the long-term implications of their changes, and who laid the groundwork for them
accordingly. Innovators who were focused on more on short-term success were
associated with administrative innovations that failed. On these three projects, the
institutionalization phase did not appear to be part of the leaders’ mental models of
the change process. They could have been influenced by the Network’s QI model,
which anticipated Phases 2 (initiation and investigation) through 5 (monitoring) of
the QI project sequence, but did not include institutionalization in its framework.
The teams and the change leaders did take some steps to institutionalize their
project’s changes. Two aspects of institutionalization were evident in these projects:
action cycles, and accountability structures. Accountability structures establish
responsibility for various tasks involved in maintaining the innovation. These
structures were often organized by actors who were not project leaders. For instance,
when the CRM made the QI Department responsible for keeping the asthma pathway
forms in stock on the units, she assigned an employee to go every other Friday to
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
170
check on the status of the forms and order more if necessary. Similarly, the
Administration initiated such a structure when they hired Alexandra (PA2) to he the
Director of Patient Access, giving her the mandate to implement and maintain the
PA team’s recommendations. The Administration later promoted Alexandra (PA2)
and left the directorship position vacant, which threatened the institutionalized
changes. This illustrates why administrative innovations need long-term support
from top management (Frost & Egri, 1991).
Accountability structures were sometimes uncertain. The asthma pathway
was one of the first clinical pathways to be successfully implemented, so there were
many institutionalization issues that had not yet been addressed. The Asthma
Pathway team members had differing expectations about who would eventually take
responsibility for monitoring and maintaining the pathway. One member of the
initial project team thought that a review board should be established to monitor all
of the clinical pathways. Some people on the later team thought that the Asthma
Case Manager would take over the management of the asthma pathway, while others
seemed to think that the CRM would perform this function, because she was in
charge of coordinating the pathways. Dr. Newman (ACP2) did not have the authority
to give that responsibility to either of these employees (the CRM worked for the QI
Department and the ACM worked for Nursing).
A member of the ACP team mentioned other institutionalization tasks or
duties that needed to be assigned. “Right now, the orders for the asthma pathway
look very different from the orders for the bronchiolitis path. Someone with authority
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
171
needs to make a decision about standardizing the format of the forms. Someone also
needs to think about how the pathways fit together [in terms of clinical decision
making].” When this study ended these issues were still unresolved.
Action cycles refer to the routinization of innovation-related behaviors so that
they become part of the recurring activities of the organization. Examples included
the addition of training on the specimen labeling policy to employees’ annual safety
training, and the addition of education on specimen labeling and on the asthma
pathway to the orientation for incoming medical residents. The Clinical Resource
Manager (CRM) also planned to establish a regular review and revision timetable for
clinical pathways once a few had been implemented.
Training and review cycles usually happen at regular intervals over the span
of months or years, whereas the essential activities required for putting the QI
innovations into practice are performed much more frequently. For instance,
employees and physicians were supposed to follow the specimen labeling policy
each and every time they collected a specimen, which happened several times a day.
The residents were supposed to use the asthma pathway forms whenever they treated
a patient with a diagnosis of status asthmaticus, except for when they were admitted
to the Intensive Care Unit, or had other complications (e.g., intubation, comorbidity).
Since asthma is extremely common among urban children in industrialized countries,
the hospital saw these patients several times a week, and even more often during
asthma season.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Ill
However, a major problem with both the MS and the ACP interventions was,
in the words of one ACP team member, “There is nothing in the process that forces
them to use it.” Lawrence, Winn and Jennings (2001) have argued that
institutionalization processes that depend on depend on episodic mechanisms—that
is, the indefinite repetition of forceful or influential acts—are inherently unstable.
Most managers and physicians who were advocates of the specimen labeling policy
believed that the policy would have succeeded if managers had enforced it by
appropriately disciplining employees who contravened it. A member of the MS team
observed that modeling was important: “New people come in and see others doing it
the wrong way.” A different member found that “a lot of [getting people to do it the
new way] has to do with just continuing to remind people.” These different
perspectives highlight how the new policy affected entire communities of practice
(Lave & Wenger, 1991) so that any relevant learning was embedded within a
complex of interactions and experiences. A combination of consistent modeling by
peers and superiors, reminders, and discipline may have been needed to ensure that
employees and residents adopted the new practices.
In order to mobilize these widespread changes in behavior, it would have
been crucial for the project leaders to enlist the help of managers at the upper,
middle, and supervisory levels in modeling, emphasizing and enforcing the policy.
Both projects would have needed the help of the attending physicians to model,
emphasize and enforce the new practices. There is evidence that middle managers
and supervisors play a critical role in the implementation and institutionalization of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
173
organizational changes through their everyday communication with subordinates and
others in their network (Fairhurst, 1993; Huy, 2001a; Larkin & Larkin, 1994; Young
& Post, 1993). These key groups were all but ignored during the investigation,
recommendations, and implementation phases of the MS and ACP projects.
If Karen (MS2) and the administrators had discussed the policy with the
directors and their managers (instead of just presenting it to them), they might have
clarified expectations about implementation and institutionalization, beyond
employee training. For example, in one meeting that took place ten months after the
policy had been adopted, a director admitted that she had recently counseled an
employee in her department who had mislabeled a specimen, rather than giving her a
written warning. Karen (MS2) explained that this was fine: it did not contravene the
specimen labeling policy, because discipline is not mandatory. The director
suggested that some directors may not have understood that The Emergency
Department’s divergent interpretation of the policy might also have been identified
sooner if there had been more discussion with its medical and nursing directors. Data
analyses in Chapter Four indicated that authority was a necessary resource for
establishing communications with directors, managers, and supervisors in the
hospital. One implication of this is that top management may have needed to initiate
discussions about the policy with managers in the different departments.
The Patient Access changes appeared to have a more lasting effect than the
implementation efforts on the other two projects. “[Wjhereas episodic forms of
power need to be repeated on a continual basis in order to sustain institutional
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
174
change, discipline is able to create stable institutions because the external pressures
from which it stems are internalized and become a routine, ongoing part of the
targeted organizations” (Lawrence et al., 2001, p. 637). The Patient Access project
appeared to achieve such internalization through a series of changes that involved
mutually adapting the organizational structure and the new practices that were
introduced (Bikson & Eveland, 1998; Leonard-Barton, 1988a). New jobs were
created, enabling some Patient Access employees to devote their entire workday to
scheduling, pre-certifying or pre-registering patients. This helped ease the workflow
in Admissions. Employees were cross-trained so that they could move between
inpatient Admissions, Diagnostics registration, and the new Patient Access Center,
depending on the volume of patients. Admissions was moved closer to Diagnostics
registration to facilitate employees’ movements. Thus, Alexandra (PA2)’s structural
changes largely determined the practices embedded in the work system.
It is interesting to note that Karen (MS2) eventually decided that technology
would offer an ideal solution for effecting changes in employees’ specimen labeling
behavior. According to Lawrence, Winn and Jennings (2001), material technologies
“support patterns of social action in an ongoing way and without the complicity of
those on whom they act” (p. 637). They can be implemented relatively quickly and
tend to have more lasting effects on change targets’ behavior than institutionalization
approaches that rely on episodic power. By the end of this study, Karen (MS2) had
come to believe that the hospital should adopt a bar coding system that would be
used for specimen labeling and other applications (e.g., administering medications).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
175
Whereas the existing practices made it difficult to detect if policies were being
observed (e.g., checking all patient armbands to make sure they match the specimen
labels) unless errors occurred, the bar coding technology integrated documentation,
error detection, and error reporting into the mechanism for collecting specimens and
administering medications. It represented a “whole change of process” that would
eliminate the need for continually reinforcing a series of practices among numerous
employees. There was a major obstacle to implementing this new technology:
allocative resources. The Network and hospital’s antiquated IS was not compatible
with the bar coding technology, and system was very expensive to replace.
Ultimately, the bar coding would have to be part of a larger IS upgrade that could
only be authorized at the executive level of the Network.
Job Attributes of Change Leaders
“The role of the top executive sponsor has been shown to be a critical one in
facilitating and ensuring the success of an innovation (Blau & McKinley, 1979; Nutt,
1986)” (Frost & Egri, 1991, p. 267). Frost and Egri note two functions of for top
managers in change efforts: they provide access to allocative resources, and they
communicate new expectations to employees that motivate or elicit change. They are
able to enact these resources because of their formal authority in the organizational
hierarchy, which grants them considerable influence.
At HFC, top managers were most involved in the QI projects during the
initiation phase (when they appointed two of the project leaders) and the recommen
dation phase (when they received and approved the team’s recommendations). The
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
176
administrators also facilitated implementation by installing replacement leaders for
the Patient Access and Asthma Pathway projects. Aside from these actions, the top
managers were not directly involved in project activities, such as communicating
about the changes with affected stakeholders. From the recommendations phase
onward, the administrators rather passively received information from the project
teams. They tended to get involved only to help with troubleshooting, (e.g., the
Medical Director intervened to revive the defunct asthma pathway, and the COO met
with the MS team to discuss their difficulties instituting the specimen labeling
policy).
Top management’s involvement in appointing project leaders appears to have
been related to the progress made by the projects. The two projects that were
initiated by top management began relatively straightforwardly (Mislabeled
Specimens and Patient Access), whereas the bottom-up clinical pathway initiative
faced more start-up difficulties. Similarly, the two projects where top management
appointed replacement leaders to implement the changes accomplished more (in
terms of measurable outcomes) than the MS team, where the second leader had
emerged informally. The QI literature suggests that bottom-up projects tend to
struggle because they are under-supported; that is, because they receive too few
resources (Blumenthal & Edwards, 1995; Lewis, 2000b). Numerous studies have
found that organizational innovations often fail due to insufficient resources
(Delbecq & Mills, 1985, cited in Frost & Egri, 1991), and indeed some employees
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
177
did report that the initial asthma pathway effort had lacked resources, especially for
data collection.
Progress may also have been affected by the job attributes of the project
leaders, particularly those affecting status. Three of the six project leaders (i.e.,
departmental directors) were upper managers (see Table 5.2). All three of these
leaders seemed to make progress in moving their projects forward. There was
widespread agreement that a physician needed to lead the Asthma Clinical Pathway,
because no other clinical professional would have sufficient expertise to oversee the
design of the pathway (knowledge resources), nor would they have the clout
(authority resources) to get physicians to use it. Thus, both leaders on the ACP were
physicians. However, the subsequent leader got more administrative support
(allocative resources) than the initial leader did. By the time that Dr. Newman
(ACP2) took over the project, the QI Department had been restructured, the Medical
Director had made clinical pathway implementation a priority and had given the
Clinical Resource Manager responsibility. This kind of support was key for every
phase of the process, particularly for enacting information (e.g., data collection,
scheduling meetings).
The second leader of the Mislabeled Specimens project was neither a director
nor a physician. Karen (MS2) felt that her lower position in the organizational
hierarchy impeded her ability to successfully implement the new specimen labeling
policy. Her lack of authority probably most affected the implementation and
institutionalization phases, but the effects were most likely indirect. For example,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
178
being in upper management would not have enabled Karen (MS2) to force other
directors to adopt and enforce the policy in their departments. But had she been a
Table 5.2 Project Leader Attributes
Project Leader Order
& Name
Title Professional
Training
Leadership
Origins
Mislabeled
Specimens
Initial:
Jane
Director of
Decision
Support
health
administration
appointed by
Administration
Subsequent:
Karen
Quality
Improvement
Coordinator
nursing chosen by team
members
Asthma
Clinical
Pathway
Initial:
Dr. Song
attending
physician
(Professor of
Pediatrics)
medicine
(subspecialty
in pediatric
allergies)
recruited by
Clinical
Pathway
Steering
Committee
Subsequent:
Dr. Newman
attending
physician
(Associate
Professor of
Pediatrics)
medicine
(general
pediatrics)
recruited by the
hospital’s
Medical
Director (i.e.,
Administration)
Patient
Access
Initial:
Madeleine
Director of
Rehabilitation
Services
physical
therapy,
public
administration
appointed by
Administration
Subsequent:
Alexandra
Director of
Patient Access,
Decision
Support
accounting,
business
administration
hired by
Administration
director, she would probably have had more opportunities to communicate with other
directors and managers about the policy, and thereby build support for her initiative.
She also would likely have had much closer ties to the next level in the hierarchy,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
179
since directors reported to a vice-president or administrative director, who have
authority over multiple directors.
Project Teams and the Dynamics o f Representation
Throughout the life span of a project, ongoing information is produced about
the status of the change relative to the ultimate goal. Project leaders are responsible
for managing such information, which includes data collection and analysis,
producing and distributing progress reports, and presenting recommendations to
decision makers. Phases 2 through 6 (investigation, recommendations,
implementation, monitoring, and revision) require the most attention to information
flow between the project leader and the rest of the organization. The project team
functioned as the main interface between the leader and the rest of the organization
for these phases (except for in the later stages of the Patient Access project, since the
team was inactive after it made its recommendations).
The popular management literature on quality improvement focuses on the
use of technical tools and the adoption of a customer-service philosophy, but it has
little to say about how QI project teams are supposed to operate. This “gap” in the
literature means that “there is very little concern for how [teams] function as a social
unit and a decision-making body set within a larger organizational context”
(Fairhurst & Wendt, 1993, p. 443). More specifically, there is little guidance as to
how the teams should relate to the organization’s existing authority structure (p.
445).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Crosby (1979) says that once an organization commits to adopting a QI
program, a planning and implementation team should be assembled, consisting of a
chairperson (team leader) and representatives from each department. “These should
be people who can speak for their department in order to commit that operation to
action. (Preferably, the department heads should participate—at least on the first go-
around.)” (p. 113). This implies that authority is an important resource for team
members (not just for the project leaders). When project team participants are lower
level employees (i.e., line staff), progress can be slowed by the need for
representatives to take information back to their directors for their input. Without
their input, important project decisions cannot be made. A leader team who was
frustrated by this lag said, “The people who come to the meeting have to be
empowered and know what their boss wants, or what they want to get out of it, so
that they don’t have to keep running back.” Of course, if the upper managers are
directly involved, it saves this step.
Meetings. Meetings are sensemaking events that can generate and even constitute
social structures (Huff, 1988; Schwartzman, 1989). Communication with team
members about the three QI projects occurred almost exclusively in meetings, so
meetings were extremely important for information and knowledge sharing, as well
as decision making. Strikingly, however, all three projects experienced problems
with meeting attendance. Schwartzman (1989) has asked “How can we explain the
appearance of participants at meetings? How and why do participants choose to go to
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
181
particular meetings?” (p. 311). The following analysis explores those questions with
respect to the project team meetings.
Occupational and organizational cultures can shape perceptions of meetings.
For instance, a study of an urban crisis center found that employees resented
“wasting time” in meetings because they would rather be spending their time directly
helping their clients (Mansbridge, 1983, cited in Schwartzman, 1989). Research in a
British women’s center (Brown & Hosking, 1986) revealed differences in members’
perceptions of their meetings. The paid workers in the center found the meetings
“largely a waste of time,” and wanted a more conventional meeting format (with a
formal agenda and meeting minutes). Some volunteer participants saw the meetings
as “good for seeing other people and keeping in touch,” as well as discussing issues.
The women’s center eventually decided to alternate business meetings with issue-
based discussions. These two examples show that employees’ attitudes toward
meetings are shaped by their occupational values or priorities, their goals for the
meeting, and their perceptions of the meeting’s form and content.
At the Hospital for Children, employees and physicians expressed two views
of meetings. According to the first perspective, meetings were seen as something to
be minimized, as an activity that took resources (i.e., staff time) away from patient
care or helping subordinates. This negative view of meetings was expressed by
individuals at every level of the hospital’s hierarchy. It could be seen in
administrators’ efforts to consolidate committees and work groups in order to
minimize the number of meetings, as described in Chapter Four. A second view
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
182
considered meetings to be essential places for conveying and obtaining information.
This could be seen in the frustration that two managers felt at their lack of access to
operational meetings at different management levels, and in the frustration of the two
project leaders who sent messages to their respective team members in an effort to
improve attendance. The tension between these opposing attitudes played out in the
dynamics surrounding the project meetings.
The project teams started with large team rosters: 24 people on the Patient
Access project, 15 on the initial Asthma Pathway team, and 11 on Mislabeled
Specimens. The number of team members who consistently attended meetings
ranged from four to seven (including the leaders) on the three projects. Leaders’
reactions to the problem of meeting attendance were discussed in Chapter Four.
Their responses included ignoring the issue, stepping in to personally replace the
missing team member, sending messages admonishing representatives to attend, and
eventually seeking replacements for former team members. When information flow
to departments broke down due to problems with representation, some individuals in
departments affected by the changes later felt resentment, perceiving the QI projects
as top-down impositions.
Figure 5.1 shows the relationships among several key factors that affected
meeting attendance. This figure was derived from qualitative analysis of interviews
with HFC employees who participated in some or all of the meetings for the three QI
projects. Access to the project meetings was controlled by three actors: the project
leader, the employee, and the employee’s boss (their director, or for directors their
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
183
vice-president). Staffing shortages affected employees’ availability both directly
(through the employee’s discretion) and indirectly (through their directors’
decisions). Many of the hospital’s departments were short-staffed, and some team
members had workloads that made it difficult to consistently participate in meetings.
This was especially true for clinical workers (e.g., nurses, respiratory therapists).
Some people observed that it was easier for managers to attend meetings than it was
for clinical staff, because managers’ days had more time available for such activities.
Figure 5.1 Factors Affecting Meeting Attendance and Implications for
Information Flow
Staffing
Shortages
Directors’
Support
Information to
team members
Member
Willingness
Mediated Project
Communication
Information to
departments/
physicians
Project
Meeting
Attendance
Meeting
Attributes
• content
• length
• timing
• frequency
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
184
Directors’ support for their employees’ participation was affected by their
sense of their department’s resources (i.e., whether they could spare this employee or
not), and by their sense of whether the meetings were a good use of the employee’s
time—which was largely influenced by the employees’ own reports. Members’
attitudes and behavior toward the meetings were influenced by their perceptions of
the meetings’ attributes, including their content, their length, their timing, and their
frequency. A physician explained that she had stopped attending meetings for one of
the projects because they were always held at an inconvenient time. A Patient
Access team member said that she had to miss a couple of project meetings because
they were held so frequently that occasionally they conflicted with other meetings.
Aside from meeting timing and frequency, meeting content was a significant
factor. Some employees applied one or more of Grice’s (1975) conversational
maxims to the discourse of the project meetings. The Maxim of Quantity tells
participants to make their contributions as informative as is required for the purposes
of the exchange, but no more so. A Patient Access team member complained about
having to sit through several two-hour meetings where employees mapped out
process flow charts for the different departments from scratch, one person describing
their department’s registration process while someone else drew the diagram on a
flip chart. Shortly after these charting meetings began, overall attendance on the
project did decline. Discussing the finalized flow charts with the team might have
been sufficient; having 15 to 20 people witness the process of deriving the flow
charts seems to have felt excessive.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
185
Another such conversational maxim is the Maxim of Relation: “Be relevant.”
Employees expected meeting discourse to be relevant to the goals and tasks of the
project, and relevant to their department. A team member complained that discussion
in the meetings on one of the projects was sometimes allowed to go off topic, into
areas that were beyond the mandate of the project. A manager in a different
department described how he and his director had decided that he did not have the
time to attend the meetings for one of the projects because that project did not
sufficiently affect their department.
The Maxim of Relation points to the relationship between meetings and their
context. “Context continually emerges out of patterned behavior offering an ongoing
frame for individual behaviors and restricting future interaction” (Adams, 1997, p.
90). The preceding examples point to two levels of context for these meetings: the
context of the project, and the context of the larger organization. A communicative
acts exists in dynamic relation to its context, and this relationship may assume
different forms. According to Branham and Pearce (1985), a “subversive loop” is a
pattern in which a communicative act and its context are “mutually invalidating” (p.
23). In such cases, the relation between the act and its context is “irreparably
breached” (p. 24). The dynamic of poor meeting attendance is an example of a
subversive loop. Team members are either pulled or pushed away from the meetings
by a poor fit with their context They are unable to reconcile ongoing attendance with
their other duties, either because human resource issues in their departments leave
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
186
them with too heavy a workload, or because the meetings themselves have failed to
meet expectations of timing or relevance.
The model in Figure 5.1 suggests several implications for project leaders.
First, it indicates that meeting management skills are very important for these
leaders. These skills are necessary for the competent management of meeting
attributes, such as content and length. Project team members noted that the initial
leader of MS and the subsequent leader of ACP both had strong meeting
management skills. Jane (MSI) ensured that all team members had a chance to speak
in her meetings; Dr. Newman (ACP2) structured her meetings around assigning and
reporting on tasks.
This model also suggests that, if there are attendance problems, the project
leader should consider at least two possible explanations: the member and/or her
director has decided that the meetings are inconvenient or irrelevant, or the member
(and her department) is overwhelmed with work. Several solutions may be inferred:
changing the content or timing of the meetings, including members through other
communication channels (e.g., email), requesting member attendance only for
specific meetings where their input is needed, or requesting that the member’s
director assign a different employee to the team. The next section addresses two
relevant aspects of employees’ participation in the projects: their role expectations,
and their levels of participation.
Participation. Stohl (1995) has observed that the word “participation” is ambiguous,
and that it suggests multiple possible definitions. “It may denote, among other things,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
187
membership in a group or general involvement in group activities, or it may refer
more narrowly to involvement in decision making” (Stohl, 1995, p. 134). Part of the
problem with employee attendance may have been due to a gap between the
representatives’ perceived roles (what they themselves thought their duties were on
the project) and their expected roles (what the team leaders and active group
members expected of them) (Wofford et al., 1979, cited in Stohl, 1995). For instance,
the PA team member mentioned earlier may not have felt that watching others
construct flow charts was part of her job as a team member. As this gap became
evident over time, the representatives may have decided to stop attending, shifting
their enacted role from participant to non-participant.
Employees may be involved in change projects to varying degrees, with
inclusion levels ranging from very low to very high, depending on how much
information is shared and exchanged. DeLuca (1984) notes that those who attend
project meetings may do so as “observers,” “consultants,” or “participants” (i.e.,
team members), with expectations for attendance and communication that vary
accordingly. For example, on the Patient Access project, the hospital’s Risk Manager
was not formally a team member, but she did attend a few meetings to give advice on
certain issues (e.g., consent forms). Alexandra (PA2) was working as the director of
Patient Access at a children’s hospital in a different city when Madeleine (PA1)
invited her to speak at a meeting of HFC’s Patient Access team. It is up to the project
team leader to manage the appropriate level of involvement of the different actors in
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
188
order to ensure information flow in and out of the group for decision making
purposes, and to ensure that affected constituencies are adequately represented.
As change projects move through their different phases, it is also normal for
team members to change their participation levels. For example, a team member who
was part of the initial Asthma Pathway team became less involved after Dr. Newman
(ACP2) implemented the pathway forms, in part because she thought that “it’s
probably kind of to the point now where it’s not as necessary to have somebody
[from my department] there all the time.” Project leaders who discuss these role
transitions with their team members can help clarify expectations and understandings
about the status of the project.
Leaders Respond to Project Challenges
The seven project phases listed in Table 5.1 did not always go smoothly for the three
QI teams. Each project encountered difficulties that were not predicted in the
organization’s life-cycle model of QI change initiatives. These developments created
challenges for leaders, and raised this research question:
Research Question 5: How do change leaders manage emergent social
conditions (e.g., transitions in leadership) that are not addressed informal
models o f the change process? What role does communication play in the
management o f unexpected developments and unintended consequences?
The leaders on the three QI projects encountered similar sorts of challenges in the
process of designing and implementing their changes. These challenges, which are
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
189
summarized in Table 5.3, included conflict, bureaucratic obstacles, and changes in
project leadership.
Table 5.3 Unplanned Challenges in Quality Improvement Projects
Unplanned
Challenges
Relevant
Dimensions
Patient Access Mislabeled
Specimens
Asthma
Pathway
Conflict Locus
Project Phase
Focus
Within Team
Investigation
Problem Definition
& Solution Design
With Emergency
Dept
Monitoring
Problem
Definition &
Solution Design
Within First
Team; With
Emergency Dept
Investigation
Problem
Definition &
Solution Design
Changes in
Project
Leadership
"W hen Old
Leader Left
New Leader’s
Origins
Mandate of
New Leader
After
Recommendations
Hired by
Administration
Implement
Recommendations
After
Recommendations
Informally
appointed by team
Implement
Recommended
Policy
After Attempted
Implementation
Appointed by
Medical Director
Implement
Pathway
Bureaucratic
Obstacles
Level
Phase
Issue
Network
Implementation
Delays with
Network’s IS Dept
Delays with job
descriptions
Hospital
Implementation
Delays in
approval o f
hospital-wide
policy
Department
Implementation
Forms not kept in
stock
Conflict
Conflict may be defined as “the interaction of interdependent people who
perceive opposition of goals, aims, and values, and who see the other party as
potentially interfering with the realization of these goals” (Putnam & Poole, 1987, p.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
190
552). Organizational change initiatives frequently affect multiple stakeholders with
different agendas, and often some stakeholders will act to advance their own interests
with respect to the changes (DeLuca, 1984; Frost & Egri, 1990a; Frost & Egri,
1990b; Frost & Egri, 1991). These political maneuvers sometimes develop into
conflicts. Dabbah (1993) has observed that “conflicts in projects are generally subtle,
covert, and insidious” (p. 199). In the absence of open communication about
disagreements, conflicts may remain undetected or unacknowledged. There is
evidence that when leaders delay intervening in organizational conflicts, the
outcomes are generally undesirable (Phillips & Cheston, 1979, cited in Putnam &
Poole, 1987).
Leaders are responsible for managing organizational conflicts. Conflict
management entails steering a conflict so that it may achieve constructive
consequences (e.g., revealing insights, problem resolution, enhanced group
productivity, commitment, and unity), and avoid destructive consequences (e.g.,
polarization; factionalism; decreased group productivity, satisfaction, and
commitment) (Barge, 1994). Leaders need strong conflict resolution skills for
working with stakeholders on either interpersonal or task-related conflicts (Barge,
1996). Constructive conflict resolution helps foster a safe environment for airing
future conflicts and helps establish norms for managing them. Since conflicts are
social processes that occur through communication (e.g., Barge, 1994; Nicotera,
1995; Putnam & Poole, 1987), conflict management is essentially a communicative
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
191
undertaking. It involves defining issues, clarifying viewpoints, exploring
alternatives, and negotiating solutions, all of which involve communication.
In the three QI projects, the conflicts concerned defining the problems or
designing solutions. Two of the projects (Patient Access and the Asthma Pathway)
experienced extended disagreements among team members. The initial asthma
pathway was revised repeatedly (someone said they “must have changed it 15
times”), in part because the three specialists who were working on it could not agree
on its content. According to one subsequent team member, the pathway spent “a year
in limbo,” being passed among the three physicians (allergists and immunologists),
each adding changes that contradicted the other two doctors. One of these physicians
was Dr. Song, the initial project leader. It does not appear that she took any steps to
directly address the conflict.
Similarly, the conflict on the Patient Access team was not actively addressed.
Several team members reported that directors on the team did not want to
recommend the centralization of the Patient Access function. They wanted to
recommend standardization, but they opposed centralization because they feared that
they would lose their employees who were doing registration together with other
duties. Eventually, however, the team did recommend centralization. A team
member suggested two explanations for how this happened. One factor was attrition:
those who most strongly objected to centralization stopped attending the meetings.
The other reason was that the recommendation was made in response to pressure
from above. The Administration was advocating bedside registration, which the team
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
192
did not feel was feasible. Their response to was to recommend centralization first, to
establish a structure that would support the eventual introduction of bedside
registration (in fact, Alexandra [PA2] later did come up with a plan for implementing
bedside registration). Thus, the conflict was resolved due to shifts in the balance of
power, rather than active conflict management techniques.
The directors of the Emergency Department (ED) were involved in conflicts
with two of the projects (the Asthma Pathway and Mislabeled Specimens). The ED
initially refused to participate in the ACP project. This was an immovable obstacle
for the team until the Medical Director of the ED left and was replaced with a
physician who was open to the pathway. At that point, the ED took over the design
of their piece of the pathway, which was later integrated into the remainder of the
path. In this case, neither the project leader nor the adm inistrators took any step to
resolve the conflict—it resolved itself through turnover.
The conflicts discussed above occurred when the teams were still
investigating their problems and deciding on recommendations, whereas the
Mislabeled Specimens project’s conflict with the ED occurred after the new
specimen labeling policy had been implemented. One employee commented on this
timing, remarking that the whole purpose of the QI process is for the
multidisciplinary team to reach an agreement on a re-engineered process, which is
then approved and implemented. Indeed, other MS team members had registered
disagreements with the policy while it was being drafted, and these had been
addressed. The conflict with the ED became political and personal, because Karen
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
193
(MS2) had been an assistant nursing director in the ED, and she had been the ED’s
representative on the MS project team before moving to the QI Department and
assuming the leadership of the MS project.
The conflict centered around the fact that Nursing Director of the ED did not
want medical residents to label their own specimens. This was a policy that she and
the previous Medical Director of the ED had developed years before. The rationale
was that the Nursing Director had no control over training or discipline for the
residents. She did not have the allocative resources to collect the data to show that
this policy actually reduced the number of mislabeled specimens. She also did not
see any inconsistency between the ED’s policy and the specimen labeling policy. For
instance, “it doesn’t even say in the hospital policy that the person has to attach the
label.” However, the MS team had intended for the line at the top of the policy that
said “Responsibility: Person Collecting Specimens” to indicate that all of the actions
in the policy were to be performed by that individual, unless it stated otherwise.
After Karen (MS2) brought the disagreement to the Medical QI Council
(enlisting the authority of the physician leadership), and the Medical Director of the
Laboratory got involved, a compromise was reached whereby the residents would
use a marker to temporarily label any specimens they collected, to which others
would later add paper labels. Ultimately, this compromise did not satisfy any of the
stakeholders. ED employees continued to resent, misunderstand, or ignore the
official specimen labeling procedures, and the MS team continued to be frustrated by
ongoing problems with mislabeling throughout the hospital, including the ED.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
194
The Hospital for Children’s Medical Director believed that conflict
management skills are essential for leaders because conflict is “one of the biggest
problems you have to learn how to deal with” as a leader. “Physicians tend to shy
away from conflict. And that’s probably the most important part of this job, is
dealing with conflict.” The Mislabeled Specimens project was the only case in which
a leader actively intervened to resolve the situation. On the other projects, conflicts
were allowed to resolve themselves through historical accident. In the case of the
Asthma Pathway, the conflict dragged out over at least a year, depleting team
members’ morale.
Project Leadership Transitions
Much has been written on the need to carefully manage transitions in chief
executive leadership, but the literature on lower levels of leadership (group
leadership, project management) has largely ignored the issue of managing
transitions. Jane (MSI) and Madeleine (PA1) both left their projects after they
presented the team’s recommendations to the Administration; Dr. Song (ACPI) was
replaced after her team attempted to implement the pathway, without success. HFC’s
top management hired Alexandra (PA2) to be the Director of the Patient Access
Department, and the Medical Director asked Dr. Newman (ACP2) to lead a new
initiative to re-implement the asthma pathway. After their initial leader left the
organization, the Mislabeled Specimens team chose one of their teammates (Karen)
to be their new leader.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
In all three projects, the subsequent leaders were tasked with implementing
changes advocated by the initial leaders. They referred to the documents left behind
by the previous project leaders, especially summary documents (e.g., the description
of the asthma pathway, the Patient Access recommendations). There seems to have
been little direct communication between the initial leaders and the subsequent
leaders during or after the transition; there was none on MS, because the new leader
was chosen only after the old one had been gone for some time.
The overall pattern across the three projects demonstrates minimal or no
transition management. The difficulties faced by Karen, the MS team’s emergent
leader, have already been discussed. Karen’s transition would have benefited from
stronger links to top management (to supplement her weak formal authority), and
from immediate training in QI knowledge. Top management saw its role as ensuring
continued leadership on two of the projects, and the administrators used their
authority and allocative resources to do so. But the administrators did not convey any
expectations to the new leaders with regard to updating project team members from
earlier phases of the projects. The new leaders did not know or think to transmit this
information, leaving those members uninformed about the status of their projects. In
a more formal leadership transition, the subsequent project leaders on ACP and PA
also might have contacted their predecessors in order to access knowledge or lessons
about the projects that were not contained in the summary documents.
permission of the copyright owner. Further reproduction prohibited without permission.
196
Bureaucratic Obstacles
Just as project leaders need conflict management skills, they also require a
well developed ability to navigate bureaucracy. All three projects experienced
bureaucratic obstacles during the implementation phase. These obstacles occurred at
different levels: the Network, the hospital, and the department.
Karen (MS2) and Alexandra (PA2) both experienced implementation delays
because of bureaucratic approval processes. The hospital took months to approve the
specimen labeling policy, and the Network also took several months to approve new
job descriptions for the Patient Access Department. Neither leader took any action to
speed up this process; they were preoccupied with other implementation tasks. Karen
(MS2) later felt that, if faced with such a problem again, she would be relentless in
following up. She would “stay on top of it, keep asking, ‘Did you get that policy?’,
ask the secretary if she’s given it to [the COO].”
Alexandra (PA2) took a different approach when the IS department’s work
schedule threatened to delay the launch of the new Patient Access Center by several
weeks. Chapter Four described how she used her informal network to get the
necessary work done right away. The Clinical Resource Manager, who was Dr.
Newman (ACP2)’s lieutenant on the ACP project, also sidestepped bureaucracy in
order to solve a problem with institutionalizing the asthma pathway in clinical
departments. The units were not reordering pathway forms when they ran low. First
she tried to get the unit secretaries to order more forms by asking them to call her,
she also posted signs as reminders. When those efforts were unsuccessful, the CRM
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
197
allocated QI Department funds to cover the cost of the forms and assigned a QI
Department employee to check and restock the forms on the units.
Sometimes the project leaders reacted passively to bureaucratic challenges,
but other examples show that the leaders or their “seconds-in-command” could be
quite creative in responding to these problems. Informal networks and the authority
to allocate resources were important aspects of responding effectively to these
challenges.
Summary
Change projects take place over extended periods of time, and involve
cyclical events that affect a team’s progress. On these three projects, top managers
were not involved on an ongoing basis. They initiated the projects by appointing
leaders and certain team members (e.g., directors); they received, revised, and
approved recommended changes; and they monitored the projects’ progress. After
the project teams made their recommendations, the administration “managed by
exception” (Bass, 1990), intervening only when there were problems. All three
projects experienced transitions in leadership. The top managers found new leaders
for two of the projects, but they did not appear to manage the transitions for the
teams. Communication was largely unidirectional, tending to flow upward to top
management, and rarely coming down to team members or employees in
departments affected by the changes.
The project leaders conducted their teams to investigate the problems, design
solutions, implement the changes, monitor these changes, and revise the innovations.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
198
Two of the projects had trouble institutionalizing their changes. Institutionalization
may not have been part of the mental models of the leaders, who often seemed to
overlook the action cycles and accountability structures that were needed to maintain
the new practices. The leaders apparently underestimated the quality and quantity of
communication with upper managers, middle managers, and supervisors that was
probably needed to institutionalize the new practices.
Some project leaders and certain team members could deal with unexpected
challenges posed by bureaucracy or conflict quite effectively, but often these
challenges were not addressed at all. As a result of conflict avoidance, the projects
suffered from delays, disillusioned team members, and frustrated employees in
affected departments. Two project leaders chose impersonal directives as their mode
of communication with team members about attendance problems; two leaders acted
as substitutes for missing team members, instead of communicating about the issue.
These approaches did not address the causes of the truancy and were only somewhat
effective. In sum, the three projects would have benefited from stronger and more
active conflict management.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
199
CHAPTER 6: DISCUSSION & CONCLUSION
Health care organizations are still struggling with quality improvement.
Organizational change is often difficult, and communication problems are among the
most common and challenging difficulties encountered by implementers (Lewis,
1999). The QI literature emphasizes the role of executive leaders in implementing QI
programs, but there is almost no research on project leadership or communication
during change initiatives. This study investigated how change leaders enacted
resources for three organizational change projects in an American pediatric hospital.
Four types resources were considered: information, knowledge, authority, and
allocative resources. The analyses focused on how leaders enacted these resources
and circulated them through organizational structures and systems, how patterns of
enactment changed over the project phases, and how leaders dealt with challenges
that arose over time.
Each of the three projects experienced a transition in leadership, so each
project had an initial and a subsequent leader. Research on QI and planned change
suggests that top managers can play an important sponsorship role in supporting such
innovations, so this study also examined how the hospital’s administrators
contributed to the projects. The analyses looked at how patterns of resource
enactment varied in different project phases, according to job attributes of the
different leaders, and in response to unexpected challenges.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
200
This chapter begins with a summary of the study’s findings, and then
discusses its contributions to the literature on organizational communication,
organizational resources, and leadership. The next section addresses some of the
limitations of the study. This is followed by a discussion of its implications for
practitioners. The chapter concludes with recommendations for future research, and
some remarks about my hopes for this study.
Summary of Findings
Certain patterns that were observed in this study were consistent with
previous research and with the practitioner literature. Top management involvement
was essential for project success. That involvement included attending meetings to
show support for an initiative, formalizing expectations for the process (e.g., setting
deadlines for deliverables), and dedicating allocative resources to the projects (e.g.,
appointing leaders, assigning skilled support staff). Emergent project leaders had
more difficulty than those appointed by top management. Project managers who
came to the project with more formal or professional authority, and with more
change management knowledge, were more successful than those who didn’t. Strong
meeting management skills, performance measurement skills, project management
expertise, media selection skills, and past experience with implementing similar
innovations were all beneficial forms of knowledge.
This study gives an unprecedented longitudinal picture of the roles and
interactions of different leaders throughout planned change initiatives. Over the
project life cycle, top managers and project leaders interacted at four main points: (1)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
201
when the administrators appointed a project leader, (2) when the project leader
reported progress to the administrators, (3) when the team made recommendations to
top management, and (4) when the project leader solicited help from top
management. After the early project phases, top managers appeared to adopt a style
of “managing by exception,” monitoring team performance, but only intervening
when things went wrong (Bass, 1990). Project leadership mainly involved managing
internal communication on the teams (e.g., meetings, data collection activities), and
external communication with stakeholder groups (especially during implementation,
revision, and institutionalization). From the investigation phase through the revision
phase, project leaders enacted information resources through the hospital’s
communication media and its information systems. Top managers consumed this
information; they were not involved in its production or dissemination.
Information flow within and beyond the project team was affected by a
number of factors, including the project leaders’ communication modalities.
Meetings were critical for communication among team members, since none of the
leaders distributed information to team members outside of project meetings. Email
tended to work well for planning and scheduling communication within the team, but
it was less effective for publicizing the changes that were implemented, and was only
weakly effective as a medium for directing team members to come to meetings.
Repetition of messages through multiple media was sometimes effective, but its
efficacy in part depended on the different media’s capacity to reach all the
stakeholder groups. In sum, most of the leaders were observed to experience some
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
202
communication difficulties that were related to their choice of media, though their
media selections also interacted with the style and content of their messages.
All three projects experienced problems with meeting attendance, which
negatively affected information flow to the representatives’ departments. The project
leaders took four approaches to the problem: (1) they ignored it, (2) they filled in for
the missing team member and communicated directly with the relevant stakeholders,
(3) they sent written messages to the team demanding their attendance, (4) they
eventually recruited a replacement from the lapsed member’s department. None of
these approaches dealt directly with the dynamics surrounding team attendance. Two
major factors affected team members’ attendance: the workload of the member’s
home department (this affected their availability), and the nature of the project
meetings (timing could determine availability, and content affected willingness to
participate). When team members perceived that meetings were relevant to the
project’s goals and to their departments’ concerns, they were more likely to
participate.
Beyond the team, most of the project leaders experienced serious difficulties
implementing their innovations. This is probably not surprising, given what is known
about the difficulty of implementing organizational change (see Frost & Egri, 1991).
Knowledge transfer from the team to the departments during implementation was
generally quite challenging. Employee and resident education on new policies and
practices was suboptimal on the two projects that required clinicians to change their
behavior. The project leaders who attempted to implement the specimen labeling
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
203
policy or the asthma pathway did not acknowledge that their innovations affected
entire communities of practice (Lave & Wenger, 1991), not just individuals. They
also failed to recognize that their changes relied on an unstable mechanism for
institutionalization: episodic power (Lawrence et al., 2001). As a result, they
underestimated their need for the support and cooperation of authority figures (e.g.,
directors) and influential groups (e.g., attending physicians) to change behaviors that
were so embedded in eveiyday work routines. Consultation with these individuals
and groups at key decision making moments might have helped gain their
cooperation, and thereby avoided their negative reactions toward the changes.
The project leaders missed other opportunities—mostly communication
opportunities—that could have helped their projects. Communication networks were
important in emergent processes (e.g., initiating a project, finding a new leader), but
the project leaders rarely used their networks for strategic purposes, to overcome or
pre-empt implementation challenges. This may have been because the project
leaders’ networks were inadequate. Some project leaders displayed underdeveloped
skills in data analysis or meeting management or both. Some lacked conflict
management skills: only one of the four major conflicts on the projects was
successfully resolved by a project leader; it appeared that two of the conflicts were
not even addressed.
The project leaders also faced organizational obstacles, including inert
bureaucracy and belt-tightening. At the time of this study, the hospital and its
Network were experiencing financial pressures resulting in budget cuts, as well as
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
204
staffing shortages, so allocative resources were scarce. The Network’s outdated
information system made data collection time consuming for the Asthma Pathway
project, and precluded the technological innovations that project leaders wanted the
hospital to adopt. When faced with bureaucratic problems, the project leaders
sometimes took creative steps to circumvent the bureaucracy. At other times they
simply waited out bureaucractic delays, which impeded the implementation process.
There were also times when the administrators’ inaction slowed the progress
of the QI projects, or resulted in poor communication. Top managers did little to
manage the transitions in project leadership that all three projects experienced. They
did not appoint a new leader for the Mislabeled Specimens project, and the team
stagnated for months in confusion until a leader emerged informally. Top managers
did appoint new leaders for the other two projects, but they did not clarify
expectations regarding project-related communication. As a result, two of the new
leaders did not update their projects’ initial team members.
On two occasions, administrators’ actions even appeared to undermine the QI
projects. The second Asthma Pathway team stalled after introducing their new
pathway forms because the COO had asked the team to merge into the almost
inactive Asthma Group. Also, by promoting the Director of Patient Access and then
leaving her position vacant, the administrators threatened to neglect the very
structure that had been recommended by the PA team and endorsed by the
Administration. External constraints (tight budgets, scarce human resources) may
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
205
have influenced top managers to act in ways that delayed project work and that
threatened the institutionalization of project changes.
In the preceding examples, the adm inistrators made decisions that had major
implications for the projects, apparently without consulting the project leaders.
Problems with meeting attendance, information flow to the departments, and
institutionalization of the changes suggest that both internal and external
communication could have been more effective. The leaders at all levels faced
serious constraints, but some of these (e.g., inadequate social networks, poor conflict
management skills) may have been within their power to change.
Contributions to Theory
In today’s increasingly complex organizations and challenging competitive
and regulatory environments, the concept of bounded agency is more relevant than
ever to the everyday reality of leadership. The project leaders in this study were
constrained at multiple levels. At the Network level, they faced fiscal constraints and
bureaucracy. At the hospital level, project leaders were hindered by resistance from
various groups (e.g., upper managers) and departments (e.g., the Emergency
Department). At the level of the project team, leadership turnover and attendance
problems created communication difficulties. And at the individual level, the project
leaders who were not appointed by top management, and who lacked certain skills
(e.g., in performance measurement, meeting management, and project management),
were generally less effective than those who were appointed by top managers and did
possess those skills.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
206
This list of difficulties may seem overwhelming, but the success of the
Patient Access changes, and the partial success of the Asthma Pathway project
suggest that such constraints are not always insurmountable. Administrators can
appoint (new) project leaders, management skills can be learned, conflicts can be
managed, bureaucracy can be navigated, and substitutes can be found for technology.
This is where creativity makes a difference in leadership. Resourceful leaders are
entrepreneurial problem-solvers; they enact resources that allow them to overcome
constraints that would have impeded other leaders. A good example of this was when
Alexandra (PA2) used her informal network to overcome a bureaucratic delay in
implementation. This dimension of leadership has received scant attention from
leadership scholars. It deserves closer consideration, as it could be an important
factor affecting the outcomes of change initiatives. A resource-enactment framework
offers a starting point for both diagnosing constraints and identifying instances of
creative enactment.
There is a long tradition of research on resources and organizations in
management studies, but few researchers have adopted a perspective that looks at
resource enactment. Some have discussed resources for sensemaking (i.e.,
knowledge; e.g., schema, scripts, and stories, as discussed in Weick, 1995), while
others have looked at resources as economic or material (i.e., allocative; e.g.,
budgetary) and have considered their relationship to authority. Writers have argued
that resources are created, transformed, fixed and circulated through human
interaction (Clegg, 1989; Giddens, 1984), but this study is one of the few that looks
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
207
at how different kinds of resources are enacted in practice, through leaders’
negotiation of systems and structures. By considering how communicators constitute
resources through their social interactions, we can start to identify where and why
some leaders experience problems (e.g., transferring knowledge), and can perhaps
develop models that may help leaders better manage their change efforts.
Scholars have called for research that specifies how organizational change
“really ‘happens’” (Lewis & Seibold, 1998). Lewis and Seibold have complained
that the literature has given little empirical evidence regarding what implementers
actually do, and that it generally ignores the role of lower level employees. Change
leaders in this study underestimated the need for communication with various groups
(e.g., to update members of initial project teams, to consult with upper and middle
managers regarding planning and implementation) at different times points in the
change process (e.g., during leadership transitions, at the end of phases). Although
guides do exist for managing communication on change projects (e.g., Dabbah,
1993; Project Management Institute, 1996), the lack of attention to communication
planning and conflict management at this hospital raises the question of how widely
such project management knowledge has diffused.
Little research has looked at the roles of leaders at different levels in enacting
planned change. This study helps shed light on project leaders’ interactions with top
management, as well as the roles of administrators and project leaders in each of the
project phases. In professional organizations like hospitals that have poorly
integrated silo structures, project leaders often have weak internal networks. In these
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
208
cases, the data in this study suggest that administrators may need to intervene and set
up structures to facilitate communication between the project leader and the local
leaders in the stakeholder units. This communicative function of top managers is
rarely discussed in the literature on change leadership, where supportive executives
are generally seen as “sponsors” who provide authority and allocative resources to
change “champions.”
Scholars have long recognized that leadership often involves conflict (Bass,
1990). Managers reportedly spend from a fifth to a quarter of their time dealing with
conflict (Thomas & Schmidt, 1976, cited in Putnam & Poole, 1987; Lippitt, 1983,
cited in McAfee & Champagne, 1987), and research has found that managers who
spend more time managing conflict are more successful (Luthans, Rosenkrantz, &
Hennessey, 1985, cited in Nicotera, Rodriguez, Hall, & Jackson, 1995). This study
confirms past research on leaders’ need for conflict management skills, including the
ability to diagnose conflicts and their causes, the ability to encourage and conduct
open communication about disagreements, and the ability to convert conflicts into
collaborative problem solving processes (Bass, 1990).
By showing how organizational power and politics interacted with leaders’
conflict management strategies and the ultimate resolution of the conflict, this
research extends the small but growing empirical literature on change leadership in
health care organizations (e.g., Cook, 1995; Denis et ai, 1996; Denis, Lamothe, &
Langley, 1999; Lammers et al., 1996; Weiner et al., 1996; Weiner et al., 1997), as
well as the older body of work on negotiated order in health care settings (e.g.,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
209
Barley, 1986; Barley, 1990; Geist, 1995; Geist & Hardesty, 1992; Schatzman &
Bucher, 1964; Strauss, 1978; Strauss, Schatzman, Ehrlich, Bucher, & Sabshin,
1963). Three conflict management styles were observed: avoidance, forcing, and
compromising. The dominant strategy seemed to be conflict avoidance. Intra-team
and extra-team conflicts were sometimes resolved by power shifts rather than active
intervention. Another common approach was forcing, “use of position power,
assertiveness, verbal dominance, and perseverance” (Putnam & Poole, 1987, p. 556).
Two leaders used forcing to try to improve meeting attendance, and many managers
and physicians believed it to be the best way to get employees to follow the new
specimen labeling policy. Compromising was used to resolve the conflict between
the Emergency Department and the Mislabeled Specimens team. Karen (MS2), who
was one of the least powerful leaders, enlisted the help of administrators and the
physician leadership in managing conflicts on her project team, and with others
outside the team. These conflicts illustrate how the project leaders managed the
interface between their teams and the larger organization, which is an area where
research has been lacking (Frey, 1994).
Communication researchers have looked at the use of different
communication channels for disseminating information to organization members,
and the channels used for soliciting their input for planned change (Lewis, 1999).
They have also considered the use of mission statements, internal newsletters, and
feedback mechanisms (Lewis, 2000b), as well as the role of everyday conversations
among lower-level employees (Fairhurst, 1993) in implementing quality
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
210
improvement programs. This study is one of the first to consider the role of meetings
in organizational change processes. Meeting characteristics (timing, length,
frequency, content), affected team member attendance, which in turn affected
information flow to the departments. Although all the project leaders ran their team’s
meetings, some leaders were better than others at managing those meetings. These
findings suggest that practitioners and researchers who seek to better manage or
understand planned changes should pay more attention to meetings.
The findings also suggest that a key function of leadership is to manage
employee involvement in change processes. Several authors have called for
communication researchers to examine how employee participation is realized
(Lewis & Seibold, 1998; Seibold & Shea, 2000). An important implication of this
study is that leaders play a critical role in managing the participation levels of
various stakeholders. Although most team com m unication took place in meetings,
leaders sometimes accommodated other individuals by using alternate
communication channels (e.g., email, phone, one-on-one meetings) in order to ensure
their input. For example, the ACP team consulted with the director of Education only
outside of team meetings, often via email, to coordinate the new asthma pathway
forms with the Education Department’s Continuity of Care forms. Some project
leaders had difficulty managing the communication from certain team members to
their constituent groups. Some representatives were frequently unreliable liaisons to
their home departments. Representation is a common issue in organizations
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
211
(especially on councils and committees), and the communication practices involved
in representing organizational constituencies deserve further study.
We still know little about the relative effectiveness of different change
communication strategies (Lewis, 2000b, p. 153). Proposed contingency models
have included situational characteristics such as the need for communication
efficiency and the need for consensus building (Lewis, Hamel, & Richardson, 2001),
or the volume and ambiguity of environmental information (Dafts, Bettenhausen, &
Tyler, 1993, cited in Eisenberg et al., 1999). This study raises two other innovation-
related considerations for change communication strategies: the mode of
institutionalization, and the degree of embeddedness in communities of practice.
Certain communication strategies may be better suited to organizational changes that
rely on episodic mechanisms of institutionalization (e.g., punishments or rewards),
whereas other strategies may be better suited to innovations that depend on structural
mechanisms (e.g., formal hierarchies or material technologies) (Lawrence et al.,
2001).
Different communication strategies may also be indicated for innovations that
affect only individual employees (e.g., handwashing, identification badges), as
compared to innovations that affect processes performed in communities of practice
(Lave & Wenger, 1991). QI efforts aim to change processes, which means that these
initiatives are usually complex in that they involve numerous and diverse members
of the organization (Leonard-Barton, 1988b). Embeddedness adds an additional layer
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
212
of complexity, since the norms and role relationships of one or more communities
are also involved.
Limitations
Every study suffers from limitations; this one is no exception. First, although
this study generated an abundance of data, some data that could have been useful
were not available. The study design involved eight site visits (totaling more than
240 hours) over the course of a year, so many of the events and activities involved in
the change projects could not be directly observed. Direct observation of team
meetings on all of the projects and of implementation-related activities (e.g., special
presentations to directors at division meetings, employee instruction on a new policy
during department meetings) would have helped triangulate and flesh out the
accounts of team members. The timing of the field research also limited my access to
data, because the projects had already transferred from the initial leaders to the
subsequent leaders. Ideally, I would have liked to interview all three of the initial
leaders, not just Madeleine, but that was not possible.
A second limitation concerns the problem of causality. This study was not a
controlled experiment, so there is greater uncertainty in making causal claims about
the effects of different variables on the outcomes of the change initiatives.
Furthermore, the complexity of the organizational change system, with so many
factors at different levels of analysis affecting the change processes, would suggest
that the outcomes are overdetermined. In mathematics, overdetermined systems are
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
213
algebraic systems that have more equations than unknowns. In such a system, how
can we know that leadership is making a difference?
First, it is clear from the data that leadership is a necessary, if insufficient,
element for changes like QI projects in organizations like hospitals: when leaderless,
these projects stalled completely. Second, the study design allowed for the collection
of data that strongly suggested the effects of leaders’ actions on the outcomes of
planned change. The study compared three different projects that represented three
different outcomes ranging from very successful to unsuccessful, which helped
identify which of the leaders’ actions were more or less effective. Previous research
has shown the importance of some of these activities (e.g., networking, correctly
using performance data) for successful change leadership. Also, although there was
no experimental control group, the longitudinal approach enabled observation of the
effects of leaders’ actions over days, weeks, and months. In sum, although one could
imagine other decisive factors for change initiatives (e.g., organizational crises), the
data and the literature make a strong case that leaders play a critical role in achieving
organizational changes.
Finally, there is the issue of generalizability. This study looked at three
projects within the same hospital, so how relevant are the findings for other
hospitals, or for other kinds of organizations? Case study research cannot and does
not aim to make the kinds of generalizations that are possible in nomothetic designs.
Furthermore, the level of analysis in this study is the project, not the organization. As
a comparative case study of three QI projects, the analyses determined patterns
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
214
regarding change leadership within the same organization. By specifying these
processes and practices, the study invites comparison with planned change processes
in other organizations. It is through systematic comparison of findings across many
studies (qualitative and quantitative) and that we will improve our theories of
organizational change. It is only when a body of research has been analyzed that we
can feel some confidence in our generalizations.
Implications for Practitioners
One of the most important messages of this research is that leaders need to
recognize how important communication is for their organizational change
initiatives. In hospitals, the centrality of communication for managing change is
often overlooked. Communication planning during the earliest stages is crucial, so
that leaders can determine key stakeholders, their level of involvement, and how they
will participate. This planning should be done by top management in consultation
with the project leader. The com m unication plan should be updated as the project
evolves. Some of my findings support Weick’s (1995) “perverse” advice that some
organization members should meet more often to engage in sensemaking together,
but they also imply that project meetings need to be managed effectively.
These three projects, for instance, might have been better served by different
levels of involvement of actors at various levels. Top management could have been
more involved more often, to facilitate com m unication with upper management and
to convey their support for the MS and ACP projects. Also, if the directors and
attending physicians had participated in the development and implementation of the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
215
specimen labeling policy and the asthma pathway, it probably would have facilitated
the acceptance and institutionalization of these innovations. Others (Fairhurst, 1993;
Huy, 2001a; Larkin & Larkin, 1994; Young & Post, 1993) have noted the
importance of involving middle managers and supervisors in change processes. This
study suggests that these individuals need to be actively involved in implementation,
revision, and institutionalization (in addition to analyzing the problem and designing
the solutions). Finally, it probably would have made sense to involve lower level
clinical employees more selectively, since attendance was most problematic for these
individuals.
The Medical Director at HFC was very interested in learning how to select
the appropriate leader for clinical QI efforts like the asthma pathway. This study
suggests several things to look for in a project leader, such as strong skills in
managing projects, meetings, and conflict, well-developed measurement and data
analysis skills, and robust interdepartmental and interdisciplinary social networks. It
also suggests that project leaders may need a m inim al level of authority to implement
change. For physicians, this authority may derive mainly from professional
reputation or expertise, whereas other employees may derive their authority from
their formal position in the organizational hierarchy. Top management should
carefully assess the candidates for the job of project leader. They should attempt to
choose a leader who has the necessary skills, authority, and social connections, but
any deficiencies or weak areas should be identified and supported through other
mechanisms. These mechanisms could include a larger role for top management, co
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
216
leadership, or establishing bodies to ensure involvement of key groups (as HFC tried
to do with the Oversight Council, though there did not appear to be follow-through).
Another important lesson is that project leadership transitions need to be
actively managed. All three of these projects experienced changes in leadership, and
given employee turnover rates in hospitals, this may not be so unusual. At a
minimum, administrators need to monitor change projects closely enough that they
can react in a timely manner if the project leader leaves. If top management is
committed to an initiative, it cannot allow a team to drift for months without a leader
like the Mislabeled Specimens project. Since administrators are responsible for
finding new project leaders, they are also the ones that should explain expectations
regarding communication with the project team, previous team members, and other
stakeholders.
Many of the employees at HFC complained that the QI projects took too
long, and they have not been alone in that complaint (1994; Bigelow & Arndt,
1995b; Blumenthal & Edwards, 1995; Koch & Fisher, 1998; Zbaracki, 1998). Some
quality experts have responded by developing techniques that allow for more rapid
QI projects (e.g., Berwick, 1998), and as my study was ending the Network was
about to introduce a new process model that represented a more efficient approach to
QI. The research at HFC suggests that change leaders who want faster results should
ensure that their projects have a strong, clear mandate from above, with a formally
appointed project leader who has the authority needed to implement change,
adequate support for data collection, and a deadline for deliverables.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
217
The project teams in this study were most productive and efficient when they
consisted of a core group of four to eight active members and held frequent, task-
focused meetings. The small groups literature indicates that although larger groups
have more human resources, they also experience more problems with coordination
and member motivation (Arrow, McGrath, & Berdahl, 2000; Davidson, 1995;
McAfee & Champagne, 1987). Research has shown that people are most comfortable
participating in discussion groups and decision making groups with three or five
members (Davidson, 1995; Nixon, 1979), and that the average number of
participants in organizational meetings is five (Barry, 1986, cited in Davidson,
1995). Early studies suggested that the ideal span of control is four to eight
subordinates for upper level supervisors and ten to fifteen employees for lower-level
supervisors (Eackles, Carmichael, & Sarchet, 1981, cited in McAfee & Champagne,
1987). Work groups larger than twelve persons tend to naturally break down into
smaller subgroups, and have a greater need for formal structures, such as leadership
(Arrow et al., 2000) or facilitating technology (Davidson, 1995).
This study suggests that leaders working on change implementation need to
think ahead and envision what will be needed to institutionalize their innovation over
the long term. Several of the leaders involved in these projects gave too little thought
to institutionalization. Leaders should tailor their implementation approach to the
nature of their innovation, as well as their audience. New behaviors are harder to
teach than straight factual information. New policies are highly equivocal (i.e., open
to misunderstanding) and therefore should be conveyed in the richest media (face-to-
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
218
face) and in multiple media (Daft & Lengel, 1986; Daft et al., 1987; Lengel & Daft,
1988). Again, that means having meetings where the changes can be discussed,
which enables sensemaking and aids retention of the information (Hernandez, 1998).
Discussion, diagrams, demonstrations, and other techniques may be needed to
convey the new practice. Leaders also need to consider what action cycles and
accountability structures are needed to sustain their innovation over the long term.
Organizational processes are situated within communities of practice (Lave &
Wenger, 1991); implementation strategies must take this social and political reality
into account. New behaviors will need to be normalized, so individuals with
authority (e.g., managers, attending physicians) will need to model the behaviors,
remind their staff, and reinforce practices with punishments and rewards. Top
management support is also needed throughout the institutionalization phase.
Directions for Future Research
Communication researchers need to devote more attention to leadership and
communication in organizational change projects. Because organizational
communication processes involved in change initiatives are often highly complex,
involving multiple actors, stakeholder groups, and media over extended time periods,
study designs must be appropriately sophisticated.
There is a need for longitudinal, in-depth studies of leaders’ discourse in
project meetings (with team members, and with other stakeholders), and during
implementation-related events (e.g., training sessions). Mini-ethnographies that
involve extensive observation of real-time interactions would help illuminate some
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
219
of the dynamics revealed in this study surrounding meeting attendance, stakeholder
representation, and conflict. Ideally, such studies would be multiple cases, involving
several projects in several organizations. It would be particularly useful if these cases
included network analyses of social relationships and communication patterns among
team members, administrators, and stakeholders in the change projects. Network
analysis could help explain information flow on the projects, and could uncover
evolving political configurations that affect the progress of change. It would also
allow for comparison of the project leaders’ networks, and how they leverage those
relationships in enacting resources for change at various stages of the change
process.
There are very few existing quantitative studies of change initiatives and their
leadership. A large-scale survey of a sizable population of organizations would allow
for a number of useful comparisons: among sectors, among industries, and among
different sizes of organization. Ideally, this survey would be administered by one or
more researchers on site, to ensure timely, accurate, and maximally complete
responses from all relevant individuals. Respondents should include the top
managers, the head of QI, the head of organizational development, the head of
training and education, all known project leaders, the team members from a
successfully completed project, the members of an unsuccessful project team, and
key representatives from stakeholder groups affected by the two projects. The
questionnaire should include items that follow up on the patterns observed at HFC
regarding job attributes, implementation practices, and attributes of the innovation.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
This study underscores Lewis’s (2000a) call for increased attention to
cooperation dynamics between change implementers and other stakeholders. In
particular, there needs to be more research into the patterns of interaction between
change leaders and managers at different organizational levels. What tactics do
change leaders use in communicating with directors, middle managers, and
supervisors about change? Does the source of the message (e.g., project leader, top
management) make a difference? Does the timing? What kinds of involvement
strategies work best for different groups? These questions are very applied, but they
might be addressed as part of the quantitative or qualitative designs discussed above.
For example, network research into change projects might suggest how to refine
project management techniques, including the selection of team representatives. It
would also be useful if researchers conducted action research studies to
systematically test some of the recommendations discussed in this chapter.
Future research could explore the theoretical framework of this study in other
organizations and settings. Resource enactment might serve as a departure point for
developing a theory of communication competence for leaders—especially for
change leaders. The categories used in this study are not comprehensive. There may
be other kinds of resources or additional systems and structures that are important for
change leadership. Future research should also examine how organizational and
professional cultures shape resource enactment.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
221
Conclusion
Health care continues to face fiscal austerity. In the first six months of
George W. Bush’s presidency, the U.S. federal government cut S35 million in
funding for advanced pediatric medical training, and enacted an 86% reduction in the
Community Access Program for public hospitals, clinics, and care providers for
people without insurance. Quality improvement is not an easy task in the face of
shrinking financial resources, but it is still a necessary one. In such a context, there is
a tendency for QI to shift away from focusing on customers or patients and to
become fixated on cost reduction. Thus, it may not be surprising that the most
expensive and most successful of the three projects in this study concerned a process
that directly affected the hospital’s revenue. Health care organizations need increased
revenues so they can improve their staffing levels and devote the necessary resources
to clinical quality. It will take leaders who are entrepreneurial and skillful
communicators to shift away from a mindset of cost-cutting and into a revenue
generating mode. They will have to dream up private sources of income, and to
convince policy makers and the public of the need to invest in health care quality.
I undertook this research because I have an abiding interest in organizational
change, and because I want to help hospitals to improve themselves more effectively.
I hope that this dissertation helps leaders think differently about resources for
change. I hope it helps them realize that change efforts require an investment in
communication structures and practices, because communication is important at
every step of the change process. I hope it inspires researchers to study leadership
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
222
and communication in organizational change initiatives. Finally, I hope that it
empowers my readers, so that you might think about new and creative ways of
enacting resources for your own initiatives.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
223
REFERENCES
Adams, K. I. (1997). Pragmatism and context: A matter of pattern. In J. L. Owen
(Ed.), Context and communication behavior (pp. 81-95). Reno, NV: Context
Press.
Adler, P. S., & Kwon, S.-W. (2002). Social capital: Prospects for a new concept. The
Academy o f Management Review, 27(1), 17-40.
Alemi, F., Allemango, S. A., Goldhagen, J., Ash, L., Finkelstein, B., Lavin, A.,
Butts, J., & Ghadiri, A. (1996). Computer reminders improve on-time
immunization rates. Medical Care, 54(10 (Supplement)), OS45-51.
Anacona, D. G., Okhuysen, G., & Perlow, L. A. (2001). Taking time to integrate
temporal research. Academy o f Management Review, 26(4), 512-529.
Arndt, M., & Bigelow, B. (1995). The implementation of total quality management
in hospitals: How good is the fit? Health Care Management Review, 20(3). 7-
14.
Arrow, H., McGrath, J. E., & Berdahl, J. L. (2000). Small groups as complex
systems: Formation, coordination, development, and adaptation. Thousand
Oaks, CA: Sage.
Arthur Andersen. (1999). National Hospital Quality Improvement Survey
(Knowledge Leadership Series, Issue 3). Chicago: Arthur Andersen.
Bandura, A. (1997). Self-efficacy: The exercise o f control. New York: W. H.
Freeman and Company.
Bantz, C. R. (1993). Understanding organizations: Interpreting organizational
communication cultures. Columbia, SC: U of South Carolina.
Barber, B. (1963). Some problems in the sociology of the professions. Daedalus, 92,
669-89.
Barge, J. K. (1994). Leadership: Communication skills fo r organizations and groups.
New York: S t Martin’s Press.
Barge, J. K. (1996). Leadership skills and the dialectics of leadership in group
decision making. In R. Y. Hirokawa & M. S. Poole (Eds.), Communication
and group decision making (2nd ed., pp. 301-342). Thousand Oaks, CA:
Sage.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
224
Barley, S. R. (1986). Technology as an occasion for structuring: Evidence from
observations of CT scanners and the social order of radiology departments.
Administrative Science Quarterly, 31, 78-108.
Barley, S. R. (1990). The alignment of technology and structure through roles and
networks. Administrative Science Quarterly, 35, 61-103.
Barsness, Z. I., Shortell, S. M., Gillies, R. R., Hughes, E. F. X., O’Brien, J. L., Bohr,
D., Izui, C., & Kralovec, P. (1993a). The quality march: National survey
profiles quality improvement activities. Hospitals & Health Networks,
67(23), 52,54-5.
Barsness, Z. I., Shortell, S. M., Gillies, R. R., Hughes, E. F. X., O’Brien, J. L., Bohr,
D., Izui, C., & Kralovec, P. (1993b). The quality march: Part two of a
national survey of quality improvement activities. Hospitals & Health
Networks, 67(24), 40-42.
Barsness, Z. I., Shortell, S. M., Gillies, R. R., Hughes, E. F. X., O’Brien, J. L., Bohr,
D., Izui, C., & Kralovec, P. (1994). The quality march: Part three of a
national survey of quality improvement activities. Hospitals & Health
Networks, 65(1), 45-48.
Bass, B. M. (1990). Bass & Stogdill’ s handbook o f leadership: Theory, research,
and managerial applications (3rd ed.). New York: The Free Press.
Bennis, W. G., Berkowitz, N., Affinito, M., & Malone, M. (1958). Authority, power,
and the ability to influence. Human Relations, 11, 143-155.
Berwick, D. M. (1989). Continuous improvement as an ideal in health care. The New
England Journal o f Medicine, 320(1), 53-56.
Berwick, D. M. (1998). Developing and testing changes in delivery of care. Annals
o f Internal Medicine, 128(B), 651-656.
Berwick, D. M., Godfrey, A. B., & Roessner, J. (1990). Curing health care: New
strategies fo r quality improvement. San Francisco: Jossey-Bass.
Bigelow, B., & Arndt, M. (1995a). Authors’ response to the commentaries. Health
Care Management Review, 20(4), 32-33.
Bigelow, B., & Amdt, M. (1995b). Total quality management: Field of dreams?
Health Care Management Review, 20(4), 15-25.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
225
Bikson, T. K., & Eveland, J. D. (1998). Sociotechnical reinvention: Implementation
dynamics and collaboration tools. Information, Communication & Society,
1(3), 270-290.
Blumenthal, D., & Edwards, J. N. (1995). Involving physicians in total quality
management: Results of a study. In D. Blumenthal & A. C. Scheck (Eds.),
Improving clinical practice: Total quality management and the physician (pp.
229-266). San Francisco: Jossey-Bass.
Blumenthal, D., & Kilo, C. M. (1998). A report card on continuous quality
improvement. The Milbank Quarterly, 76(4), 625-648.
Blumer, H. (1969). Symbolic interactionism: Perspective and method. Englewood
Cliffs, NJ: Prentice Hall.
Boerstler, H., Foster, R. W., O’Connor, E. J., O’Brien, J. L., Shortell, S. M., Caiman,
J. M., & Hughes, E. F. X. (1996). Implementation of total quality
management: conventional wisdom versus reality. Hospital & Health
Services Administration, 41(2), 143-159.
Branham, R. J., & Pearce, W. B. (1985). Between text and context: Toward a
rhetoric of contextual reconstruction. Quarterly Journal o f Speech, 71 ,19-36.
Brass, D. J. (1984). Being in the right place: A structural analysis of individual
influence in an organization. Administrative Science Quarterly, 2 9 ,518-539.
Brown, M. H., & Hosking, D. M. (1986). Distributed leadership and skilled
performance as successful organization in social movements. Human
Relations, 39(1), 65-79.
Biyman, A. (1996). Leadership in organizations. In S. R. Clegg, C. Hardy, & W. R.
Nord (Eds.), Handbook o f organization studies (pp. 276-292). Thousand
Oaks, CA: Sage.
Cabana, M. D., Ebel, B. E., Cooper-Patrick, L., Powe, N. R., Rubin, H. R., & Rand,
C. S. (2000). Barriers pediatricians face when using asthma practice
guidelines. Archives o f Pediatrics & Adolescent Medicine, 154,685-693.
Calomeni, C. A., Solberg, L. L, & Conn, S. A. (1999). Nurses on quality
improvement teams: How do they benefit? Journal o f Nursing Care Quality,
13(5), 75-90.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
226
Carman, J. M., Shortell, S. M., Foster, R. W., Hughes, E. F., Boerstler, H., O’Brien,
J. L., & O’Conner, E. J. (1996). Keys for successful implementation of total
quality management in hospitals. Health Care Management Review, 21(1),
48-60.
Clegg, S. R. (1989). Frameworks o f power. Newbury Park, CA: Sage.
Coleman, J., Katz, & Menzel. (1966). Medical innovation: A diffusion study.
Indianapolis: Bobbs-Merrill Co.
Conger, J. A. (1991). Inspiring others: The language of leadership. Academy o f
Management Executive, 5(1), 31-45.
Connolly, T. (1977). Information processing and decision making in organizations.
In B. M. Staw & G. R. Salancik (Eds.), New directions in organizational
behavior (pp. 205-234). Chicago: St. Clair Press.
Cook, J. V. (1995). Innovation and leadership in a mental health facility. The
American Journal o f Occupational Therapy, 49(7), 595-606.
Counte, M. A., Glandon, G. L., Oleske, D. M., & Hill, J. P. (1992). Total quality
management in a health care organization: How are employees affected?
Hospital & Health Services Administration, 37(4), 503-518.
Crane, D. (1972). Invisible colleges: Diffusion o f knowledge in scientific
communities. Chicago: University of Chicago Press.
Cronen, V. E., Pearce, W. B., & Harris, L. M. (1982). The Coordinated Management
of Meaning: A theory of communication. In F. E. X. Dance (Ed.), Human
communication theory: Comparative essays (pp. 61-89). New York: Harper
&Row.
Crosby, P. (1996). The absolutes o f leadership. San Diego: Pfeiffer and Company.
Crosby, P. B. (1979). Quality is free. New York: New American Library.
Crosby, P. B. (1984). Quality without tears: The art o f hassle-free management. New
York: McGraw-Hill.
Cyert, R. M. (1990). Defining leadership and explicating the process. Nonprofit
management and leadership, 7(1), 29-38.
Dabbah, R. (1993). Total project management: Strategies and tactics fo r the
healthcare industries. Buffalo Grove, IL: Interpharm Press.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
227
Daft, R. L., & Lengel, R. H. (1986). Organizational information requirements, media
richness and structural design. Management Science, 3 2 ,554-571.
Daft, R. L., Lengel, R. H., & Trevino, L. K. (1987). Message equivocality, media
selection, and manager performance: Implications for information systems.
Management Information Systems Quarterly, 11(3), 355-366.
Daft, R. L., & Weick, K. E. (1984). Toward a model of organizations as
interpretation systems. Academy o f Management Review, 9(2), 284-295.
Davenport, T. H., & Prusak, L. (1998). Working knowledge: How organizations
manage what they know. Boston, MA: Harvard Business School Press.
Davidson, R. (1995). Socio-psychological aspects o f group processes (hrformation
Systems Department Working Paper 95/03). Hong Kong: City University of
Hong Kong. Last retrieved November 17,2002 from
http://www.is.cityu.edu.hk/Research/WorkingPapers/paper/9503.pdf
DeLuca, J. R. (1984). Managing the socio-political context in planned change
efforts: Elements of a framework for practitioners. In A. Kakabadse & C.
Parker (Eds.), Power, politics, and organizations: A behavioural science
view. Chichester, UK: John Wiley & Sons.
Deming, W. E. (1982, 1986). Out o f the crisis. Cambridge, MA: Massachusetts
Institute of Technology.
Denis, J.-L., Lamothe, L., & Langley, A. (1999). The dynamics o f collective
leadership and strategic change in pluralistic organizations. Unpublished
manuscript.
Denis, J.-L., Langley, A., & Cazale, L. (1996). Leadership and strategic change
under ambiguity. Organization Studies, 17(4), 673-699.
Eisenberg, E. M., Andrews, L., Murphy, A., & Laine-Timmerman, L. (1999).
Transforming organizations through communication. In P. Salem (Ed.),
Organizational communication and change (pp. 125-147). Cresskill, NJ:
Hampton Press.
Ellwood, P. M. (1988). Outcomes management: A technology of patient experience.
The New England Journal o f Medicine, 375(23), 1549-1556.
Eubanks, P. (1992, June 5). The CEO experience: TQM/CQL Hospitals, 24-36.
Fairhurst, G. T. (1993). Echoes of the vision: When the rest of the organization talks
total quality. Management Communication Quarterly, 6(4), 331-371.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
228
Fairhurst, G. T., & Sarr, R. A. (1996). The art o f framing: Managing the language o f
leadership. San Francisco: Jossey-Bass.
Fairhurst, G. T., & Wendt, R. F. (1993). The gap in total quality: A commentary.
Management Communication Quarterly, 6(4), 441-451.
Ferris, T. G., Dougherty, D., Blumenthal, D., & Perrin, J. M. (2001). A report card
on quality improvement in children’s health care. Pediatrics, 707(1), 143-
155.
Foucault, M. (1976/1980). Two lectures. hi C. Gordon (Ed. & Trans.), L. Marshall,
J. Mepham, & K. Soper (Trans.), Power/knowledge: Selected interviews &
other writings, 1972-1977 (pp. 78-108). New York: Pantheon.
Frey, L. R. (Ed.). (1994). Group communication in context: Studies o f natural
groups. Hillsdale, NJ: Lawrence Erlbaum Associates.
Frost, P. J., & Egri, C. P. (1990a). Influence of political action on innovation: Part L
Leadership & Organization Development Journal, 77(1), 17-25.
Frost, P. J., & Egri, C. P. (1990b). Influence of political action on innovation: Part H
Leadership & Organization Development Journal, 77(2), 4-12.
Frost, P. J., & Egri, C. P. (1991). The political process of innovation. In B. M. Staw
(Ed.), Research in organizational behavior (Vol. 13, pp. 229-295).
Greenwich, CT: JAI Press.
Gann, M. J., & Restuccia, J. D. (1994). Total Quality Management in health care: A
view of current and potential research. Medical Care Review, 57(4), 467-500.
Geist, P. (1995). Negotiating whose order? Communicating to negotiate identities
and revise organizational structures. In A. M. Nicotera (Ed.), Conflict and
organizations: Communicative processes (pp. 45-64). Albany, NY: State
University of New York Press.
Geist, P., & Hardesty, M. (1992). Negotiating the crisis: DRGs and the
transformation o f hospitals. Hillsdale, NJ: Lawrence Erlbaum Associates.
Giddens, A. (1979). Central problems in social theory: Action, structure and
contradiction in social analysis. Berkeley: U of California P.
Giddens, A. (1984). The constitution ofsociety: Outline o f the theory o f
Structuration. Berkeley. University of California Press.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
229
Goldsmith, M. (1996). Ask, leaxn, follow up, and grow. In F. Hesselbein, M.
Goldsmith, & R. Beckhard (Eds.), The leader o f the future (pp. 227-237). San
Fransisco: Jossey-Bass.
Gouldner, A. W. (1957). Cosmopolitans and locals: Toward an analysis of latent
social roles. Administrative Science Quarterly, 2(3), 281-306.
Graves, J. R., & MacDowell, N. M. (1994). Mapping out the road to quality. The
Journal o f Long-Term Care Administration, 22(4), 12-17.
Grayson, M. A. (1992, June 5). Benchmark TQM survey tracks a new management
era in administration. Hospitals, 26-27.
Graz, B., Vader, J. P., Bumand, B., & Paccaud, F. (1996). Quality assurance in Swiss
university hospitals: A survey among clinical department heads. International
Journal fo r Quality in Health Care, 8(3), 271-277.
Greenfield, S. (1995). Continuous improvement and the physician: Building bridges
with outcomes research. In D. Blumenthal & A. C. Scheck (Eds.), Improving
clinical practice: Total quality management and the physician (pp. 51-71).
San Francisco: Jossey-Bass.
Grice, H. P. (1975). Logic and conversation. In P. Cole & J. L. Morgan (Eds.),
Syntax and semantics HI: Speech acts (pp. 41-58). New York: Academic
Press.
Hackman, J. R. (in press). Leading teams: Setting the stage fo r great performances.
Boston: Harvard Business School Press.
Hackman, J. R., & Wageman, R. (1995). Total Quality Management: Empirical,
conceptual, and practical issues. Administrative Science Quarterly, 40, 309-
342.
Harris, T. E. (1993). Applied organizational communication. Hillsdale, NJ:
Lawrence Erlbaum Associates.
Hernandez, J. B. (1998). Evaluating a multi-hospital quality improvement strategy to
implement clinical guidelines fo r radiographic contrast agents (Doctoral
dissertation, RAND Graduate School, Santa Monica, CA; NTIS No. PB99-
117277).
Homer, C. J., Kleinman, L. C., & Goldman, D. A. (1998). Improving the quality of
care for children in health systems. Health Services Research, 35(4 (Part II)),
1091-1109.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
230
Hosking, D.-M., & Morley, I. E. (1988). The skills of leadership. In J. G. Hunt, B. R.
Baliga, H. P. Schler, & C. A. Schriesheim (Eds.), Emerging leadership vistas
(pp. 88-106). Lexington, MA; Toronto: D. C. Heath and Company.
Hospitals/ServiceMaster survey tracks TQM/CQI. (1992, June 20). Hospitals, 42-50.
Howell, J.-P., Dorfman, P. W., & Kerr, S. (1986). Moderator variables in leadership
research. Academy o f Management Review, 11(1), 88-102.
Huberman, A. M., & Miles, M. B. (1994). Data management and analysis methods.
In N. K. Denzin & Y. Lincoln (Eds.), Handbook o f qualitative research (pp.
428-444). Thousand Oaks, CA: Sage.
Huff, A. S. (1988). Politics and argument as a means of coping with ambiguity and
change. In L. R. Pondy, R. J. Boland, Jr., & H. Thomas (Eds.), Managing
ambiguity and change (pp. 79-90). Chichester, UK; New York, USA: John
Wiley & Sons.
Hutchins, G. (1999, Nov.). Project management: The next big thing. Quality
Progress, 32, 118-119.
Huy, Q. N. (2001a). In praise of middle managers. Harvard Business Review, 79(8),
72-79.
Huy, Q. N. (2001b). Time, temporal capability, and planned change. Academy o f
Management Review, 26(4), 601-623.
Institute of Medicine. (2001). Crossing the quality chasm: A new health system fo r
the 21st century. Washington, DC: National Academy Press.
Janesick, V. J. (1994). The dance of qualitative research design: Metaphor,
methodolatry, and meaning. In N. K. Denzin & Y. Lincoln (Eds.), Handbook
o f qualitative research (pp. 209-219). Thousand Oaks, CA: Sage.
Jennings, K., & Westfall, F. (1994). A survey-based benchmarking approach for
health care using the Baldrige quality criteria. Joint Commission Journal on
Quality Improvement, 20,500-509.
Juran, J. M. (1989). Juran on leadership fo r quality: An executive handbook. New
York: The Free Press.
Kaluzny, A., McLaughlin, C. P., & Jaeger, B. J. (1993). TQM as a managerial
innovation: Research issues and implications. Health Services Management
Research, 6(2), 78-88.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Kaluzny, A. D. (1995). Quality improvement: Beyond the institution. Hospital &
Health Services Administration, 40(1), 172-188.
Kanter, R. M. (1983). The change masters: Innovation & entrepreneurship in the
American corporation. New York: Touchstone.
Kaushal, R., Bates, D., Landrigan, C., McKenna, K. J., Clapp, M. D., Federico, F., &
Goldmann, D. A. (2001). Medication errors and adverse drug events in
pediatric inpatients. Journal o f the American Medical Association, 285(16),
2114-2120.
Koch, J. V., & Fisher, J. L. (1998). Higher education and total quality management.
Total Quality Management, 9(8), 659-668.
Kotter, J. P. (1982). The general managers. New York: The Free Press.
Kotter, J. P. (1995). Leading change: Why transformation efforts fail. Harvard
Business Review, 73(2), 59-67.
Kotter, J. P. (1996). Leading change. Boston: Harvard Business School Press.
Rrackhardt, D. (1990). Assessing the political landscape: Structure, cognition, and
power in organizations. Administrative Science Quarterly, 3 5 ,342-369.
Kreps, G. (1988). The pervasive role of information in health and health care:
Implications for health communication policy, hi J. Anderson (Ed.),
Communication Yearbook (Vol. 11, pp. 238-276). Newbury Park, CA: Sage.
Laffel, G., & Blumenthal, D. (1989). The case for using industrial quality
management science in health care organizations. Journal o f the American
Medical Association, 262(20), 2869-2873.
Lammers, J. C. (1992). Work autonomy, organizational autonomy, and physicians’
job satisfaction. Current Research on Occupations and Professions, 7 ,157-
175.
Lammers, J. C., Cretin, S., Gilman, S., & Calingo, E. (1996). Total Quality
Management in hospitals: The contributions of commitment, quality councils,
teams, budgets, and training to perceived improvement at veterans health
administration hospitals. Medical Care, 34(5), 463-478.
Larkin, T. J., & Larkin, S. (1994). Communicating change: Winning employee
support fo r new business goals. New York: McGraw-Hill.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
23 2
Lave, J., & Wenger, E. (1991). Situated learning: Legitimate peripheral
participation. Cambridge England; New York: Cambridge University Press.
Lawler, E. E., & Mohrman, S. A. (1985). Quality circles after the fad. Harvard
Business Review, 63, 65-71.
Lawrence, T. B., Winn, M. I., & Jennings, P. D. (2001). The temporal dynamics of
institutionalization. Academy o f Management Review, 26(4), 624-644.
Lee, J. (1997). Leader-member exchange, the “Pelz Effect,” and cooperative
communication between group members. Management Communication
Quarterly, 11(2), 266-287.
Lekan-Routledge, D., Plamer, M. H., & Belyea, M. (1998). In their own words:
Nursing assistants’ perceptions of barriers to implementation of prompted
voiding in long-term care. The Gerontologist, 38(3), 370-378.
Lengel, R. H., & Daft, R. L. (1988). The selection of communication media as an
executive skill. Academy o f Management Executive, 11(3), 225-232.
Leonard-Barton, D. (1988a). Implementation as mutual adaptation of technology and
organization. Research Policy, 17(5), 251-267.
Leonard-Barton, D. (1988b). Implementation characteristics of organizational
innovations: Limits and opportunities for management strategies.
Communication Research, 15(5), 603-631.
Leonard-Barton, D. (1990). The intraorganizational environment: Point-to-point
versus diffusion. In F. Williams & D. Gibson (Eds.), Technology transfer: A
communication perspective (pp. 43-62). Newbury Park, CA: Sage.
Levin, D. Z., & Shortell, S. M. (1996). Desire to implement and ability to implement:
The case o f total quality management. Paper presented at the Academy of
Management, Cincinnati, OH.
Lewis, L. K. (1999). Disseminating information and soliciting input during planned
organizational change: Implementers’ targets, sources, and channels for
communicating. Management Communication Quarterly, 75(1), 43-75.
Lewis, L. K. (2000a). “Blindsided by that one” and “I saw that one coming:” The
relative anticipation and occurrence of communication problems and other
problems in implementers’ hindsight. Journal o f Applied Communication
Research, 28(1), 44-67.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
O '" * '■»
Lewis, L. K. (2000b). Communicating change: Four cases of quality programs. The
Journal o f Business Communication, 128-155.
Lewis, L. K., Hamel, S. A., & Richardson, B. K. (2001). Communicating change to
nonprofit stakeholders: Models and predictors of implementers’ approaches.
Management Communication Quarterly, 75(1), 5-41.
Lewis, L. K., & Seibold, D. R. (1998). Reconceptualizing organizational change
implementation as a communication problem: A review of literature and
research agenda, Communication Yearbook (Vol. 21, pp. 93-151). New
Brunswick, NJ: International Communication Association.
Li, L. X. (1997). Relationships between determinants of hospital quality
management and service quality performance—A path analytic model.
Omega, 25(5), 535-545.
Lievrouw, L. A. (1986). The communication network as interpretive environment:
"Sense-making’ ’ among biomedical research scientists. Unpublished doctoral
dissertation, University of Southern California, Los Angeles.
Martin, J., Felman, M. S., Hirsch, M. J., & Sitkin, S. B. (1983). The uniqueness
paradox in organizational stories. Administrative Science Quarterly, 28,438-
453.
McAfee, R. B., & Champagne, P. J. (1987). Organizational behavior: A manager’ s
view. St. Paul, MN: West Publishing Company.
McCall, M. W., Jr., & Bobko, P. (1990). Research methods in the service of
discovery. In M. Dunnette & L. Hough (Eds.), Handbook o f industrial and
organizational psychology (Vol. 1, pp. 381-418). Palo Alto, CA: Consulting
Psychologists Press.
McPhee, R. D., & Poole, M. S. (2000). Organizational structures and configurations.
In F. M. Jablin & L. L. Putnam (Eds.), The new handbook o f organizational
communication: Advances in theory, research, and methods (pp. 503-543).
Thousand Oaks, CA: Sage.
Miles, M. B., & Huberman, A. M. (1994). Qualitative data analysis: An expanded
sourcebook. Thousand Oaks, CA: Sage.
Miner, J. B. (1982). The uncertain future of the leadership concept: Revisions and
clarifications. Journal o f Applied Behavioral Science, 18(3), 293-307.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
234
Mitchell, T. R., & James, L. R. (2001). Building better theory. Time and the
specification of when things happen. Academy o f Management Review, 25(4),
530-547.
Monge, P. R., & Contractor, N. S. (2000). Emergence of communication networks.
In F. M. Jablin & L. L. Putnam (Eds.), The new handbook o f organizational
communication: Advances in theory, research, and methods (pp. 440-502).
Thousand Oaks, CA: Sage.
Monge, P. R., & Eisenberg, E. M. (1987). Emergent communication networks. In F.
M. Jablin, L. L. Putnam, K. H. Roberts, & L. W. Porter (Eds.), Handbook o f
organizational communication (pp. 304-42). Newbury Park, CA: Sage.
Morse, J. M. (1994). Designing funded qualitative research. In N. K. Denzin & Y.
Lincoln (Eds.), Handbook o f qualitative research (pp. 220-235). Thousand
Oaks, CA: Sage.
Motwani, J., Sower, V. E., & Brashier, L. W. (1996). Implementing TQM in the
health care sector. Health Care Management Review, 27(1), 73.
Mulley, A. G. (1995). Industrial quality management science and outcomes research:
Responses to unwanted variation in health outcomes and decisions. In D.
Blumenthal & A. C. Scheck (Eds.), Improving clinical practice: Total quality
management and the physician (pp. 73-107). San Francisco: Jossey-Bass.
Mumby, D. K. (2000). Power and politics, hi F. M. Jablin & L. L. Putnam (Eds.),
The new handbook o f organizational communication: Advances in theory,
research, and methods (pp. 585-623). Thousand Oaks, CA: Sage.
Ngwenyama, O. K., & Lee, A. S. (1997). Communication richness in electronic
mail: Critical social theory and the contextuality of meaning. MIS Quarterly,
27(2), 145-167.
Nicotera, A. M. (1995). Thinking about communication and conflict. In A. M.
Nicotera (Ed.), Conflict and organizations: Communicative processes (pp. 3-
15). Albany, NY: State University of New York Press.
Nicotera, A. M., Rodriguez, A. J., Hall, M., & Jackson, R. L., II. (1995). A history of
the study of communication and conflict. In A. M. Nicotera (Ed.), Conflict
and organizations: Communicative processes (pp. 17-41). Albany, NY: State
University of New York Press.
Nixon, H. L., II. (1979). The small group. Englewood Cliffs, NJ: Prentice-Hall.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
235
Northouse, P. G. (1997). Leadership: Theory and practice. Thousand Oaks, CA:
Sage.
Nutt, P. C. (1986). Tactics of implementation. Academy o f Management Journal,
29(2), 230-261.
Palmer, R. H., Hargraves, L., Orav, E. J., Wright, E. A., & Louis, T. A. (1996a).
Leadership for quality improvement in group practices. Medical Care, 34(9
(Supplement)), SS40-SS51.
Palmer, R. H., Louis, T. A., Peterson, H. F., Rothrock, J. K., Strain, R., & Wright, E.
A. (1996b). What makes quality assurance effective? Results from a
randomized, controlled trial in 16 primary care group practices. Medical
Care, 34(9 (Supplement)), SS29-SS39.
Parker, V., Wubbenhorst, W., Young, G., Desai, K., & Chams, M. (1999).
Implementing quality improvement in hospitals: The role of leadership and
culture. American Journal o f Medical Quality, 14(1), 64-69.
Pfeffer, J. (1977). Power and resource allocation in organizations. In B. M. Staw &
G. R. Salancik (Eds.), New directions in organizational behavior (pp. 235-
265). Chicago: St. Clair Press.
Pfeffer, J. (1978). The ambiguity of leadership. In M. W. McCall, Jr. & M. M.
Lombardo (Eds.), Leadership: Where else can we go? (pp. 13-34). Durham,
NC: Duke University Press.
Pondy, L. R. (1978). Leadership is a language game. In M. W. McCall, Jr. & M. M.
Lombardo (Eds.), Leadership: Where else can we go? (pp. 89-99). Durham,
NC: Duke University Press.
Powell, T. C. (1995). Total Quality Mangement as competitive advantage: A review
and empirical study. Strategic Management Journal, 1 6 ,15-37.
Project Management Institute. (1996). A guide to the project management body o f
knowledge. Newtown Square, PA: Project Management Institute.
Putnam, L. L., & Poole, M. S. (1987). Conflict and negotiation. In F. M. Jablin, L. L.
Putnam, K. H. Roberts, & L. W. Porter (Eds.), Handbook o f organizational
communication (pp. 549-599). Newbury Park, CA: Sage.
Reardon, K., & de Pillis, E. G. (1996). Multichannel leadership: Revisiting the false
dichotomy. In M. B. Salwen & D. W. Stacks (Eds.), An integrated approach
to communication theory and research (pp. 399-407). Mahwah, NJ:
Lawrence Erlbaum.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
236
Reichers, A. E., Wanous, J. P., & Austin, J. T. (1997). Understanding and managing
cynicism about organizational change. Academy o f Management Executive,
77(1), 48-59.
Rogers, E. M. (1995). Diffusion o f innovations (4th ed.). New York: The Free Press.
Ryan, S. R., Jean, M., & Capozzalo, G. (1992). Cultivating quality. Health Progress,
3 ,42-48.
Savitz, L. A., & Kaluzny, A. D. (2000). Assessing the implementation of clinical
process innovations: A cross-case comparison. Journal o f Healthcare
Management, 45(6), 366-380.
Schatzman, L., & Bucher, R. (1964). Negotiating a division of labor among
professionals in the state mental hospital. Psychiatry, 27(3), 266-277.
Schwartzman, H. B. (1989). The meeting: Gatherings in organizations and
communities. New York: Plenum Press.
Seibold, D. R., & Shea, B. C. (2000). Participation and decision making. In F. M.
Jablin & L. L. Putnam (Eds.), The new handbook o f organizational
communication: Advances in theory, research, and methods (pp. 664-703).
Thousand Oaks, CA: Sage.
Senge, P. (1992). The real message of the quality movement: Building learning
organizations. Journal fo r Quality and Participation, 15(2), 30-38.
Senge, P. (1996a). Leading learning organizations. Executive Excellence, 13(4), 10.
Senge, P. (1996b). Leading learning organizations. In F. Hesselbein, M. Goldsmith,
& R. Beckhard (Eds.), The leader o f the future (pp. 41-57). San Fransisco:
Jossey-Bass.
Senge, P. M. (1990). The leader’s new work: Building learning organizations. Sloan
Management Review, 32(1), 7-25.
Senge, P. M. (1997). Communities of leaders and learners. Harvard Business
Review, 75(5), 30-31.
Shortell, S. M., Bennett, C. L., & Byck, G. R. (1998). Assessing the impact of
continuous quality improvement on clinical practice: What will it take to
accelerate progress. The Milbank Quarterly, 76(4), 593-624.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
237
Shortell, S. M., Levin, D. Z., O’Brien, J. L., & Hughes, E. F. (1995a). Assessing the
evidence on CQI: Is the glass half empty or half full? Hospital & Health
Services Administration, 40(1), 4-24.
Shortell, S. M., O’Brien, J. L., Carman, J. M., Foster, R. W., Hughes, E. F. X.,
Boerstler, H., & O’Connor, E. J. (1995b). Assessing the impact of continuous
quality improvement/total quality management: Concept versus
implementation. Health Services Research, 30(2), 377.
Shortell, S. M., O’Brien, J. L., Carman, J. M., Foster, R. W., Hughes, E. F. X.,
Boerstler, H., O’Connor, E. J., & Gillies, R. R. (1995c). Physician
involvement in QI: Issues, challenges, and recommendations. In D.
Blumenthal & A. C. Scheck (Eds.), Improving clinical practice: Total quality
management and the physician (pp. 205-228). San Francisco: Jossey-Bass.
Signer, J. M. K. (2001). Performance improvement capability case study report:
Hospital B. Unpublished report, University of Southern California, Los
Angeles.
Simpson, K. N., McLaughlin, C. P., & Kaluzny, A. D. (1994). Planning, organizing,
and leading continuous improvement. In C. P. McLaughlin & A. D. Kaluzny
(Eds.), Continuous quality improvement in health care: Theory,
implementation, and applications (pp. 169-186). Gaithersburg, MD: Aspen
Publishers.
Smith, S. D. (1995). Commentary on “The implementation of total quality
management in hospitals: How good is the fit?” Health Care Management
Review, 20(4), 26-27.
Solberg, L. I., Brekke, M. L., Kottke, T. E., & Steel, R. P. (1998). Continuous quality
improvement in primary care: What’s happening? Medical Care, 36(5), 625-
635.
Solberg, L. I., Kottke, T. E., Brekke, M. L., Magnan, S., Davidson, G., Calomeni, C.
A., Conn, S. A., Amundson, G. M., & Nelson, A. F. (2000). Failure of a
continuous quality improvement intervention to increase the delivery of
preventive services: A randomized trial. Effective Clinical Practice, 3 ,105-
115.
Stake, R. E. (1994). Case studies. In N. K. Denzin & Y. Lincoln (Eds.), Handbook o f
qualitative research (pp. 236-247). Thousand Oaks, CA: Sage.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
238
Steel, R. P., Mento, A. J., Dilla, B. L., Ovalle, N. K., & Lloyd, R. F. (1985). Factors
influencing the success and failure of two quality circle programs. Journal o f
Management, 77(1), 99-119.
Stohl, C. (1995). Organizational communication: Connectedness in action.
Thousand Oaks, CA: Sage.
Strauss, A. (1978). Negotiations: Varieties, contexts, processes, and social order.
San Fransisco: Jossey-Bass.
Strauss, A., & Corbin, J. (1998). Basics o f qualitative research: Techniques and
procedures fo r developing grounded theory (2nd ed.). Thousand Oaks, CA:
Sage.
Strauss, A., Schatzman, L., Ehrlich, D., Bucher, R., & Sabshin, M. (1963). The
hospital and its negotiated order. In E. Freidson (Ed.), The hospital in modem
society (pp. 147-169). New York: Free Press of Glencoe.
Strauss, A. L. (1987). Qualitative analysis fo r social scientists. Cambridge, NY:
Cambridge University Press.
Thayer, L. (1988). Leadership/communication: A critical review and a modest
proposal. In G. M. Goldhaber & G. A. Bamett (Eds.), Handbook o f
organizational communication (pp. 231-263). Norwood, NJ: Ablex.
Thinking barriers to Total Quality Management (1994). The Journal o f Long-Term
Care Administration, 22(4), 14-15.
Thompson, J. W., Chesney, R. W., Stocks, R. M., Shmerling, J., & Herron, P.
(1999). Pediatric hospitals’ and physician strategies for the 21st century.
Clinical Pediatrics, 38(5), 259-263.
Waldman, D. A., Lituchy, T., Gopalakrishnan, M., Laframboise, K., Galperin, B., &
Kaltsounakis, Z. (1998). A qualitative analysis of leadership and quality
improvement. Leadership Quarterly, 9(2), 177-201.
Weech-Maldonado, R., Zinn, J. S., & Brannon, D. (1999). Managerial implications
of corporate board involvement and perceived market competition for quality
improvement in nursing homes. Journal o f Healthcare Management, 44(5),
382-396.
Weick, K. (1978). The spines of leaders. In M. W. McCall, Jr. & M. M. Lombardo
(Eds.), Leadership: Where else can we go? (pp. 37-61). Durham, NC: Duke
University Press.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
239
Weick, K. E. (1979). The social psychology o f organizing (2nd ed.). New York:
McGraw-Hill.
Weick, K. E. (1980). The management of eloquence. Executive, 5(3), 18-21.
Weick, K. E. (1995). Sensemaking in organizations. Thousand Oaks, CA: Sage.
Weick, K. W., & McDaniel, R. R., Jr. (1989). How professional organizations work:
Implications for school organization and management. In T. J. Serviovanni &
J. H. Moor (Eds.), Schooling fo r tomorrow: Directing reforms to issues that
count. Boston: Allyn and Bacon.
Weiner, B. J., Alexander, J. A., & Shortell, S. M. (1996). Leadership for quality
improvements in health care: Empirical evidence on hospital boards,
managers, and physicians. Medical Care Research and Review, 53(4), 397-
416.
Weiner, B. J., Shortell, S. M., & Alexander, J. (1997). Promoting clinical
involvement in hospital quality improvement efforts: The effects of top
management, board, and physician leadership. Health Services Research,
32(4), 491-510.
Westphal, J. D., Gulati, R., & Shortell, S. M. (1997). Customization or conformity?
An institutional and network perspective on the content and consequences of
TQM adoption. Administrative Science Quarterly, 4 2 ,366-394.
Yin, R. K. (1994). Case study research: Design and methods (2nd ed.) (Vol. 5).
Thousand Oaks, CA: Sage.
Young, M., & Post, J. E. (1993). Managing to communicate, communicating to
manage: How leading companies communicate with employees.
Organizational Dynamics, 22, 31-43.
Zbaracki, M. J. (1998). The rhetoric and reality of Total Quality Management.
Administrative Science Quarterly, 43, 602-636.
Zinn, J. S., Weech, R. J., & Brannon, D. (1998). Resource dependence and
institutional elements in nursing home TQM adoption. Health Services
Research, 33(2 Part I), 261-273.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
APPENDIX
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission.
Leaders’ Involvement throughout the Three Change Projects
Project
Phases
Projects
Patient Access Mislabeled Specimens Asthma Clinical Pathway
1. Problem
Awareness
& Definition
The Network’s top management
became aware of the loss of
revenue resulting from problems
with patient access.
The Risk Manager and the QI Dept, noted
the increase in incidents o f mislabeled
specimens. They reviewed the data and
could not locate a trend by department or
person or job class, which suggested that
it was a process problem. So they took it
to the QI committee (which included the
hospital’s top management) and asked
that a team be formed to look at the
specimen labeling process.
The Director of Nursing and tire Assistant
Director of Pharmacy each learned about
new techniques for disease management.
Together with the Director of Respiratorv
Therapy and an Education Department
employee, they started an initiative to
develop clinical pathways for several
conditions, including asthma. The
hospital’s top managers then ioined their
Clinical Pathway Steering Committee.
2. Project
Initiation &
Problem
Investigation
Hospital for Children’s ton
management appointed
Madeleine!, the Director of
Rehabilitation Services to lead
the project. Directors from
various departments sent
representatives to serve on the Q1
project team. Top management
gave the team 9 months to
research and prepare their
recommendations.
MFC’s top management appointed Jan e1 ,
the Director of Decision Support, to lead
the project. Directors from several
departments sent representatives to be on
the project team.
The team’s survey found widespread
ignorance of the current specimen
labeling policies, and ignorance of the
problem with mislabeled specimens. They
designed a new hospital-wide policy for
labeling specimens.
The Steering Committee had trouble
recruiting a leader to develop and
implement the asthma pathway.
Eventually, Dr. Song1 , an allergist, agreed
to lead the initiative.
The content o f the pathway was highly
contested both internally (among
physicians on the ACP team) and
externally (the Medical Director of the
Emergency Dept, rejected the concept of a
clinical pathway for asthma).
Key: underline = top m anagem ent; wavy underline = upper m anagem ent; bold = project leader; 1 = initial project leader;
2 = subsequent project leader
to
Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission.
L eaders’ Involvem ent throughout the T hree C hange P rojects (continued)
Project
Phases
Projects
Patient Access Mislabeled Specimens Asthma Clinical Pathway
3. Team
Recommen
dations
MiUlcldJJc! and selected team
members presented
recommendations to MFC’s top
management. The main
recommendation was to
centralize the patient access
function under one department,
headed by one director.
After the recommendations were
presented, the team never met
again.
Jan e1 and several team members
presented recommendations to MFC’s top
management. The main recommendation
was to implement their new, hospital-
wide policy for labeling specimens.
Sometime after the presentation, Jane left
the organization.
With the advent of a new Medical Director
in the ED. ED became more amenable to
working on the pathway, and (hey
developed the “front end” o f the path,
which the team then integrated with the
inpatient part of the pathway. The
document describing the pathway
standards was presented to the Clinical
Pathway Steering Committee, and was
given approval for piloting.
4. Implemen
tation
Hospital top management hired
A lexandra2 to be the Director of
Patient Access, per the CIP
team’s recommendations.
The new Director led the
implementation effort, working
one-on-one with managers in
other hospital departments, and
with people in several
departments at the Network level.
The project team chose KarenTrom
among their own ranks to be their new
leader. She had been an assistant director
in the Emergency Dept and was now
working in the QI Dept.
She submitted the new, hospital-wide
specimen-labeling policy to top
management for approval. Once it was
approved, Karen coordinated the
education efforts for the policy.
Dr. Song1 and her team attempted to
implement the asthma pathway, with little
effect. Few people at the hospital knew
about the pathway; it existed only “on
paper.” not in practice. The hospital’s
Medical Director nut the 01 Dept.’s
Clinical Resource Manager in charge of
the clinical pathways, and asked Dr.
Newman2 to lead a new asthma pathway
team for implementing it. Dr. Ncwnian
educated the residents on the pathway.
Key: underline = top m anagem ent; wavy underline = upper m anagem ent; bold = project leader; 1 = initial project leader;
2 = subsequent project leader
Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission.
Leaders’ Involvement throughout the Three Change Projects (continued)
Project
Phases
Projects
Patient Access Mislabeled Specimens Asthma Clinical Pathway
5. Monitoring A lexandra2 monitored the
effects of various changes
through reports obtained from the
Network. She processed these
data into reports that were
distributed to MFC’s top
management. When data
concerned specific departments,
the Director would discuss them
with tire relevant managers and
directors.
To monitor employees’ adherence to the
policy, K aren2 collected and processed
data from the Laboratory and the
departmental incident reports. She
reported on MS at the Medical 01
Committee fwhich included the hospital’s
Medical Director), and later at the
hospital’s top management Operations
meetings.
Although a list of items to monitor existed
under Dr. Song1 , it is unclear what data
were collected on the pathway, or who had
access to it.
Under Dr. Newman2, the QI department
collected both current and comparative
data on the pathway. These data were
occasionally presented to the Medical 01
Committee fwhich included the Medical
Director).
6. Revision
gatekeeper policy (i.e., no
admission of non-emergency
cases without insurance pre
certification) to accommodate
certain cases (e.g., those in need
of brain tumor operations). She
emphasized to her employees
that they were to ensure that
patients were not “caught in the
middle” of insurance-related
bureaucracy.
'ITeEjti5igency.Dep_attJm\tl§,Nursing
Director interpreted the policy in a wav
that ran counter to the QI team’s
intentions. Karen brought this issue to the
Medical OI Committee. After some
discussion among ED management and
the MS team, the policy was revised. The
final procedure for collecting specimens
represented a compromise between the
ED and the MS team’s positions.
Under Dr. Song1 , the pathway only existed
in abstract (“on paper”). Dr. Newman2
instantiated the pathway through
developing and instituting standardized
order forms. After implementation, she
talked informally with the residents and
incorporated their feedback into later
versions of the forms.
The units were not re-ordering the ACP
forms when they ran out. The Clinical
Resource Manager made the QI Dept,
responsible for keeping them in stock.
Key: underline = top m anagem ent; wavy underline ~ upper m anagem ent; bold = project leader; 1 = initial project leader;
2 = subsequent project leader
Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission.
L ead ers’ Involvem ent th roughout the T hree C hange P rojects (continued)
Project
Phases
Projects
Patient Access Mislabeled Specimens Asthma Clinical Pathway
7. Institution
alization
Dramatic improvements in the
rate of pre-certification, and the
rate of insurance claim denials
were achieved. But these gains
were threatened when top
management promoted
AlSMhUoi! to Administrative
Director of the Division, and the
directorship over Patient Access
was left vacant.
A year after implementation of the new
policy, no improvements in the rate of
mislabeled specimens had been achieved.
The K aren2 began working more closely
with top management to ensure
enforcement o f the policy. She also began
advocating technological solutions.
Regular training on the pathway was
instituted for each year’s incoming medical
residents. At the end o f this study, the team
was being reconstituted to address aspects
o f the ACP that had not been adequately
implemented. People on the team that
implemented the ACP wondered who
would be responsible for monitoring and
updating it over the long term.
Key: underline = top m anagem ent; wavy underline = upper m anagem ent; bold = project leader; 1 = initial project leader;
2 = subsequent project leader
244
INFORMATION TO USERS
This manuscript has been reproduced from the microfilm master. UMI films
the text directly from the original or copy submitted. Thus, some thesis and
dissertation copies are in typewriter face, while others may be from any type of
computer printer.
The quality of this reproduction is dependent upon the quality of the
copy submitted. Broken or indistinct print, colored or poor quality illustrations
and photographs, print bleedthrough, substandard margins, and improper
alignment can adversely affect reproduction.
In the unlikely event that the author did not send UMI a complete manuscript
and there are missing pages, these will be noted. Also, if unauthorized
copyright material had to be removed, a note will indicate the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by
sectioning the original, beginning at the upper left-hand comer and continuing
from left to right in equal sections with small overlaps.
ProQuest Information and Learning
300 North Zeeb Road, Ann Arbor, Ml 48106-1346 USA
800-521-0600
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UM I Number 3093975
Copyright 2002 by
Signer, Jordana Kanee
All rights reserved.
___ ®
UMI
UMI Microform 3093975
Copyright 2003 by ProQuest Information and Learning Company.
All rights reserved. This microform edition is protected against
unauthorized copying under Title 17, United States Code.
ProQuest Information and Learning Company
300 North Zeeb Road
P.O. Box 1346
Ann Arbor, Ml 48106-1346
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Strategic alliances and networks of financial relationships in Hollywood

PDF

Storytelling community: communication infrastructure and civic engagement in urban spaces

PDF

The globalization of mobilization: protest strategies of global justice movements, 1999-2001

PDF

Transforming supply-chains: infrastructures for collaborative knowledge creation

PDF

Lab coats in the dream factory: science and scientists in Hollywood

PDF

Community-based rehabilitation in coastal Oaxaca, Mexico: Projecto Piña Palmera

PDF

Understanding the shared situation awareness process: a communication framework for improved team performance

PDF

A typology of corporate environemental strategy and its driving factors in jultinational corporations

PDF

Loan portfolio information and accounting in banks

PDF

Performance management in health care in Iceland

PDF

Coevolution of technology and organization: A case study of BrokerHouse

PDF

Alternate models of women's health care policy in the United States

PDF

Control, health perceptions, and information seeking as correlates of risky behavior in the context of AIDS

PDF

Changing health-care delivery organizations in Costa Rica: an inter-institutional approach

PDF

Loose coupling in global teams: tracing the contours of cultural complexity

PDF

Evaluation of the Care Advocate Program: Bridging managed care and home community -based services

PDF

Satellite cell proliferation: two models and two anabolic stimuli

PDF

A study of employee health plan choice and medical cost: Panel data probit regression and sample selection model

PDF

Seniors in cyberspace: the determinants and consequences of empowered Internet use among senior citizens

PDF

Effects of a formulary expansion of the use of SSRIs and health care services by depressed patients in the California Medicaid program

Asset Metadata

Creator Signer, Jordana Kanee (author)

Core Title Leading change initiatives: communication and bounded agency in a health care organization

School Graduate School

Degree Doctor of Philosophy

Degree Program Communication

Degree Conferral Date 2002-12

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag business administration, management,health sciences, health care management,OAI-PMH Harvest,speech communication

Language English

Contributor Digitized by ProQuest (provenance)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-273060

Unique identifier UC11334652

Identifier 3093975.pdf (filename),usctheses-c16-273060 (legacy record id)

Legacy Identifier 3093975.pdf

Dmrecord 273060

Document Type Dissertation

Rights Signer, Jordana Kanee

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA