Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Implementation of unique device identification in the medical device industry: a survey of the change management experience
(USC Thesis Other)
Implementation of unique device identification in the medical device industry: a survey of the change management experience
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
IMPLEMENTATION OF UNIQUE DEVICE IDENTIFICATION IN THE
MEDICAL DEVICE INDUSTRY:
A SURVEY OF THE CHANGE MANAGEMENT EXPERIENCE
by
Aimee Greco
A Dissertation Presented to the
FACULTY OF THE USC SCHOOL OF PHARMACY
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF REGULATORY SCIENCE
May 2020
Copyright 2020 Aimee Greco
ii
DEDICATION
To all the past, present and future Regulatory Science students who were first curious
about Regulatory Science and then ended up making a career out of that wonder.
Don’t ever stop being curious.
iii
TABLE OF CONTENTS
LIST OF TABLES….. ....................................................................................................... vi
LIST OF FIGURES…….. ................................................................................................ vii
ABSTRACT……….. ......................................................................................................... ix
CHAPTER 1. Overview .............................................................................................. 1
1.1 Introduction .................................................................................................... 1
1.2 Statement of the Problem ............................................................................... 6
1.3 Purpose of the Study ...................................................................................... 8
1.4 Importance of the Study ................................................................................. 8
1.5 Limitation, Delimitations, Assumptions ........................................................ 9
1.6 Organization of Thesis ................................................................................. 11
CHAPTER 2. Literature Review .............................................................................. 12
2.1 Evolution of Unique Identification Coding.................................................. 13
2.1.1 Medical Device Amendments of 1976 ........................................... 14
2.1.2 The Safe Medical Devices Act ....................................................... 17
2.1.3 Device Tracking in the 21
st
Century ............................................... 19
2.2 FDA’s Unique Device Identification (UDI) Program .................................. 22
2.3 Implementing UDI change ........................................................................... 24
2.3.1 Tactical Requirements for UDI Labeling ....................................... 25
2.3.1.1 Developing a Unique Identification Number ..................... 25
2.3.1.2 Incorporating the Unique Identification Number into a
Central Database ............................................................... 28
2.3.2 Strategies for UDI-driven System Changes .................................... 29
2.3.3 Quality System Infrastructure in Companies .................................. 30
2.3.3.1 Design Control .................................................................. 32
2.3.3.2 Records/Documents/Change Control ................................ 34
2.3.3.3 Production, Process and Equipment Controls .................. 36
2.3.3.4 Materials Control .............................................................. 36
2.3.3.5 Post-production Oversight and Corrective Actions ........... 37
2.3.3.6 Facilities and Equipment Control ..................................... 38
2.3.4 Management Control ...................................................................... 40
2.3.5 Managing Change ........................................................................... 41
2.4 Theories of Change and its Corollary, Implementation ............................... 42
2.4.1 The Change Management Model of Lewin .................................... 43
iv
2.5 The Fixsen Model of Implementation .......................................................... 47
2.6 Studying Change .......................................................................................... 52
CHAPTER 3. Methodology ...................................................................................... 55
3.1 Overview ...................................................................................................... 55
3.2 Creation of the Survey Instrument ............................................................... 55
3.3 Focus Group and Validation of the Survey Instrument ............................... 56
3.4 Administration, Data Collection and Analysis............................................. 58
CHAPTER 4. Results................................................................................................ 60
4.1 Focus Group ................................................................................................. 60
4.2 Survey Instrument Distribution .................................................................... 61
4.3 Analysis Survey Respondents ...................................................................... 61
4.4 Views on Behavior and Culture Related to Change ..................................... 65
4.5 Possible Areas of Opportunity During Change ............................................ 68
4.6 Findings by Research Questions – Exploration Phase ................................. 72
4.7 Findings by Research Questions – Execution .............................................. 78
4.8 Findings by Research Questions – Implementation ..................................... 84
4.9 Findings by Research Questions – Sustainability ........................................ 85
4.10 Findings by Research Questions – Time Constraints .................................. 90
CHAPTER 5. Discussion .......................................................................................... 93
5.1 Introduction .................................................................................................. 93
5.2 Consideration of Methods ............................................................................ 93
5.2.1 Internal Validity .............................................................................. 94
5.2.2 External Validity............................................................................. 97
5.3 Culture and Change ...................................................................................... 99
5.4 Unfreezing: Exploration and Installation ................................................... 100
5.5 Changing (Execution and Implementation) ............................................... 104
5.5.1.1 Inadequate Process Workflows........................................ 104
5.5.1.2 Label Changes ................................................................. 105
5.5.1.3 Data Management and Computer Systems:..................... 106
5.5.1.4 Personnel Development and Capabilities ....................... 109
5.5.1.5 Management .................................................................... 112
5.5.1.6 Timelines .......................................................................... 113
5.5.1.7 Small Companies versus Large Companies .................... 114
5.6 Freezing: Sustainability and Maturity ........................................................ 115
5.7 Future Direction and Recommendations .................................................... 117
v
REFERENCES….. ......................................................................................................... 120
Appendix A. .................................................................................................................... 130
Appendix B. .................................................................................................................... 131
Appendix C. .................................................................................................................... 138
Appendix D. .................................................................................................................... 170
vi
LIST OF TABLES…..
Table 1: Focus Group Participants ................................................................................57
Table 2: Comments on the Utility of Sources in the Exploration Phase ......................75
Table 3: Comments on Challenges or Unforeseen Issues from the Early
Exploration to the Completion of Implementation of UDI. ...........................77
vii
LIST OF FIGURES……..
Figure 1: Trends of FDA Warning Letters in Response to QMS
Noncompliance .................................................................................................4
Figure 2: Example of National Drug Code ....................................................................22
Figure 3: Unique Device Identifier Device Label Bar Code Exemplar .........................27
Figure 4: UDI Data Package Framework Submission Method ......................................29
Figure 5: Application of Design Controls to Waterfall Design Process ........................33
Figure 6: Common Industry Packaging Practices ..........................................................40
Figure 7: Top 10 Barriers to Change Management ........................................................45
Figure 8: UDI Exploration Framework: Phase 1: Lewin’s ‘Unfreeze Stage” ................49
Figure 9: UDI Implementation Framework: Fixsen’s Implementation Phase ...............50
Figure 10: UDI Implementation Framework: Phase 3: Lewin’s Re-Freezing
Stage ...............................................................................................................52
Figure 11: Which Statement Best Describes Your Organization? ...................................62
Figure 12: Functional Roles Held by Respondents ..........................................................63
Figure 13: Job Levels Held by Respondents ....................................................................63
Figure 14: How is Your Organization Best Described? ...................................................64
Figure 15: What Class(es) of medical products does the organization sell for the
US market? .....................................................................................................64
Figure 16: Culture for Change: Using UDI as an Example, how well did their
organizations handle change? .........................................................................66
Figure 17: Based on your experience, how long would you estimate that it took
to complete the following stages of UDI implementation? ...........................68
Figure 18: Considering your experience with UDI, how well do the following
statements (considered as problems by others during change)
describe your initial efforts at implementation? .............................................70
viii
Figure 19: Considering your experience with UDI, how well do the following
statements (considered as problems by others during change)
describe your initial efforts at implementation? .............................................72
Figure 20: Exploration: Please rate the utility of sources when using the
following industry references to aid UDI implementation planning. .............74
Figure 21: Portion of the project was conducted by the consultant(s). ............................76
Figure 22: In order of importance please rank the following issues that were
prioritized when incorporating UDI into your organization. ..........................79
Figure 23: What were the top 3 areas in change management that you found
caused the most difficulty? .............................................................................80
Figure 24: How would you describe the time needed for the process compared
to initial expectations? ....................................................................................82
Figure 25: Overall, how was your experience when implementing UDI? .......................83
Figure 26: With UDI implemented, do you think your organization is ready for
an FDA inspection that includes UDI? ...........................................................83
Figure 27: I am confident that we made the correct changes to incorporate UDI
successfully.....................................................................................................84
Figure 28: Implementation Phase: Views on UDI Implementation .................................85
Figure 29: Sustainability: Please let us know your level of agreement with the
following statements. ......................................................................................87
Figure 30: Sustainability: In retrospect, which areas of UDI implementation
would you wish to improve? ..........................................................................90
Figure 31: Please estimate the time it took to implement UDI? ......................................91
Figure 32: Did the time to implement UDI: Exceed, equal, fall short of
expectations or did not have any idea how long UDI implementation
would take? .....................................................................................................92
ix
ABSTRACT………..
The commodities produced by the medical device industry are amongst the most
rigorously regulated on the market. Change to such products must be carried out in a
way that maintains regulatory compliance, protects productivity, assures cost-
effectiveness and most importantly, mitigates unanticipated risk to patients. However,
change can be difficult to study systematically because most product-related changes will
occur at different times and for different reasons. The present analysis took advantage of
the new requirements by the FDA for unique device identification, which had to be
introduced concurrently across all medical device companies, to understand how medical
device companies approached a complex change that involved multiple departments. A
survey tool, built on the implementation model of Lewin supplemented by that of Fixsen,
was directed at industry professionals with roles in medical device development. Results
suggest that this particular change has been demanding logistically. Overall, companies
appeared to struggle particularly with early exploration and installation phases of the
implementation, which often took longer than anticipated and was challenged by the
paucity of detailed guidance to navigate the later stages. Nonetheless, this unfreezing
stage seemed to lack the emotional overtones often associated with detachment from one
state to another in the literature. Most companies reported that strong leadership from
senior corporate leadership was not typical, but such leadership did not seem to handicap
the implementation of new requirements mandated by law. What did seem to be
important to improve the early planning stages were the availability of detailed support
materials and standardized practices for logistics and communication, including a
well-organized training program as early as possible.
1
CHAPTER 1. OVERVIEW
1.1 Introduction
Two basic rules of life are: 1) Change is inevitable.
2) Everybody resists change.
W. Edwards Deming
In companies that make medical devices, change is a critical part of commercial survival.
Products must change in order to compete in a global market whose worth has been
estimated to exceed $260 billion dollars (WHO, 2010).
New products, introduced to
meet the needs of patients, are soon modified to incorporate features that improve their
efficacy and safety. Change must also occur in the processes and systems that govern the
manufacture and sale of those products. For example, manufacturing engineers may
develop more efficient manufacturing processes or companies may move production
from one manufacturing site to another. Forces outside of the traditional sphere of
engineering design or manufacturing can also drive change such as a marketing team
choosing to “rebrand” a product with new packaging or a launch a new advertising
campaign.
In a highly-regulated industry such as the medical device industry, Change Management
is generally considered to be one of the greatest challenges when trying to ensure quality
systems and products. Ford, the famous automobile industrialist and founder of the Ford
Motor Company, was famously quoted as saying that “anything is possible given enough
time and money” (Ford, 2013). But change can be perilous in a regulated industry that
must follow elaborate systems and procedures for change management to ensure that the
2
changes do not introduce new risks or quality problems that could impact the safety and
efficacy of the medical device.
Change management implies a systematic approach, with a method and a plan, for
dealing with change both from the perspective of an organization and its affected
individuals. Thus, over the past 30 years, quality experts in academia, industry and
regulatory agencies have attempted to develop organized approaches that can systematize
and document the change process. An extensive literature review suggests that a robust
quality-system framework is needed to meet the challenges of what can be a complex
process, especially when the change has broad-ranging implications across several
company divisions or manufacturing sites. In a regulated environment, the complexity of
the process is amplified because the managers of change must understand the regulatory
requirements to capture and document change(s) to a medical device. All of this requires
expertise, resources, and time. If change management is not implemented correctly,
activities carried out to plan and implement a change may introduce risk to the product,
incur the unnecessary cost or result in regulatory non-compliance with attendant
monetary penalties and damage to corporate reputation. Because timelines can be critical,
success is often measured by how quickly and completely change can be implemented
with cost as one of the highest considerations.
In the United States (US), perhaps the most important starting point in change control for
a medical product are the regulations in the Food and Drug Administration (FDA)
3
Quality System Regulation (QSR) under 21 CFR Part 820.30, Medical Devices; Quality
System Regulation (FDA, 2015b).
Design changes. Each manufacturer shall establish and maintain
procedures for the identification, documentation, validation or
where appropriate verification, review, and approval of design
changes before their implementation (21CFR820.30).
A similar set of requirements is described in the European Union (EU) Medical Device
Directives (MDD), and also the International Standards Organization (ISO) 13485
standard, (Medical Devices – Quality Management Systems; Requirements for
Regulatory Purposes) that governs product that will be marketed in Europe (ISO, 2004).
However, the quality management systems (QMS) for both the US and EU are generic in
their nature. Thus, it is left to individual companies to fashion specific methods by which
they can make and communicate changes in their organization. The interpretation of
change requirements requires extensive expertise to implement the regulations and assess
conformance successfully. The fact that this process is not easy is reflected in the
increasing number of warning letters that have been issued by the FDA to companies in
which QMS methods have been found to be insufficient or badly applied (FDA, 2018).
Clearly, despite an abundance of guidance from regulators, consultants, and regulatory
experts, it remains difficult to maintain such quality systems.
4
Figure 1: Trends of FDA Warning Letters in Response to QMS Noncompliance
Note the increase since the FDA escalated the rate of industry inspections from 2009
onward (Emergo, 2016).
Reproduced with permission.
Regulatory agencies responsible for medical devices expect that product quality will not
be sacrificed on the altar of financial expediency. However, no patient is helped if a
medical device company becomes insolvent and can no longer provide those needed
products. The challenge to every company, then, is to effect changes in a way that not
only assures control over product design and manufacturing but also does it in a way that
respects operational budget constraints. Methods may vary depending on the size of the
enterprise, the previous experience of its regulatory and quality professionals, and the
culture and resources of the company. Large companies typically have more resources to
5
hire experts in quality methods and to implement changes without “breaking the bank”.
Yet, there are times when a small medical device company may be able to handle some
types of change with more success than a large company because fewer people with a
smaller geographic distribution are involved in communication chains and operations.
One particularly challenging set of changes currently occupies the attention of US
medical device companies is the new requirement to implement unique device
identification (UDI), a tracking system to ensure the traceability of medical products as
they are distributed in the marketplace. Such a program was seen to be needed to reduce
medical errors, improve supply chain management, facilitate adverse event reporting and
increase the capabilities to identify and reduce product counterfeiting. Some regulations
to require device tracking were introduced in August 1993 under 21 CFR part 821
“Medical Device Tracking Requirements”. The legal basis for such requirements,
updated on February 1998 under the Food and Drug Administration Regulatory
Modernization Act of 1997, Modernization Act, gave the FDA additional authority to
order manufacturers of certain types of Class II or Class III devices to initiate a program
that would track their medical devices down to the patient level (US-105th-Congress,
1998).
Subsequent regulations developed by the FDA set up a standardized process to which all
manufacturers, distributors and hospital organizations were required to comply within a
specified timeframe. Nevertheless, it was clear when UDI regulations were announced
that it would be difficult for companies to accomplish this change in a reasonable length
6
of time. For this reason, a staged implementation was recommended by the FDA so that
companies had a period of time in which to transition. As part of this staging, a set of
dates were identified as deadlines for the implementation of UDI changes across all
companies big and small. The fact that large groups of enterprises were required to
implement the same change requirements simultaneously offered an interesting
opportunity to study change management across a diversity of companies all
implementing the same change in lock-step.
1.2 Statement of the Problem
If change management is to be improved in medical device companies, it is important to
understand the biggest hurdles that stand in the way of effective change management and
make companies vulnerable to regulatory noncompliance and product problems. At the
same time, change management systems can be difficult to study. In part, this is because
a specific class of change may be carried out by individual companies at different times
and for different reasons. Thus, it is difficult to compare the response of one company
with one type of change to that of another company with either a different type of change
or a similar type of change made at a different time and place. However, the recent
requirements for UDI implementation within a preset period of time offered a unique
chance to study the change practices and problems of a variety of companies that all must
accomplish a similar task within a single epoch.
The opportunity to examine challenges faced by several different companies that must all
make a singular type of change could remove some of the confounding factors that exist
7
when different types of change are studied, or when the same type of change is studied at
different times when variations in the external environment may affect the results. At the
outset of the study, we knew that all of the medical device companies would implement
the UDI change because to do otherwise would prevent the sale of their marketed
products. What we did not know is the nature of vulnerabilities exposed by the stress of
such a multifaceted change. It was likely that a large company with many products could
face a more challenging task of coding and labeling dozens or possibly even hundreds of
product lines and models within a specific timeframe, whereas a small company with a
few products may have faced different problems. We do not know whether the biggest
pressure points experienced by large companies are shared with small companies.
To date, most of what has been written about UDI implementation has been anecdotal.
From trade journals and blogs, it was possible to glean some idea of different types of
hurdles that one or another author regarded as important. However, no systematic
approach had been used to capture the experience of a large number of companies. Thus,
we did not know how change management was typically accomplished in organizations
of different sizes and which hurdles in change management proved most vexing to them.
In the face of UDI requirements, we wished to understand better whether they were able
to affect the change efficiently using the change control processes that they had in place,
or whether they identified weaknesses. Identification and understanding of the
weaknesses might help to identify where extra resources are needed, training is
insufficient or regulatory hurdles are unnecessarily restrictive. Further, they might help to
8
recognize what companies do to remedy weaknesses that they identify during the
process.
1.3 Purpose of the Study
The purpose of the study was to gain insight into the challenges faced by the medical
device industry as it implements a significant and stressful change that tests their risk
management and change management systems. A review of the literature related to
change management and UDI requirements is first presented. The information was used
to guide the development of a novel survey tool that is framed by two frameworks, the
change management framework of Lewin expanded using the implementation framework
of Fixsen (Fixsen, Blase, Betz & Van Dyke, 2015). The survey was critically evaluated
by a focus group of professionals experienced in industry practices and/or survey
development prior to its dissemination to a variety of individuals responsible for UDI
change management in US-based medical device companies. An electronic survey
platform, Qualtrics, was be employed to deliver, collect and analyze the results.
1.4 Importance of the Study
This study provided insight into current and best industry practices with regard to the
implementation of change management. Its results should, therefore, be useful to a
range of stakeholders responsible for medical device development and
commercialization, including regulators who may wish to understand how a new
regulation has affected the industry. First, by identifying where some of the biggest
hurdles to implementation existed, it may be possible to help companies anticipate and
9
intervene to improve their change management controls. Even a small improvement
often has the positive effects of increasing productivity, maintaining compliance,
decreasing costs and improving communication. Results are also potentially helpful to
regulators. When policies are developed, those policies can experience resistance and
unexpected consequences related to the challenges of implementation. Knowledge of
areas of particular sensitivity might allow regulators and companies to work together to
adjust the requirements in a way that improves communication and compliance.
Further, the lessons learned during this example of change implementation and control
will potentially aid in the development of effective future regulations, by informing those
involved with regulatory activities about specific areas of vulnerability.
1.5 Limitation, Delimitations, Assumptions
This research analysis is delimited to the medical device industry, including not only
companies that make “stand-alone” medical devices but also pharmaceutical companies
that have medical devices or combination products in their portfolios. It does not restrict
its respondent pool only to those involved in the regulation of certain classes of products,
but demographic questions at the beginning of the survey will be used to stratify
companies with different product lines or areas of focus. It will direct its focus to US and
EU companies or subsidiaries of multinational companies that are have worldwide
operations that include the US, and thus are required to carry out FDA mandated UDI
implementation.
10
This study explored current practices through survey responses of medical device
professionals with experience in areas of Research and Development, Quality Assurance
and Regulatory Affairs. However, it excluded individuals who might also have a
tangential role in change management, in areas such as finance and sales.
Methodological procedures with respect to the change control requirements of QMS were
delimited to those relevant to US FDA 21 CFR 820 regulations, and International
Organization for Standardization (ISO) 13485: 2003 standards. Thus, it was assumed that
the medical products in the company’s portfolio were properly classified in an
appropriate regulated category, have gone through or going are going through the life
cycle development processes typical from discovery to market approval and had received
or are planning to pursue appropriate regulatory approval.
The survey was expected to face limitations associated with the availability and
cooperation of respondents who are in mid- to senior levels in medical products
companies. These professionals are in highly demanding technical fields that require
substantial active engagement, so they may have limited time and motivation to
participate in the survey. Further, the respondents sometimes may not feel confident to
provide accurate, honest answers that might present them or their companies in an
unfavorable manner, particularly if they view those answers as potentially disclosing
proprietary or sensitive information. The survey approach assumed that the respondents
have the appropriate experience and knowledge in the industry to provide adequate
feedback to provide relevant information. If a number of individuals failed to respond to
11
certain questions, bias could have been introduced into the cumulative answers to that
question.
The survey itself may also have been biased or insufficient. This could happen if the
research model selected to structure the survey is found in retrospect to be too narrow, or
if the guidance of the focus group proves to be insufficient. Additionally, this study
presupposed that the literature review would be sufficiently substantive to provide a
theoretical and experiential basis for an appropriate research instrument.
1.6 Organization of Thesis
The dissertation has five chapters. Chapter 1 introduces the reader to the rationale and
design of the research to follow. Chapter 2 examined the history of the regulations
pertinent to UDI and reviewed the tactical requirements to implement UDI operations
into a company’s processes. Additionally, stages of change defined theoretically are
related to the types of activities that are to be expected in a change of the magnitude of
UDI implementation. Chapter 3 outlined the methods that were used to conduct the
research and Chapter 4 examined the results of the survey, by looking not only at the
sampled respondents as a group but also by sub-stratifying subgroups of respondents to
look for differences in their views and concerns. Finally, Chapter 5 concludes with a
discussion and formal assessment of results, possible implications, and conclusions and is
presented with suggestions for further research.
12
CHAPTER 2. LITERATURE REVIEW
For more than a century, devices designed for medical use have been viewed as a special
product class, subject to numerous rules by government regulators in order to protect
public health. The layman’s definition of a Medical Device is a product used for medical
purposes in patients in order to diagnose or treat their medical problems. If applied to the
body, the effect of the medical device is primarily structural, in contrast to the actions of
pharmaceutical drugs that exert their actions through metabolic pathways (Merriam-
Webster Inc., 2005). The regulatory definition of a medical device from the US Food,
Drug and Cosmetic Act (FD&C) is more specific. There, a device is defined as:
…an instrument, apparatus, implement, machine, contrivance,
implant, in vitro reagent, or other similar or related article,
including a component part, or accessory which is:
recognized in the official National Formulary, or the
United States Pharmacopoeia, or any supplement to
them,
intended for use in the diagnosis of disease or other
conditions, or in the cure, mitigation, treatment, or
prevention of disease, in man or other animals, or
intended to affect the structure or any function of the
body of man or other animals, and which does not
achieve any of its primary intended purposes through
chemical action within or on the body of man or other
animals and which is not dependent upon being
metabolized for the achievement of any of its primary
intended purposes (FDA, 2014).
The medical device sector has grown hugely over the past 50 years, from a cottage
industry with simple products such as crutches and syringes to a global enterprise
currently estimated to be worth approximately US $302 billion in 2017 with a Compound
13
Annual Growth Rate (CAGR) of 6.1% during next six years (2011-2017) (Lucintel,
2012). The United States is the largest of these medical device markets, accounting for
about 38 percent of its total value, approximately $110 billion in 2012, and a projected
value of $133 billion by 2016 (SelectUSA, 2016). U.S. exports of medical devices in key
product categories identified by the Department of Commerce (DOC) exceeded
$44 billion in 2012, a more than 7 percent increase from the previous year.
As global populations grow and age, the medical-device sector is projected to become
even larger and more important to meet the needs for specific therapies. However, with
such growth comes the need for change in the way that the devices are designed and
produced. Those changes may include evolution in the processes and facilities used to
make such products and in the roles of individuals who execute the changes. In addition,
regulatory requirements are changing in ways that often require companies to alter their
approaches to managing products. The management of change is a key challenge for
medical device companies. This dissertation considers how medical device companies
are dealing with one such change associated with a new regulatory requirement, the
requirement to incorporate unique identification codes on all their regulated products.
2.1 Evolution of Unique Identification Coding
A century ago, medical devices had relatively few rules for labeling. The present system
of regulatory oversight stems from the Food Drug and Cosmetic Act (FD&C Act) passed
by Congress in 1938. However, that Act and its subsequent amendments through the first
14
half of the twentieth century were aimed primarily at managing the safety and efficacy of
drugs (Kleinfeld, Kaplan & Weitzman, 1976). Medical devices became regulated more
systematically under a more recent piece of legislation, the Medical Device Amendments
of 1976.
2.1.1 Medical Device Amendments of 1976
Regulators of the 1960s were faced with a changing environment as novel devices with
much more risk than had been typical previously, began to cause problems in the
marketplace. For example, when intrauterine devices were responsible for injuring at
least 900,000 women in the United States (Hicks, 1994), the Cooper Committee,
commissioned by Congress to deal with issues of device safety, recognized the need for
device regulations. It recommended a new regulatory framework that balanced the
rigorous review of high-risk devices with more modest requirements for most other
products in order to strike a balance between innovation and safety (Merrill, 1994). Their
recommendations presaged the classification tiers used to regulate medical devices today
and were a first step toward the passage of the game-changing Medical Device
Amendments of 1976 (O'Keefe, Spiegel & Murphy,1976).
The Medical Device Amendments and associated regulations detailed in 21 CFR 812
acknowledged the fact that medical products varied greatly in their design and functional
characteristics. They introduced a classification system that categorized each medical
device into one of three classes according to risk. The lowest risk, Class I medical
15
devices, would be subject to only modest registration requirements and filing fees so long
as certain basic manufacturing requirements are met. However, a submission to the FDA
would be required before a Class II or Class III device is marketed. Class II devices
typically would need an abbreviated “clearance” known as a “510(k)” pre-market
notification, whereas the highest risk, Class III products would require a more extensive
Pre-Market Approval Application (PMA). A full reporting of these differences is beyond
the scope of the dissertation here (see Rados, 2006 for a more detailed review).
However, they affect tangentially the way that device tracking and unique device
identification ultimately have been developed, as will be apparent later.
The Medical Device Amendments and its associated regulations not only classified
devices but also set in place requirements for commercialization. One of those
requirements was the need to identify and respond to adverse events particularly if those
adverse events would require a product recall. Approaches to adverse event reporting
immediately after the promulgation of the Medical Device Amendments were modeled
on the way that drugs were managed and focused on provisions to govern labeling and
misbranding (Merrill, 1994) without providing specific methods to accomplish medical
device surveillance (Committee on the Public Health Effectiveness, 2011). This approach
was soon found to be insufficient at preventing important public health challenges related
to device safety.
16
One concern that caused a return to regulatory policy setting was the problem of tracking
individuals with defective medical implants. The incidents attracting this regulatory
attention presented between 1969 and 1981, when patients with cardiac disease were
injured or died as the results of adverse effects associated with a particular type of
Björk-Shiley convexo-concave (BSCC) prosthetic heart valve (Wieting, 1996). This
particular type of valve caused patient injuries and deaths after they had been implanted
for weeks or months when devices fractured and failed as a result of metal fatigue in the
struts holding the mechanical valve in place (Blot et al., 2005). As fatalities increased
with time, it became apparent that patients with Björk-Shiley heart valves would need
appropriate monitoring. However, such monitoring was very difficult to carry out
because the nearly 55,000 patients who received the fragile devices could not be
differentiated from those with other safer valves. Although the FDA required the
manufacturer to monitor and track patients, an effective tracking system was not in place.
As a beginning measure to find a solution, in 1991, the heart valve manufacturer hired the
non-profit firm, MedicAlert (Berkman, 1991), to develop and maintain a database of
members' medical information that could be made available to medical authorities in the
event of an emergency. The now distinctive metal bracelet or necklace tag that carries
critical medical data for medical professionals has since become well-accepted by the
public to identify a variety of conditions in case of emergency. However, this effort was
only mildly successful for the type of problem presented by the Shiley heart valve,
because it relied on a voluntary registration process to which many patients did not
17
subscribe. Thus, many patients were not represented in an appropriate registry of
implanted products.
2.1.2 The Safe Medical Devices Act
The tracking challenges associated with the Björk-Shiley heart valve were a driving force
behind the passage of the Safe Medical Devices Act in 1990. Part of the Act was aimed
specifically at assuring that the FDA could be informed quickly about dangerous medical
products so that they could then notify recipients of the recalled product. To do this
effectively the FDA introduced 4 major changes that went beyond the requirements for
any other regulated product type, including pharmaceuticals (US-101-Congress, 1990).
First, high-risk products introduced after January 1
st
, 1991 had to be subject to an
FDA-approved protocol that outlined a specific plan for post-market surveillance.
Second, manufacturers were required to adopt methods such as the use of supply chain
software and lot nomenclature that would permit high-risk devices to be tracked. Third,
companies were required to report product corrections or removals to FDA when they
took an action such as a product recall to reduce health risks. Finally, hospitals, nursing
homes, ambulatory surgery centers and outpatient treatment facilities (but not doctors’
offices) were required to report within ten working days to FDA or the product’s
manufacturer whenever a device was suspected to cause or contribute to an adverse event
(FDA, 2015a). These changes later were expanded to base the FDA’s more systematized
Adverse Event Reporting System (AERS), which was designed to capture adverse events
as part of required post-market surveillance programs.
18
In 1993, the FDA took the additional step of establishing a new vigilance program called
the FDA’s Safety Information and Adverse Event Reporting Program, or more simply,
the MedWatch program. This program, further explained in a guidance document based
on the final Medical Device Reporting (MDR) rule and published in 1995, was designed
to expedite and broaden voluntary reports of serious AEs by HCPs and manufacturers
(FDA, 2009). It incorporated changes mandated by the Medical Device Amendments of
1992 and added the new tracking regulations (FDA, 1996). Ultimately, the main
accomplishment of the MedWatch program was to provide a database in which adverse
events could be recorded and monitored for signals, called “sentinel” events that might
indicate a novel problem with a medical product.
The effectiveness of the MedWatch system left much to be desired. Critics noted that
much of the reporting was voluntary and often incomplete. Further, the level of control
and oversight designated by the 1976 Amendments varied from one device Class to
another (Congress, 1976). Medical professionals were often confused about what should
be reported (Schremp, 1992). Sometimes, they found it difficult to ascertain if a product
was or was not a medical device and how to find the classification of a device (which is
often the role of a regulatory professional and not a clinician). This type of challenge can
be illustrated by examining orthodontic brackets. Traditionally, metal dental braces were
classified as Class I, low-risk medical devices with few requirements for adverse event
reporting. Health Care Practitioners (HCPs) thus were not accustomed to reporting
adverse events associated with such products. However, clear ceramic dental braces were
19
introduced later with enhanced regulatory expectations. Ceramic braces were placed into
a Class II category by FDA because the high-temperature glass press and grinding
process by which the braces were made was felt to make such braces vulnerable to
shattering when improperly removed at the end of an orthodontic procedure (Travess,
Roberts-Harry, & Sandy, 2004). Any problems with ceramic braces were, however,
underreported, presumably because HCPs did not realize that ceramic braces were in a
different class for which reporting was necessary (HHS-OIG, 2012). Further confusion
about the need to report adverse events became apparent because some types of facilities
were covered by the reporting requirements, but others were not. For example,
physicians’ offices were exempt from reporting adverse events, but surgery centers such
as those run by many ophthalmologists were not. Finally, the procedures to report
adverse events took time and incurred cost for practitioners (Berniker, 2001). For all of
these reasons, underreporting was common. It has been estimated that only about 10% of
adverse events were captured by the MedWatch system at that time (Kessler, Natanblut,
et al., 1993).
2.1.3 Device Tracking in the 21
st
Century
By 2000, a rigorous set of rules was established for drug and device vigilance
(FDA, 2015a). For devices, the well-established MedWatch reporting system required
covered entities, including manufacturers and healthcare facilities, to send unstructured
data regarding serious, unanticipated device adverse events (SUDAEs) to FDA within
10 days, if identified during a clinical trial, or 30 days, if the product was already on the
20
market. Manufacturers were also required to report performance failures of marketed
devices if those failures could have a serious effect on patient safety. The SUDAEs and
performance failures were listed in a publicly accessible database called MAUDE (an
acronym for Manufacturer and User Facility Device Experience). For class III devices
entering the market through a premarket approval route, an additional level of adverse
event reporting was required in annual reports that summarized a broader range of
adverse events that might not fall into the serious, unexpected category.
Nevertheless, many stakeholders continued to voice concerns about the quality and
comprehensiveness of data collected by retrospective device tracking systems. Three
issues, in particular, appear to have driven policy analysts to suggest further
improvements to the device vigilance system.
1) Expansion of registries: FDA required registries for certain high-risk devices, but little
similar tracking was carried out for Class I and II medical devices that made up about
90% of all medical devices (FDA, 2015a). A more extensive system of tracking was seen
to be required in order to find devices in lower classes if they were recalled. In a recent
report (Pew, 2014) designed to boost FDA’s effort in this area, several objectives were
outlined including establishing criteria to determine if a registry was the appropriate tool
for post-market surveillance and methods to streamline the collection of registry data in
ways that would reduce time and cost (Durkin, 2014). The proposed ideas about the
expanded use of registries foreshadowed a broader plan to harness Unique Device
21
Identifiers and adverse event reporting to strengthen post-market surveillance for medical
devices.
2) Identification of counterfeit products: The increased presence of contraband, that is,
goods that have been imported or exported illegally (Merriam-Webster, 2005), and
counterfeit products, that is, products made in exact imitation of something valuable or
important with the intention to deceive or defraud (Oxford, n.d.), added new concerns
regarding the ability to recognize and track unapproved products. A UDI system was
seen to have value in efforts to counter contraband activities so that a user could use the
identifier to distinguish a counterfeit device from one that was legally marketed (Ament,
2006).
3) Introduction of electronic submissions and structured datasets: The expansion of
technologically advanced electronic data capture provided tools to increase the content of
information in supply chains and patient records. Prescription drugs began to embrace
the use of electronic reporting with structured datasets that could rapidly populate
electronic databases such as those associated the FDA’s Sentinel Initiative or the Vaccine
Adverse Event Reporting System (VAERS) program, co-managed by the U.S. Centers
for Disease Control and Prevention (CDC) and the FDA. In the developing “Sentinel”
system, for example, FDA would be able to query new sources of adverse events from a
more diverse range of healthcare data holders, such as electronic health record systems in
hospitals, administrative and insurance claims databases, and registries. To use health
insurance claims or other electronic data sources, however, medical products would need
22
to be coded (Figure 2). Prescription drugs already had such codes, called NDC (National
Drug Code) codes (Rising & Moscovitch, 2014). However, no such codes existed for
medical devices.
Figure 2: Example of National Drug Code
(Source: Food and Drug Administration)
All of the above challenges could be mitigated if devices of different types could be
coded and tracked. Some system of unique device identification appeared to be needed
that could facilitate early detection and recall of defective devices, improve the analysis
of adverse events statistically and provide a pedigree system by which problems in the
sourcing and distribution chain could be traced.
2.2 FDA’s Unique Device Identification (UDI) Program
The perceived need for better tracking led Congress to require unique identification codes
(UDI) on medical devices as part of the Food and Drug Amendments Act in 2007
(FDA, 2013). Such a UDI system was seen to be important because it would give
companies and regulators the capabilities to detect problematic devices more efficiently
and facilitate device recalls (Ament, 2006). It would eventually provide a platform by
23
which customers or users could gain access to safety and efficacy information at the point
of use. It would also assist the FDA in data gathering and communication with other
health authorities (FDA, 2013).
However, the need to convert this law to more actionable regulations was challenging
because it entailed a major change in device management, not only for manufacturers
who would have to place the UDI codes on their products but also for distributors and
users who would have to adapt their systems to use the UDI information. For about
5 years, FDA held public workshops and requested comments from stakeholders about
how such a system should be constructed. These activities between industry stakeholders
and the FDA guided the development of a proposed rule in 2012. The proposed rule
instigated a new round of consultations and comment submissions, in this case to critique
the proposed regulations. Feedback was wide-ranging, from enthusiastic “think tanks”
that viewed the regulation as a “landmark step”, to more cautious industry stakeholders
who identified that “implementing it effectively would not be easy” (Hall, 2014). In
2013, the FDA issued its final regulations to establish this UDI system (Sullivaninstitute,
2014).
The new UDI legislation represented a substantial shift in the then-current paradigm of
medical device management and company operations. The new regulations imposed
dates on which UDI was expected to be implemented. Implementation deadlines for
Class I and Class II medical devices were identified to be September 2017, and
24
September 2016 respectively. The Class III device requirements had an implementation
deadline of September 2015. Thus, it created an unusual situation in which companies
with products in the US market were all required to make large changes within a similar
time frame. Such a synchronized set of “change” activities provides an opportunity to
study change management through the examination of a unique “intervention” across the
regulated medical device industry. The specifics of the rules are presented below as a first
step in understanding the nature of that change.
2.3 Implementing UDI change
UDI regulation is significant because it requires that firms develop and implement
methods to accomplish two main requirements, changing the labels on marketed products
to incorporate the UDI codes and uploading the data that interpret the assigned codes in a
centralized database. These requirements are outlined in section 2.6. As part of these
activities, the UDI framework must be inserted into a variety of company activities, such
as the documentation of the design control process, the management of labeling, the
modification of tracking methods and equipment, and the revision of adverse event
reporting methods. The implications of these changes are discussed below in
section 2.3.2.
25
2.3.1 Tactical Requirements for UDI Labeling
2.3.1.1 Developing a Unique Identification Number
UDI regulations require that medical device labels conform to specific identification
requirements that start with identifying a barcode for each product. This unique, 14-digit
product identification number encodes 65 specific attributes of the product (Figure 3). To
do this, data describing the product in “human readable text” are translated to UDI data
by using one of two bar code forms- either a GS1-128 linear bar code or a 2-dimensional
(2D) bar code (also known as a QR or “quick read” code). The code has two components,
a Device Identifier (DI) and a Production Identifier (PI). The DI encodes data to identify
the manufacturer, make, model, and globally unique product code for the finished
medical device. The PI encodes production-specific information whose nature will
change depending on attributes such as serial or lot numbers, expiration dates,
manufacturing dates and even data that could be unique for a singular device. For
example, the PI could code information that links a product to an intended patient who
may have specific requirements, such as a prosthetic implant with a certain size and
shape. Thus, every device variant will have its own number. If a product is modified by
altering such features as its packaging, quantity, or device version, the UDI information
must be modified as well. The PI data is required on all levels of packaging including the
outer package (also called the “shipper”) and the inner label on the package that houses
the medical device itself (known as the “primary”). Component suppliers are also
affected if they must mark certain parts of a product with the UDI, a requirement that is
26
often important for reusable devices. In some cases, a distributor or another business may
wish to purchase finished devices in order to re-sell them under a new name. This action
classifies them as a “re-labeler”. Under the new regulation, these organizations would
also be affected by UDI requirements.
In addition, a manufacturer that produces devices at different sites must also specify a
separate manufacturer identification known as a DUNS number as part of the UDI
dataset. A DUNS number is a nine-digit identification number unique to each physical
location of a business that is issued by the private entity, Dun & Bradstreet, to facilitate
commerce with banks, governments, and foreign entities. The DUNS number is
particularly important if a medical device company wishes to participate and bid on
government contracts and tenders, particularly in countries that have subsidized
healthcare programs.
27
Figure 3: Unique Device Identifier Device Label Bar Code Exemplar
(Thompson, 2014)
Reproduced with permission
Several bar code formats are theoretically available to the manufacturer, but FDA has
accredited three issuing agencies that sell blocks of UDI numbers to manufacturers at a
cost. These are the GS1, the Health Industry Business Communications Council
(HIBCC) and the International Council for Commonality in Blood Banking Automation,
Inc. (ICCBBA/ISBT-128). GS1 is an international non-profit organization that was
known until 2004 as European Article Numbering-Uniform Code Council or EAN.UCC
System. It created global standards for product barcodes across multiple sectors of global
commerce (Hamilton, 2010). The acronym, GS1, is now known as the numbering
standard. HIBCC is a similar nonprofit standards development organization but has a
narrower focus on barcoding for the healthcare industry. The third issuing agency,
ICCBBA/ISBT-128, sells codes that are specifically designed for the blood banking
28
industry under the current ISO 15459 blood banking standard specific to the laboratory
health market. Manufacturers work with one or more of these agencies to create the
globally unique, standardized UDIs for their products (BSI, 2013). Typically, they
purchase a block of numbers which are then used to assign a unique Global Trade Item
Number (GTIN) to individual product models, created as bar code sequences in the
format used by the issuing agency (GS1, 2013).
2.3.1.2 Incorporating the Unique Identification Number into a Central Database
Once a number has been assigned to a product, the manufacturer must upload associated
information about the device into the FDA’s Global Unique Device Identifier Database
(GUDID) that is maintained by the FDA’s Center for Devices and Radiological Health
(CDRH). In order to interact with the system, the manufacturer must first make an
official request to create an account with the FDA by providing their facility’s
identification. Once access to the database portal is granted, a UDI data package that
contains the collection of data describing the unique product can be “uploaded” into the
system.
The UDI data-package is the collection of data as a single master file to manage large
datasets, whose development and maintenance requires significant resource allocation.
Once the dataset is uploaded, the user must wait for that data to be validated, upon which
time it becomes part of a publicly searchable web-visible database, open indefinitely to
the regulators and the public. Figure 4 below illustrates an example of the workflow
29
process by a device manufacturer. Although the FDA calls the database a “global”
database, it is important to note that the database is specific to medical devices marketed
for the United States only.
Figure 4: UDI Data Package Framework Submission Method
(Source: Food and Drug Administration)
2.3.2 Strategies for UDI-driven System Changes
The implementation of UDI poses a multifaceted challenge for the production, software
and quality systems that govern the entire portfolio of products made by a company. The
changes are particularly difficult for companies that make class II and III products
because UDI must fit into a management strategy that includes design controls. Within
30
the design control framework, labeling would be an integral part of the device, subject to
same design control constraints as the device itself. The difficulty of assigning UDI codes
to all affected products grows with the size of a company’s product portfolio
(Lentz, 2015). To illustrate the complexity of UDI implementation, a short description of
some of the affected areas is given below. A fuller accounting of these challenges can be
found in various references elsewhere, from both governmental and trade publications
(McClellan et al., 2014; Weiker, 2014). What is more germane to this dissertation is how
this large and synchronized requirement for change across all US companies can be used
to explore the change process itself, especially for areas in which changes of this
magnitude are most likely to be problematic. Areas that are likely to be affected can be
dissected by looking at the quality system frameworks in which such change would be
embedded.
2.3.3 Quality System Infrastructure in Companies
All medical device companies have some type of quality system, built around a
systematic framework of policies and processes. For US-marketed medical devices, that
system must meet the requirements of the “Quality Systems Regulations” (QSRs)
described in 21CFR820 (FDA, 2018). The scope of the QSRs is further reflected in an
FDA quality management (QM) program (FDA, 2014), which is a five-system audit
framework that the FDA uses to guide its inspection of the quality processes governing a
company’s operations. These five areas include design controls, production and process
31
controls, facilities and equipment controls, records/documents/change controls, material
controls, and corrective and preventive actions.
In Europe, manufacturers adhere to a comparable quality system, whose requirements are
outlined in the standard, “ISO 13485:2003 – Medical devices - Quality management systems -
Requirements for regulatory purposes”. To satisfy this standard, the organization must
demonstrate that its medical devices and related services consistently meet customer and
applicable regulatory requirements (Myhrberg & Raciti, 2013). Key sections are often
described using a five-element framework that includes 1) QMS Requirements 2)
Management Responsibility, 3) Resource Management, 4) Product Realization, 5)
Quality System Evaluation such as measurement, analysis, and improvement.
The FDA regulations and the ISO standards cover similar areas of design and production,
and both emphasize lifecycle management and control of changes as key elements for the
success of a quality system. Implementation of UDIs will challenge these systems on
both the design and manufacturing fronts. Further, when changes involve modifications
in product design, design control requirements may have to be satisfied, depending on the
classification of the device. When the changes involve manufacturing, production
controls and often other areas, such as equipment, materials, and records, will be
affected.
32
2.3.3.1 Design Control
Design controls represent a method to systematize the design process to be sure that the
design of a device meets all the needs of its users. The system is typically described as
series of steps arranged as a “waterfall” (Figure 5), but the activities form a loop, in
which activities start by identifying user needs and end by assuring that those user needs
are met. The needs of the user provide the “inputs” of the system. As the design process
proceeds, the needs of the user are transformed into specifications that are then built into
the final product. The final product, or “design output” is reflected in a gradually fine-
tuned master file, called the Device Master Record (Loeb & Richmond, 2003). When a
change such as the requirement for UDI labeling is introduced, this change will be
characterized as a new “user need”. It will change the specifications for the product and
its labeling and will require a set of activities to verify and validate that the change has
met the new “user need” without compromising other aspects of the safe and effective
design (FDA, 2011).
33
Figure 5: Application of Design Controls to Waterfall Design Process
(Source: FDA, 1997)
A change in labeling will, therefore, require activities at many levels of the design control
process. Documentation to describe the new requirements must be added to the user
needs and design specifications. Personnel must then redesign the label to incorporate
the bar code and to assure that the correlated data has been uploaded to the GUDID.
They must also verify the integrity of this change, and then update the design output file
and the corresponding device master record. These decisions will ultimately be
transferred as changes to the production areas to alter the “build” of the product. Thus,
activities at the design stage must provide a solid foundation that can be integrated with
tracking methods used in manufacturing and distribution, as described below.
Subsequent changes to the design of the product, such as the introduction of a new
34
variant or model, will also require a change to the UDI code and quite possibly a return to
the design history waterfall.
2.3.3.2 Records/Documents/Change Control
In a regulated environment, documentation underpins all aspects of design and
manufacturing. Thus, the FDA demands a system in which many significant documents
are “controlled”, by assuring that those documents are approved, stored and distributed
appropriately. It further demands that:
Changes to documents shall be reviewed and approved by an
individual(s) in the same function or organization that performed
the original review and approval, unless specifically designated
otherwise. Approved changes shall be communicated to the
appropriate personnel in a timely manner. Each manufacturer
shall maintain records of changes to documents.
(Source: FDA, 2015b)
Maintaining a compliant documentation system is challenged at various levels, from
design and production to distribution and post-market management, when multiple parts
of the system must be adjusted to include the UDI code on relevant documentation. For
example, travelers that accompany the device as it passes through different
manufacturing stations may have to be adjusted to include this new information.
Distribution forms and electronic spreadsheets will also need to be adapted to add new
fields. Specifically:
35
Documentation (Records) Requirements of the QSR. Subpart M—
Records, has five sections. The general section requires that records
be maintained at a location accessible to responsible officials, made
available for review and copying by FDA inspectors, be legible, etc.
The FDA also can require an employee with executive responsibility
to certify that management reviews and quality audits were
performed and formally documented – and required corrective
action has been undertaken.
The Device Master Record: [Contains the information needed to
replicate the device.]
Device History Record: [Contains information related to
manufacture and acceptance of each device, by serial number.]
Quality System Record: [The sets of records not specific to
particular types of device(s), while conforming to provisions for
Document Controls.]
(Source: FDA, 2006)
Additional documentation will be required to describe policies and procedures related to
the incorporation of UDI. Rules must be developed to govern how various essential
elements required for the UDI code are captured on a metadata level, what software or
documentation process will be used to house the data, and what processes will be
employed when changes are needed to the UDI code or registration. New software
systems may need to be installed and integrated with other parts of the secure electronic
systems that are typically unique to the company. The UDI infrastructure must ensure
that any software systems used to manage the change meet specific technological
requirements to assure that electronic data are secure from unintentional corruption and
software hacking and are managed in compliance with the rigorous requirements
demanded to satisfy 21 CFR Part 11 regulations. Personnel must be trained on the
36
modified requirements and systems. These activities will require significant resource
allocation.
2.3.3.3 Production, Process and Equipment Controls
The need to identify UDI codes on final products and packaging has significant
implications for manufacturing. Expense and planning will be associated with the
installation of UDI into automated processes and tracking systems. New printing or
embossing machines to apply the bar codes may be required. Alternatively, if such
marking is possible with current equipment, software may need to be changed. Further,
barcode readers may have to be installed at critical points along the production path, so
that tracking can occur more efficiently.
2.3.3.4 Materials Control
For many devices, product components are obtained from vendors. Some of these
components are critical to the function of the device, and some represent advanced
subassemblies that form the core of the operation of the device. For many products, then,
barcoding will first be implemented as early as the stage at which materials are sourced,
so that the UDI can be associated with incoming materials that will be put into specific
products. Further, not all products are assembled in one location or by one manufacturer.
If parts of a device are made in different locations or by different vendors, the company
must interact with those vendors to assure that the UDI is managed appropriately as the
37
product moves from the original equipment manufacturer to the holder of the market
registration.
2.3.3.5 Post-production Oversight and Corrective Actions
An ongoing challenge experienced by medical device manufacturers and regulators is
that of knowing where devices go after they leave the manufacturing plant. Once the
products leave the loading dock, they become more difficult to track. The UDI is
anticipated to play a central role in improving the ongoing oversight of distribution as
products move from the manufacturer to the distributor, hospital or clinic, and ultimately
to the patient. However, device tracking will only succeed if recipients of the product
can use the UDI information effectively. The end-user therefore must have the
equipment necessary to read the barcodes, and to capture the data on their paper and
electronic records, so that the UDI becomes part of the permanent history of the patient
under sections 520(f) (21 U.S.C. 360j(f)) and 701(a) of the FD&C Act. As end-users
begin to use the tracking capabilities of UDI, companies must respond by redesigning
their complaint handling systems and adverse event reporting methods to incorporate the
UDI tracking information.
Another challenge in post-production is the oversight and maintenance of several coding
systems. A large company may need to use several coding systems such as a GS1 for
their medical device portfolio but then may need to implement ICCBBA bar codes if they
also distribute administration kits for In Vitro diagnostic testing and thus, double their
production maintenance and QC processes.
38
2.3.3.6 Facilities and Equipment Control
UDI implementation affects equipment management because equipment will be needed
to affix the bar codes on devices, labels, and records. Many companies must purchase
new equipment if their systems cannot handle UDI. Even the FDA recognized that this
might impose a financial burden- “with UDI costs to be as much as 1%+ of annual sales
and …the final rule may have a significant economic impact on a substantial number of
small entities that label medical devices” (FDA, 2013). That new equipment must be
integrated into the production workflow and must be validated, secure and reliable.
The re-design of facilities and equipment control can be time consuming and costly. For
example, durable medical devices, which are devices that are used more than once, may
need to be etched with the UDI code. In a study charted by the Eastern Research Group
(ERG) for the FDA, ERG estimates that some medical device labelers will bear costs
greater than 1% of revenues because they will need to buy costly laser direct marking
equipment (Barlow, 2013).
If the initial high-level approach to the equipment issues is not well planned, the change
can disrupt the production of devices. Thought must also be given to the needs for future
expansion and flexibility if the equipment must be adapted to production change or be
integrated with changing software that might later be introduced to modify or automate
manufacturing or tracking processes.
39
Some of the challenges when considering both the equipment and documentation needs
of UDI marking are not obvious to those without considerable experience in device
labeling and management. Device packaging is complicated by the way that product is
sold. For example, a “unit of sale” in some packaging cases, may need to be identified.
However, a “unit of use” and what is considered a “case” of identical product or a
multiple unit package to a distributor may require a different level of UDI identification
(Engler, 2016). This is additionally complicated by who (such as a distributor, for
example) may open the box to sell individual items to different customers since
distributors are also required to place their own UDI if they engage in labeling (i.e.
private labeling). Further, requirements may become complicated when devices and other
items are configured into kits, whose elements may be sold individually as well. The kit
may not only require a GTIN numbers for each item, but also a master GTIN number. As
part of change management, UDI assignments must be integrated with the product
scheme prior to product development.
40
Figure 6: Common Industry Packaging Practices
(Source: GSI, 2013)
2.3.4 Management Control
The UDI changes outlined above are intrusive across many activities that a company
must perform. They cannot be carried out without careful consideration of the
organization’s structure, ways of assigning accountabilities and allocation of resources.
In order for change to be successful, most who have written about such change identify
that senior management must champion the change and communicate that commitment
(Heathfield, 2016). Management must be involved in deciding if the current policies and
procedures are able to absorb the UDI program effectively, and if not, in determining
how to remediate the shortcomings. The exact extent and types of activities that would
result in an effective, persistent integration sufficient for compliance are, at best, vaguely
41
stated, so that companies must have flexibility to integrate the UDI in ways that best fit
the organization. Management will need to be educated about the needs of UDI
incorporation because they must allocate the money and personnel to implement the
changes operationally under Subpart B--Quality System Requirements, 21 CFR 820.20
Management Responsibility. Further, implementation of the new UDI system will take
time, so that decisions need to be made about the extent to which expertise is outsourced
or developed in-house so that the company can meet relevant deadlines. Regardless of
the chosen approach, delays can be encountered- training can take time, equipment must
be evaluated and installed, and software solutions and coding need to integrate the UDI
into multiple layers of documentation. However, without clear guidance, management
may find it hard to estimate the cost and human capital needed to affect the changes in
time to satisfy the imposed deadlines for implementation. The extent of these activities
often is not appreciated fully, forcing the company to adopt suboptimum or poorly
considered approaches to meet deadlines.
2.3.5 Managing Change
When an organization makes a change, the success of that change is well-known to be
affected by the unique culture, structure, and systems of processes that characterize the
organization. Thus, an extensive literature has centered on change in companies of
different sizes and types, most of which are beyond the scope of the work here (see, for
example, McCalman, Paton, & Siebert, 2016). Such literature typically draws attention to
the difficulty of change management and its complexity when required in a highly
42
regulated industry that is prone to resist change. Notably, a recent study has found that
only 25 percent of change management initiatives are successful over the long term
(Towers Watson, 2013), while a global study by IBM noted that nearly 60 percent of
projects aimed at achieving business change do not fully meet their objectives (IBM,
2008) impacting return on investment (ROI). Both studies noted that some of the
predominant traits such as the lack of management initiative and how managers were
trained to manage the change contributed to the lack of effective change. Further, the
change to a subsystem often also requires a change in the culture to sustain the
appropriate level of quality and avoid returning to old habits (Brandt, 2016).
In the case of medical device companies, UDI labeling is, however, a change that cannot
afford to fail. When a change must be made in a short space of time to comply with the
law, that change becomes riskier and costlier. Failure to implement UDI by the required
deadlines can, for example, result in delayed product shipments, as well as fines and
other legal actions related to product misbranding.
2.4 Theories of Change and its Corollary, Implementation
Change management models have been studied for many decades but are still relevant for
the twenty-first century (Pryor, Taneja, Humphreys, Anderson & Singleton., 2008).
Perhaps the best known of the many change models that have been created over the last
century is that of Kurt Lewin (1890-1947), a German physicist and social scientist whose
seminal work laid the foundation of change management theory.
43
2.4.1 The Change Management Model of Lewin
If you want to truly understand something, try to change it.” - Kurt Lewin
The change management model of Lewin likens the process of change to that of
modifying the shape of a block of ice. He identifies three stages of change: unfreezing,
changing, and freezing (Lewin, 1943). Each of these stages has a number of associated
activities. The first, “unfreezing” stage is regarded as a preparatory stage in which
activities are carried out to help the organization become receptive to change by
“dismantling” the established operative norm. The second stage is the stage in which new
policies and processes are developed and implemented. The final stage is that of
“freezing”, during which the changed system is embedded as a permanent part of the
organization. Attention to long-term sustainability is often needed to prevent later failures
after new objectives distract attention and resources from the changed system.
Each of the three stages of change has its own challenges. Lewin considers that the
dynamics of change at each stage occur in a “force field” in which some forces either
drive the change toward its goal or hinder the change from achieving its goal
(Lewin, 1951) (Swanson & Creed, 2014). In the “unfreezing” stage, individuals are often
predisposed to resist change (Dalessio, 1996) and may exhibit anxiety, reluctance, and
fear (George, 2016). Change is troubling for employees; they must cope with uncertainty
and stress associated with the changing work context and increased work demands
(Bordia, Hobman, Jones, Gallois, & Callan, 2004). New initiatives are met with concerns
44
related to insufficient resources, excessive costs, or other aspects of implementation that
they worry will limit success or change the working environment. Thus, planners must
recognize that resistance must be anticipated, and employees must be supplied with new
information, new behavioral models or new ways of looking at a situation (Sims, 2010)
that can either align the employees with the desired goal or can allow behaviors to oppose
the desired change. The first challenge of change, then, is overcoming psychological
resistance (Torben, 2011) in order to transition to the second stage of implementation.
The greater the discomfort and the less rewarding the activity, the more likely it is that
resistance will be encountered. These feelings may be magnified if personnel also feel
that failure will be associated with penalties (Dievernich, 2014).
In a recent evaluation of the 10 top obstacles to change, resistance to change was
identified as the most important hurdle seen to be a problem by 82% of polled companies
(Figure 7). Some have described this challenge as being one of “behavioral economics”,
in which thought, and behavior affects financial decisions and resource allocation
(Minton, 2014). This field of study helps to illuminate why lasting organizational change
is so difficult to assure; the psychological bias to maintain the status quo represents what
is known as a “loss aversion” strategy (Tversky & Kahneman, 1992). Thus, in any type of
change, the odds are already “stacked” against change from happening. In the instance of
UDI implementation, it is not productive to resist a change that is mandatory under the
law. Nevertheless, Lewin’s theory would anticipate some level of foot-dragging in
response to UDI implementation.
45
Figure 7: Top 10 Barriers to Change Management
(Source: Cisco, 2013) Reproduced with Permission
The second stage of “changing” can also be challenging. In the case of UDI
implementation, heightened complexity might be anticipated because the work of UDI
implementation spans so many different systems that may need to be interconnected. For
example, the way that documentation systems are structured will have implications for
the way that changes are embedded into standard operating procedures, tracking
documents and manufacturing records. Changes in production methods to accommodate
UDI may require a return to design control to be sure that the changes are tested
appropriately so that they do not introduce mistakes or new risks. Regional needs may
also add complexity. While the research here is focused on the changes associated with
FDA deadlines for UDI compliance, other countries have also introduced requirements
for UDI identification that may be consonant or dissonant with those of the US. These
varying regulations can affect internal processes and support systems (Lentz, 2015).
46
There is no question that once UDI has been implemented, the maintenance phase will
also be challenging. Even the regulators have not yet ironed out all of the “wrinkles”
associated with the continued maintenance of UDI as products change (McClellan et al.,
2014). Manufacturers must cope with the ongoing assurance that UDI systems keep pace
with those changes.
If Lewin’s model were to be applied to the UDI change process, certain tasks can be
associated with each stage. Because medical device manufacturers have change control
procedure embedded in their Quality Management Systems (QMS), the “unfreezing”
process in change management is particularly time consuming. Typically, any change
must be requested and documented to establish and memorialize requirements,
justifications, and permissions for that change before moving forward. The unfreezing
stage will require that personnel be convinced that such a change is necessary and
desirable and be educated about the required actions for such a change (Pheney, 2013).
The UDI requirements affect the entire portfolio of the company, and it is likely that
much of the “unfreezing” process will require a top-down approach in which Regulatory
Affairs teams partner with the Quality Assurance teams to integrate activities holistically.
To facilitate effective change, the organization needs a clear vision of the change as part
of the “new normal”. Lewin notes that the “unfreezing” process has a higher likelihood
of success if the organization focuses on creating the motivation to change amongst
stakeholders, rather than dictating terms and instructions.
47
The change process gains momentum after an organization has accepted that change
needs to occur. At this stage, it is necessary to investigate all of the places and activities
described above in section 2.3, where a UDI change may have impact. Controlled
documents from design to post-production must be adjusted to take into account the new
tracking information and procedures must be written that, for example, guide the creation
of new labels and labeling that then must be added to the electronic inventories
(PharmOut, 2016). It is likely that the Regulatory Affairs teams will play a central role
although it is possible that different companies have different models for structuring the
responsibilities and authorities for this stage of activity. These accountabilities may
undergo multiple changes and adjustments as the process proceeds. As the change starts
to take shape, confidence tends to rise and the organization becomes comfortable with the
new system, even though further changes will no doubt be required over time.
The “freeze” process reflects the fact that an intended change will only be successful if
the new paradigm is adopted permanently by assuring changes become part of the
ongoing methods of doing business. Change is established by helping employees to reach
the stage in which the changed system is a part of their normal way of doing things. This
gives the employees the chance to exhibit new behaviors, which are then reinforced
(Kinicki, 2005).
2.5 The Fixsen Model of Implementation
The Lewin model has much in common with another model, proposed initially by Fixsen
and his colleagues (Fixsen et al., 2015), that concerns itself with the implementation of a
48
change, as part of a field of study recently characterized as implementation science. In
that model, the focus appears to be more restricted to the “changing” process (Bertram et
al., 2015). Its primary concerns are the description and identification of factors that could
facilitate the implementation phase of change management.
As shown in Figure 8, the activities defined in the Fixsen model occur in five stages:
exploration and adoption; program installation; initial implementation; full operation;
innovation; and sustainability. The “Exploration Phase” appears to overlap with what
Lewin would call the “unfreezing stage”, a critical phase to ensure a proper workflow
downstream. Some of the planning and communication activities that would typically
take place in this phase when undertaking UDI exploration and assessment are shown in
Figure 8.
49
Figure 8: UDI Exploration Framework: Phase 1: Lewin’s ‘Unfreeze Stage”
Copyright © 2018 Aimee Greco
During the implementation phases, Fixsen breaks what Lewin characterized as a single
stage into a set of stages involving program installation, initial implementation, full
operation and innovation (Fixsen et al., 2015). The map in Figure 9 attempts to illustrate
some of the stakeholders and activities that may have to occur in tandem from a high
level.
50
Figure 9: UDI Implementation Framework: Fixsen’s Implementation Phase
Copyright © 2018 Aimee Greco
Lewin’s final stage of “freezing”, corresponding to Fixsen’s “Sustainability” phase, also
has challenges. A change towards a higher level of group performance is frequently
short-lived, after which group performance soon degrades. Permanency at a new level, or
for a desired period, should be included as a defined objective of the change process
(Lewin, 1943). However, UDI management can never be frozen completely. A
sustainable system must not only assure the continued implementation of the changes
51
now required by the FDA but must anticipate that other requirements may have to be
implemented to “change the change”. At the time of writing, for example, discussions in
several governments are taking place about expanding UDI content to include data
related to reimbursement and device warranties (Thompson, 2016). Thus, management
will have to continue to be involved in the continued efforts to maintain UDI compliance.
52
Figure 10: UDI Implementation Framework: Phase 3: Lewin’s Re-Freezing
Stage
Copyright © 2018 Aimee Greco
2.6 Studying Change
It is rare that an opportunity exists to examine the change control approaches of many
companies that are all presented with the same challenge, in this case to implement UDI
53
in a specified timeframe. It is a particular stressor to make such changes in the face of
regulatory deadlines, so in such a situation, it should be easier to identify where change
systems have vulnerabilities. While this study is not intended to be a prescriptive process
to implement UDI, the UDI implementation process provides a “vehicle” in
understanding change management science in the regulated industry.
The goal of the present study is to understand how medical device companies navigate
the change process when faced with a multifaceted problem like UDI implementation that
spans several functional divisions. A question of particular interest can perhaps be
stated: “What is the phase of change that causes the most consternation and why?” A
subordinate question is: “Do change management challenges such as those posed by UDI
affect smaller companies differently than larger ones?”
The present study proposes the use of a survey to gain insight into UDI change through
the eyes of industry professionals working in areas responsible for some aspect of UDI
change. In order to study the change process systematically, the Lewin model will be
used to structure the survey around the three phases of change. The implementation
model of Fixsen will be referenced in order to expand the intermediate, “change”
segment of the Lewin model into additional segments and to provide a lexicon for
describing these different phases of change. Of particular interest, will be the challenges
associated with change management in three areas; 1) Processes: challenges associated
with the “how” of UDI change 2) People: challenges associated with the “who”, the
54
employees of the organization, and 3) Facilities and equipment: challenges associated
with the “where”, the infrastructural requirements of change.
55
CHAPTER 3. METHODOLOGY
3.1 Overview
Survey methods were used to examine the challenges involved in change management
during UDI implementation. A novel survey instrument was developed and subjected to
critical review by a focus group of experts, then disseminated to industry professionals
who were involved with UDI implementation in their organizations. The survey data was
collected, analyzed and summarized in a subsequent section of this study. Finally, the
survey was published for public dissemination using Qualtrics software, an internet
web-based analytics survey company that specializes in “experience management” with
quantitative statistical analysis capabilities.
3.2 Creation of the Survey Instrument
A survey instrument consisting of approximately 30 questions was constructed that could
be completed by most respondents in less than 20 minutes. An initial set of demographic
questions identified the characteristics of the respondent and his or her organization.
Further questions were directed at exploring the challenges that respondents experienced
throughout the lifecycle of UDI implementation. To structure these questions, a
framework developed by Lewin and expanded by Fixsen was used, with a particular
emphasis on three areas, the processes that were developed or modified (how to achieve
implementation) the people involved in that implementation (who is responsible for
implementation) and the facilities and equipment that required to be changed (where
changes must be carried out to permit implementation). Further, the survey attempted to
56
identify where challenges were localized in different stages of the change from
unfreezing to freezing.
An electronic survey platform, Qualtrics (qualtrics.com) was chosen to both disseminate
and analyze the survey. The Qualtrics platform enables questions to be developed in a
variety of formats, including yes/no, multiple choice, matrix and ranked preference
formats, as well as open-text boxes. It also allows the survey to be disseminated in a way
that protects the anonymity of respondents. This capability was viewed as important
because respondents may be cautious about revealing what they believe to be proprietary
information.
3.3 Focus Group and Validation of the Survey Instrument
A selected group of industry professionals and academics experienced in various aspects
of UDI implementation and/or survey development were approached to serve on a focus
group to advise on the working draft of the survey instrument. Their comments regarding
the content and format of the survey instrument were used to modify the draft survey in
order to formulate a final document ready for general dissemination. The final
respondent pool was collected from medical device and biotechnology companies whose
product portfolios require the incorporation of the new UDI regulation and therefore must
go through their established change management process. Inclusion criteria for
participants focused on individuals who have participated in at least one exercise of UDI
implementation in any capacity. Potential respondents were expected to have experience
57
as active participants involved in UDI implementation and have job responsibilities in
Quality Assurance, Regulatory Affairs, Research and Development and/or Project
Management. They were contacted by email in order to brief them about the purpose of
the study and the intent to distribute the anonymized survey within a specified time
frame.
The group comprised of both industry and academic professionals as noted in Table 1.
Table 1: Focus Group Participants
Participant Profile
Institution Job Title
Large Company (1) Vice President of Regulatory Affairs
and Quality Assurance
(1) Vice President of Regulatory Affairs
(2) Director of Regulatory Affairs
(3) Regulatory Affairs Manager
Medium Company (1) Director Regulatory Affairs and
Quality Assurance
(1) Quality Assurance Lead
Small Company (1) CEO, Medical Device Company
Government (1) Former FDA Consumer Safety Officer
USC Professors (1) Professor and Chair, Department of
Regulatory and Clinical Sciences
(1) Director and Assistant Professor,
International Center for Regulatory
Science
(2) Professor School of Pharmacy,
Department of Regulatory and
Clinical Sciences
58
Total Number of Participants (15) Fifteen
The focus group was conducted mid-week during business hours in an academic setting
via both in-person and via teleconference with approximately half the participants being
physically present. The focus group lasted approximately 90 minutes. The researcher
provided an overview of the newly drafted UDI regulations and a highlight of the change
management and implementation models and finally, a proposed structure of the survey
instrument. A significant portion of the focus group meeting was devoted to discussion
and validation of each individual survey question.
3.4 Administration, Data Collection and Analysis
The finalized survey was first sent to a small number of faculty in ICRS who examined
the survey to assure that it displayed and captured answers appropriately. Email
addresses of the appropriate potential participants were then compiled into a survey
panel. The survey was sent to the identified panel, with a personal note to assure the
potential respondents that the results would only be used in a grouped way that protects
their individual identities. The panel included professionals in the medical device
industry and/or professionals who worked in companies with combination products or
varied portfolios that included medical devices. Particular effort was made to find
individuals in quality or regulatory affairs that had experience in the manufacturing
setting because that is the environment in which many aspects of UDI implementation
lies. A total of 312 invitations to participate in the survey was sent in early August 2017
to the end of March 2017. The surveyed respondents were recruited through multiple
59
outlets including industry organizations, industry referrals, and industry networking
amongst conference settings, academia and professional networks such as online industry
forums. The identity of the respondents was de-identified in the survey results.
Responses to the survey were collected and stored electronically. The survey data was
then collected, analyzed and summarized in a subsequent section of this study. While
some questions had binomial “yes” or “no” choices, other forms of multiple-choice and
matrix formats were utilized that allowed for graphing and simple statistical analysis.
Results were then incorporated into the thesis with a discussion and conclusion of data
analysis. Finally, the survey was published for public dissemination.
60
CHAPTER 4. RESULTS
4.1 Focus Group
The focus group was held on Friday, July 21
st
, 2017 at the University of Southern
California’s Health Science Campus in the conference library of the International Center
for Regulatory Science. After a brief discussion of the aims of the survey, the focus
group members reviewed each question sequentially. They discussed how to fine-tune
the questions to sub-stratify respondents with logistical UDI experience (such as
challenges in manufacturing, for example), from respondents in senior management with
a more strategic role. Thus, in the survey, questions were tailored to differentiate tactical
versus strategic activities related to UDI implementation. The focus group discussed
how perspectives might differ at different organizational levels, and how this might relate
to values, norms, and behaviors in change management.
Participants also suggested that differences might exist between companies of different
sizes. They suggested that sub-stratification of large and small companies would depend
on structuring the early demographic questions to capture this information appropriately.
Additionally, they discussed the differences between change control and change
management, and more specifically, the ambiguity of their definitions in regulated
industry and the need for sensitivity to these terms in the questions. The wording of each
question was scrutinized to assure that regulatory language was specific and that the
structures of the questions were appropriate. Each question was evaluated to assure that
it would yield useful information. Attention was given to articulating the questions in a
way that would not introduce bias to the respondent. Participants suggested that
61
questions should try to tease out which departments and functions in the companies were
primarily responsible for UDI management.
4.2 Survey Instrument Distribution
The survey was distributed on September 15, 2017. Three hundred and twenty-seven
potential participants were invited to participate in the survey. Many potential recipients
did not receive the invitation as indicated by a software server message that the message
had bounced back and a further third of the e-mails were either being transferred to a
“junk” folder or eliminated by the server security systems. In these latter cases, I resent
the web-link to different, usually personal, e-mail addresses when possible. In some
cases, recipients received the invitation but declined to participate. Seventeen respondents
communicated to my personal email that they either did not have the time to participate
or did not wish to participate for reasons that were not disclosed.
The survey results were locked sixty-days after the first distribution. A total of
55 recipients participated in the survey (see Appendix 2 for the consolidated results of the
survey). Some respondents did not answer all the questions.
4.3 Analysis Survey Respondents
As shown in Figure 11, respondents came from organizations of different sizes; a
majority (43/55; 78 %) of respondents indicated that they worked in small companies
with under 500 employees, seven (7/55; 13%) in medium-size companies with 3,000 -
49,999 employees and (5/55; 9%) in large companies with over 50,000 employees. The
distribution of company sizes is consistent with the medical device industry. The 2012
Economic Census and the 2016 ITA Medical Devices Top Markets Report (ITA, 2016)
62
reported that the medical device industry employed more than 356,000 people in the
United States, at over 5,800 establishments. Most are small and medium-sized
enterprises; 80 percent of these are estimated to have fewer than 50 employees.
Figure 11: Which Statement Best Describes Your Organization?
N=55
Most commonly, the respondents had a principal functional role in Regulatory Affairs
(RA) (22/55; 40%) (Figure 12). The second most common choice was Quality Assurance
(QA) (10/55; 18%). Only a few respondents self-identified with research and
development, manufacturing, government or consulting. However, the “other” category
contained about a third of respondents (17/55; 31%). These respondents identified that
they held a variety of roles. One respondent was an Executive Vice President. Another
described a blend of Quality Assurance and Regulatory Affairs. A few respondents
reported other diverse roles including technical service, marketing, and product
management. No participants came from supply chain management. One respondent
noted that he /she worked at a very small company and had multiple roles.
63
Figure 12: Functional Roles Held by Respondents
N=55
As shown in Figure 13, a majority of the survey respondents were seasoned professionals
(27/35; 77%) in upper or middle management. Six (6/35; 17%) were at specialist or
associate levels and two self-identified as “other”, one specifying a position as
Founder/Owner and one as President.
Figure 13: Job Levels Held by Respondents
N=35
Most survey respondents (26/36; 72%) indicated that they worked at companies that only
developed medical devices but about a quarter had both device and pharmaceutical
64
products in their product lines (8/36; 22%) (Figure 14). No company produced only one
class of medical devices (Figure 15). Most respondents worked with a blend of medical
devices in different classes. A single respondent identified that products included both
medical devices and other beauty products, and one identified that portfolios included
prescription devices and drugs.
Figure 14: How is Your Organization Best Described?
(If you are a consultant, think of one company with which you work and which you will
reference to answer most questions).
N = 36
Figure 15: What Class(es) of medical products does the organization sell for the
US market?
N=36
65
4.4 Views on Behavior and Culture Related to Change
Respondents were asked to identify their level of agreement with several preliminary
statements designed to explore the culture for change in the companies at which they
worked. Most agreed (strongly: 15/35, 43%; somewhat agreed: 17/35; 49%) that their
departments handled change management well; two neither agreed nor disagreed (2/35;
6%) and only one strongly disagreed (1/35; 3%) (Figure 16). Most also agreed (strongly:
18/35, 51%; somewhat; 9/35, 26%) that their managements supported systematic
methods of change management; only two (2/35; 6%) disagreed and six (6/35; 17%)
neither agreed nor disagreed. Most respondents also agreed (strongly: 14/35, 40%;
somewhat: 13/35, 37%) that change management is effectively embedded in the quality
systems in their organization. However, a quarter (8/35; 23%) neither agreed nor
disagreed. Similarly, when respondents were asked if sufficient resources were allocated
for regulatory changes, most agreed (strongly: 9/35, 26%; somewhat: 13/35, 37%), but a
quarter disagreed (somewhat: 7/35, 20%; strongly: 2/35; 6%) and a few neither agreed
nor disagreed: 4/35, 11%).
66
Figure 16: Culture for Change: Using UDI as an Example, how well did their
organizations handle change?
N=35
Respondents had a range of views about how long it would take to implement UDI into
their organizations (Figure 17) from “identification of the need to change” to “resource
allocation”. Four respondents (4/34; 12%) indicated that it would take one month or less,
eleven (11/34; 32%) that it would take 1-3 months, and six (6/34; 18%), 4-6 months.
About a third (11/34; 32%) estimated that the activity would take six-months or more. A
few (2/34; 6%) could not answer the question.
67
A second phase of implementation from “resource allocation” to “assignment of bar
codes” also had a range of estimates. Four respondents (4/35; 11%) anticipated that it
would take one month or less, nine (9/35; 26%) 1-3 months, eleven (11/35; 31%)
4-6 months, and eight (8/35; 23%) 6 months or more. Three (3/35; 9%) could not answer
the question.
For a third phase, from “assignment of bar codes” to “full data submission on GUDID”,
seven respondents (7/35; 20%) estimated that the activity would take one month or less.
Eleven (11/35; 31%) expected it to take 1-3 months, seven (7/35; 20%) that it would take
4-6 months, and nine (9/35; 26%) that it would take 6 months or more. One respondent
(1/35; 3%) was not able to answer the question.
Finally, from the “beginning of equipment modification” to “completion of equipment
validation”, four (4/35; 11%) of respondents indicated the activity would take one month
or less. Twelve respondents (12/35; 34%) indicated the activity would take 1-3 months,
four (4/35; 11%) indicated 4-6 months and seven (7/35; 20%) indicated 6 months or
more. Eight (8/35; 23%) could not answer the question.
68
Figure 17: Based on your experience, how long would you estimate that it took to
complete the following stages of UDI implementation?
N=34-35
4.5 Possible Areas of Opportunity During Change
Respondents were asked to critique the current state of project management for UDI in
their organizations. When asked whether their organizations were resistant to change
(Figure 18), a majority disagreed (strongly: 14/34; 42%; somewhat: 7/34; 21%). Only one
respondent strongly agreed (1/34; 3%), seven (7/34; 21%) somewhat agreed and five
(5/34; 15%) neither agreed nor disagreed (Figure 18).
Responses were mixed when asked whether management had unrealistic expectations of
the timelines and costs associated with UDI implementation. Four (4/34; 12%) strongly
agreed, ten (10/34; 29%) somewhat agreed and seven (7/34; 21%) neither agreed nor
disagreed; two (2/34; 6%) somewhat disagreed and eleven (11/34; 32%) strongly
disagreed.
69
Most respondents disagreed (somewhat: 8/34, 23%; strongly: 15/34, 44%) that UDI
implementation was hampered by poor project management. Fewer agreed that
implementation was hampered by poor project management (strongly: 2/34, 6%;
somewhat: 5/34, 15%) and four (4/34; 12%) neither agreed nor disagreed. When asked if
the initial case for change was not presented in a compelling way, most respondents
disagreed (strongly: 13/32, 41%; somewhat: 6/32, 19%). Fewer agreed (strongly: 1/32,
3%; somewhat: 6/32; 19%), six (6/32; 19%) neither agreed or disagreed.
When respondents were asked if the project team lacked the necessary skills for UDI
implementation, most disagreed (strongly: 13/34, 38%; somewhat: 8/34, 24%), but nearly
a third agreed (strongly: 1/34, 3%; somewhat agreed: 9/34, 27%) and three (3/34, 9%)
neither agreed nor disagreed.
70
Figure 18: Considering your experience with UDI, how well do the following
statements (considered as problems by others during change) describe
your initial efforts at implementation?
N=32, 34
Respondents were presented with certain problems that had previously been identified as
potential problems associated with UDI implementation (Figure 19). When asked if
available sources of guidance for UDI implementation were insufficient, nearly half
either strongly (6/34; 18%) or somewhat (9/34; 26%) agreed. However, nearly as many
(somewhat 8/34, 24%; strongly: 4/34, 12%) disagreed, and seven (7/34, 21%) neither
agreed nor disagreed. However, most respondents disagreed (strongly: 18/34, 53%;
somewhat: 7/34, 21%). that they lacked clear SOPs for change control; only six either
71
strongly (3/34, 9%) or somewhat agreed (3/34, 9%) and three (3/34, 9%) neither agreed
or disagreed. When asked if “…computer systems were not integrated in a way that
made UDI changes easy”, more respondents agreed (strongly: 9/34, 27%; somewhat:
7/34, 21%) than disagreed (strongly: 8/34, 24%; somewhat: 4/34, 12%). A smaller group
of five respondents (5/34; 15%) neither agreed or disagreed and one respondent did not
know.
When asked if the scope of the project was unclear at the beginning, more respondents
agreed (strongly: 1/34; 3%; somewhat: 13/34, 38%) than disagreed (somewhat: 6/34,
18%; strongly: 5/34, 15%) and about a quarter neither agreed or disagreed (9/34, 27%).
72
Figure 19: Considering your experience with UDI, how well do the following
statements (considered as problems by others during change) describe
your initial efforts at implementation?
N=34
4.6 Findings by Research Questions – Exploration Phase
To support the introduction of UDI, the FDA and a variety of consultants, bar code
providers, and industry news agencies have provided guidance documents and other types
of documentation regarding UDI implementation. Respondents were asked about the
usefulness of these different resources when they initially explored how to implement
UDI (Figure 20). Most commonly respondents identified FDA guidance documents or
information on the FDA website as the most preferred resource (13/35; 37% and 6/35;
17% respectively) and were considered as very or moderately useful by most others (see
Figure 20 for detailed distribution of rankings); unusually, one individual ranked FDA
guidance documents as “not at all useful”. Respondents also typically ranked the
materials provided by UDI code suppliers as very or moderately useful (15/35; 43% and
73
11/35; 31% respectively) and four (4/35; 11%) identified these documents as their most
useful source. Three (3/35; 9%) ranked materials from UDI code suppliers as only
slightly useful. A less uniform set of views were seen for other sources of guidance, such
as blogs or articles in trade journals. None of the respondents saw these sources as their
most useful resources, but they were nevertheless ranked as very or moderately useful by
most (very, 7/35, 21%; moderately, 15/35, 44%; slightly, 9/35, 26%) and only one (1/35;
3%) ranked them as not useful at all. Ten respondents who chose “Other” further
clarified their choices with comments (Table 2).
74
Figure 20: Exploration: Please rate the utility of sources when using the
following industry references to aid UDI implementation planning.
N=35
75
Table 2: Comments on the Utility of Sources in the Exploration Phase
Please rate the utility of sources when using the following “guidances” or comment on
other research sources.
Small organization and relatively few products. UDI team consisted of two persons, so dynamics was not an issue.
It was very helpful to have an excellent consultant who had worked on UDI at FDA
We have a slightly different perspective - we were launching a product that was all-sterile and incorporated UDI
from the start - we were not trying to retrofit an older system without UDI.
Because it was a new concept there were many unknowns at the beginning. It turned out to be not an issue at all.
We already had barcodes for our devices, but the difficult was understanding from FDA how to put that
information to GUDID. After it was figured out it really wasn't that difficult
External training was absolutely necessary to complete this project successfully.
Project costs grew as we became more knowledgeable
It was very helpful to have an excellent consultant who had worked on UDI at FDA
We followed the FDA guidance which was very informative.
I didn’t know much about buying bar codes.
For many companies, and like FDA said:" UDI will change the way we manage data." We were not / and still remain
unorganized with respect to data management.... Data resides in 5-7 departments who have not routinely worked
together ever before.
The biggest problem with implementation of UDI in or organization was not due to change management problems.
The biggest issues were trying to learn what UDI meant and how to implement an entire new system with no
previous background or very clear direction from FDA in how it has to be applied in industry.
Respondents were asked to comment on their initial approaches to UDI implementation.
The survey asked whether they had expected at the outset to have sufficient personnel
with the knowledge to implement UDI without bringing additional staff into the
organization. About half (19/36; 53%) replied yes and about half (17/36; 47%) replied
“no”. (Figure not shown). When asked if they engaged the services of a consultant, one-
third of respondents (12/36; 33%) said “yes”, but most (23/36; 64%) said “no”; one
recipient did not know (Figure not shown).
Of the companies that did use a UDI consultant, half (4/8; 50%) identified that the
consultant handled 1-25% of the project and a further third (3/8; 38%) noted that they
handled 26-50% of the project. None handled more than 50% of the project (Figure 21).
76
When asked if they were satisfied with the way that the consultants interfaced with
departmental staff, a majority (6/8; 75%) of the respondents were satisfied with interface
activities, whereas two (2/8; 25%) were neither satisfied nor dissatisfied. No respondents
identified that they were dissatisfied. (Figure not shown). Twenty-one respondents who
chose “Other” further clarified their choices with comments (Table 3).
Figure 21: Portion of the project was conducted by the consultant(s).
N=8
77
Table 3: Comments on Challenges or Unforeseen Issues from the Early
Exploration to the Completion of Implementation of UDI.
Please comment on challenges or unforeseen issues (Example: Did it take longer than
you thought?)
Software devices were challenging
Difficulty bringing contract manufacturers up to speed, equipment needs, significance if label changes.
It took much longer than anticipated. It was an evolution of improved insight the further along we were in the
process.
Took longer than we thought and there are still areas of gray that need to be fully addressed...i.e. procedure packs
It took way longer (6-9 additional months) than we all thought.
Getting into the database took a little longer than I thought. We have several thousand-part numbers and it should
have taken about 10 days and it took longer. We had to start with 10-part numbers and work our way up to 500+.
Once we got the first submission in and we learned how they wanted each of the 50 categories formatted with
data, things went very smooth.
such a big thing with many different issues. We have become UDI experts in self-defense.
Master data within ERP systems was often not complete nor accurate. Catalogue identifiers were often inaccurate
or obsolete.
Yes! my mistake was that I contracted with World Sync, which is suggested by the
GS1 label formatting requirements are contained in massive guidance documents that are often difficult to
decipher if you are not familiar with label and barcode creation, making the implementation of physical labeling
much more difficult than the UDI and GUDID database information gathering process.
The symbols had to be modified to also meet EU and Asian regulations
Went reasonably well, though with nearly 40 years’ experience working in the regulated industries, I may have had
a slight advantage in understanding the objectives and methods employed by the Agency.
Longer than we thought... Still lack a seamless approach for capturing, storing and submitting data
This took much longer than though. One thing which was not clear initially is that the barcode printed also needs
to pass quality standards and we had to find readers which could grade the barcode for acceptability. Additionally,
changes in labeling required notifications and alignment with our customer chain.
My organization has well over 10,000 unique product codes that are subject to 801.45 (direct part marking) so the
sheer number of people needed to even update product drawings is overwhelming and has been difficult to get
funding for
Pretty much everything we did for UDI involved a lot of interpretation on our own, pieced together with free
webinars from advertisers wanting to sell us special software that was the most helpful with practical advice on
how to implement UDI. I think there were a lot of things included in the UDI/GUDID that didn't apply to us that we
didn't even need to worry about, but we didn't know that at first. Once we started entering things in the GUDID it
became clearer the intention of the UDI ruling. The intention of the rule is a lot simpler than the way it was
explained I think the FDA made things a lot harder than they had to.
No.
Once in contact with GS1, implementation was simpler.
GS1 was making changes simultaneously with our implementation of UDI which was unexpected and caused great
confusion
Cost of the UDI numbers - codes purchased for identifying your project is yet another expense to medical devices.
Orthopedic Sets in the field pose a huge challenge, as well as Class I devices packaged together.
GUDID seemed not to be stable in the beginning making upload of data cumbersome
78
4.7 Findings by Research Questions – Execution
Respondents were asked to rate, from highest to lowest, several challenges previously
identified to hinder execution (Figure 22). The three areas identified as most challenging
were insufficient business processes and workflows and a lack of expertise. Insufficient
business processes and workflows were identified by nine (9/28; 32%) as the most
challenging of the issues. Nonetheless, most areas were seen very challenging by at least
some respondents as shown in Figure 22. The area of least concern appeared to be that of
leadership and management support, by about half of the respondents with an opinion, as
shown in Figure 22.
“Not enough focus on how to implement the change” was the second greatest area of
concern. Four respondents (4/28; 14%) rated this the highest, five (5/28; 18%) rated it
second, nine (9/28; 32%) rated it third; one (1/28; 4%) rated it fifth. “A lack of expertise”
was rated highest by seven (7/28; 25%) second highest by 9 (9/28: 32%); and third by six
(6/28: 21%). Six respondents placed this lower in ranking, three as fourth and 3 as fifth
(both 3/28: 11%).
Other areas on the lower scale of concern included lack of leadership or management
support and whether the government regulation or “guidances” were unclear. With
respect to leadership or management support, four (4/28; 14%) ranked the issue as
highest, (second 3/28: 11%; third 2/28: 7%; fourth 5/28: 18%), but nearly half (13/28;
46%) ranked this area fifth, and one (1/28; 4%) ranked it as the area of least concern.
79
With respect to the clarity of government regulations or guidance documents, four (4/28;
14%) ranked this area as highest, but responses were highly mixed (second 6/28: 21%;
third 4/28: 14%; fourth 7/28: 25%; fifth 6/28: 21%: least 1/28: 4%).
Figure 22: In order of importance please rank the following issues that were
prioritized when incorporating UDI into your organization.
When asked to rank the three most important areas of implementation in which difficulty
had been faced, respondents most commonly selected “understanding the regulations to
determine changes” (20/28; 71%), followed by “creating new labels and new equipment
for each device” (16/28; 57%) and “housing master data syndication” (15/28; 54%)
(Figure 23). Two other areas, “implementing the manufacturing changes” (11/28; 39%)
and “determining which department should own the UDI process” ran closely behind
(10/28; 36%). Much less frequently chosen were “supplier management” (6/28; 21%) and
“quality control” (1/28; 4%) (Figure 23).
80
Figure 23: What were the top 3 areas in change management that you found
caused the most difficulty?
(Note that because three areas were selected, the percentages will total more than 100%.)
Respondents were asked about the concordance between the expected and required time
to complete certain required process changes associated with UDI. In a previous
81
question, shown in Figure 17, respondents had anticipated that implementation for most
UDI activities would be completed in less than six months. When asked if those
expectations were too cautious, respondents most commonly found that three areas
required significantly more time to implement: how to access the FDA’s system to
establish a GUDID account and set up the security protocols and data entry (15/27; 55%);
the time it took to author SOPs, work instructions and personnel training (14/27; 52%);
the initial research phase of regulatory requirements (14/28; 50%). In two areas, time
requirements were identified to be “much longer”: the initial research into regulation
requirements, and the time needed to communicate with FDA (Figure 24).
82
Figure 24: How would you describe the time needed for the process compared to
initial expectations?
Many respondents did not characterize the implementation of UDI as either difficult or
not difficult (12/28: 43%). Most of the remaining respondents found the implementation
to be somewhat difficult (10/28: 36%) or very difficult (3/28: 11%). A few respondents
found UDI implementation to be “somewhat easy” (3/28; 11%). (Figure 25)
83
Figure 25: Overall, how was your experience when implementing UDI?
When asked whether their organization was ready for an FDA inspection that would
examine the incorporation of UDI into their quality systems, nearly 80% (22/28, 79%)
said “yes”, about 20% were not sure (5/28; 18%) and one (1/28; 4%) said “no”
(Figure 26).
Figure 26: With UDI implemented, do you think your organization is ready for
an FDA inspection that includes UDI?
N=28
Respondents were asked if they were confident in retrospect that they made the correct
changes to incorporate UDI successfully. Over eighty-five percent of the respondents
(24/28; 86%) replied “yes”, and only two replied “no” (2/28; 7%). Two respondents
(2/28; 7%) identified that they could not tell (Figure 27).
84
Figure 27: I am confident that we made the correct changes to incorporate UDI
successfully.
N=28
4.8 Findings by Research Questions – Implementation
Respondents were asked to agree or disagree with certain statements that might
characterize certain specific aspects of implementation related to documentation and
training (Figure 28). About equal numbers of those who could comment either disagreed
(strongly: 8/29; 28%; slightly: 5/29; 17%) or agreed (strongly: 6/29; 21%; slightly: 6/29;
21%) that it was not difficult to update the procedures, electronic records and electronic
signatures (ERES) to satisfy regulatory traceability requirements. Only four (14%) could
not comment (Figure 28). However, more than 40% (12/29; 41%) of the same
respondents could not comment on the difficulty of converting and publishing in HL7
SPL XML format. Of the remaining individuals who did have an opinion, seven
indicated that they agreed (strongly: 5/29; 17%; slightly: 2/29; 7%) and ten disagreed
(slightly: 3/29; 10%; strongly: 7/29; 24%).
A mix was also seen when respondents were asked about UDI training. Twelve agreed
(strongly:4/29; 14%; slightly, 8/29; 28%), twelve disagreed (strongly: 4/29; 14%;
85
slightly: 8/29; 28%), and five (17%) could not comment whether the training process was
straightforward. Most agreed that sufficient tools or materials were available for UDI
training (strongly: 3/29; 10%; slightly; 15/29; 52%), but seven disagreed (slightly: 3/29;
strongly: 4/29; 14%) and four (14%) could not comment.
Figure 28: Implementation Phase: Views on UDI Implementation
4.9 Findings by Research Questions – Sustainability
Respondents were asked to reflect on the effectiveness and sustainability of UDI
implementation in their organizations (Figure 29). Asked whether personnel were
sufficiently knowledgeable about UDI after implementation, a few felt that their
expectations were exceeded (far exceeded, 2/28: 7%; exceeded 3/28: 11%). However,
most identified that expectations were equal (18/28: 64%) and a few identified that the
knowledge of the personnel fell short (fell short: 4/28: 14%; fell far short 1/28: 4%).
(Figure 20).
When respondents were asked if UDI maintenance was an integrated part of their
operations (Figure 29), a majority indicated that their expectations were matched (about
86
equal 13/28: 46%). Of the others, more found that their expectations were exceeded (far
exceeded 4/28: 14%; exceeded 7/28: 28%) while four (4/28; 14%) responded that their
SOPs fell short of expectations. No respondents indicated that their research fell far short
of expectations.
When specifically asked if the documents issued by the FDA were adequate to support
the sustainable management of UDI, only one (1/28; 4%) respondent indicated that their
expectations were far exceeded, and two that they exceeded expectations (2/28; 7%).
Most commonly the respondents indicated that their expectations were about equal to the
adequacy of the documents (17/28; 61%). Six (6/28; 21%) estimated that their knowledge
fell short of expectations, and two that they fell far short (2/28; 7%).
Standard operation procedures (SOPs) are a set of required documents that outline the
steps for key processes. Since SOPs are the first tool used to ensure consistency and
quality in a company’s activities, respondents were asked if their companies had SOPs
sufficient for the sustainable management of UDI databases. Responses varied, but about
half found that their SOPs conformed with their expectations whereas the others were
distributed more commonly in the positive rather than negative direction (far exceeded:
3/11: 28%; exceeded 6/28: 21%; about equal: 14/28: 50%; fell short 3/28: 11%).
Similarly, resources are a general term in an organization that outlines assets necessary to
accomplish goals or operate effectively. When asked if resources, such as materials,
personnel or financial assets, were sufficient to support the change process over time, two
(2,28; 7%) identified that their expectations were far exceeded, whereas three identified
87
that their expectations were exceeded or matched (3/28: 18%; matched 13/28: 46%). A
smaller number indicated that expectations fell short (6/28: 21%; far short 2/28: 7%).
Figure 29: Sustainability: Please let us know your level of agreement with the
following statements.
88
When respondents were asked in retrospect about the areas of UDI that would they wish
to improve, the documentation process, employee training, and data housing were chosen
most commonly as the highest choices (Documentation processes 6/29; 21%; Training:
7/29; 21%; Data housing: 7/29; 24%). However, assessments were mixed for all of these
choices as well as the others, that included equipment procurement and or revision, initial
UDI needs assessment and level of management support. Figure 30 shows the detailed
distribution of responses. A majority rated the documentation process the highest area
that they wished to improve, while a smaller majority, nine (9/29; 31%) ranked this area
as the second highest as well. However, six (6/29; 21%) ranked this same area in third
place, while seven (7/29; 24%) ranked it in fourth place. One respondent (1/29; 3%) felt
that this area did not need improvement.
Expectations were similarly mixed when it came to training. While six (6) ranked this
area as the highest, only two (2/29; 7%) ranked this area second, seven respondents (7/29;
24%) ranked it in third, and seven (7/29; 24%) ranked it fourth in order of importance.
Six respondents (6/29;21%) ranked it fifth, while one (1/29; 3%) did not feel it was an
area that needed improvement.
Six (6/29; 21%) indicated that data housing was most important to improve, and four
(4/29; 14%) ranked this area second. Seven (7/29; 24%) ranked this area third, and one
(1/29; 3%) ranked it fourth. Eight (8/29; 28%) and three (3/29; 10%) ranked this area
fifth and sixth respectively.
89
Areas of sustainability that respondents did not expect to need as much improvement
included equipment procurement and/or revision, initial UDI needs assessment and level
of management support Figure 30.
A majority of respondents rated equipment procurement on the lower end of the scale.
Only two (2/29; 7%) ranked this area the highest rank, while four (4/29; 14%) ranked this
area second, and three (3/29; 10%) ranked this area third. Additionally, six (6/29; 21%)
ranked this area fourth, and five (5/29; 17%) ranked this fifth, nine (9/29; 31%) rated this
area sixth in importance,
Initial needs assessment for UDI was ranked highest by five respondents (5/29; 17%),
five as second and five as third in importance (5/29; 17%). Three (3/29; 10%) indicated
this choice was fourth, eight (8/29; 28%) as fifth and three respondents (3/29; 10%) felt it
was not an area that needed improvement.
Finally, the level of management support did not seem to be an area that needed
improvement. Only four respondents (4/29; 14%) ranked this area first in the area
needing most improvement, whereas five (5/29; 17%) ranked this area second, and one
(1/29; 3%) ranked this area third. Five (5/29; 17%) ranked it fourth, while one (1/29; 3%)
ranked it fifth, and the largest number of recipients, thirteen (13/29; 45%) ranked level of
management support as the area that needed least improvement Figure 30.
90
Figure 30: Sustainability: In retrospect, which areas of UDI implementation
would you wish to improve?
4.10 Findings by Research Questions – Time Constraints
Respondents were also asked about the time required from the point at which the
organization became aware of the regulation and associated need to incorporate the
requirements into their processes to the point at which the company was able to produce
the first UDI label (Figure 31). The average reported time was between 8 and 12 months,
but responses ranged widely from 2-28 months.
91
Figure 31: Please estimate the time it took to implement UDI?
N=29
Respondents were also asked about the concordance between the time that they expected
to take and the time that they actually did take to implement UDI (Figure 32). About half
indicated that the time to implement was similar to that predicted (15/29: 51%), but about
one-third were split between taking a shorter or longer time than expected (took more
time: 5/29: 17%; took less time: 5/29: 17%) A few did not know (4/29: 14%).
92
Figure 32: Did the time to implement UDI: Exceed, equal, fall short of
expectations or did not have any idea how long UDI implementation
would take?
93
CHAPTER 5. DISCUSSION
5.1 Introduction
It is not the strongest of the species that survive, nor the most intelligent, but the
one most responsive to change. (Charles Darwin)
Response to change depends not only on the nature of the organization undergoing the
change but also on the environment into which the change is introduced. Thus, even
when two companies must make the same type of change, the ability to respond
effectively can require different resources and time frames for the adjustment. The length
of each stage and the difficulty of the path will be influenced by the magnitude and
importance of the change. It can be particularly difficult when the change needs to be
implemented within a constrained period of time in order to avoid significant
consequences. By using UDI as a case study, this research attempts to open the black
box of change by examining the challenges faced by companies of different sizes and
product areas when all attempting to implement the same type of change at the same
time.
5.2 Consideration of Methods
The present research took advantage of the new requirements by the FDA for unique
device identification, which must be introduced across the entire spectrum of medical
device companies. The first part of this study involved the examination of literature
associated with UDI implementation, an approach that has been instrumental not only to
bring together diverse voices but also possibly identify gaps in knowledge very much like
an “extended conversation” (Locke, Spirduso & Silverman, 2000). Extending and
94
confirming those insights requires a more systematic methodology that solicits feedback
on specific issues from appropriate individuals in the field (National Board of Certified
Counselors, 1993). Survey methods were considered to be the most appropriate way to
explore how change management is implemented in a range of companies producing
medical devices because it provides a relatively efficient way of to gather information
from a large, yet specific cohort (Fincham & Draugalis, 2013). The opportunity to ask the
same questions to a sufficiently large population of experts is central to assuring to the
greatest extent possible that the data are representative and relevant to the activity, and
thus to increase confidence in the “validity” of the results.
There are two types of validity that assume importance in a study such as this. Internal
validity is the degree to which the results reflect accurately the relationships and views
for which the survey was designed. External validity is the ability of the study to
generalize beyond the particular populations studied at a particular time to populations
more generally across time, place and job description. Each of these will be affected by
the limitations and delimitations of the study. Both aspects of validity are relevant to this
discussion.
5.2.1 Internal Validity
The strength of data in any type of survey research depends on its design. A robust
design, in turn, depends on the use of a theoretical framework and solid construction of
relevant questions (Polland, 1998). Thus, before the survey was disseminated, it underwent
testing to optimize its usefulness by using a focus group of experts to conduct a formal
content review. For surveys that are directed at a specialized respondent pool, focus
95
groups have been shown to be useful in order to assess the face validity of the survey
before it is disseminated widely. In this study, it helped to refine the results of the initial
heuristic approach used to choose the survey questions. After consulting with a group of
experts familiar with more structured survey development and /or UDI requirements and
logistics, the survey could then be remodeled to be sure that the questions were
appropriate and clear. The focus group could also provide a certain level of affirmation
that the survey would add value to the rather scanty information so far available from the
literature review.
Nonetheless, the construction of a good survey is a challenging task. A long survey
discourages participants from finishing a survey (Tomaskovic, Leiter, & Thompson, 1999).
Further, even those survey participants who completed the survey did not always answer all
of the questions and often indicated that they did not know one or more answers. This
might not be surprising. The UDI change process has several facets, and some
professionals may not have a full view of the change process. For example, when
recipients were asked if they felt that sources from UDI code suppliers such as GS-1,
H1BCCC, or ICCBBA were adequate, about a third of respondents were not able to
answer. The assignment of codes is a very specific task in which many respondents are
unlikely to participate so their inexperience with documents from code providers may be
understandable.
Other limitations can result from logistical challenges on a more tactical front. For
example, during the survey distribution, some respondents failed to receive the survey,
either because the available email address was no longer functional (often a challenge in
96
an industry in which personnel change companies frequently) or because the survey was
rejected by a spam filter. Others appeared to receive the survey but did not participate.
Many previous survey developers have reviewed some of the reasons why potential
respondents might choose not to complete the survey (Cannell & Kahn, 1968). Heavy
workloads are well known to be typical for those in the middle and senior management of
regulatory products companies. For them, surveys may be seen as one more unnecessary
distraction. Alternatively, the respondent pool targeted here is privy to highly sensitive
corporate secrets and may fear that the survey may expose proprietary information.
Additionally, they may decline to open a survey like this from “an abundance of caution”.
Because this challenge and limitation were recognized prior to survey distribution, extra
time was spent to develop a focused list of potential participants and to explain the goals
and confidentiality limits to them, in order to optimize both the “inputs” of respondent
experience and the “output” of securing responses of high quality. The length of the
survey was of particular concern because indifference or survey fatigue could cause the
respondent to provide “straight line” responses, that is, selecting the same intermediate
choices that might be typical of uncommitted participant (Herzog & Bachman, 1981).
Thus, development of the survey instrument was specifically designed to permit
respondents to skip a question and still continue to participate in the survey. In such a
case, missing data may not necessarily be a limitation (Marino, 2018). Since the survey
was designed to allow questions to be skipped, an individual could contribute their
valuable expertise and views in one area, but not in others where the level of expertise is
insufficient to provide a good answer.
97
5.2.2 External Validity
The fundamental assumption of a good survey is that it represents the views of a broader
constituency from which the respondent sample was derived. For this reason, external
validity is affected by the nature of delimitations placed on the respondent pool. Thus, in
order to properly disseminate a survey sample, it was imperative that a subset of
respondents represented a general population of interest (Rea, Parker & Wallace, 2014).
The research presented here was limited to the medical device industry that must
implement the new UDI regulation because its products are sold in the United States.
Thus, the survey results may not represent the views of those in other geographies whose
regulations regarding UDI are different or still evolving. Further, it was imperative that
the selected participants provide a representative sample of the broader target population
of interest (Martinez-Mesa, Gonzalez-Chica, Duquia, Bonamigo & Bastos, 2016).
Challenges to the representativeness of the data may come from an insufficient or poorly
chosen sample of respondents. However, reaching individuals with experience in the
implementation of UDI can be a challenge. A “wide net” was cast to reach as many
experts in this area as possible, by using not only personal contacts but also referrals and
professional groups to add to the list of contacts. In this effort, I approached individuals
who had job functions such as change management, management operations, process
operations, Quality Systems or regulatory affairs, for example. These individuals are
most directly involved in the UDI process but might have somewhat different views from
those different work experiences. It must be recognized that the sample of respondents
represented here were primarily working in regulatory and quality sciences and not in
98
more peripheral areas such as manufacturing or information technology. These
individuals might have different views about the success and challenges of UDI specific
to their domains.
Another aspect of representativeness that also must be considered is the degree to which
the sizes of the medical device companies matched those in the overall US medical
device environment. As noted earlier, the 2012 Economic Census and the 2016 ITA
Medical Devices Top Markets Report (ITA, 2016) reported that the medical device
industry employed more than 356,000 people in the United States, at over 5,800
establishments. Most are small and medium-sized enterprises; 80 percent are estimated to
have fewer than 50 employees. In this study, the predominance of smaller companies
would appear to be consistent with that in the medical device industry as nearly 80% of
respondents from this survey indicated that they belonged to a “small” organization
(under 500 employees).
The concerns about representativeness may also be allayed to some degree by the results
showing that a new, wide encompassing policy such as UDI posed similar challenges of
interpretations and implementation for both large and small companies. For example,
cross-tabulation of the data showed that for companies of all sizes, change management
was at times, “confusing”, “vague”, “incomplete” and “indistinct”.
This research study was further delimited to a timeframe that represents a slice of
experiences and views in the middle of an evolving process. We might anticipate that
regulations and guidance documents will continue to evolve. For example, the
International Medical Device Regulators Forum (IMDRF) has released, in 2019, a
99
harmonized guidance on UDI implementation that was not available at the start of this
work. In accordance with these ongoing changes, we might anticipate that
implementation procedures will also evolve, and those changes may affect the views of
practitioners. It would be intriguing to see if the experiences reflected here are typical of
respondents five years in the future after the implementation of UDI is more advanced
and respondents have had more opportunity to understand the requirements of
implementation. Such a study would be particularly interesting since, for many of these
companies, the “sustainability “stage of implementation has not yet been reached.
However, it is clear that the requirements to change the bar coding on devices undergoing
transition to a new model or from a new manufacturing facility will require ongoing
systems to maintain the currency of the GUDID database for example.
5.3 Culture and Change
The study examined change management using a framework based on Lewin’s model of
change, a model in which the change process has three stages: unfreezing a current state;
making the change and freezing the altered state that has incorporated the change.
Within each of these three stages can be identified certain substages identified by Fixsen
that are specific to the implementation of change. In the “Unfreezing” stage, it proved
useful to place Fixsen’s substages of “exploration”, and “installation”, because some of
the challenges appeared to be specific to one or another of the substages. The “change”
stage appears congruent with Fixsen’s initial and full “implementation” substages. The
“Freezing” stage appears to be equivalent to the “sustainability” substage of Fixsen. By
100
using such a framework. it was easier to map the challenges experienced by companies as
the change progressed.
5.4 Unfreezing: Exploration and Installation
Because UDI affects many departments, respondents indicated that the most difficult
processes at first appeared to be understanding the nature of the changes that were needed
across the organization. In particular, at least half of the respondents reported that the
company expended significant time and effort in three activities during the
exploration/installation phase: doing research on regulatory requirements; learning how
to establish a GUDID account and to set up the security protocols and validations prior to
data entry; and developing SOPs, work instructions, and training. Many companies
appeared to have an incomplete understanding of the scope of the changes. Thus, of all
of the implementation stages, this one seemed to be the longest. The fact that it took
most companies more than 4 months, and often more than 6 months, to move from the
identification of the need for the change to resource allocation may reflect this lack of
clarity. For some respondents, it appeared to be difficult to develop a comprehensive
logistical plan for implementation, so that later activities required more resources than
initially estimated. This was perhaps captured best by two of the respondent comments,
below:
It took much longer than anticipated. It was an evolution of
improved insight the further along we were in the process.
(Anonymous Respondent)
101
The biggest issues were trying to learn what UDI meant and how to
implement an entirely new system with no previous background or
very clear direction from FDA in how it has to be applied in
industry… (Anonymous Respondent)
In the exploration phase, guidance documents were the preferred source of consultation
for UDI information and were considered moderately to very useful by most respondents.
Nevertheless, more respondents agreed than disagreed that the available resources were
insufficient. Further, some noted that the regulations were difficult to understand during
the exploratory phase, and alternate resources provided to industry were also too general
to guide the change from a tactical perspective.
It is perhaps not surprising that the FDA documents consulted most widely would be
seen to have limitations. The perspective of a regulatory agency is to identify at a high
level what is needed to comply with the law. However, implementation in a specific
company may require further exploration to understand particular requirements related to
the specific manufacturing, labeling and regulatory environment unique to that company.
As stated by FDA in the preamble to the final rule for UDI, “We received many
comments that requested guidance or suggested a need for guidance on various aspects of
the rule” (FDA, 2013). The preamble also contained a large number of comments
concerning the need to clarify expectations for specific situations or products, ranging
widely from convenience kits to stem cells.
Further information to assist implementation can also be obtained from trade materials or
bar code providers, both of which were frequently used as sources by the respondents
here. Nevertheless, it seems clear that the sources of useful information could be
102
improved to facilitate the more efficient planning for UDI introduction and management.
Literature contains anecdotal comments that corroborate this view. For example, Jay
Crowley, who previously served at FDA for 26 years, most recently in the role of senior
advisor for patient safety at CDRH, discussed the aspects of UDI that have and have not
worked, along with the challenges ahead. He suggests that the language of UDI
regulation is not completely clear, and more work continues to facilitate the process even
though FDA guidance has been finalized (Fontanazza, 2019).
Part of unfreezing is preparing the organization for change. Although the literature
suggests that an organization is often resistant to change at the beginning (Jones, Aguirre
& Calderone, 2004), the survey responses suggested that only a small number of
respondents (approximately 20%) somewhat agreed with the statement that their
organizations were resistant to change. Perhaps the cooperative way in which most
companies appear to accept the new requirement reflects the fact that the medical device
industry is often confronted by changes in regulatory requirements, and these legally
forced changes are quite widely accepted as part of doing business in this sector. It would
be interesting to know if the situation differs when a change is not made to satisfy
regulatory requirements, such as a merger or plant relocation, particularly if such changes
were to threaten the affected personnel with more personal risks such as job change or
even job loss (Rouse & Frame, 2009). For example, if a medical device company divests
parts of its operation such as its sterilization operation, is this change accepted as easily?
(Fontanazza, 2019).
103
As exploration evolves into installation, resources must be acquired so that the
organization is well prepared for implementation. Key departments and stakeholders are
identified, funding is secured, and the teams begin to coalesce to drive the
implementation. One of the key factors in this stage is assuring adequate manpower and
expertise. As described by one of the respondents,
Such a big thing with many issues. We have become UDI experts in self-defense.
(Anonymous Survey Respondent)
Yet most of the respondents were not dissatisfied with the level of resource allocation to
the UDI project, suggesting that companies understand the need to meet the regulatory
requirements at least at the level of financial support. Nonetheless, providing financial
resources is generally regarded as only one aspect of management support. The literature
has suggested that the success of a change also depends on management’s engagement to
communicate optimism and urgency in order to set the tone (Prosci, 2014). In this study,
however, more than a third of respondents strongly disagreed that the case for UDI
change had been presented in a compelling way by the management. This finding
suggests that for changes mandated by the FDA, the case for compliance and urgency
may have to be communicated in other ways. For example, the company’s regulatory or
quality department may have to communicate and champion the requirement, or the
regulatory agency itself may have sufficient visibility to the key individuals
implementing the change that the championship role of senior management is diminished.
104
5.5 Changing (Execution and Implementation)
Changes were companywide to house and manage, it was not a simple endeavor.
(Anonymous Survey Respondent).
The survey illuminated additional challenges occurring later in the implementation
phases of the change. In the case of UDI implementation, execution is about more than
introducing new processes. It is about purging an old process to make way for a new one
(AchieveIt, 2018). During the period of UDI implementation, respondents identified at
least four broad areas that seemed most challenging: insufficient process workflows;
labeling challenges; introduction or modification of equipment and computer systems;
and insufficient understanding or expertise to conduct the change efficiently.
5.5.1.1 Inadequate Process Workflows
Once companies began to implement the necessary changes to assure a robust UDI
program, they found numerous challenges related to developing a systematic workflow.
At the highest level, respondents appeared to feel confident that they had good systems
and Standard Operating Procedures (SOPs) for change control, but specific SOPs for
certain aspects of bar coding and database management were less mature. This caused
many to identify the management of documentation as one of the areas that they felt in
retrospect was given insufficient attention at the outset. This appeared to be particularly
true for the more difficult processes such as the sustainable management of UDI
databases. Challenges with workflow and with related equipment modifications and
validations were particularly troublesome for activities related to label change and
management of computer systems, and these are areas of particular discussion below.
105
Nevertheless, the fact that robust SOPs are required in regulated industry (FDA, 2015b)
makes it likely that companies will bring the workflow needs into compliance over time
if only to reduce the likelihood of quality failures and expensive product recalls (Fuhr,
2013).
5.5.1.2 Label Changes
Respondents as a group identified that labeling changes were a particular challenge
during the UDI transition, second only to understanding what the UDI change entailed
overall. Some indicated that this challenge led the company to misestimate the budget
allocation for labeling changes because it was not clear at the outset that new label
printers might be needed to apply variable barcodes, for example. Such mistakes can
have a serious impact, given that a label change for a single medical device has
previously been estimated to cost $80,000 (Eastern Research Group, 2012). This cost
includes not only development and validation changes but also the costs of scrapped
inventory and disposition of nonconforming materials (as required under FDA 21 CFR
820.90). Respondents suggested in their comments that considerable attention was
needed to the way that new labels were designed, and bar codes were assigned. One
individual noted that his/her company had to apply more than 10,000 unique product
codes to its diverse collection of products; this complexity has been previously identified
as a problem not only for process design but also for database management. Not
surprisingly then, costs of implementation that would be incurred by all domestic medical
device labelers have previously been estimated to be $396.3 million in the first year and
$73.6 million in subsequent years (Eastern Research Group, 2012).
106
“[We had] difficulty bringing manufacturers up to speed, equipment needs,
significance (if there were) label changes.” (Anonymous Survey Respondent).
One area that also provided many companies with unexpected expense and distraction
was that of procuring and validating new “fit for purpose” equipment. Renewing such
equipment can depend on budget allocations unforeseen at earlier stages when financial
planning was undertaken. The costs of procuring the correct equipment are not trivial- it
has been estimated in the tens of thousands of dollars (Eastern Research Group, 2012).
With the FDA’s new UDI labeling requirements, manufacturers also had to evaluate and
assess whether they needed to redesign their labeling process and medical device UDI
labels (Allen, 2017). As an example, custom label making capabilities were needed for
the wide variety of labels that required flexible 1200 dpi print quality, in order to fit the
new narrow labels needed to reduce cost and waste. At the same time, the custom labels
had to comply with the FDA’s UDI label content, design, GUDID upload, and print
regulations.
5.5.1.3 Data Management and Computer Systems:
A third top concern in change management for this particular group of respondents was
that of managing the data required as part of bar code documentation. As identified by
one respondent,
Our computer systems were not integrated in a way that made UDI changes easy.
(Anonymous Survey Respondent).
To introduce UDI, software systems must be able to accommodate new, large sets of
data, particularly if the company must label multiple products from different
107
manufacturing sites. UDI implementation experts have suggested that proper control of
the associated data is especially important because labeling is managed and used by
various departments. For example, the activities of those developing the labels must be
synchronized with activities in supply chain logistics throughout the product cycle (Allen,
2017). In addition, for the more complex products, bar coding also will have effects at the
level of design control as well as manufacturing, because labeling is an integral part of
the product as a whole entity (FDA, 1989) and often must be applied to the device itself
as well as its packaging (FDA, 1989).
Syndicated data, i.e., data that is not used by a single client or department, can be also be
challenging to house and access. Global Data Synchronization (GDSN) systems must be
accessible not only to internal users but often to certain third parties who may or may not
have permission to change or supplement the data. Any such changes must be 21 CFR
11 compliant (FDA, 2003). Privacy issues may be involved if the bar codes include
individually identifiable health information associated with specific patients through
electronic registries (Gliklich, Dreyer, & Leavy, 2014). Thus, both the housing and
distribution of this data can be problematic particularly when dealing with third parties.
These complications were recognized by many respondents, who noted that the
construction and management of computerized systems for data management were
challenging and expensive.
Data management is not only a challenge for the manufacturer but also for its many
vendors and contractors. One respondent who worked as a vendor to medical device
manufacturers in China, for example, found that understanding how UDI rules
108
specifically affected their products sold in the US became critical to the vendor’s ability
to engage in those business partnerships. Essentially, the foreign manufacturer’s
challenges became the vendor’s challenges. Because medical products are sold in many
countries, most of which have unique bar-coding systems and requirements, it would be
interesting in the future to understand how U.S. regulations have affected foreign
manufacturers and importers, which may have language and cultural difficulties in
understanding their responsibilities with respect to UDI. Challenges may also extend to
users such as healthcare facilities, where appropriate changes would need to be made to
existing EHR systems to facilitate the collection and integration of UDIs with other
patient data (Wilson & Drozda, 2013). Thus, modifications to existing electronic systems
were seen to require extensive coordination at every stage as discussed elsewhere
(Simplor, 2017).
The complexity of a system is defined as the number of elements and relationships of the
system (Funke, 2003). The more features in the process, the more interdependencies are
likely to exist (Dörner, 1989). Thus, it is not surprising that almost half of the respondents
agreed that their computer systems were not integrated to absorb UDI easily. It is not
only important that the company’s internal systems for data management are robust.
Respondents commonly pointed to the frustrations and delays introduced by trying to
enter required data into the GUDID database.
GUDID seemed not stable… making upload of data cumbersome… (Anonymous
Survey Respondent).
109
Preparing for GUDID data upload required multiple steps, such as establishing the
business entity’s business DUNS number which may take up to 30 days (NiceLabel,
2016). Additionally, the GMDN code number must be retrieved from the GMDN
Agency, and then FDA must conduct a GUDID data-verification process that can take a
week or more (FDA, 2019a). These bureaucratic delays in data processing can affect the
timeliness of product launch.
5.5.1.4 Personnel Development and Capabilities
Key to implementation are the abilities of personnel to make the necessary changes to
processes and products. Because the UDI requirements are new, individuals must
continue to seek clarification about regulatory requirements and logistical issues as
implementation proceeds. Most respondents considered that the available resources were
insufficient in this regard. Further, they identified that much of the FDA information
included in the UDI/GUDID guidances did not necessarily apply to their company’s
product, but the relevance of those reference materials to their products was not always
clear at the outset of planning. Thus, significant resources and energy were required to
teach the organization new skills, processes and at some level, to institute a cultural
change.
“There were always questions and none [no one] to really bounce ideas off of… I
felt like I was always chasing answers.” (Anonymous Survey Respondent).
“…anything and everything that involved UDI required internal interpretation,
particularly in “piecing” together information from free webinars from
advertisers that sold special software.” (Anonymous Survey Respondent).
110
The need to understand implementation details led some respondents to identify specific
areas in which more guidance would be helpful. One problem that arose early in the
implementation phase appeared to be that of choosing an FDA UDI issuing agency.
Further, the FDA guidance relative to direct part marking was seen by some to be
somewhat unclear and confusing. Some also commented that the guidance documents
gave good timelines, but other documents were needed to find more prescriptive
information related to actual implementation details. Some of these challenges have in
fact been predicted by those who gave comments addressed in the preface to the final rule
for UDI (FDA, 2013).
One way that some companies found to improve their capabilities was by using
experienced consultants. Of those companies using consultants, most had a positive
experience and their appreciative comments reflected the value that they brought. This
may not be surprising because consultants or subcontractors specializing in industrial
installations of UDI and other process enhancements bring expertise. However, their
costs may be prohibitive to smaller firms (Mohammed, 2014).
“I was the one tasked with implementing UDI and had no clue what I was doing.
I had to educate myself, go to conferences and glean whatever from colleagues,
guidance documents, GS1 provider, free webinars… there was much trial, and
error involved.” (Anonymous Survey Respondent).
The relative paucity of good educational materials does not help to improve the capability
of company personnel. As stated by one consultant,
111
…there was a great deal of confusion on roles and learning how to work with
UDI because they [the consultants/advisors] needed to understand their client’s
problems to service them properly, however, the clients did not understand their
own problems, because the regulation was simply not prescriptive enough to
determine whose responsibility was to do what. (Anonymous Survey Respondent)
This sentiment resonated in other questions and other comments. The fact that about half
of respondents did not agree that they had appropriate personnel may also reflect the
challenges discussed above in the “Unfreezing” stage. Certain members of the
implementation teams, such as those responsible for IT, were particularly vulnerable
because they “…were only trained in operations systems and nothing that had anything
to do with FDA compliance”.
Respondents were split on their views regarding the sufficiency of their teams for
implementing UDI. Further, a small majority indicated that the project teams lacked the
needed skills to implement UDI into their organization. This is an interesting set of
responses, suggesting variability of experience amongst the industry professionals. Cross
tabulation to stratify the data by company size did not indicate that inadequate skill-sets
were specific to either large or small companies but suggested a lack of experience and
training to an unfamiliar task in which all companies were learning at the same time.
“This was new to the industry. We all learned together. There were no SME’s…
when we all started on the change.” (Anonymous Survey Respondent).
“No one had the right people, but you must do it…We did not have the right
people, it was hard at times, but implemented on time anyway.” (Anonymous
Survey Respondent).
112
It would be interesting as a future study to characterize more fully the unrealized cost of
such a change in a company and the usefulness of installing subject matter experts in
order to increase speed and effectiveness of implementing the changes, that might then
translate into value in financial terms.
5.5.1.5 Management
Most respondents agreed that their companies handled activities related to change
management well. Most also agreed that their management supported their needs in
making this change. However, two aspects of managing change seem significant. First is
the aforementioned observation that management gave resources but did not get too
involved as champions of the change. It would be interesting in future research to
determine who in fact was the champion of the change process. Assigning a project
champion is a key to implementation success for many projects (Pinto & Slevin, 1989).
A second aspect was the reaction of the respondents to questions about whether their
organization was ready for an FDA inspection of their new UDI systems. The fact that
nearly 80% said “yes” suggests that the companies were able to make the change
effectively to their systems, at least through the eyes of the respondents, regardless of the
presence or absence of leadership from the highest levels of the company. At the same
time, a number of respondents were not confident about their ability to pass an audit.
These two observations seem to be at odds. It may suggest that the respondents from
those companies either did not have the expertise in house or the financial resources to
engage subject matter experts who could bring the company to a level of compliance that
they felt would be needed. It is also possible that certain respondents simply did not have
113
enough knowledge of the implemented system at the time of the survey to make a
determination.
5.5.1.6 Timelines
“This is incredibly inefficient.” (Anonymous Survey Respondent).
One of the biggest challenges in UDI implementation was its division into many small
projects with different deliverables, all of which had to be completed concurrently to
meet a looming deadline. Despite these challenges, respondents commonly identified that
they were able to predict the time that it would take for most of the activities, such as
implementing the process into their design controls or into multiple manufacturing
facilities. Fewer, but still at least half, identified that it also took about the same amount
of time as they predicted to validate manufacturing equipment. This was perhaps
unexpected given the complexity of these logistical tasks and suggests that they
understood their capabilities well. However, other tasks, especially those related to new
processes, did introduce delays at the later stage of implementation. For example, several
respondents noted that the GUDID subphase took longer than anticipated. Some of the
text answers suggested that at least two aspects caused this delay. First were the
complicated steps of reviewing the material, registering as a manufacturer, and entering
the data. Second was the time taken to satisfy the validation steps and wait for the FDA’s
data-validation check. One respondent noted that the GUDID software did not seem
stable, making data uploads cumbersome. Further observations in the literature also
indicated confusion amongst GUDID users such as inefficiencies in requesting a GUDID
account or delayed responses in clarification questions for the account (Elsmar, 2016).
114
5.5.1.7 Small Companies versus Large Companies
There are approximately 5,300 to 5,600 U.S. companies in the medical device industry,
with approximately 330,000 to 365,000 employees (BMI Research, 2015; International
Trade Administration, 2010) with a majority based in California, Massachusetts, and
Minnesota. However, most are small medical device companies. Of U.S companies whose
primary activity is making medical supplies and equipment, 83 percent had less than $1
million in assets, and 95 percent had less than $10 million in assets (Gravelle & Lowry,
2015). These smaller companies are engaged primarily in the development of new
medical technologies and are often focused on relatively narrow therapeutic areas. In this
study, the distribution of respondents from small and large companies reflected these
statistics. Thus, only a small proportion of respondents came from large companies.
This outcome was positive for external validity as discussed earlier, but it was not helpful
to answer questions about differences between large and small companies. Cross
tabulations were attempted, as shown in Appendix D, but the numbers were too small to
provide convincing evidence of differences between companies of different sizes. They
did however give a few insights into areas that might deserve more thorough
investigation in future. For example, the larger companies appeared to take slightly
longer to ensure personnel training, whereas most small companies appeared to take
slightly longer than expected to set up their GUDID account than did the larger
companies. However, with such small numbers, the results would only provide a
direction for further investigation.
115
5.6 Freezing: Sustainability and Maturity
When making a change that will disrupt the usual workflows so widely, it is critical to
reduce risk by reaching a stable, operating model as quickly as possible by freezing the
new model (Heathfield, 2019). Respondents were asked about their expectations in
several areas related to the sustainability of the system. About two-thirds of the
respondents considered that personnel weres sufficiently knowledgeable about UDI to
maintain the system, and nearly as many identified that the government guidance was
adequate for sustainable management of UDI. However, at this point, sustainability has
not yet been testable, given the newness of the requirements. What is not clear, then, is
whether the system will degrade over time in ways that have not yet been anticipated. A
better understanding of the process, once mature, may be possible if a study such as this
were to be conducted in about 5 years after the deadline for implementation has passed
and the system has been in place for a few years. Studies of change management have
indicated that reinforcement is vital to the success of an initiative because people have a
natural tendency to revert back to what they know and what makes them comfortable
(Prosci, 2014). It is likely that the rigorous FDA oversight of the UDI systems of
companies will motivate the companies to cement the changes effectively. Nevertheless,
we might anticipate that the ongoing management of change will yield surprises. For
example, supply chain tiering in large companies often requires the use of hundreds of
suppliers. Any changes in the supply chain must be captured in the data management
systems and the UDI labels because errors could trigger a device recall (Fuhr, 2013).
116
Failure to make even small changes, such as updates to the GUDID database, for
example, could have far-reaching effects.
Ongoing attention to systemic improvements will be necessary in the future because the
implementation of UDI is not a “one-off” change. Not only will continued actions be
needed as product modifications must be reflected in new UDI labels and associated
product databases. Requirements for UDI are now expanding internationally and some of
these regulations and expectations elsewhere differ from the rules that have been
established by the FDA. For example, the European Parliament recently voted to release
its own version of the unique device identification requirements using a different bar-
coding system and to publish the associated data in a separate European Database of
Medical Devices (EUDAMED), a system similar to the US GUDID counterpart.
Similarly, Australia’s Therapeutic Goods Administration (TGA) is asking the industry to
comment on draft guidance that proposes a unique device identification system of its
own. Sponsors of medical devices supplied in Australia would be required to ensure their
devices carry UDIs, and that relevant information is entered into the Australian UDI
database (AusUDID). It is likely that almost all countries will adopt some form of UDI in
the next decade, in order to trace the movement of devices across global markets.
However, dissonant multiple systems will further complicate the management of UDI
labels and associated documentation. Companies producing combination products will
also be challenged by requirements of the FDA Drug Supply Chain Security Act
mandates, that by 2024 will require labeling to permit the tracking of drugs through the
117
supply chain at the individual package level to assure its authenticity using a digitized
electronic chain of custody (FDA, 2019b).
How all of the competing requirements will affect the ability to sustain a resilient system
is still not clear. It will challenge quality departments in particular, since that department
is likely to be responsible for assuring continued compliance and change control
(O’Leary, 2015). Device companies could be in a unique position to advise on
approaches that would foster the integration of the necessary changes within quality
management frameworks. The identification of specific challenges and solutions at
different stages of implementation may help companies to anticipate problems and to
adopt best practices for dealing with them.
5.7 Future Direction and Recommendations
It is not often that the FDA promulgates a regulatory requirement that impacts the
industry in such a broad and dynamic manner that it will affect the way of doing business
for years to come. The new UDI regulation was intended to increase the safety and
efficacy of the supply chain for the patient and provide the public with better access to
device-related information. It was also created to make it easier for the FDA to monitor
compliance and share information with other regulatory agencies. This survey study did
not look to see if the promulgation of this regulation was successful in achieving these
objectives, but rather the impact of its implementation.
A review of the literature and the survey results here suggest that while industry appears
to make a strong effort to implement requirements expeditiously, the change has been
118
demanding and difficult. The results reported in this study could possibly assist the
development of better guidance and standardized practices for each company by
illustrating areas in which companies have had difficulty. One important finding in this
study was the degree of difficulty associated with UDI implementation at the earliest
stages when approaches were researched and individuals were trained, an area not often
highlighted in considerations of such implementations. One recommendation, then, is to
recognize that more upfront attention is better. An effective exploratory team that rapidly
assembles a collection of essential educational materials and guidance and a well-
organized training program as early as possible might go a long way of accelerating the
implementation and easing the rough edges of the later stages.
It is likely that the lessons learned here may generalize not only to other international
UDI implementations but also to other major regulatory changes that are on the horizon,
such as the implementation in the EU of the new Medical Device Regulations (WHO,
2010). Results may suggest that well-crafted comprehensive guidances and training
programs are essential elements in smoothing these expensive transitions, in order to
prepare industry personnel better for the new work that they must undertake. One
important recommendation to the regulatory agencies is to attend not only to the
requirements but also to more practical assistance in assuring that their documentation is
as clear and comprehensive as possible, while at the same time making sure that targeted
requirements for specific product classes are identified early. This would ensure that
industry personnel can sort the information appropriate to their situation and product lines
more quickly. The types of guidance documents that most respondents identified to be
119
insufficient were those at transactional rather than legal or outcomes levels. Additionally,
creating a forum where stakeholders are able to discuss questions related to their strategy
and tactics might help new individuals to avoid pitfalls already experienced by those who
have already conquered the system.
120
REFERENCES…..
AchieveIt. (2018). Strategy execution is change management in disguise. Retrieved from
https://www.achieveit.com/resources/blog/strategy-execution-is-change-
management-in-disguise. (Accessed 1 Feb 2019).
Allen, D. (2017). 5 ongoing challenges for UDI in unique device identification. Retrieved
from https://www.packagingdigest.com/unique-device-identification/pmp-5-
ongoing-challenges-for-udi-170523. (Accessed 18 June 2018).
Ament, L. (2006). AHA calls for medical device identification system. AHA News,
42(23), 2.
Barlow, R.D. (2013). UDI it's sync or swim time: FDA carves out anticipated regulations,
motivating industry to move. Healthcare Purchasing News, 37(4), 12. Retrieved
from //www.hpnonline.com/inside/2013-04/1304-SF-UDI.html (Accessed 30 Apr
2017).
Berkman, L. (1991, January 20) Shiley heart-valve recipients are focus of worldwide
search. Los Angeles Times. Retrieved from https://www.latimes.com/archives/la-
xpm-1991-01-20-fi-986-story.html (Accessed 12 June 2016).
Berniker, J. (2001). Spontaneous reporting systems: achieving less spontaneity and more
reporting. Retrieved from https://dash.harvard.edu/handle/1/8846816 (Accessed
12 July 2018).
Bertram, R.M., Blase, K.A., & Fixsen, D.L. (2015). Improving programs and outcomes:
implementation frameworks and organization change. Research on Social Work
Practice, 25(4), 477-487.
Blot, W.J., Ibrahim, M.A., Ivey, T.D., Acheson, D.E., Brookmeyer, R., Weyman, A.,
Defauw, J., Smith, J. K. and Harrison, D. (2005). Twenty-five-year experience
with the Björk-Shiley convexoconcave heart valve: a continuing clinical concern.
Circulation, 111(21): 2850-2857.
BMI Research. 2015. United States medical devices report: Q3 2015. New York, NY:
Business Monitor International.
Bordia, P., Hobman, E., Jones, E., Gallois, C., & Callan, V. J. (2004). Uncertainty during
organizational change: types, consequences, and management strategies. Journal
of Business and Psychology, 18(4), 507-532.
Brandt, C. (2016). System sustainability. Quality Progress, 49(6), 22.
121
BSI. (2013). What you need to know about the FDA’s UDI system final rule. Retrieved
October 4, 2015, from http://www.bsigroup.com/en-GB/our-services/medical-
device-services/BSI-Medical-Devices-Whitepapers/ (Accessed 4 Oct 2015).
Cannell, C., & Kahn, R. (1968). Interviewing. In G. Lindzey and E. Aronson (eds),
Handbook of Social Psychology (2nd edn), Volume 2. Reading, Mass.: Addison-
Wesley.
Cisco. (2013). Change management drives customer savings. Retrieved from
https://www.cisco.com/c/dam/en/us/solutions/collateral/workforce-
experience/change-management-june.pdf (Accessed 25 Aug 2017).
Committee on the Public Health Effectiveness. (2011). Medical devices and the public's
health: The FDA 510(k) clearance process at 35 years. Retrieved from
http://www.nationalacademies.org/hmd/~/media/Files/Report%20Files/2011/Med
ical-Devices-and-the-Publics-Health-The-FDA-510k-Clearance-Process-at-35-
Years/510k%20Clearance%20Process%202011%20Report%20Brief.pdf
(Accessed 11 Nov 2016).
Congress, U.S. (1976). Regulatory reform [microform]: Hearings before the
subcommittee on oversight and investigations of the Committee on Interstate and
Foreign Commerce, House of Representatives, Ninety-fourth Congress, first and
second sessions. Committee on Interstate and Foreign Commerce. Washington:
GPO.
Dalessio, A.T. (1996). Technimanagement: The human side of the technical organization
(Vol. 49). Durham: Wiley Subscription Services, Inc.
Dievernich, F. (2014). Change management and the human factor: Advances, challenges
and contradictions in organizational development (Vol. 2015). Cham,
Switzerland: Springer.
Dörner, D. (1989). Die logik des misslingens. Strategisches denken in komplexen
situationen [Logic of failure. Strategic thinking in complex situations]. Hamburg:
Rowohlt.
Durkin, E. (2014). Experts -- following talks with FDA, CMS -- push to integrate
registries. InsideHealthPolicy.Com's FDA Week, 20(36).
Eastern Research Group. (2012). “Final report: Unique Device Identification (UDI) for
medical devices.” White Paper. Task Order No. 24. Submitted to: Food and Drug
Administration Office of Policy & Planning. May, 2012.
Elsmar. (2016). Request an FDA GUDID Account – Problems. US Medical Device
Regulations. The Elsmar Cove. World-wide discussion Forums for Industry
Professionals. Retrieved from
122
https://elsmar.com/elsmarqualityforum/threads/request-an-fda-gudid-account-
problems.68300/ (Accessed 15 Sept 2019).
Emergo. (2016). FDA data shows increase in warning letter, designing and conducting
survey research: a comprehensive guide to medical device firms since 2005.
Retrieved from http://www.emergogroup.com/blog/2013/05/fda-data-shows-
increase-warning-letters-medical-device-firms-2005 (Accessed 7 Feb 2016)
Engler-Modic, E. (2016). Preparing for upcoming UDI compliance. Retrieved from
http://magazine.todaysmedicaldevelopments.com/article/june-2016/preparing-for-
upcoming-udi-compliance.aspx (Accessed 17 Sept 2019).
FDA. (1989). CDRH. Labeling; regulatory requirements for medical devices. Retreived
from: https://www.fda.gov/media/74034/download (Accessed 07 March 2017).
FDA. (1996). Medical device reporting regulation history. Retrieved from
https://www.fda.gov/medical-devices/mandatory-reporting-requirements-
manufacturers-importers-and-device-user-facilities/medical-device-reporting-
regulation-history (Accessed 21 Sept 2019).
FDA. (1997). Design control guidance for medical device manufacturers. Retrieved from
https://www.fda.gov/regulatory-information/search-fda-guidance-
documents/design-control-guidance-medical-device-manufacturers (Accessed 21
Sept 2019).
FDA. (2003). Part 11, electronic records; electronic signatures - scope and application.
Retrieved from https://www.fda.gov/regulatory-information/search-fda-guidance-
documents/part-11-electronic-records-electronic-signatures-scope-and-application
(Accessed 22 Sept 2019).
FDA. (2006). Quality system regulation. Retrieved from https
https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/CFRSearch.cfm?fr=820.
180 (Accessed 28 Sept 2019).
FDA. (2009). MedWatch. Retrieved from
https://www.fda.gov/Safety/MedWatch/default.htm (Accessed 07 March 2017).
FDA. (2011). Guidance for industry. Process validation: General principles and
practices. Retrieved from https://www.fda.gov/media/71021/download.
(Accessed 7 March 2017).
FDA. (2013). Unique device identification system, Federal register. September 24, 2013.
Retrieved from https://www.federalregister.gov/documents/2013/09/24/2013-
23059/unique-device-identification-system (Accessed 22 May 2019).
123
FDA. (2014). Is this product a medical device? Retrieved from
https://www.fda.gov/medical-devices/classify-your-medical-device/product-
medical-device (Accessed 07 March 2017).
FDA. (2015a). Medical device reporting. Retrieved from
https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfCFR/CFRSearch.cfm?CFR
Part=803 (Accessed 07 Dec 2018).
FDA. (2015b). Quality systems regulation. Retrieved from
https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/CFRSearch.cfm?fr=820.
40 (Accessed 30 Apr 2017).
FDA. (2018). Quality system (QS) regulation/medical device good manufacturing
practices. Retrieved from https://www.fda.gov/medical-devices/postmarket-
requirements-devices/quality-system-qs-regulationmedical-device-good-
manufacturing-practices (Accessed 10 Sept 2019).
FDA. (2019a). Guidance for industry. Unique device identification system (UDI system).
Retrieved from https://www.fda.gov/medical-devices/device-advice-
comprehensive-regulatory-assistance/unique-device-identification-system-udi-
system (Accessed 10 Oct 2019).
FDA. (2019b). Guidance for industry. Drug supply chain security act (DSCSA).
Retrieved from https://www.fda.gov/drugs/drug-supply-chain-integrity/drug-
supply-chain-security-act-dscsa (Accessed 15 Sept 2019).
Fincham, J. E., & Draugalis, J. R. (2013). The importance of survey research standards.
American Journal of Pharmaceutical Education, 77(1), 4.
Fixsen, D., Blase, K., Metz, A., & Van Dyke, M. (2015). Implementation science.
International Encyclopedia of the Social and Behavioral Sciences, 11, 695-702.
Fontanazza, M. (2019). UDI implementation: progress, challenges, where we are now.
Retrieved from https://www.medtechintelligence.com/news_article/udi-
implementation-progress-challenges-where-we-are-now/ (10 Sept 2019).
Ford, H. (2013). The life of Henry Ford. Retrieved from
http://www.thehenryford.org/exhibits/hf/default.asp. (Accessed 30 Apr 2017).
Fuhr, T., George, K., & Pai, J. (2013). The business case for medical device quality.
Retrieved from
https://www.mckinsey.com/~/media/McKinsey/dotcom/client_service/Public%20
Sector/Regulatory%20excellence/The_business_case_for_medical_device_quality
.ashx (Accessed 27 June 2019).
Funke, J. (2003). Problem Solving Thinking. Stuttgart: Kohlhammer.
124
George, M. (2016). Basic strategies for avoiding and overcoming resistance in change
management. Retrieved from https://www.isixsigma.com/implementation/change-
management-implementation/basic-strategies-avoiding-and-overcoming-
resistance/ (Accessed 27 Nov 2016).
Gliklich R.E., Dreyer N.A., & Leavy M.B., (Eds.) (2014) Registries for evaluating
patient outcomes: a user's guide [internet]. 3rd edition. Rockville, MD: Agency
for Healthcare Research and Quality (US); Principles of Registry Ethics, Data
Ownership, and Privacy. Retrieved from
https://www.ncbi.nlm.nih.gov/books/NBK208620/ (Accessed 1 August 2019).
Gravelle, J., & S. Lowry. 2015. The medical device excise tax: economic analysis.
Washington, DC: Congressional Research Service.
GS1. (2013). Are you ready for UDI? Unique Device Identification for medical devices.
Retrieved from http://www.gs1.org/docs/healthcare/UDI_Leaflet_Final.pdf.
(Accessed 18 Oct 2015).
Hall, S.D. (2014). Brookings Roadmap Outlines UDI Implementation. Newspaper
Article, Questex Media Group, LLC Delaware. Retrieved from
https://www.fiercehealthcare.com/it/brookings-roadmap-outlines-udi-
implementation (Accessed 11 Nov 2019).
Hamilton, D., Michael, K., & Wamba, S. F. (2010). Overcoming visibility issues in a
small-to-medium retailer using automatic identification and data capture
technology: An evolutionary approach. International Journal of E-Business
Research, 6(2), 21-44.
Heathfield, S. (2016). Executive support and leadership in change management .
Retrieved https://www.thebalancecareers.com/executive-support-and-leadership-
in-change-management-1917803 (Accessed 26 Nov 2016).
Heathfield, S. (2019). Your organization's future demands agility: Changing quickly is
key to your organization's future success. Retrieved from
https://www.thebalancecareers.com/your-organization-s-future-demands-agility-
1918814 (Accessed 26 Nov 2016).
Herzog, A.R., & Bachman, J.G. (1981). Effects of questionnaire length on response
quality. Public Opinion Quarterly, 45(4), 549-559.
HHS-OIG. (2012). Hospital incident reporting systems do not capture most patient harm.
Retrieved April 2017 from the Department of Health and Human Services Office
of Inspector General Reports website: https://oig.hhs.gov/oei/reports/oei-06-09-
00091.asp.
125
Hicks, K.M. (1994). Surviving the Dalkon shield IUD: women v. the pharmaceutical
industry. Teachers College PR, New York, NY.
International Trade Administration, Department of Commerce. 2010. Medical devices
industry assessment. Washington, DC: Department of Commerce.
http://ita.doc.gov/td/health/
medical%20device%20industry%20assessment%20final%20 ii%203-24-10.pdf.
IBM. (2008). Making change work…while work keeps changing. Retrieved from
https://www.ibm.com/downloads/cas/WA3NR3NM (Accessed 9 Sept 2019).
ISO. (2004). Medical devices: quality management systems: guidance on the application
of ISO 13485:2003, ISO/TR 14969. C.F.R.
ITA. (2016). 2016 top markets report - Medical devices: a market assessment tool for
U.S. exporters. Retrieved from
https://www.trade.gov/topmarkets/pdf/Medical_Devices_Top_Markets_Report.pd
f (Accessed 22 June 2019).
Jones, J., Aguirre D. & Calderone, M. (2004). 10 principles of change management.
Organizations & People. (originally published by Booz & Company): Retrieved
from https://www.strategy-business.com/article/rr00006?gko=643d0 (Accessed
14 Apr 2019).
Kessler, D.A., Natanblut, S., Kennedy, D. and et al. (1993). Introducing medwatch: a
new approach to reporting medication and device adverse effects and product
problems. JAMA 269(21): 2765-2768.
Kinicki, A. & Kreitner, R. (2005). Organizational behavior: Key concepts, skills, and
best practices by Angelo Kinicki. Noida area of Delhi/NCR:McGraw-Hill
Education.
Kleinfeld, V., Kaplan, A.H., & Weitzman, S.A. (1976). Federal food, drug, and cosmetic
act: judicial record, 1969-1974. Food and Drug Law Institute series. Publisher
unknown.
Lentz, H. (2015). UDI lessons learned: Q1 medical device conference. Retrieved from:
http://knowledge.reedtech.com/fda-industry-updates/udi-lessons-learned-q1-
medical-device-conference. (Accessed 10 Jul 2016).
Lewin, K. (1943). Defining the "field at a given time". Psychological Review, 50(3), 292-
310.
Lewin, K.(1951) Cartwright, D. (Ed.). Field theory in social science: Selected theoretical
papers (Vol. 1st). New York: Harper and Brothers.
126
Locke, L.F., Spirduso, W.W., & Silverman, S.J. (2000). Proposals that work: A guide for
planning dissertations and grant proposals. 4th ed. Thousand Oaks, Calif: Sage
Publications.
Loeb, G., & Richmond, F. (2003). Making design controls useful for R&D.. Retrieved
from http://www.mddionline.com/article/making-design-controls-useful-
rd?page=48 (Accessed 21 Apr 2017).
Lucintel. (2012). Global medical device industry 2012-2017: Trends, profits, and forecast
analysis. Nanotechnology Weekly, 254.
Marino, M., & Pagano, M. (2018). Role of survey response rates on valid inference: an
application to HIV prevalence estimates. Emerging Themes in Epidemiology,
15(1), 6.
Martinez-Mesa, J., Gonzalez-Chica, D.A., Duquia, B.P., Bonamigo, R.A. & Bastos, J.L.
(2016). Sampling: how to select participants in my research study? Anais
Brasileiros de Dermatologia, 91, 326-330.
McCalman, J.A., Paton, R.A., & Siebert, S.A. (2016). Change management: a guide to
effective implementation (Fourth ed.). Los Angeles: Sage Publications.
McClellan, M.B., Gardina, S., Daniel, G.W., Deak, D., Bryan, J. & Streit, C. (2014).
Unique device identifiers (UDI): A roadmap for effective implementation.
Retrieved https://www.brookings.edu/research/unique-device-identifiers-udis-a-
roadmap-for-effective-implementation/ (Accessed 1 May 2016).
Merriam-Webster’s Collegiate Dictionary. (2005). Springfield, MA: Merriam-Webster
Incorporated.
Merrill, R. (1994). Regulation of drugs and devices: an evolution. Health Affairs, 13(3).
Minton, E.A. (2014). Belief systems, religion, and behavioral economics: Marketing in
multicultural environments. New York: Business Expert Press.
Mohammed, Y. (2014). UDI: What you need to know about timelines, compliance
and submissions. Retrieved from:
https://www.medtechintelligence.com/feature_article/udi-what-you-need-to-
know-about-timelines-compliance-and-submissions/ (Accessed 19 Oct 2019).
Myhrberg, E.V., & Raciti, J. (2013). A practical field guide for ISO 13485. Milwaukee,
WI: American Society for Quality, Quality Press.
National Board of Certified Counselors, Loesch, L.C. and Vacc, N.A. (1993). A work
behavior analysis of professional counselors. Muncie, IN: Accelerated
Development Inc.
127
NiceLabel. (2016). The NiceLabel guide to UDI compliance. Retrieved from
https://www.nicelabel.com/resources/files/doc/resources/local/en_US/Steps_to_U
DI_Compliance_Guide.pdf (Accessed 19 Oct 2019).
O'Keefe, D.F., Spiegel, R.A., & Murphy, E.G. (Eds.). (1976). An analytical legislative
history of the medical device amendments of 1976: an amendment to the Federal
Food, Drug, and Cosmetic Act. Food and Drug Law Institute.
O’Leary, D. (2015). Updating your quality management system (QMS) for UDI.
Retrieved https://www.meddeviceonline.com/doc/updating-your-quality-
management-system-qms-for-udi-0001. (Accessed 9 Oct 2019).
Oxford. (n.d.) Retrieved from Oxford English Dictionary website: https://www.oed.com/
Pew. (2014). Medical device registries: Recommendations for advancing safety and
public health. Retrieved from https://www.pewtrusts.org/en/research-and-
analysis/reports/2014/09/medical-device-registries (Accessed 10 Oct 2019).
PharmOut. (2016). Comparison of medical device standards & regulations ISO
13485:2003 & US FDA 21 CFR part 820. Retrieved from
https://www.pharmout.net/downloads/white-paper-comparison-ISO13485-and-
US-FDA-CFR-part-820.pdf (Accessed 16 Oct 2016).
Pheney, M. (2013). Using the GS1 system for FDA unique device identification (UDI)
requirements. Retrieved from https://www.gs1us.org/what-we-do/about-gs1-
us/media-center/press-releases/detail/articleid/46/gs1-healthcare-us-releases-
implementation-guideline-for-using-gs1-standards-to-address-fda-unique-device-
identification. (Accessed 16 Nov 2015).
Pinto, J.K. & Slevin, D.P. (1989). The project champion: key to implementation success.
Project Management Journal, 20(4), 15–20.
Polland, R. J. (1998). Essentials of survey research and analysis. Retrieved from
http://www.pointk.org/resources/files/Guidelines_on_Essentials_of_Survey_Rese
arch.pdf (Accessed 5 Oct 2005).
Prosci. (2014). Prosci(R) unveils interactive change management eBest practices report.
PR Newswire.
Pryor, M. G., Taneja, S., Humphreys, J., Anderson, D. and Singleton, L. (2008).
Challenges facing change management theories and research. Delhi Business
Review 9(1): 1-8.
Rea, L.M., Parker, R.A. and Wallace, C. (2014). Designing and conducting survey
research : a comprehensive guide. 4th ed. San Francisco, California: Jossey-Bass.
128
Rising, J. and B. Moscovitch (2014). The Food and Drug Administration's unique device
identification system: Better postmarket data on the safety and effectiveness of
medical devices. JAMA Internal Medicine, 174(11): 1719-1720.
Rouse, T., & Frame, T. (2009). Bain brief: The 10 steps to successful M&A integration.
Retrieved from http://www.bain.com/publications/articles/10-steps-to-successful-
ma-integration.aspx (Accessed 19 Oct 2019).
Schremp, P. (1992). Safe medical devices act: reporting requirements and risk
management concerns. Retrieved from https://www.omic.com/safe-medical-
devices-act-reporting-requirements-and-risk-management-concerns-2/ (Accessed
2 May 2016).
SelectUSA. (2016). The medical device industry in the United States. Retrieved from
http://selectusa.commerce.gov/industry-snapshots/medical-device-industry-
united-states (Accessed 11 Jan 2018).
Sims, R.R. (2010). Reforming (transforming?) a public human resource management
agency: The case of the personnel board of Jefferson County, Alabama:
(Contemporary human resource management: Issues and challenges). Charlotte,
North Carolina: Information Age Publishers.
Simplor. (2017). UDI: Industry benefits and challenges. Retrieved from
http://simplor.com/articles-1/2017/4/22/udi-industry-benefits-and-challenges
(Accessed 1 Oct 2019).
Sullivaninstitute. (2014). Unique device identifiers: facilitating the capture and
transmission of UDI. Retrieved from https://www.pewtrusts.org/-
/media/assets/2014/04/07/wedi_udi_white_paper_04072014.pdf (Accessed 10
Oct 2015).
Swanson, D., & Creed, A. (2014). Sharpening the focus of force field analysis. Journal of
Change Management, 14(1), 28-47.
Thompson, H. (2016). UDI and reimbursement: what we know so far about Medicare’s
plan to implement UDI. Retrieved from
http://www.medicaldesignandoutsourcing.com/udi-reimbursement-know-far-
medicares-plan-implement-udi/ (Accessed 6 Nov 2016).
Tomaskovic, D., Leiter, J., & Thompson, S. (1999). Organizational survey nonresponses.
Administrative Science Quarterly, 39(3), 439-457.
Torben, R. (2011). Top 12 reasons why people resist change. Retrieved from
https://www.torbenrick.eu/blog/change-management/12-reasons-why-people-
resist-change/ (Accessed 6 Nov 2016).
129
Towers Watson. (2013). Only one-quarter of empoyers are sustaining gains from change
management initiaitives, Towers Watson survey finds. Retrieved from
https://www.businesswire.com/news/home/20130829005188/en/One-Quarter-
Employers-Sustaining-Gains-Change-Management-Initiatives (Accessed 10 June
2019).
Travess, H., Roberts-Harry, D., & Sandy, J. (2004). Orthodontics. Part 6: Risks in
orthodontic treatment. Br Dent J, 196(2), 71-77.
Tversky, A., & Kahneman, D. (1992). Advances in prospect theory: cumulative
representation of uncertainty. Journal of Risk and Uncertainty, 5(4), 297-323.
US-101-Congress. (1990). Conference report on H.R. 3095, safe medical devices act of
1990. Retrieved April 30, 2017from https://www.congress.gov/bill/101st-
congress/house-bill/3095 (Accessed 30 Apr 2017).
US-105th-Congress. (1998). Implementation of the food and drug administration
modernization act of 1997: hearing before the committee on commerce, house of
representatives, one hundred fifth Congress, second session, October 7, 1998.
The United States Congress.
Weiker, M. (2014) Incorporating the unique device identifier (UDI) into healthcare
delivery. Retrieved from
https://www.wedi.org/forms/uploadFiles/4331300000441.toc.UDI_Webinar-
Part_2.pdf (Accessed 30 May 2019).
Wieting, D.W. (1996). The Björk-Shiley Delrin tilting disc heart valve: historical
perspective, design and need for scientific analyses after 25 years. The Journal of
Heart Valve Disease, 5 Suppl 2: S157-68.
WHO. (2010). Medical devices managing the mismatch: an outcome of the priority
medical devices project. (pp. xiii, 129 p. col. ill.). Retrieved from
https://libproxy.usc.edu/login?url=http://site.ebrary.com/lib/uscisd/Doc?id=10430
905 (Accessed 1 Oct 2019).
Wilson, N.A., & Drozda, J. (2013). Value of unique device identification in the digital
health infrastructure. JAMA, 309(20), 2107-2108.
130
APPENDIX A.
Survey Instrument Solicitation
131
APPENDIX B.
Survey Instrument
Part A: Demographic Information:
132
Part B: Culture and Change Questions:
133
Part C: UDI Process Ability Questions:
134
Part C: UDI Process Ability Questions: (continued)
135
Part C: UDI Process Ability Questions: (continued)
136
Part C: UDI Process Ability Questions: (continued)
137
Part C: UDI Process Ability Questions: (continued)
138
APPENDIX C.
Survey Results
1. Which statement best describes your organization?
# Answer % Count
1 Small (1-499 employees) 78.18% 43
2 Medium (500-49,999 employees) 12.73% 7
3 Large (50,000+ employees) 9.09% 5
Total 100% 55
139
2. Which statement best describes your functional role in your organization?
# Answer % Count
1 Regulatory Affairs 40.00% 22
2 Quality Assurance 18.18% 10
3 R&D 5.45% 3
4 Manufacturing 3.64% 2
5 Consultant 0.00% 0
6 Government 1.82% 1
7 Other (Please Specify) 30.91% 17
Total 100% 55
140
3. At what level is your current position?
# Answer % Count
1 VP 17.14% 6
2 Director/Sr. Director 22.86% 8
3 Manager/Sr. Manager 37.14% 13
4 Specialist/Associate 17.14% 6
5 Other 5.71% 2
Total 100% 35
141
4. How is your organization best described? (If you are a consultant, think of one
company with which you work and which you will reference to answer most
questions).
# Answer % Count
1 Medical Devices only 72.22% 26
2 A blend of devices and pharmaceutical/biotech 22.22% 8
3 Other 5.56% 2
Total 100% 36
142
5. What class(es) of medical products does the organization sell for the US market?
(You may choose multiple answers)
# Answer % Count
1 Class I 36.23% 25
2 Class II 52.17% 36
3 Class III 11.59% 8
Total 100% 69
143
6. Culture for Change: Using UDI as an example, how well do the following
statements describe your company?
144
145
7. Based on your experience, how long would you estimate that it took to complete
the following stages of UDI implementation?
146
8. Considering your experience with UDI, how well do the following statements
(considered as problems by others during change) describe your initial efforts at
implementation?
147
9. Considering your experience with UDI, how well do the following statements
(considered as problems by others during change) describe your initial efforts at
implementation?
148
10. Do you have any comments about your INITIAL approaches to plan UDI
implementation?
149
11. Exploration: Please rate the utility of sources when using the following
guidances to aid UDI implementation planning.
150
12. Did you find the government guidance documents on UDI adequate to
implement UDI into your organization? (If "no", could you please provide a
reason?)
# Answer % Count
1 Yes 62.86% 22
2 No 37.14% 13
Total 100% 35
151
13. Can you comment on any challenges or unforeseen issues from the early
exploration to the completion of implementation of UDI? (Example: Did it take
longer than you thought?)
Software devices were challenging
Difficulty bringing contract manufacturers up to speed, equipment needs, significance if label changes.
It took much longer than anticipated. It was an evolution of improved insight the further along we were
in the process.
Took longer than we thought and there are still areas of gray that need to be fully addressed...i.e.
procedure packs
It took way longer (6-9 additional months) than we all thought.
Getting into the database took a little longer than I thought. We have several thousand-part numbers
and it should have taken about 10 days and it took longer. We had to start with 10-part numbers and
work our way up to 500+. Once we got the first submission in and we learned how they wanted each of
the 50 categories formatted with data, things went very smooth.
such a big thing with many different issues. We have become UDI experts in self-defense.
Master data within ERP systems was often not complete nor accurate. Catalogue identifiers were often
inaccurate or obsolete.
Yes! my mistake was that I contracted with World Sync, which is suggested by the
GS1 label formatting requirements are contained in massive guidance documents that are often
difficult to decipher if you are not familiar with label and barcode creation, making the implementation
of physical labeling much more difficult than the UDI and GUDID database information gathering
process.
The symbols had to be modified to also meet EU and Asian regulations
Went reasonably well, though with nearly 40 years’ experience working in the regulated industries, I
may have had a slight advantage in understanding the objectives and methods employed by the
Agency.
Longer than we thought... Still lack a seamless approach for capturing, storing and submitting data
This took much longer than though. One thing which was not clear initially is that the barcode printed
also needs to pass quality standards and we had to find readers which could grade the barcode for
acceptability. Additionally, changes in labeling required notifications and alignment with our customer
chain.
My organization has well over 10,000 unique product codes that are subject to 801.45 (direct part
marking) so the sheer number of people needed to even update product drawings is overwhelming and
has been difficult to get funding for
Pretty much everything we did for UDI involved a lot of interpretation on our own, pieced together
with free webinars from advertisers wanting to sell us special software that was the most helpful with
practical advice on how to implement UDI. I think there were a lot of things included in the UDI/GUDID
that didn't apply to us that we didn't even need to worry about, but we didn't know that at first. Once
we started entering things in the GUDID it became clearer the intention of the UDI ruling. The intention
of the rule is a lot simpler than the way it was explained I think the FDA made things a lot harder than
they had to.
No.
Once in contact with GS1, implementation was simpler.
GS1 was making changes simultaneously with our implementation of UDI which was unexpected and
caused great confusion
Cost of the UDI numbers - codes purchased for identifying your project is yet another expense to
medical devices.
Orthopedic Sets in the field pose a huge challenge, as well as Class I devices packaged together.
152
GUDID seemed not to be stable in the beginning making upload of data cumbersome
14. Did you feel you had the right personnel with the knowledge to implement UDI
without bringing people into the organization? (If "no", please elaborate on
reason if you can).
# Answer % Count
1 Yes 52.78% 19
2 No 47.22% 17
Total 100% 36
153
15. Did you use consultants for the UDI implementation?
# Answer % Count
1 Yes 33.33% 12
2 No 63.89% 23
3 Don't Know 2.78% 1
Total 100% 36
154
16. About what proportion of the project was conducted by the consultant (s)?
# Answer % Count
1 none 12.50% 1
2 1-25% 50.00% 4
3 26-50% 37.50% 3
4 51-75% 0.00% 0
5 more than 75% 0.00% 0
6 100% 0.00% 0
Total 100% 8
155
17. Were you satisfied with the way that the consultants interfaced with departmental
staff?
# Answer % Count
1 Satisfied 75.00% 6
2 Neither satisfied nor dissatisfied 25.00% 2
3 Dissatisfied (Please comment below) 0.00% 0
Total 100% 8
156
18. Implementation: in retrospect what are your views on the following statements
regarding UDI implementation?
157
19. Sustainability: Please let us know your level of agreement with the following
statements.
158
159
20. In retrospect, which areas of UDI implementation would you wish to improve
the most? Please rank in order of importance with 1 being the most important
and 7 the least important.
160
21. From the time your organization was aware of the regulation, to the point you
were able to produce the first UDI label on a product, how long did it take to
implement UDI from start to finish?
# Field Minimum Maximum Mean
Std
Deviation
Variance Count
1
Please estimate the
months it took to
implement UDI
2.00 26.00 12.45 7.60 57.83 29
161
22. Did the time to implement UDI:
# Answer % Count
1 Exceed expectations? 17.24% 5
2 Equal expectations? 51.72% 15
3 Fall short of expectations? 17.24% 5
4 We had no idea how long it would take 13.79% 4
Total 100% 29
162
23. In order of importance please rank the following issues that were prioritized
when incorporating UDI into your organization.
Question 1 2 3 4 5 6 Total
Insufficient
business process
and workflow
32% 9 18% 5 25% 7 14% 4 11% 3 0% 0 28
Not enough focus
on how to
implement the
change
14% 4 18% 5 32% 9 32% 9 4% 1 0% 0 28
Lack of
leadership or
management
support
14% 4 11% 3 7% 2 18% 5 46% 13 4% 1 28
Lack of expertise
25% 7 32% 9 21% 6 11% 3 11% 3 0% 0 28
Unclear
government
regulation or
guidance
14% 4 21% 6 14% 4 25% 7 21% 6 4% 1 28
Other 0% 0 0% 0 0% 0 0% 0 7% 2 93% 26 28
163
24. What were the top 3 areas in change management that you found caused most
difficulty?
# Answer % Count
1 Understanding the regulation to determine changes 71.43% 20
2 Creating new labels and new equipment for each device 57.14% 16
3 Housing master data syndication 53.57% 15
4 Supplier management 21.43% 6
5 Quality Control 3.57% 1
6 Determining who/which department owned the UDI implementation process 35.71% 10
7 Implementing manufacturing changes 39.29% 11
8 Other (Please specify) 0.00% 0
Total 100% 28
164
25. Would you like to comment on aspects of the implementation that were more
difficult than you expected?
Our laboratory works for mfrs. working on UDI. We do not mfr., but need to
understand their issues which is difficult when they didn’t
None
Direct Part Marking of readable bar code on unreadable surfaces-- OK...
Synchronizing with FDA
Implementing the changes by the Chinese OEM-manufacturers
Company-wide changes to house and manage data
The equipment and validation therefore was unexpectedly challenging
Multiple items packaged together
165
26. How would you describe the time needed for the process compared to initial
expectations?
166
27. Overall, what was your experience when implementing UDI?
# Answer % Count
2 Somewhat easy 10.71% 3
3 Neither easy nor difficult 42.86% 12
4 Somewhat difficult 35.71% 10
5 Very difficult 10.71% 3
Total 100% 28
167
28. With UDI implemented, do you think your organization is ready for an FDA
inspection that includes UDI?
# Answer % Count
1 Yes 78.57% 22
2 Not sure 17.86% 5
3 No 3.57% 1
Total 100% 28
168
29. I am confident that we made the correct changes to incorporate UDI
successfully.
# Answer % Count
1 Yes 85.71% 24
5 No 7.14% 2
6 Cannot tell 7.14% 2
Total 100% 28
169
30. As result of UDI, is there a lesson in change management that you would like to
share?
Start early, communicate often.
Accountability--- Make someone the Project Champion and hold them accountable.
Project management is a key factor.
Master data management to manage: (1) US FDA-USI, (2) OUT - UDI, (3) GDSN
Just because some regulatory change may seem simple at first, remember that this is
the government, and nothing is simple.
170
APPENDIX D.
171
172
173
174
Abstract (if available)
Abstract
The commodities produced by the medical device industry are amongst the most rigorously regulated on the market. Change introduced to such products is challenging because it must be carried out in a way that maintains regulatory compliance, protects productivity, assures cost-effectiveness and most importantly, mitigates unanticipated risk to patients. However, change can be difficult to study systematically because most product-related changes that companies must make will occur at different times and for different reasons. The present survey analysis took advantage of the new requirements by the FDA for unique device identification, which had to be introduced across the entire spectrum of medical device companies. In this research, I examined the current practices and processes that reflected how medical device companies executed change management for a complex change that involved several different departments. A survey tool was structured to explore change management using a framework based on the implementation models of Lewin supplemented by that of Fixsen. The survey was directed at regulatory professionals with an industry role in medical device development. Results suggest that while industry appears to make a strong effort to implement requirements expeditiously, the change has been demanding. The results reported here could assist the development of better guidance and standardized practices by illustrating areas in which companies have had difficulty. More upfront attention including a well organized training program as early as possible might go a long way toward accelerating implementation and easing later stage challenges.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Organizational communication of regulatory intelligence: a survey of the medical device industry
PDF
Risk management and recalls: a survey of medical device manufacturers
PDF
Implementation of risk management in medical device companies: a survey analysis of current practices
PDF
Design control for software medical devices: an industry survey of views and experiences
PDF
Computerized simulation in clinical trials: a survey analysis of industry progress
PDF
An industry survey of implementation strategies for clinical supply chain management of cell and gene therapies
PDF
Views on global harmonization of pharmacopeial standards: a survey of key stakeholders
PDF
Use of natural colors: experience and views in pharmaceutical and dietary supplement industries
PDF
Challenges in the implementation of Risk Evaluation Mitigation Strategies (REMS): a survey of industry views
PDF
Reprocessing of single-use medical devices: a survey investigation comparing the views of three unheard stakeholders
PDF
Experience with breakthrough therapy designation: an industry survey
PDF
Using telemetry to ensure safe and reliable medical device operation: experience with defibrillators and infusion pumps
PDF
Current practices in biocompatibility assessment of medical devices
PDF
A survey of US industry views on implementation of decentralized clinical trials
PDF
Regulatory agreements for drug development collaborations: practices in the medical products industry
PDF
Effect of GDUFA legislation on the development and approval of generic drugs: a survey of industry views and experiences
PDF
Regulation of pediatric cancer drug development: an industry perspective
PDF
Regulatory team development in post-merger integration: a survey of views from medical product companies
PDF
Use of electronic health record data for generating clinical evidence: a summary of medical device industry views
PDF
Validation master plans: progress of implementation within the pharmaceutical industry
Asset Metadata
Creator
Greco, Aimee
(author)
Core Title
Implementation of unique device identification in the medical device industry: a survey of the change management experience
School
School of Pharmacy
Degree
Doctor of Regulatory Science
Degree Program
Regulatory Science
Publication Date
02/10/2020
Defense Date
12/13/2019
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
change management,design control,medical devices,OAI-PMH Harvest,regulatory affairs
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Richmond, Frances (
committee chair
)
Creator Email
aimee.greco@yahoo.com,greco@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c89-268112
Unique identifier
UC11673242
Identifier
etd-GrecoAimee-8158.pdf (filename),usctheses-c89-268112 (legacy record id)
Legacy Identifier
etd-GrecoAimee-8158.pdf
Dmrecord
268112
Document Type
Dissertation
Rights
Greco, Aimee
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 a...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Tags
change management
design control
medical devices
regulatory affairs