University of Southern California Dissertations and Theses

Convergence of United States regulatory and reimbursement policies impacting patient access to humanitarian use devices (HUD)

(USC Thesis Other)

Convergence of United States regulatory and reimbursement policies impacting patient access to humanitarian use devices (HUD)

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

CONVERGENCE OF UNITED STATES REGULATORY AND REIMBURSEMENT
POLICIES IMPACTING PATIENT ACCESS TO HUMANITARIAN USE DEVICES
(HUD)
by
Chin-Wei Soo

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 2014

Copyright 2014 Chin-Wei Soo

2
Dedication

To my parents for their unconditional love and sacrifices.

This dissertation is also dedicated to those who suffer from rare diseases.
May your voice be heard.

3
Acknowledgements

First and foremost, I wish to thank my dissertation advisor, Dr. Frances
Richmond, for her steady support, encouragement, and guidance over the last three years
to bring this dissertation to fruition. I would like to thank my committee members
including Dr. Gerald Loeb, Dr. Yaniv Bar-Cohen, and Dr. Grant Lawless for their
feedback and insight to ensure the quality of this dissertation. I also wish to acknowledge
Winifred C. Wu, RPh, FRAPS and Jeanmarie Sales, JD for introducing me to the rare
disease field. Their dedications and passions in this field inspire me everyday. Finally, I
would like to thank the 43 regulatory and reimbursement informants who participated in
this research study by sharing their insights and experience. This dissertation would not
have been possible without the guidance and support of many people to whom I am
forever grateful.

4
Table of Contents
Dedication ........................................................................................................................... 2
Acknowledgements ............................................................................................................. 3
List of Tables ...................................................................................................................... 7
List of Figures ..................................................................................................................... 8
Abstract ............................................................................................................................... 9
Chapter 1: Overview of the Study .................................................................................... 10
1.1 Introduction ...................................................................................................................... 10
1.2 Statement of the Problem ................................................................................................. 15
1.3 Purpose of the Study ........................................................................................................ 16
1.4 Importance of the Research Study ................................................................................... 17
1.5 Delimitations, Limitations, and Assumptions .................................................................. 19
1.6 Organization of the Study ................................................................................................ 20
1.7 Definitions and Acronyms ............................................................................................... 20
1.7.1 Definitions ................................................................................................................ 20
1.7.2 Acronyms ................................................................................................................. 22
Chapter 2: Literature Review ............................................................................................ 24
2.1 Introduction ...................................................................................................................... 24
2.2 Legislative and Regulatory History ................................................................................. 25
2.2.1 Safe Medical Device Act of 1990 ............................................................................ 25
2.2.2 Food and Drug Administration Modernization Act of 1997 .................................... 27
2.2.3 The Pediatric Medical Device Safety and Improvement Act of 2007 ..................... 29
2.3 Two-Step HDE Regulatory Pathway ............................................................................... 31
2.3.1 Designating Humanitarian Use Devices .................................................................. 31

5

2.3.2 Approved Humanitarian Device Exemptions .......................................................... 35
2.4 Reimbursement Requirement ......................................................................................... 41
2.4.1 Statutory Boundaries on Medicare Coverage ........................................................... 43
2.5 Views and Activities of Key Stakeholders ..................................................................... 46
2.5.1 Food and Drug Administration (FDA) ..................................................................... 48
2.5.2 Congress ................................................................................................................... 50
2.5.3 Centers for Medicare and Medicaid Services (CMS) .............................................. 51
2.5.4 Patient Advocacy Groups ......................................................................................... 52
2.5.5 Physician and Healthcare Professional Groups ........................................................ 53
2.5.6 Industry .................................................................................................................... 55
2.6 Institute of Medicine (IOM) and Government Accountability Office (GAO) ................. 60
2.7 Systematic Evaluation of HDE Regulatory Pathway ....................................................... 62
2.8 Recent Regulatory and Reimbursement Initiatives ......................................................... 64
2.8.1 The Food and Drug Administration Safety and Innovation Act of 2012 ................. 64
2.8.2 FDA/CMS Collaboration and Parallel Review ........................................................ 65
2.8.3 Coverage with Evidence Development (CED) Program .......................................... 67
2.9 Conceptual Framework to Base Further Investigation .................................................... 70
2.9.1 Origins and General Characteristics of General System Theory ............................. 70
2.9.2 Applications of System Theory to Policy Analysis.................................................. 74
2.9.3 Proposed Research Conceptual Framework ............................................................. 80
2.10 Summary and Research Direction .................................................................................... 82
Chapter 3: Methodology ................................................................................................... 84
3.1 Introduction ...................................................................................................................... 84
3.1.1 Content Analyses ..................................................................................................... 85
3.1.2 Expert Review of Interview Tool ............................................................................. 89

6

3.1.3 Deployment and Analysis of Informant Interviews ................................................. 91
Chapter 4: Results ............................................................................................................. 95
4.1 Research Approach 1: Content Analysis of Public Dockets ............................................ 95
4.1.1 Content Analysis: Parallel Review of Medical Products ............................................. 95
4.1.2 Content Analysis: CMS 2012 Draft CED Guidance Document ................................ 101
4.2 Research Approach 2: Expert Review of Interview Questions...................................... 105
4.3 Research Approach 3: Key Informant Interviews ......................................................... 106
4.3.1 Profiles of Key Informants ..................................................................................... 106
4.3.2 Constituent of Policy-Making: Causation .............................................................. 111
4.3.3 Constituent of Policy-Making: Intervention .......................................................... 125
4.3.4 Constituent of Policy-Making: Evaluation ............................................................ 141
Chapter 5: Discussion ..................................................................................................... 145
5.1 Methodological Considerations ..................................................................................... 145
5.2 Considerations of the Results ........................................................................................ 150
5.3 Conclusion and Recommendation ................................................................................. 168
References ....................................................................................................................... 170
Appendix A: Analysis of Approved HDEs ..................................................................... 189
Appendix B: Regulatory Interview Guide ...................................................................... 196
Appendix C: Reimbursement Interview Guide .............................................................. 204
Appendix D: Examples of Invitation Emails to Informants ........................................... 212

7
List of Tables
Table 1: HUD regulations implemented as required by FDAMA of 1997 .................................... 29
Table 2: HUD requested vs. designated from FY2006 - FY2013 ................................................. 33
Table 3: Mean and range of progression of HUDs to HDEs (months) FY1997 – FY2014 .......... 40
Table 4: Public comment packages used for content analysis ...................................................... 87
Table 5: Public comment packages used for content analysis ...................................................... 88
Table 6: Description of experts’ professional background ........................................................... 90
Table 7: Grouping of questions in public docket using a modified triad of constituents ............. 95
Table 8: Themes and subthemes in the “Parallel Review of Medical Product” public docket ..... 98
Table 9: Results of the content analysis of CMS’ draft CED guidance document ..................... 102
Table 10: Impact of FDASIA 2012 HDE lifting of profitability restriction provision ............... 116
Table 11: Factors that contribute to failure of progression of HUD designation to HDE .......... 120
Table 12: Factors that contribute to lack of reimbursement of HUDs. ....................................... 123
Table 13: Extent to which pursue of HDEs impacted by FDA policies/requirements ............... 129
Table 14: What could FDA do to spur innovation in HUDs? ..................................................... 131
Table 15: Importance of CED characteristics ............................................................................. 133
Table 16: Importance of parallel review characteristics ............................................................. 140

8

List of Figures
Figure 1: HUD decision flowchart ................................................................................................. 32
Figure 2: HUDs explicitly indicated for use in pediatric patients for FY2001 - FY2011 ............. 34
Figure 3: HUD requests, designations, and approvals .................................................................. 36
Figure 4: Number of HUD requests, designations, approvals from FY1997- FY2014 ................ 38
Figure 5: Progression from HUD designation to HDE approval ................................................... 40
Figure 6: Key stakeholders of HUDs ............................................................................................ 48
Figure 7: Application of system thinking to three constituents of policy-making ........................ 77
Figure 8: Causal feedback loop ..................................................................................................... 78
Figure 9: Conceptual framework to study patient access of HUDs .............................................. 81
Figure 10: What is your current job title? ................................................................................... 107
Figure 11: What is/was the annual revenue of your place of business? ...................................... 108
Figure 12(a): How many years of medical device experience do you have? .............................. 109
Figure 12(b): How many years of HUD regulatory experience do you have? ........................... 109
Figure 13: What is your current job title? ................................................................................... 110
Figure 14(a): How many years of experience do you have in the medical device industry? ...... 111
Figure 14(b): How many years of experience do you have in HUD reimbursement? ................ 111
Figure 15: Legislative incentives to promote the development of HUDs ................................... 118
Figure 16: Can you tell more about your firm’s decision to pursue HDE pathway? .................. 130
Figure 17: How effectively the parallel review benefits patient access to HUDs? ..................... 139
Figure 18: Potential FDA evaluation methods ............................................................................. 142
Figure 19: Causal loop diagram governing patient access to HUDs .......................................... 157

9

Abstract
As regulatory requirements of the Food and Drug Administration (FDA) and the Centers
of Medicare and Medicaid Services (CMS) continue to increase, concerted effort between
these agencies is essential to ensuring patient access to Humanitarian Use Devices
(HUD). The purpose of this exploratory research was to explore the industry’s views on
policy issues related to each agency, and on the degree of cohesiveness and alignment
needed between the agencies in achieving a state of convergence to support patient access
of HUDs. A conceptual framework was developed that included a triad of constituents of
policy-making – causation, intervention, and evaluation – embedded in a framework
based on system theory. The study results suggested that HUD policy convergence
between FDA and CMS remains a challenge. Differing statutory mandates impede the
alignment between the agencies governing patient access to HUDs as a “system.” The
well-intentioned statutory regulatory requirements for HUDs, including the need for IRB
oversight and the lower requirements to demonstrate “probable benefit,” seem to have
had unintended reimbursement consequences that impede patient access to HUDs. From
the perspective of policy evaluation, the results suggested that methods or metrics to
evaluate the effectiveness of the “system” for HUD approval and marketing appears not
to be present. Collectively, these insights underline the need for statutory changes that
can enhance compatibility between the agencies, and thus encourage a coordinated,
responsive solution to the problems of patient access to HUDs.

10

Chapter 1: Overview of the Study
1.1 Introduction
In today’s society, most deaths are caused by a small number of prevalent
diseases, such as heart disease, cancer, and stroke. The fact that few families are
untouched by these particularly common diseases makes them particularly visible to the
American populace. Not so visible are the many other diseases that occur rarely but have
similarly destructive consequences and often-significant social costs. A disease is
considered “rare” if the disease affects fewer than 200,000 Americans at any given time,
according to the definition used by the Food and Drug Administration (FDA) and the
National Institutes of Health (NIH) (National Institutes of Health). Recent testimony
given by Diane Edquist Dorman, Vice President for Public Policy of National
Organization for Rare Diseases (NORD) indicated at the FDA Part 15 Public Hearing on
July 21, 2010 that between 6,000 and 7,000 diseases are considered to be rare in the
United States (National Organization for Rare Diseases, 2011). Although each individual
rare disease affects no more than 200,000 people, and some affect only a few hundred or
even a few dozen, rare diseases in the aggregate affect an estimated 25 million to 30
million Americans. Importantly, approximately 50% of people affected by rare disease
are children, according to NORD’s physician guides (National Organization for Rare
Diseases).
The fact that certain diseases occur so infrequently does not minimize the
importance that they have for the patients and their families, yet treatment options are
often limited. A recent editorial in the journal, Nature, noted that among patients

11
afflicted with any of the 350 most common rare diseases, 27% will not live to see their
first birthday (Anonymous, 2010). For this reason, rare diseases are also called orphan
diseases because they appear to have been abandoned by industry and funding agencies.
The low incidence of each particular disease poses a serious problem for the development
of treatments. It is very expensive to develop new devices and drugs; manufacturers of
medical products are unlikely to recover the costs that would be incurred by the necessary
research, preclinical trials, clinical trials, and quality system requirements normally
demanded under current laws and regulations. Such a problem has not gone unnoticed by
most developed societies. In an attempt to accelerate the availability of effective
treatments, regulatory relief of various forms has been considered or implemented to
reduce development times and costs, as well as provide incentives to industry for such
activities. In the United States, the Orphan Drug Act (ODA) and was enacted for
drugs/biologics in 1983, and the Safe Medical Device Act (SMDA) for medical devices
in 1990. The Safe Medical Device Act of 1990 provided for a Humanitarian Device
Exemption (HDE) regulatory pathway to encourage the discovery and use of devices that
are intended to benefit a particularly small number of patients. The law authorized the
FDA to develop regulations that would carry out the provisions of the Safe Medical
Device Act. To this end, the FDA issued a final rule on June 26, 1996 regarding
Humanitarian Use Device (HUD). As defined in 21 CFR 814.3(n), a HUD is a medical
device intended to benefit patients in the treatment or diagnosis of a disease or condition
that affects or is manifested in fewer than 4,000 individuals in the United States per year
(U.S. Food and Drug Administration, 2010b, 2011e). Generally speaking, the intent of
the HUD provisions was to provide incentives for the development of devices intended

12
for treatment or diagnosis, in small patient populations where otherwise a device
manufacturer’s research and development costs could exceed market returns (U.S. Food
and Drug Administration, 2010c).
The HDE regulatory pathway for commercializing a HUD is a two-step process
that consists first of obtaining a HUD designation from the Office of Orphan Product
Development (OOPD) of the FDA. For the HUD designation, the applicant must provide
authoritative references to demonstrate that the device under consideration is intended for
one of the following two conditions: (a) a rare disease, which affects or is manifested in
fewer than 4,000 individuals in the United States per year, and (b) a specifically
identified subset of a non-rare disease, which affects or is manifested in fewer than 4,000
individuals in the United States per year. In the recently published HUD designation
guidance document in year 2013, OOPD describes a “medically plausible subset” as the
“subset of individuals with a non-rare disease or condition on whom use of a device is
appropriate, where use of the device on remaining individuals with that disease or
condition would be inappropriate given some intrinsic feature of the device” (U.S. Food
and Drug Administration, 2013a). For example, an intrinsic feature of an implantable
device may be its serious adverse event profile associated with the implantable process.
The serious adverse event profile may preclude its use on patients who have not failed
less risky medical treatment such as drug therapy. Therefore, the “medically plausible
subset” of the disease or condition may be patients who have previously failed drug
therapy.
In the second step, a HDE application is submitted to gain marketing approval. A
HDE is a marketing application that is similar to a premarket approval (PMA)

13
application, but exempt from the effectiveness requirement. Specifically, a HDE is not
required to contain the results of valid scientific evidence demonstrating that the device is
effective for its intended purpose, but the applicant must provide sufficient information
for the FDA to determine that the device does not pose a significant risk of illness or
injury to patients, and that the probable benefit to health outweighs the risk of using the
device (U.S. Food and Drug Administration, 2010b). However, there is no regulatory
definition for the term “probable benefit.” The FDA has yet to specify the type or level
of evidence required to fulfill “probable benefit” requirement. In most cases, the
regulatory standards for safety and probable benefit are met through combination of
bench testing, non-clinical study, in-vitro animal testing, and human clinical study. The
FDA encourages the applicant to collaborate with the review center to develop an
appropriate level of evidence to substantiate “probable benefit” for the device. Moreover,
the applicant must demonstrate that no comparable devices are available to treat or
diagnose the disease or condition other than another device approved under HDE
pathway or a device approved under an Investigational Device Exemption (IDE).
Importantly, the FDA will not approve an HDE application once a comparable device
with the same indication for use is marketed through either the Premarket Approval
(PMA) or the Premarket Notification (510(K)) process (U.S. Food and Drug
Administration, 2010b).
Regulatory approvals of HDE are bound by several key statutory conditions.
First, HUDs can only be used in a facility after an Institutional Review Board (IRB) has
approved their use in that facility, except in certain emergencies (U.S. Food and Drug
Administration, 2010b). Second, the labeling of a HUD must clearly identify the product

14
as an HUD, and must state that the effectiveness for the approved indication has not been
demonstrated. Third, according to the Food and Drug Safety and Innovation Act
(FDASIA) enacted on July 9, 2012, an approved HDE indicated for condition that does
not occur (or only rarely occurs) in pediatric patients can be sold for a profit. However,
the Annual Distribution Number (ADN) is restricted to the number of devices needed to
treat, diagnose or cure 4,000 individuals in the U.S. during any calendar year, unless the
FDA grants a request to modify the ADN. HDEs approved prior to enactment of
FDASIA 2012 may qualify for the profitability exemption (Arnold and Porter LLP,
2012). Prior to enactment of the FDASIA, HUDs could not be sold for a profit, if the
devices were to sell for more than $250, with one exception – an exemption might be
granted for HUDs indicated for pediatric populations or a pediatric subpopulation
approved on or after September 27, 2007.
Concerted efforts between the FDA and the Centers of Medicare and Medicaid
Services (CMS) are essential to ensuring patient access of HUD technologies. The FDA
approval of a HDE is only the first hurdle. Equally important, if not more important, is
that the HUD can be reimbursed by a patient’s health care coverage. According to statute
42 United States Code (U.S.C) §1395y, Medicare adheres to the “reasonable and
necessary” determination to limit payment to items or services for the diagnosis and
treatment of illness or injury or to improve the functioning of a malformed body member.
The interpretation of “reasonable and necessary” is that a service or item must be safe
and effective, medically necessary and appropriate, and not experimental in order to
qualify for reimbursement (Institute of Medicine, 2011). Although a HUD that has been
approved via the HDE regulatory pathway is no longer “experimental” according to the

15
FDA, CMS often views the level of clinical evidence as being insufficient to meet the
evidentiary standards of “reasonable and necessary” determination.
1.2 Statement of the Problem
The HDE regulatory pathway has been established to assist the treatment of rare
diseases, but it is still considered to be a challenging and burdensome approach for
bringing therapies and devices to the market. Since October 1996, the FDA received
more than 300 Humanitarian Use Device requests, of which 200 were designated as
HUDs (U.S. Food and Drug Administration, 2013c). 58 of these have been approved but
one ultimately pursued a PMA for broader indications or other business related reasons,
and five of the HDEs were withdrawn. Most importantly, the rate of HDE approval
appears to have decreased over time since the commencement of HDE regulatory
pathway in 1996. Thus, while the intent of the HDE regulatory pathway is to ensure that
patients with rare diseases are provided access to the medical devices, it is often not a
pathway sought by device manufacturers. The reasons why this pathway is not used
more frequently is not, however, well understood. It may relate to multiple factors
including financial considerations, technological challenges presented by the diseases, or
regulatory impediments that affect timeline and cost. More specifically, the HDE
“probable benefit” regulatory requirement is not defined in the statute and regulations;
therefore the type or level of evidence needed to meet such a requirement is often not
well understood.
As noted previously, the requirement that the device meet the stringent requirements of
CMS may influence the ability of such devices to gain traction in the market. Because
approval of an HDE by the FDA does not require clinical evidence to demonstrate

16
effectiveness, the reimbursement process with CMS can become particularly burdensome
and lengthy. This problem could potentially lead to delayed access of HUDs to patients
who need them. Further, the reimbursement decisions can at times appear arbitrary, non-
transparent, and are typically dealt with based on a case-by-case basis. Some private
health plans have authorized coverage for specific uses of one HUD and rejected it for
others HUDs (Institute of Medicine, 2011). For example, Cigna and Harvard Pilgrim
Health Care covers Enterra
TM
gastric electrical stimulation (HDE # H990014) as
medically necessary when provided according to the HDE approved indications for the
treatment of chronic, intractable nausea and vomiting secondary to gastroparesis of
diabetic or idiopathic etiology (Cigna, 2011b; Harvard Pilgrim HealthCare, 2010).
Conversely, Health Plan of Nevada deems the Ascension Proximal Interphalangeal (HDE
# H010005) implant as not medically necessary due to insufficient effectiveness
evidence, and does not provide coverage for the device (Health Plan of Nevada, 2011).
In addition, it is not uncommon for a treating physician to seek determination of medical
necessity through a prior-authorization process from a health care plan, as in the case of
the ThereSphere® (HDE # H980006) coverage policy from the UPMC Health Plan
(UPMC Health Plan, 2011).
1.3 Purpose of the Study
As regulatory requirements of the FDA and the CMS continue to rise, concerted
efforts between these agencies are essential to ensure patient access of HUDs in fulfilling
unmet medical needs. At this point, these agencies appear to be operating in a
disconnected manner with known and unknown barriers embedded within their respective
policy frameworks. The purpose of this exploratory research using mixed methods of

17
content analysis and informant interviews was to study the impact of FDA and CMS
policies to the access of HUDs. More specifically, I conducted this exploratory study to
(a) query the industry about the specifics of the FDA and CMS policies themselves and
whether potential influences exist that might modify, assist or impede the development of
effective policies, (b) explore the industry’s views with respect to the effectiveness of the
current strategies for policy implementation in each agency as well as whether changes
are needed to improve the interconnectedness of the FDA and CMS as they implement
policies, and (c) explore the nature and level of awareness regarding current evaluative
methods, the degree to which these evaluation methods are sufficient, and the way in
which the two agencies share evaluations to understand how together their actions might
improve policy design and policy implementation.
1.4 Importance of the Research Study
Although the number of people affected by a particular rare disease is relatively
small – ranging from tens to thousands – the reality is that rare diseases are not rare, at
least in the aggregate (Institute of Medicine, 2011). Patients suffering from rare diseases
are at a disadvantage, particularly when there is no effective medical treatment.
Therefore significant unmet medical needs exist and these unmet needs result in
increased morbidity and mortality for affected patients.
In the FDA draft document titled “Strategic Priorities 2011- 2015: Responding to
the Public Health Challenges of the 21
st
Century,” the agency acknowledged a serious
need to expand the options for special populations. Specifically, the agency stated that
market incentives might be insufficient to drive the investment needed to develop
medical products to prevent, diagnose, and treat rare disease conditions, and the agency

18
was taking active steps to meet the needs of these rare disease patients (U.S. Food and
Drug Administration, 2011f). To support the FDA’s initiative, it is hoped that this
exploratory study would increase the awareness of the magnitude and extent of potential
barriers to the HDE pathway, which could be valuable to policy makers to assure that the
solutions proposed match the problem to be solved. In addition, a better understanding of
the potential impediments may ultimately improve the regulatory environment for the
HDE pathway, either through changes in regulation or its associated interpretation, so-
called “soft law.” This research study revealed some of these potential impediments.
From the patient-access perspective, it is a significant concern if the current
regulatory and reimbursement environment places impediments in the path of needed
treatments. The problem can only be addressed if the root cause or causes of the
regulatory and reimbursement environment are understood individually, and importantly
interconnections between these environments are addressed holistically. This research
study contributed to an understanding of the causations and the interconnections of the
regulatory and reimbursement environments to facilitate the development of innovative
medical devices so that patient advocacy groups and consumers can better recognize
where efforts to influence policy may increase access.
From the point of view of medical device industry, this study provides specific
insight into elements that affect the pursuit of an HDE as a strategic regulatory pathway.
Specifically, the research study revealed potential factors that may modify, assist or
impede the development of effective regulatory and reimbursement policy to ensure
patient access to HUDs. This knowledge may allow the industry to plan proactively and
to avoid investments in areas where problems may prove insurmountable. An

19
understanding of these barriers allows the industry to implement effective risk mitigation
early during the product development phases.
1.5 Delimitations, Limitations, and Assumptions
This research study, like any other research, has its delimitations, limitations, and
assumptions. This study was delimited to the evaluation of HUDs since the inception of
the Safe Medical Device Act of 1990, which promulgated the HDE regulatory pathway.
Specifically, the research study examined HUD designations and HDEs in the following
categories – approved HDEs for which the devices currently remain in commercial
distribution in the United States, approved HDEs for which the devices are no longer
sold, and withdrawn HDEs.
There are several limitations pertaining to the research study. First, the HUD
populations are inherently small and the number of individuals experienced with the HDE
pathway is correspondingly low in number. Accordingly, it was challenging to involve a
representative sample of industry regulatory and reimbursement participants in the
research study. Second, unlike orphan drug designations in which information about the
holders of the designations are published, the FDA does not publish the identities of
HUD holders. This was a major limitation, which was mitigated by performing extensive
research in the public domain or filed a Freedom of Information (FOI) request with the
FDA. On September 21, 2012, FOI Services (www.foiservices.com) filed a FOI request
on my behalf to seek information with regard to HUD designation owners.
Unfortunately, to date (January 2014) the FDA has yet to respond to the FOI request
(Soo, personal observation).

20
1.6 Organization of the Study
This study was organized into five chapters. Chapter 1 describes the research
problems, questions, and lays out the reasoning behind them. Chapter 2 reviews what has
already been written in the field of HUDs, specifically the legislative and regulatory
development of the HDE regulatory pathway, current landscape of HUDs, perspectives
and positions of various stakeholders, provisions of Medicare reimbursement, and an
analysis of current literature. Next, Chapter 3 defines the proposed research
methodology, which includes content analysis and informant interviews. The results of
the research are presented in Chapter 4, including specific observations and findings
related to the research questions. Further, Chapter 5 provides a discussion of the overall
results, implications, conclusions, and considerations for further research.
1.7 Definitions and Acronyms
1.7.1 Definitions
Annual Distribution Number (ADN) – An ADN is calculated by multiplying the number
of devices reasonably necessary to treat, diagnose, or cure an individual by 4,000
(Covington and Burlington LLP, 2012).
Humanitarian Use Device (HUD) – A HUD is a medical device intended to benefit
patients in the treatment or diagnosis of a disease or condition that affects or is
manifested in fewer than 4,000 individuals in the United States per year (U.S. Food and
Drug Administration, 2011a).

21
Humanitarian Device Exemption (HDE) - A HDE is a marketing application that is
similar to a premarket approval application, but is exempt from the effectiveness
requirements of sections 514 and 515 of the Food, Drug, and Cosmetic Act (U.S. Food
and Drug Administration, 2011a).
Orphan Subset – A subset of individuals with a non-rare disease or condition on whom
use of a device is appropriate, where use of the device on remaining individuals with that
disease or condition would be inappropriate given some intrinsic feature of the device
(U.S. Food and Drug Administration, 2013a).
Progression of HUD to HDE – Time interval from HUD designation and HDE approval.
Premarket Approval (PMA) – FDA regulatory review to evaluate the safety and
effectiveness of Class III medical devices (U.S. Food and Drug Administration, 2012c).
Investigational Device Exemption (IDE) – An application submitted to the FDA seeking
regulatory approval to conduct clinical investigation in order to collect safety and
effectiveness data (U.S. Food and Drug Administration, 2009b).
Institutional Review Board (IRB) – An organizational group that has been formally
designed to review, approve the initiation of, and to conduct periodic review of
biomedical research involving human subjects (U.S. Food and Drug Administration,
2009b).

22
Informed Consent – A process in which a human subject learns key facts about a
biomedical research, including potential risks and benefits, before making participation
decision in the research (U.S. Food and Drug Administration, 2011d).
Current Procedural Terminology (CPT) – The CPT is the most widely accepted medical
nomenclature used to report medical procedures and services under public and private
health insurance programs (American Medical Association).
Dignostic Related Group (DRG) - The DRG is a scheme that provides a means for
classifying patients into clinically cohesive groups that demonstrate similar patterns of
costs (length of stay and resource consumption) incurred by the hospitals (Hallas, 2012).
Institute of Medicine (IOM) – An independent, nonprofit organization that operates
outside government to provide unbiased and authoritative advice to decision makers and
the public (Institute of Medicine).
Government Accountability Office (GAO) – An independent, nonpartisan agency that
supports congress to meet its constitutional responsibilities and ensure performance and
accountability of the federal government for the benefit of the American people (U.S.
Government Accountability Office, 2012)
1.7.2 Acronyms
AdvaMed Advanced Medical Technology Association
AMA American Medical Association
CED Coverage with Evidence Development
CDRH Center for Devices and Radiological Health

23
CPT Current Procedural Terminology
FDCA Food, Drug, and Cosmetic Act
FDAMA Food and Drug Administration Modernization Act of 1997
FDASIA Food and Drug Administration Safety and Innovation Act of 2012
FOI Free of Information
DRG Diagnosis Related Group
HHS Department of Health and Human Services
HELP U.S. Senate Committee on Health, Education, Labor, and Pensions
LMRP Local Medical Review Policies
NORD National Organization for Rare Diseases
NIH National Institute of Health
NCD National Coverage Determination
OOPD Office of Orphan Product Development
ODE Office of Device Evaluation
PMDSIA Pediatric Medical Device Safety and Improvement Act of 2007
SMDA Safe Medical Device Act of 1990

24
Chapter 2: Literature Review
2.1 Introduction
Humanitarian Use Devices (HUDs) were introduced in the late 1990s, and thus
have a relatively recent history compared to many of the regulated medical product
classes. In the following chapter, I will first introduce the legislative and regulatory
history of the Humanitarian Device Exemption (HDE) regulatory pathway. Using the
publicly available information on HDEs, I will then evaluate the current HDE landscape
and the overall trends regarding approvals through this pathway. Second, I will discuss
the history and trends of reimbursement for HUDs. Coverage policies of HUDs from
various major insurance providers are also presented to give a broader perspective of
current coverage decisions and their requirements. I will examine several recent HDE-
related provisions of the FDA Innovation and Safety Act of (FDASIA) 2012, particularly
the new rules that modify the profitability restriction for HDEs, and how these changes
are predicted to affect regulation and reimbursement. Third, I will describe the
perspectives of various stakeholders – medical device industry, trade association, patient
advocacy groups, healthcare professional groups, Food and Drug Administration (FDA),
Centers of Medicare and Medicaid Services (CMS), and Congress – to the extent that
they are currently understood. Finally, I will describe the conceptual framework that had
been identified to guide the systematic analysis of interconnectedness between FDA and
CMS policy in supporting patient access of HUDs.

25
2.2 Legislative and Regulatory History
2.2.1 Safe Medical Device Act of 1990
HUD regulations were introduced in the late 1990s, and thus have a relatively
recent history compared to many of the regulated medical device classes. The foundation
for HUD regulations was laid by the Safe Medical Device Act (SMDA) of 1990, which
was enacted on November 28, 1990 by President George Bush. The Act was intended to
enhance the Medical Device Amendments of 1976, the principal federal law regulating
the safety and effectiveness of medical devices in the U.S. (Raubicheck & Weigel, 1976).
One of its enhancements was the addition of section 520(m) to the Food, Drug, and
Cosmetic Act (FDCA), which essentially defined HUDs as a distinct device category
with an accelerated approval path. According to this section of the FDCA, the following
conditions must be met for a device to be considered as HUD: (a) it must be intended to
diagnose or treat a disease or condition affecting 4,000 or fewer patients in the U.S., (b) it
has not been marketed under another approval route, (c) it must be the only device of its
type available, and (d) it must not expose patients to an unreasonable or significant risk of
illness or injury from its use (U.S. Food and Drug Administration, 2010d). In addition,
section 520(m) stipulated that a HUD cannot be sold for a profit and the devices can only
be used in a facility after an Institute Review Board (IRB) has approved their use in that
facility. Furthermore, a sponsor of HDE must obtain approval from the FDA every 18
months.
The provision of the SMDA required that the FDA should implement section
520(m) of the FDCA within one year of its enactment. However, the implementation
became a lengthy and bureaucratic process that extended well beyond the designated

26
timeframe. It took the FDA approximately six years to implement a regulatory
framework for HUDs. The development of regulations was problematic because of the
potential safety issues that were associated with the abbreviated approval path. The only
way that the path for a risky device can be significantly abbreviated is by reducing the
volume of clinical trials on which decisions for commercialization are made. The
essential difference between a HUD and a class III device is the willingness to forego
efficacy studies, and rely only on sufficient clinical evidence to support basic safety
concerns. Because HUDs would likely be approved with so little clinical experience,
regulators feared that problems might be revealed unless additional mechanisms for
oversight were implemented. Thus, they took the time to consider different options for
the approval of such products. The FDA initially proposed to regulate HDEs as
premarket approval applications (PMAs) that did not require review of device
effectiveness. The FDA ultimately abandoned the approach because of a number of HUD
specific requirements, such as need for Institute Review Board (IRB) review and
approval and patient informed consent that would not apply to PMA specific regulations
(Sherman, Wentworth, & Whitaker, 2002). The FDA then published a proposed rule on
December 21, 1992, with the intent to amend the Investigational Device Exemption
(IDE) regulations (21 CFR Part 812) as a route to implement the HUD regulatory
provision, instead of creating a new premarket approval regulation. The rationale for this
revision was based on the fact that existing regulatory requirements related to IRB and
patient consent could be adapted readily to HUDs. However, the industry opposed the
inclusion of HDE provisions in 21 CFR Part 812, the section of the CFR that dealt with
investigational device applications. The industry argued that the proposed approach

27
would contradict the intent behind section 520(m) of the FDCA, which some believed
was to facilitate marketing of HUDs rather than establishing a path for clinical
investigations involving HUDs (U.S. Food and Drug Administration, 2010d).
Responding to these comments, the FDA agreed to create a new Subpart H under 21 CFR
Part 814 specifically addressing HUDs. It then published a final rule (61 FR 33233) to
introduce this new path on June 26, 1996. In this final rule, the FDA established the two-
step HDE approval process that consisted of first obtaining a HUD designation from the
Office of Orphan Product Development (OOPD) and then submitting a HDE to the
Center for Devices and Radiological Health (CDRH) of the FDA. Further, the same final
rule stipulated that an HDE is a marketing application similar to a premarket approval
application, but it would be exempted from the effectiveness requirements of sections
514 and 515 of the FDCA. The FDA believed that the exemption would reduce the cost
and burden for sponsors to bring HUDs to market. However, the regulation introduced a
major restriction, that HDE holders were not permitted to make a profit; they could
recoup only the costs of research and development, fabrication and distribution of the
HUDs (Bleicher, 2009).
2.2.2 Food and Drug Administration Modernization Act of 1997
Additional requirements and amendments for the HUD pathway were added on
June 26, 1996, when President William Clinton signed into law the Food and Drug
Administration Modernization Act (public law 105-115) on November 21, 1997 (U.S.
Food and Drug Administration, 2009a). Amongst its requirements and stipulations were
the following: (a) the FDA must issue an order approving or denying an HUD application
75 days after receiving it, (b) a physician could use an HUD for patient treatment prior to

28
local IRB approval if such approval could not be obtained in time to prevent serious harm
or death. However, the physician must notify the local IRB as soon as possible, (c) no
mandatory 18-month renewal of HDE approval regulatory would be required. Instead, the
FDA could require a sponsor to demonstrate continued compliance with the requirements
of section 520(m), if such a demonstration were judged to be necessary to protect public
health, or re-affirm the eligibility for exemption, and (d) the FDA was authorized to
suspend or withdraw an HDE approval but would then be required to provide an
opportunity for an informal hearing prior to taking such decision (Rowberg et al., 1998).
On April 17, 1998, the FDA published a direct final rule (63 FR19185) and
companion rule (63 FR 19196) to implement the changes initiated by FDAMA. The
FDA indicated that if the agency received critical comments regarding the published rule,
the agency would withdraw the rule. A number of critical comments were received,
subsequently, the FDA decided to withdraw these rules. The FDA published a new final
rule (63 FR 59217) on November 3, 1998 to implement changes required by FDAMA, as
described further in Table 1 below (Sherman et al., 2002).

29
Table 1: HUD regulations implemented as required by FDAMA of 1997
Regulation Provisions
21 CFR Part
814.100
FDA has the authority to require an HDE sponsor to
demonstrate continued compliance of section 520(m), if FDA
believes that such demonstration is necessary to protect public
health, or re-affirm the eligibility for exemption.
21 CFR Part
814.104(b)(5)
An HDE sponsor, who is charging more than $250 per HUD,
must submit an attestation by a responsible individual of an
organization (in lieu of a report by an independent certified
account), verifying that the amount charged does not exceed the
cost of research, development, fabrication, and distribution. No
report is required for HUDs sold for less than $250.
21 CFR Part
814.106, 814.108,
814.112, and
814.114
FDA must issue an order approving or denying an application
75 days after receiving it.
63 FR 19197
An 18-month term for HUD marketing and a five-year subset
provision were repealed by FDAMA.
21 CFR Part
814.124(a)
A physician can administer an HUD prior to local IRB approval
if such approval cannot be obtained in time to prevent serious
harm or death. However, the physician must notify the local
IRB within 5 days.

2.2.3 The Pediatric Medical Device Safety and Improvement Act of 2007
Prior to 2007, the legislative and regulatory development of HUDs was
nonspecific with regard to subclasses of humanitarian devices, but most such devices
were used mostly used in adult populations. The needs of children with rare diseases
were not a primary focus of the regulators and industry. However, it became clear over
time that pediatric devices for small populations were facing greater challenges and few
were coming to market. According to Bleicher, the most common barrier to the

30
introduction of HUDs for pediatric populations was the profitability restriction introduced
by previous HUD regulations. Bleicher stated:
Pediatric populations are small, diverse, and dispersed in comparison to the adult
market. Often, the costs of research and development, manufacturing, and
distribution cannot be recouped, given the size of the patient base. In addition,
problems identifying needs and technological challenges create financial as well
as practical barriers to pediatric device development. (Bleicher, 2009)
Recognizing this unmet medical need for pediatric populations, President George
W. Bush signed into law the Pediatric Medical Device Safety and Improvement Act
(PMDSIA) on September 27, 2007 (American Academy of Pediatrics, 2008). PMDSIA
removed the profitability restriction for HUDs indicated for pediatric populations, but
replaced it with another requirement. Sponsors were allowed to make a profit from
pediatric HUD sales but the number of devices distributed in any calendar year was not
allowed to exceed the Annual Distribution Number (ADN). The FDA calculated the
ADN for each HUD, but in no case the ADN would be more than 4,000. The
profitability restriction was lifted in order to provide the medical device industry with an
incentive to address unmet pediatric medical needs (Bernad, 2009). In practice, only
HUDs indicated for pediatric populations or subpopulations approved on or after
September 27, 2007 could take advantage of this new opportunity, as stipulated in an
FDA guidance document initially issued on July 18, 2006 (U.S. Food and Drug
Administration, 2010b). Only three HDEs with specific pediatric indications have been
approved following the enactment of PMDSIA in 2007.

31
2.3 Two-Step HDE Regulatory Pathway
2.3.1 Designating Humanitarian Use Devices
The first step in the process to gain marketing approval for an HDE is for the
device to be designated as an HUD. This step requires that a request for HUD
designation be directed to the OOPD, as set forth in 21 CFR 814.102. Within 45 days,
OOPD must make its designation decision through a specified process, as illustrated in
Figure 1 (U.S. Food and Drug Administration, 2011a).

32
Figure 1: HUD decision flowchart

Note: Adapted from “Guidance for Industry and Food and Drug Administration Staff: Humanitarian Use
Device (HUD) Designations,” by Food and Drug Administration (FDA), 2013, p. 14. Copyright 2011 by
FDA.

33
Unlike orphan drug designations in which information about the holders of the
designations are published, the FDA does not publish the identities of HUD designation
holders. Thus, only the aggregate number of HUD requests and designations are available
in the public domain. Through extensive search, Table 2 below summarizes the HUDs
that were requested and designated for recent years that published in the OOPD annual
budgetary reports (U.S. Food and Drug Administration, 2009c, 2012b, 2013b). Further,
the HUD information explicitly indicated for use in pediatric patients from FY2001
through FY2011 is recently published by the Government of Accountability Office
(GAO) (U.S. Government Accountability Office, 2011), as illustrated in Figure 2.

Table 2: HUD requested vs. designated from FY2006 - FY2013
FY2006 FY2007 FY2008 FY2009 FY2010 FY2011 FY2012 FY2013*
Requested 14 17 13 22 24 21 23 25
Requested
– Pediatric
2 2 4 4 9 7 NA NA
Designated 8 6 9 11 14 19 14 12
Designated
– Pediatric
2 2 4 4 6 8 NA NA
* OOPD’s estimate
Table 2 shows that the number of requested and designated HUDs has increased
from FY2006 through FY2011 for both non-pediatric and pediatric indications. This
upward trend could be correlated to the incentives provided by PMDSIA of 2007 (as a
part of FDAAA of 2007). Further, Figure 2 illustrates the HUDs information explicitly
indicated for pediatric indications from FY 2001 through FY2011.

34

Figure 2: HUDs explicitly indicated for use in pediatric patients for FY2001 - FY2011
HUD requests, HUD designations, and HDE approvals explicitly indicated for use in
pediatric patients for FY2001 - FY2011

Note: Adapted from “Report to Congressional Committees: Pediatric Medical Devices: Provisions Support
Development, but Better Data Needed for Required Reporting,” by United States Government Accountability Office
(GAO), GAO-12-225. p. 31. Copyright 2011 by the GAO.

According to GAO, the number of HUD requests and designations intended for
pediatric indications has increased since enactment of FDAAA of 2007, and HDE
approval for pediatric populations may be predicted to increase in the future (U.S.
Government Accountability Office, 2011). Because HUD designation is only the first
step of the HDE process, it is not known yet if such a prediction has merit, or whether
other factors could impede or encourage HUD progression to HDE.

35
2.3.2 Approved Humanitarian Device Exemptions
The second step in the HDE process is to submit a Humanitarian Device
Exemption (HDE) application to the appropriate device branch in the Office of Device
Evaluation (ODE) in order to gain marketing approval. An HDE applicant must include
a copy or reference to its HUD designation letter with the HDE application, as indicated
in 21 CFR 814.104(b)(1). This allows the ODE branch to determine if the device meets
the statutory criteria for a HUD set forth in section 520(m)(2) of the FDCA (U.S. Food
and Drug Administration, 2010b). The HDE regulation stipulates that the application
does not need to contain the results of scientifically valid clinical investigations
demonstrating that the device is effective for its intended purpose, but the applicant must
provide sufficient information for the FDA to determine that the device does not pose a
significant risk of illness or injury to patients, and that the probable benefit to health
outweighs the risk of using the device (U.S. Food and Drug Administration, 2010b).
Since the enactment of HDE regulations in 1996, the FDA has received over 300
HUD requests, of which 200 (67%) were designated as HUDs. Only fifty-eight (29%) of
these designations subsequently received marketing approval. One previously approved
HDE ultimately pursued a traditional premarket approval (PMA) route for broader
indications or other business related reasons, and five of the HDEs were withdrawn
(Figure 3).

36
Figure 3: HUD requests, designations, and approvals

In the article titled “Humanitarian Use Devices/Humanitarian Device Exemptions
in Cardiovascular Medicine,” Kaplan and his colleagues analyzed all approved HDEs
according to certain predefined criteria – types of device (vascular, orthopedic, urologic),
implant vs. non-implant, and pediatric vs. non-pediatric indication (Kaplan et al., 2005).
To build on that work, I have updated the analyses to include all HDEs approved to date,
and have additionally included the evaluation of the time interval from HUD designations
to HDE approvals. The results of the further analysis are illustrated in Appendix A.
According to the analyses of data in Appendix A, about 71% of the approved
HDEs are implantable devices, and thus could be characterized as high-risk devices.
Nearly 33% of the approved HDEs, the largest single group of HDEs, are cardiovascular
devices. In contrast, ophthalmic devices constitute the lowest percentage (3%) of all of
the categories of devices in the approved group of HDEs. Nearly 24% of all approved

37
HDEs are indicated for pediatric populations. Perhaps surprisingly, only three HDEs
with pediatric indications were approved following the enactment of PMDSIA in 2007,
when the profitability restriction was lifted for pediatric indications. The chart shown in
Figure 4 was constructed using the approved HDE information published on the FDA’s
website (U.S. Food and Drug Administration, 2012a) and the limited information on
designated HUDs discussed previously. From the chart, it appears that (1) the number of
HUD requests and designations follow the same trend, and (b) the number of HDE
approvals has been staying stable or trending downward while the number of HUD
designations has been increasing in recent fiscal years. It is noted that FDA’s fiscal year
covers October 1
st
of current calendar year through September 30
th
of the following
calendar year.

38
Figure 4: Number of HUD requests, designations, approvals from FY1997- FY2014

To understand the nature of the HDE landscape, another key indicator worth
evaluating is the time interval from the HUD designation to HDE approval. I have
termed this time interval as “progression of HUD to HDE.” For each approved HDE, the
“Summary of Safety and Probable Benefit” was reviewed to obtain the HUD designation
date. The progression from the HUD to HDE was calculated for each approved HDE.
The results are illustrated in Table 3 and Figure 5. From the analysis, it appears that (a)
the mean progression of HUD to HDE has been increasing, and importantly (b) the
variability of the progression of HUD to HDE time is significantly higher in recent fiscal
years, notably in 2008, 2011, and 2012. Various factors contributing to the progression
of HUD to HDE were assessed in this study. To the best of my knowledge, there is no

39
literature published to date that examines the progression of HUD to HDE and explains
the contributory factors for the observed trends.

40
Table 3: Mean and range of progression of HUDs to HDEs (months) FY1997 – FY2014
FY 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14
Mean 4 10 10 12 27 19 19 25 26 34 24 69 49 31 40 83 47 47
Range 0 14 6 18 37 38 1 79 12 8 38 138 47 0 79 102 0 0

Figure 5: Progression from HUD designation to HDE approval

41

2.4 Reimbursement Requirement
FDA approval of a HDE is only the first hurdle faced when trying to
commercialize a HUD in a way that can sustain it in the market over time. At least as
important is to ensure that patients can pay for the devices, and this mostly happens
through a patient’s health care insurance program. In the U.S., reimbursement of devices
is complicated because individual patients may receive insurance coverage through
different private and public insurance plans. One of the largest sources of such insurance
is the Centers of Medicare and Medicaid (CMS), which covers the device needs of more
than 100 million individuals, typically in low income or elderly subpopulations. Because
of its size, CMS has the resources to conduct detailed technology assessments before
making decisions on key elements important to the reimbursement process. The reach of
CMS decisions is expanded further because many smaller insurance companies without
the resources to conduct such thorough analyses look to the CMS decisions as a
benchmark when they define their own coverage decisions.
The procedure to establish a coverage decision through CMS is a complex and
time-consuming process consisting of the following three key elements – coverage,
coding, and payment. All three must be in place to ensure patient access of devices. A
coverage policy determination for Medicare support is typically made by a decentralized
process. In this process, the policy for coverage is determined by local commercial
payers that are contracted by Medicare to manage the coverage decisions locally. These
local coverage policies, also known as Local Medical Review Policies (LMRP), account
for more than 90% of the coverage policies. This decentralized process can result in

42
marked inconsistencies in coverage from one constituency to another. In time, and
subject to the receipt of a submission from a device company, CMS will consider
whether the device warrants a National Coverage Determination (NCD), a national policy
that establishes a coverage decision for a product or service for all Medicare beneficiaries
meeting the eligibility criteria. A positive decision to support reimbursement for a device
or service has wide impact because payers in all regions must adhere to this policy (Smith
& Henderson, 2003). A negative decision is equally far-reaching because it results in the
denial of coverage for such devices by all local payers. The coverage decision must also
be paralleled by a decision on pricing that can also depend on many factors including the
presence of other products of a similar nature on the market.
Once a coverage decision has been reached, a second process must be carried out
by which the device or service is identified uniquely by a code that allows a healthcare
institution to bill easily for that product or service. The Current Procedural Terminology
(CPT) coding system developed by the American Medical Association (AMA) is the
most commonly used system (American Medical Association). Often, industry
collaborates with CMS to ensure that proper coding system is established. Without a
coding decision, the billing organization has a tedious manual process to obtain
reimbursement. This can inhibit access to the product in many institutions.
After a product has codes and is covered, the product may be eligible for
payment. The contractual terms between the payers and healthcare providers determine
the payment for the product. The payment methodology depends on where the product is
used and by whom, and on the kind of treatment or service that is provided. There are
different payment methodologies for the hospitals and physicians. Examples of hospital

43
payment methodology include, but are not limited to, (a) case rate – a payment is arrange
to cover a specific procedure, technology, or diagnosis, and (b) fee schedule – a flat
payment for the patient’s admission regardless of resources used or length of stay. This
payment scheme is also called Diagnostic Resource Group (DRG). Similarly, physician
payment methodologies include, but are not limited to, (a) case rate – the physician has
contracted a fixed fee for a specific procedure, and (b) capitation – the physician is paid a
fixed amount per enrollee, not per service, per month to cover the costs of care (Hallas,
2012).
2.4.1 Statutory Boundaries on Medicare Coverage
The rules that underlie reimbursement approval differ from those underlying
marketing approval by the FDA. Bounded by section 1862(a)(1)(A) of the Social
Security Act, Medicare covers medical products or services that are included in a
Medicare benefit category, and are not statutorily excluded or limited by other statutory
provisions. The covered products must also be deemed “reasonable and necessary for the
diagnosis or treatment of illness or injury or to improve the functioning of a malformed
body member” (U.S. Centers for Medicare and Medicaid Services, 2010). However, the
term “reasonable and necessary” has not been statutorily defined. CMS has generally
interpreted that the reasonable and necessary provision can only be met by products and
services that have demonstrated clinical effectiveness, and where reliable clinical
evidence supports a finding that the benefits of the product outweigh its harms (Tunis &
Kang, 2001). This restriction is especially important for HUDs. Because clinical
effectiveness is not a regulatory requirement when the FDA grants an HDE approval, the
requisite information on effectiveness may be judged to be inadequate for coverage by

44
CMS. This misalignment between FDA and CMS has been problematic in securing
health insurance coverage for a range of HUDs.
2.4.3 HUDs Reimbursement Current State and Challenges
Currently, the CMS does not have a NCD that addresses coverage of HUDs; thus,
local payers take responsibility for the discretionary decision to provide coverage as
needed (U.S. Centers for Medicare and Medicaid Services, 2008b). In practice, the CMS
deals with reimbursement of HUDs on a case-by-case basis, so that local payers coverage
policies of HUDs can vary significantly. This variability spills over into the decisions of
insurance carriers more generally. For instance, Cigna provides coverage of deep brain
stimulation for chronic, medically intractable primary dystonia indication (HDE #
020007) but rejects coverage of this same device for obsessive-compulsive disorder
(HDE # 050003) because it is considered investigational in nature (Cigna, 2011a). In a
second example, Cigna and Harvard Pilgrim Health Care cover Enterra
TM
gastric
electrical stimulation (HDE # H990014) as medically necessary when provided for the
HDE approved indications for the treatment of chronic, intractable nausea and vomiting
secondary to gastroparesis of diabetic or idiopathic etiology (Cigna, 2011b; Harvard
Pilgrim HealthCare, 2010). However, Medica and Blue Shield of North Carolina do not
cover Enterra
TM
gastric electrical stimulation because insufficient clinical evidence was
judged to be present to support its use for the same indication (Blue Cross Blue Shield of
North Carolina, 2012; Medica, 2011). In addition, it is not uncommon for a physician to
seek a determination of medical necessity related to use of HUDs by requesting a prior
approval from a health insurance plan, as in the case of the ThereSphere® (HDE #
H980006) coverage policy from the UPMC Health Plan (UPMC Health Plan, 2011). The

45
variability in local coverage policies and absence of a NCD make it difficult for device
manufacturers to navigate the reimbursement landscape, which consequently delays the
access of much-needed HUDs to patients who need them.
Bentivoglio and his colleagues have discussed the challenges associated with the
migration of devices into the health care setting. They have suggested that the differing
standards of FDA and CMS greatly complicate the coverage of HUD devices. From the
FDA perspective, approved HUDs are not investigational; however CMS struggles with
the lack of clinical evidence to demonstrate the effectiveness of HUDs. Without such
evidence, CMS does not have sufficient information to meet the “reasonable and
necessary” statutory determination (Bentivoglio, Lorell, Scanlan, Loeb, & Weinstein,
2006).
The problem of obtaining coverage for HUDs is compounded because medical
device companies, which are smaller in size than pharmaceutical companies, usually do
not have the resources or expertise to negotiate coverage, coding, and payment with
every payer. The patients themselves are also often impotent to make compelling
arguments for coverage. This can lead to suboptimal reimbursement arrangements. The
IRB review requirements imposed on HUDs can further confuse the payers. They may
be uncertain whether HUDs should be considered as commercially released products and
therefore covered, or as investigational products and therefore not covered. Bleicher
summarized this predicament by stating, that “reimbursement systems have not kept pace
with regulatory processes for HUDs” (Bleicher, 2009). Similarly, Randel E. Richner, VP
of Government Affairs and Reimbursement from Boston Scientific Corporation,
expressed her concerns in a public docket (No. 2004S-0233) to the Department of Health

46
and Human Services related to “lag” between CMS and FDA concurrence on product
commercializability.
Congress and FDA have taken explicit actions to promote the development of these
technologies through the Humanitarian Device Exemption (HDE) process, no
corresponding mechanism has been established to promote Medicare and Medicaid
coverage and payment for these technologies. (U.S. Food and Drug Administration,
2004)
Furthermore, AdvaMed, the largest trade association representing medical device in
the U.S., urged CMS to create an expedited means of providing coverage because HUDs
involve treatments for small populations with no alternative treatment modality
(AdvaMed, 2007).
2.5 Views and Activities of Key Stakeholders
As the previous overview suggests, ensuring that patients with rare diseases gain
access to appropriate therapy is a multifaceted challenge that involves interactions of key
stakeholders in the public and private sector. In the report titled “Rare Diseases and
Orphan Products: Accelerating Research and Development,” Institute of Medicine
(IOM) recommended that the Secretary of Health and Human Services should establish a
national task force to promote, coordinate, monitor, and assess the stakeholders’
implementation of rare disease initiatives in the public and private sectors. Specific
stakeholders identified in the IOM report included a number of federal agencies including
the FDA, the National Institutes of Health, the Centers for Disease Control and
Prevention and the Department of Defense, as well as pharmaceutical and medical device
industries, patient advocacy groups, private foundations, academic institutions,

47
international agencies, and other relevant parties (Institute of Medicine, 2011). A similar
range of stakeholders has been identified in other studies of orphan drugs in the U.S. and
elsewhere. For example, in a thesis by Redington, the federal government, industry,
patients, and payers all were identified as important players to affect orphan drug reform
from 1990 to 1994 (Redington, 2009). Similarly, Eeden (2011) interviewed groups in the
Netherlands that he saw as stakeholders, including the pharmaceutical industry, ministry
of health, patients’ representatives, and healthcare insurance board, to examine flaws and
restrictions in Dutch orphan drug regulations (Eeden, 2011). These stakeholders are
identified in Figure 6. Identifying stakeholders is important, because often their views,
needs, wants, and desires on a specific policy issue can vary significantly according to
certain vested interests, a situation that has been elsewhere characterized as “stakeholder
dissonance” (Samaras & Samaras, 2010). Stakeholders also vary in their abilities to
influence the outcomes of policy related to orphan products (Redington, 2009).
Understanding these different views and abilities can help to explain the reasons why
some initiatives are pursued and why they succeed or fail. Furthermore, a broad
understanding of all views makes it possible to identify solutions, and in some instances
compromises, that can ultimately result in a common solution that meets the needs of all
stakeholders.

48
Figure 6: Key stakeholders of HUDs

2.5.1 Food and Drug Administration (FDA)
FDA is a key stakeholder in the process for defining and allowing commercial
access to HUDs by virtue of its primary role as the oversight body for the HDE approval
process. The primary position and views espoused by the current FDA with regard to
orphan products in general and devices in particular has been to foster access and
facilitate innovation. In the FDA draft document titled “Strategic Priorities 2011- 2015:
Responding to the Public Health Challenges of the 21
st
Century,” the agency
acknowledged a serious need to expand the therapeutic options for special populations.
Specifically, the agency stated that market incentives alone might be not be sufficient to
drive innovation needed to develop new treatments for rare diseases, and the agency is
actively taking steps to meet the needs of these patients with rare disease (U.S. Food and
Drug Administration, 2011f).
Jesse Goodman, Chief Scientist and Deputy Commissioner for Science and Public
Health at the FDA, also reflected FDA opinion that current market incentives are not

49
adequate motivators to drive innovation, when he testified on July 21, 2010 before
the United States Senate Committee on Health, Education, Labor, and Pensions (HELP)
regarding the many initiatives taken to foster treatments for rare and neglected pediatric
diseases. He noted that the FDA had carried out many of the directives in section 740 of
the Fiscal Year 2010 Appropriation Act, by establishing internal review groups (the Rare
Disease Review Group and the Neglected Disease Review Group), creating a new
Associate Director for Rare Diseases position to promote scientific and regulatory
innovations, and establishing a grant program to fund pediatric device consortia that
develop innovative devices for children. Furthermore, he expressed the belief that these
activities had resulted in many successes, but more work would be needed to fulfill the
unmet medical needs of pediatric population (U.S. Food and Drug Administration,
2010f).
Debra Y. Lewis, Deputy Director of Office of Orphan Product Development
(OOPD), released a statement on the FDA’s Rare Disease Patient Advocacy Day on
March 1, 2012 regarding the accomplishment of OOPD in 2011 and the need for
continued efforts in the rare diseases. She stated:
In 2011, FDA approved 26 drugs and biological products for rare diseases...FDA
also approved six medical devices for rare conditions under the Humanitarian Device
Exemption program in 2011. We are very proud of the collaborating between
industry, patient advocates and FDA that have made this a reality…There are over
400 drugs and devices approved for rare diseases, and million of lives have been
saved…however we still have much to do to meet the rare disease challenge.
(FDAVoice, 2012)

50
Consistently, it has appeared that the FDA and OOPD recognize the needs for
continued efforts with regard to therapeutics for rare diseases. However, it is unclear if
the existing and potential activities of FDA would address the underlying barriers of the
HDE pathway in order to spur innovation of HUDs.
2.5.2 Congress
Members of Congress appear to share the views of the FDA regarding the need to
foster HUD development and access. In August 2010, Senator Al Franken of Minnesota
called on Congress to create extra incentives for medical device makers to tackle rare
diseases. Franken noted that medical device manufacturers lacked the same incentives
that are currently in place for orphan drugs. Additionally, Franken requested that U.S.
Government Accountability Office (GAO) compare the incentives granted under the
Orphan Drug Act, known to enhance the viability of orphan drugs to current incentives
for medical devices and to recommend legislative changes that would encourage the
development of HUDs (Reuters, 2012; Sullivan, 2010). Furthermore, Representative
Charles Bass of New Hampshire introduced the Humanitarian Device Reform Act of
2011 to amend the FDCA by removing the profitability restriction for HUDs indicated
for adult populations and by clarifying the requirement to limit sales to 4,000 patients.
One of the findings documented in the bill is that “the potential for HDEs to foster
innovation has not been reached because of the regulatory burdens of the program” (H.R.
3211, 2011). According to Representative Charles Bass, the bill would foster innovation
in HUDs and lead to a better quality of life for the 20 million American patients living
with rare diseases (Bass, 2012).

51
2.5.3 Centers for Medicare and Medicaid Services (CMS)
Unlike FDA and Congress, the view of CMS concerning HUDs is not well
articulated as evidenced by the limited literature related to national coverage of HUDs.
Some insight into its position, however, could be identified in its 2006 review of policies
regarding national coverage determinations (NCDs) for clinical research through the
public comment process. The goals of the clinical research NCDs were to ensure that
Medicare beneficiaries receive care that might have a health benefit but for which
evidence for the effectiveness of the treatment or service was insufficient to allow for
unrestricted coverage, and to encourage conduct of clinical research studies (U.S. Centers
for Medicare and Medicaid Services, 2007). Specifically, one of the issues of addressed
in that forum was related to the Medicare policy for payment of HUD costs. CMS stated
the belief at that time that HUDs would be covered most appropriately through the
coverage with evidence development (CED) program in the context of an NCD. In such
a circumstance, CMS would issue a NCD, but would require additional evidence
collection to supplement the clinical data (U.S. Centers for Medicare and Medicaid
Services, 2007, 2012a). An example of an HUD covered under the CED process is the
AbioCor Implantable Replacement Heart (HDE # 040006), which CMS deemed as not
reasonable and necessary. Nevertheless, since the HUD had the potential to improve
health outcomes for Medicare beneficiaries, CMS issued a NCD to allow reimbursement
of the device under the CED process when beneficiaries are enrolled in clinical studies
that meet all the predetermined criteria (U.S. Centers for Medicare and Medicaid
Services, 2008a). Although the CED reimbursement pathway of HUDs appears

52
directionally positive, its impact on patient access to HUDs is rather limited to small
pools of patients enrolling in clinical studies.
2.5.4 Patient Advocacy Groups
Patient advocacy groups play a crucial role in fostering a regulatory environment
that encourages both the development and timely approval of innovative medical
technologies, and also the access needed to such technologies. Traditionally, it has done
this by representing the interests of the patients to other key stakeholders who have the
power to pass legislation or modify regulations related to product approvals or research
funding. More recently they have also become vocal about issues associated with
reimbursement. Almost always these representations criticize the slow pace with which
new treatments are introduced and the barriers to commercialization and clinical uptake
that are considered to exist. Two of these impediments were recently singled out, for
example, by Diane Edquist Dorman, Vice President for Public Policy, National
Organization for Rare Disease (NORD), in her testimony on July 21, 2010 before
the United States Senate Committee on Health, Education, Labor, and Pensions (HELP)
on treating rare and neglected pediatric diseases. Dorman pointed out that the IRB
approval of HUDs is particularly troublesome because it leads insurers to view HUDs as
experimental devices that need not be reimbursed. In addition, while pediatric HUD
developers could make a profit, this was not the case for all humanitarian devices.
Dorman stated, “the prohibition against developers profiting from these devices needs to
be lifted” (National Organization for Rare Diseases, 2011). Accordingly, NORD believes
that “while the relaxed standards for approval of HUD use is encouraging, unless
manufacturers are able to recoup costs of developing devices for humanitarian use, the

53
life-changing impact they have on patients with rare diseases will continue to be
minimal” (National Organization for Rare Diseases, 2012).
Art Kessler, the President of the Dystonia Medical Research Foundation, also
shared the view with regard to the reimbursement hurdle of HUDs at the FDA’s public
hearing titled “Considerations Regarding FDA Review and Regulation of Articles for the
Treatment of Rare Disease.” Kessler revealed that some dystonia patients who were
denied insurance coverage have approached the Foundation for assistance to deal with the
lengthy insurance appeals process. The out-of pocket cost that patients must otherwise
pay was characterized as a significant barrier to receiving treatment. Moreover, Kessler
expressed the view that the clinical trial standards for rare diseases seemed to be
equivalent to those for non-orphan diseases, and urged the FDA to exercise discretion in
its clinical trial standards for orphan indications (U.S. Food and Drug Administration,
2010a).
2.5.5 Physician and Healthcare Professional Groups
Physicians and other healthcare professionals have an obvious stake in policy
development related to the products that they can eventually offer patients with serious
disease. Insight into these views was sought by GAO when it surveyed 36 physicians
who had frequently used one particular HUD, the Melody Transcatheter Pulmonary
Valve, regarding their experience in obtaining reimbursement coverage for their patients.
The Melody Transcatheter Pulmonary Valve is a pediatric HUD that was approved after
PMDSIA was passed in 2007. The survey revealed that many physicians experienced at
least some difficulty in securing health insurance coverage for the device regardless of
the types of insurance plans (private, Medicare, or Children’s Health Insurance Program)

54
with which they had to interact. Importantly, some physicians reported that the ease of
securing coverage had influenced their clinical decisions in unfortunate ways. As one
example, some clinicians were performing open-heart surgery as an alternative to the
using the new, less invasive orphan product. Further, physicians shared some of the same
challenges as other stakeholders, that result from the absence of a CPT code and the
perceived “experimental” status of the devices by the payers (U.S. Government
Accountability Office, 2011). Clearly, reimbursement of HUDs was seen as an
impediment in the path of needed treatments.
The forgoing discussion suggests that HUD rules are overly restrictive. Less
attention has been paid to the possibility that more regulation or oversight might be a
good thing. However, most HUDs are in fact relatively sophisticated devices with
significant risk and have potential opportunities for misuse or abuse as an “easier” route
to market. This debate can be seen first in the article titled “Misuse of the FDA’ s
Humanitarian Device Exemption in Deep Brain Stimulation for Obsessive-Compulsive
Disorder” where Fins and colleagues asserted that FDA’s HDE regulatory pathway was
misused to obtain approval for deep brain stimulation for the indication of severe
obsessive-compulsive disorder (OCD). The authors: (a) challenged the status of OCD as
an orphan indication, (b) asserted that the FDA could have required the manufacturer to
pursue an investigational device exemption to assess the device’s safety and efficacy, (c)
stated that SMDA of 1990 required the safety of a HUD be assessed through the
“substantially equivalent” decision marking process, and (d) claimed that the device
status should not have been moved from “investigational” to “proven effectiveness” due
to lack of scientific evidence (Fins et al., 2011). Eydelman and Chen from the FDA’s

55
OOPD responded that the article “contains factual errors, omissions, and misconceptions
regarding the humanitarian device exemption pathway” (Eydelman & Chen, 2011). The
article highlights a misunderstanding of the HDE regulatory pathway among some
members of the healthcare community; however it is unknown if this is an isolated or
more widespread issue. Further, although practice of medicine is the primary focus of
physician and healthcare professional groups, their views and understanding of regulatory
requirements to secure HDE approval is important so that they can make informed
decisions about the use of HUD devices.
2.5.6 Industry
2.5.6.1 Views of AdvaMed (Advanced Medical Technology Association)
The primary industry trade association that represents medical device companies
in the U.S. is AdvaMed (Advanced Medical Technology Association). Similar to
governmental advocates, Advamed and its constituent companies are interested in
fostering access to HUDs, but their specific views reflect the unique role of industry as
the sector that must respond to the call for increased HUD development. On November
23, 2009, Dr. Susan Alpert, Senior Vice President of Global Regulatory Affairs for
Medtronic Inc., on behalf of AdvaMed, provided written testimony to the Institute of
Medicine’s Committee on Accelerating Rare Diseases Research and Orphan Product
Development. AdvaMed criticized the FDA for its failure to provide general guidance to
manufacturers regarding the type or level of evidence that must be developed to
demonstrate that an HDE meets the probable benefit standard. She argued that the lack
of guidance has hindered the use of the HUD program as a pathway to market for devices

56
that treat or diagnose rare disease conditions. In comments addressing pediatric device
development specifically, AdvaMed identified the following two major obstacles:
Annual market associated with specific diseases and conditions may not be
commercially viable…orphan diseases and conditions are difficult to study
because they are widely dispersed making it extremely difficult to accrue
sufficient numbers of clinical trial participants over a reasonable timeframe and
within a manageable number of investigational sites to assure an adequately
powered clinical trial to meet FDA requirements. (Institute of Medicine, 2009)
Advamed urged FDA to make better use of all forms of valid scientific evidence
rather than relying on well-controlled trials. Importantly, FDA is statutorily authorized to
exercise flexibility in utilizing valid scientific evidence, but has been reluctant to do so
(Institute of Medicine, 2009).
Michal Morton, Senior Director of Regulatory Affairs for Medtronic Inc., testified
on behalf of AdvaMed at the FDA's public workshop on "Identifying Unmet Public
Health Needs and Facilitating Innovation in Medical Device Development” on June 24,
2010. Specifically, AdvaMed viewed the following items as key to faciliate innovation
and access of HUD through the HDE regulatory pathway: (a) FDA should develop a
guidance document to specify the type or level of clinical evidence needed to meet the
probable benefit standard of HDEs, (b) FDA should develop regulatory tools and
adaptive clinical trial designs that take into consideration the reduced sample size
associated with orphan and pediatric diseases and conditions, (c) FDA should accept all
forms of valid scientific evidence other than well-controlled trials, as allowed statutorily,
(d) FDA should encourage early collaboration and communication with device sponsors,

57
(e) FDA should create a new compassionate use pediatric device provision that would
allow a mechanism to provide HUDs for super-small pediatric populations, and (f) FDA
should allow tax credit and expedite clearance or approval for HDEs indicated for
pediatric populations. AdvaMed also emphasized the need to improve Medicare’s
reimbursement pathway (U.S. Food and Drug Administration, 2010e).
The view that Medicare’s reimbursement pathway for HUDs needs improvement
was further emphasized by Janet Trunzo, Executive Vice President for Technology and
Regulatory Affairs for AdvaMed, in the public comments to docket FDA-2010-N-0237
“Identifying Unmet Public Health Needs and Facilitating Innovation in Medical Device
Development.” AdvaMed supported local Medicare coverage as a means to gain a clear
understanding of the effectiveness of HUDs, which in turn, would be used to inform
Medicare’s coverage decision in the long term. Furthermore, AdvaMed urged CMS to
develop a formal process by which temporary reimbursement codes could be assigned for
HUDs, until a permanent CPT code has been developed by AMA to support access to
HUDs (FDA-2010-N-0237, 2010). In a separate docket (FDA-2011-D-00847) published
in December 2011, AdvaMed responded to an FDA request for public comments to the
draft guidance document regarding designation of HUDs. AdvaMed stated the view that
such a guidance document would faciliate the development of HUDs; however it
requested that a statement be added to the guidance document to clarify that an HDE
constituted a marketing approval from the FDA. The reason for the request was to reduce
confusion among some payers who assumed that HDEs were investigational and
therefore denied reimbursement of the devices (FDA-2011-D-0847, 2012). Furthermore,
some industry representatives provided their views to docket FDA-2010-N-0218 titled

58
“Considerations Regarding FDA Review and Regulation of Articles for the Treatment of
Rare Diseases; Public Hearing” specifically focusing on the appropriateness of the
annual threshold requirement of 4,000 patients for HUD designation and the guidance on
the “probable benefit” requirement for HDE pathway. Because these view are germane
to the main topic of this thesis, theses positions are discussed as follows.
2.5.6.2 Appropriateness of 4,000 per Year Threshold for HUD Designation
As defined in 21 CFR 814.3(n), a HUD is a medical device intended to benefit
patients in the treatment or diagnosis of a disease or condition that affects or is
manifested in fewer than 4,000 patients in the United States per year (U.S. Food and
Drug Administration, 2010b, 2011e). Several industry representatives have challenged
the 4,000 patient threshold requirements. In particular, the Biotechnology Industry
Organization (BIO), which represents more than 1,200 biotechnology companies, viewed
the use of different threshold numbers for defining rare disease for medical devices
(4,000) and pharmaceutical products (200,000) as illogical. According to the BIO, “it is
the disease incidence not the therapy that should define the population” (FDA-2010-N-
0218, 2010c). Amgen, a pharmaceutical company, also suggested that the FDA should
broaden the HUD eligibility criteria to align with the paradigm for orphan drugs, arguing
that the “fewer than 4,000 patients” regulatory requirement of HUDs is too low and too
difficult to ascertain (FDA-2010-N-0218, 2010b). Because the information on the size of
certain rare diseases is limited, AdvaMed viewed the threshold as lacking in merit and
recommended a flexible threshold based on medical, demographic, and scientific
information provided by a sponsor (FDA-2010-N-0218, 2010a). Medtronic, a medical
device manufacturer, supported the recommendation from AdvaMed, and viewed the

59
threshold as particularly problematic for ultra-orphan patient populations where the
population size is much smaller than 4,000 patients per year, and might be too small to be
commercialized under the still restrictive requirement of the HDE pathway (FDA-2010-
N-0218, 2010d). From these views, it appeared that the 4,000 patient threshold might be
considered as a regulatory impediment.
2.5.6.3 Guidance on “Probable Benefit” Requirement of HDEs
The HDE regulations require that a HDE application contain a description of the
suitability of the device for the rare disease or condition and an explanation of why the
probable benefit from the use of the device outweighs the risk of injury or illness from its
use. However, there is no regulatory definition for the term “probable benefit”, and the
FDA has yet to define the type or level of evidence required to fulfill “probable benefit”
requirement. According to Medtronic, the amount of clinical evidence to fulfill the
“probable benefit” requirement appeared to be increasing. Medtronic viewed this trend
as particularly concerning as the demands for clinical evidence that resemble the
expectations for PMA to fulfill the “probable benefit” requirement does not support the
intent of HDEs regulations (FDA-2010-N-0218, 2010d). On the related issue, AdvaMed
viewed the “probable benefit” as suggesting a different standard than the PMA
requirement for reasonable assurance of safety and effectiveness. Thus, AdvaMed urged
the FDA to develop guidance on the types and levels of clinical evidence necessary for
HDE approval (FDA-2010-N-0218, 2010a). Without such guidance, the time needed to
generate sufficient clinical evidence might be unnecessarily long, as was considered to be
the case for the Vertical Expanded Prosthesis Titanium Rib (HDE # 030009) which
consumed 14 years of clinical trials before approval (National Organization for Rare

60
Diseases, 2007). Thus, the lack of guidance to meet the “probable benefit” requirement
is seen to be problematic in progressing to HDE application or in securing a HDE
approval. This issue was examined in the exploratory study.
2.6 Institute of Medicine (IOM) and Government Accountability Office (GAO)
The many views expressed by multiple stakeholders can be hard to bring together
into a cohesive picture without careful arms-length analysis. Such an analysis was
initiated by the IOM in a consensus report titled “Rare Diseases and Orphan Products:
Accelerating Research and Development” on October 4, 2010. The report reflected the
input from its consultations with a range of stakeholders from academic, federal agency,
industry, consumer, patient, and other organizations, collectively referred to as the
committee. The committee recommended an integrated national strategy to promote
research and development for rare diseases consisting of the following elements: (a)
active involvement and collaboration between private and public stakeholders, (b) timely
application of science and technology to accelerate efficiency and effectiveness of rare
diseases research and development, (c) appropriate implementation of innovative clinical
trial design tailored for small populations, (d) promotion of efficient sharing of research
resources and infrastructure, (e) promotion of innovation by granting reasonable rewards
and incentives for private sector, (d) adequate allocation of resources in public agencies,
including staff with expertise on rare diseases research and development, and (e)
implementation of instruments to prioritize rare disease research, establish collaborative
organization-specific goals, and evaluation methods to monitor progress toward these
goals. Importantly, the committee indicated that “elements of each already existed but
lack a coordinated focus.” Furthermore, the committee recommended that the FDA and

61
NIH should conduct an assessment of unmet device needs and priorities relevant to rare
diseases, identify impediments to meeting those needs, and consider options for
overcoming impediments and stimulating high-priority innovation. The committee
recognized the difficulty of allocating resources to therapies for small populations that
lack commercial incentives and cannot often attain full commercialization by way of
PMA approval. It articulated the views of stakeholders that the HDE approval route is a
potentially less costly route to market, but had limited appeal due to the prohibition on
profit (except in pediatric indications), difficult reimbursement mechanisms and the need
for IRB approval (Institute of Medicine, 2011).
The Government Accountability Office (GAO), an investigative and evaluation
branch for the U.S. Congress, also undertook an examination of stakeholder experience
with HUDs, with a specific aim to characterize the barriers related to pediatric HUDs. Its
report in December 2011 also summarized the results of interviews with HUD
stakeholders. Those who were interviewed complained about the lack of guidance related
to the type or level of clinical evidence necessary for approval and reimbursement and
difficulty to obtain insurance. Moreover, not surprisingly, they confirmed that the
removal of restrictions on profitability was generally believed to encourage pediatric
medical device development, but the restriction of sales to a maximum annual
distribution number of 4,000 limited the incentive. They also identified that orphan drug
incentives, such as patent protection, were not always seen as beneficial for HUDs
because of the short lifecycle of medical devices as such products are being continually
improved technologically or replaced by newer devices, and thus may not outlive the
normal patent window (U.S. Government Accountability Office, 2011).

62
2.7 Systematic Evaluation of HDE Regulatory Pathway
Despite the fact that the HDE regulatory pathway has existed for over a decade,
only a small amount of literature appears to have evaluated the effectiveness of the HDE
pathway systematically. In 2009, Bernad investigated how the HDE regulatory pathway
differed from the regulatory pathways for other medical devices and discussed the
potential advantages and disadvantages of the HDE regulatory pathway. Bernad
conducted 40 semi-structured interviews with HDE stakeholders, including sponsors of
HDEs. Sponsors of HDEs expressed the view that industry might face short-term gains
in time and costs under the current rules for approval and reimbursement, but noted that
the resource allocation needed for such work could be a long-term financial drain on the
company. Some expressed concern that HDE holders could expend enormous resources
to underwrite ventures that would never provide adequate recompense for those efforts.
Furthermore, the interviews revealed that in most cases, the decision to pursue HUD
designation and HDE approval was made after product development for the following
reasons: (a) the company viewed the HDE route as an easier regulatory pathway that
required less time and resources, (b) the company viewed the HDE as a precursor to
eventual PMA filing, (c) the company valued the opportunity to recover costs from
research and development, (d) the rules allowed small-to-medium companies with limited
resources to bring products to market, and (e) the rules provided an opportunity to exploit
a new market. Barriers to development, such as profitability restrictions, IRB
bureaucracy and risks, and reimbursement challenges were also discussed. Specifically
for reimbursement, the author summarized the current environment by stating, “it can
take from 15 months to more than 5 years for new technologies to become available to

63
Medicare beneficiaries, and it can take as long as 27 months post-launch for new
procedure code to be created” (Bernad, 2009).
Kaplan and Harvey examined the role of the HDE regulatory pathway in the
development of new cardiovascular devices. The article stated:
As the costs to gain regulatory approval for commonly used devices increased,
companies would often seek alternative ways to gain market access, including the
HDE pathway…there may be multiple legitimate and distinct indications,
including indications that meet the HUD criteria. (Kaplan et al., 2005)
However, the article emphasized that the HDE regulatory pathway should be
reserved for true orphan indications and not be viewed strategically as part of the clinical
development plan to access a large market (Kaplan et al., 2005). Furthermore, Sherman
et al examined two HDE applications submitted by different companies. Based on the
research, Sherman indicated that while the effectiveness of HUDs is waived, the
applications still must supply information regarding the risks versus the probable benefits
associated with the device.
As noted previously, HUDs can only be used in a facility after an Institutional
Review Board (IRB) has approved their use in that facility, except in certain
emergencies. Gordon and Philpott surveyed IRB chairs in the United States to ascertain
whether the IRBs have clarity regarding their FDA-mandated role in overseeing the use
of HUDs in their facilities. According to the results of the survey, many IRB Chairs were
confused about what HUDs were, how to review HUD applications, and why IRBs
should review them. Gordon and Philpott suggested that the FDA should clarify their
policies to provide the guidance necessary for IRBs to review HUD applications more

64
effectively and consistently, and thereby better protect HUD-treated patients (Gordon &
Philpott, 2008). Furthermore, Dolcimascolo discussed the challenges and potential
solutions to accelerate access to new cardiac devices indicated for pediatric populations.
According to Dolcimascolo, the HDE regulatory pathway is a less desirable business
venture from the viewpoint of a restriction on profitability. However, under the PMDSIA
of 2007, which lifted the profitability restriction for pediatric indications, the HDE
regulatory pathway would be the commonly used regulatory pathway for pediatric
cardiac devices due to limited numbers of affected patients (Dolcimascolo, 2009). It is
worth noting that since PMDSIA of 2007 only three HDEs have been approved with
pediatric indications.
2.8 Recent Regulatory and Reimbursement Initiatives
2.8.1 The Food and Drug Administration Safety and Innovation Act of 2012
Some of the concerns expressed about the HDE pathway have been addressed
legislatively when President Barack Obama signed into law the Food and Drug
Administration Safety and Innovation Act (FDASIA) (Public Law 112-144) on July 9,
2012. This law marks a significant inflection point in HUD management. In the context
of an HDE regulatory framework, Section 613 of the FDASIA implemented several
amendments to FDCA section 520(m) to change certain rules regarding HUDs. First, it
removed the profitability restriction from HDEs indicated for non-pediatric indications.
Specifically, new section of 520(m)(6)(a)(i)(II) stipulates that HDE sponsors can make
profit from devices “intended for the treatment or diagnosis of a disease or condition that
does not occur in pediatric patients or that occurs in pediatric patients in such numbers
that development of the device for such patients is impossible, highly impracticable, or

65
unsafe.” Second, the new law redefines the annual distribution number as the number
“reasonably needed to treat, diagnose, or cure a population of 4,000 individuals in the
United States” (S.3187, 2012). The expansion of the profitability provision is anticipated
to foster increased investment in research and development of HUDs. However, the
impact of the FDASIA provision on the rate of HUD innovation and progression from
HUD to HDE is not known. Also unclear is whether the provision will allow for easier
and faster access to HUDs. FDASIA is a “silo” legislative reform involving only the
Food, Drug, and Cosmetic Act. CMS, the national reimbursement agency, has yet to
mirror the type of HUD policy changes required, and to some degree sought, by the FDA.
Concerted HUD policy reform by CMS might be needed to align its approaches with the
new policies of FDA.
2.8.2 FDA/CMS Collaboration and Parallel Review
The Department of Health and Human Services (HHS) has long recognized the
need for concerted collaboration between the FDA and CMS to reduce regulatory burden
and improve patient access to innovative HUDs. In 2005, HHS launched a task force
initiative whereby the agencies would collaborate to coordinate marketing approval,
postmarket evaluations, and reimbursement approvals for HDEs. The collaboration
mechanism, according to HHS, would be performed through the following activities: (a)
FDA would expedite the availability of “Summary of Safety and Effectiveness”
information, the basis for HDE approval, to CMS if the devices were eligible for
reimbursement, and (b) CMS would make available its “real world” postmarket
surveillance data to FDA in order to improve their information base regarding the overall
safety profile of the devices (U.S. Department of Health and Human Services, 2005).

66
While the task force initiative incrementally improved the transparency between the
agencies, it is unclear if it truly addressed the underlying challenges associated with
patient access to HUDs.
In 2011, in line with the aspiration of HHS to improve patient access to
innovative devices, the agencies launched a pilot program for parallel review of certain
FDA premarket review submissions and CMS national coverage determinations. On an
annual basis, the agencies announced the intention to allow up to five innovative devices
to participate in the pilot program if they satisfied one of the following eligibility criteria:
(a) new technologies for which the sponsor has requested a pre-investigational device
exemption (IDE) or approved IDE, (b) new technologies that would be reviewed via the
PMA regulatory route or a petition for de novo review, and (c) new technologies that fall
within the scope of Part A or Part B of Medicare, and have not obtained a National
Coverage Determination (NCD) (U.S. Food and Drug Administration, 2011c). The
agencies are working toward a parallel review mechanism that would address some
potential challenges that would almost certainty arise during the interactions between the
two agencies. For example, CMS is transparent to the public regarding the initiation of
an NCD process. In contrast, FDA must keep the premarket review of submissions
confidential until a decision has been made; when and how information can be revealed
in a joint process would need to be specified carefully. Further, it is unclear when the
joint review of submissions between the agencies should begin or if a joint advisory
committee should be used (FDA-2010-N-0308, 2010e). There are many unanswered
questions regarding the implementation of the parallel review program. Nonetheless, the
parallel review program aligns with the IOM’s recommended integrated national strategy

67
to promote rare disease research in way of active involvement and collaboration between
government agencies and other stakeholders (Institute of Medicine, 2011). Accordingly,
this exploratory study assessed the perspective of HUD sponsors on the degree to which
the parallel review would improve the cohesiveness and alignment between the two
agencies in ensuring patient access of HUDs, with or without any policy reforms, and any
recent experience with it.
2.8.3 Coverage with Evidence Development (CED) Program
CMS has been concerned with innovative mechanisms to ensure patient access to
medical technologies. On July 12, 2006, the CMS published the first guidance document
titled “National Coverage Determinations with Data Collection as a Condition of
Coverage” on a specific Coverage with Evidence Development (CED) program. As the
name “CED” implies, the CED is a reimbursement scheme that allows payers to offer
conditional coverage for potentially beneficial, yet unproven, medical technologies while
requiring the development of clinical evidence. The goals of the CED program were to
generate clinical evidence to assist physicians, payers and policymakers to make
informed decisions on the best use of selected medical technologies, and, importantly, to
determine coverage decisions (Kary, Bergthold, Talbott, & Aubry, 2009). In such a
circumstance, CMS would issue a NCD, but would require additional evidence to be
collected. This new evidence would then supplement the clinical data (U.S. Centers for
Medicare and Medicaid Services, 2007, 2012a).
According to the 2006 guidance document, CMS detailed the following two types
of coverage through CED: (1) Coverage with Appropriateness Determination (CAD),
promulgated in Section 1862(a)(1)(A) of the Social Security Act, is used when clinical

68
evidence of a medical technology or service meets the “reasonable and necessary”
statutory requirement. However confirmatory data is required to assure that the medical
technology or service is provided to patients meeting the prescribed criteria specified in
the NCD. As such, providers of the medical technology or service would be required to
submit clinical data through establishment of data registries, (2) Coverage with Study
Participation (CSP), promulgated in Section 1862(a)(1)(E) of the Social Security Act, is
used when clinical evidence of a FDA-approved medical technology or service is
promising, but does not meet the “reasonable and necessary” statutory requirement.
When applying CSP, CMS required the beneficiaries to enroll in a clinical study designed
to collect supplementary clinical evidence. The clinical study must be conducted under
the oversight of the Agency for Healthcare Research and Quality (AHRQ), which is
tasked to conduct and support research on outcomes, effectiveness, and appropriateness
of medical procedures and services (Lundy et al., 2012; Trueman, Grainger, & Downs,
2010).
Over the years, CMS has applied the CED program to a variety of medical
technologies for a total of 19 specific products (Neumann & Chambers, 2013). For
example, the NCD of Implantable Cardioverter Defibrillator (ICD) required submission
of data to a national registry under the CAD scheme whereas the NCD of Transcatheter
Aortic Valve Replacement (TAVR) required beneficiary enrollment in a CMS-sponsored
clinical study for non-FDA approved indications under the CSP scheme. However, the
CED program has been criticized by various stakeholders for the following reasons: (a)
the responsibility of funding between payers and manufacturers to support additional
research is not clearly defined, (b) the standards for designing a clinical study is not

69
clearly defined, (c) the randomization scheme for assigning study participants to a control
arm may be unethical, especially when a medical technology has been provide safe and
effective, and (d) the priority-setting process for applying CED has not been clearly
established (Center for Medical Technology Policy, 2011; Trueman et al., 2010; Tunis &
Pearson, 2006). In an article titled “Improving the Quality and Efficiency of the Medicare
program Through Coverage Policy,” Tunis et al. argued that “many CED studies have
had significant design flaws, received insufficient funding, lacked adequate data
collection systems, or for other reasons, have not procedure the scientifically rigorous
data needed to formulate sound coverage policy” (Tunis, Berenson, Phurrough, & Mohr,
2011).
Responding to the criticisms and intending to improve the CED program, the
CMS published a draft guidance document titled “Draft Guidance for the Public,
Industry, and CMS Staff Coverage with Evidence Development in the Context of
Coverage Decisions” on November 29, 2012. Specifically, the CMS sought public
comments from November 29, 2012 through January 28, 2013 on the following key
topics related to implementation of CED: (a) eliminate the CAD coverage scheme with
the rationale that the scope of CED would be to generate new clinical evidence as in the
case of CSP, (b) allow AHRQ to establish public/private partnerships to financially
support CED, (c) discuss a body of good practices related to CED clinical study such as
prospectively designed, scientifically sound, etc., (d) outline the factors to which CMS
would apply CED such as relevance to health outcomes, evolution of evidence base, etc.,
(e) articulate the circumstances under which a CED clinical study would be ended, (f)
generate the formal evidentiary criteria for CED, and importantly (g) acknowledge the

70
alignment of CED with an FDA approved post approval study is an example of CED
application, rather than a new CED paradigm (Neumann & Chambers, 2013; U.S.
Centers for Medicare and Medicaid Services, 2012b). This exploratory study assessed
the degree to which the CED would be applicable to HUDs.
2.9 Conceptual Framework to Base Further Investigation
In practice, policies governing patient access to HUDs involve the interplay of the
FDA and CMS at various levels within and between the agencies. These collectively can
be viewed as a patient access “system.” From the foregoing review, it is abundantly clear
that policies and interactions between CMS and FDA are viewed as suboptimum, and
better collaboration between the FDA and CMS may be needed. Not so clear are how
some of the stakeholders view these interactions and what their experiences have been
with this collective patient access “system.” To study such a system, it is useful to
develop an organized strategy for the study, to define the terms relevant to such a study,
and to identify a framework that might be used to explore specific aspects in more detail.
For this study I explored the usefulness of general system theory as one means for
framing the study of the HUD landscape in the U.S.
2.9.1 Origins and General Characteristics of General System Theory
"System theory" attempts to understand principles and laws that underlie the
behavior of systems and that can be generalized across different types of systems
(Amagoh, 2008; Dubrovsky, 2004; Steven, 2007; Yoon & Kuchinke, 2005). The concept
of a general theory of systems has a long history dating from 1937 when Ludwig von
Bertalanffy discussed its original tenets in a philosophy seminar at the University of
Chicago (Laszlo & Krippner, 1997). Bertalanffy identified a trend of increasing

71
specialization within every field that had resulted in stratification of science into
innumerable disciplines and sub-disciplines. According to Bertalanffy, each player within
scientific community was confined increasingly to his or her private universe and “it was
difficult to get word from one cocoon to the other...” (International Society for the
Systems Sciences, 2012). Nonetheless, each element in the science “system” was
interconnected. Bertalanffy defined the aim of the general system theory as follows:
(1) There is general tendency toward integration in the various sciences, natural,
and social, (2) Such integration seems to be centered in a general theory of
systems, (3) Such theory may be an important means for aiming at exact theory in
the nonphysical fields of science, (4) Developing unifying principles running
“vertically” through the universe of the individual sciences, this theory brings us
nearer the goal of the unity of science, (5) This can lead to a much-needed
integration in scientific education. (International Society for the Systems
Sciences, 2012)
During the 1950s, system economist Kenneth Boulding, mathematician Anatol
Rapoport, and physiologist Ralph Gerard, together with Bertalanffy, met to discuss their
views on a “system theory” from the perspective of their respective disciplines. Such
discussions resulted in a shared view that systems theory could be applicable to their
respective disciplines (Laszlo & Krippner, 1997). Several thought leaders have more
recently explored the systems theory further and have extended the work of Bertalanffy.
Laszlo expanded the philosophical underpinnings of systems theory by formulating the
concept of “seeing things whole,” or more specifically, “seeing the world as an
interconnected, interdependent field continuous with itself” (Erwin, 1972). Furthermore,
Chen and Stroup described the certain constituents of system theory as follows: (a) a

72
“system” consists of assembly of interacting parts, the sum of which exhibits behavior
not possessed in its constituent part, (b) a system can be physical, biological, social,
symbolic, or comprised of one or more of these, (c) change is viewed as transformation
of the system in time without losing its identify, (d) goal-directed behavior characterizes
the changes observed in the state of the system, (e) feedback is the mechanism that
mediates between the goal and system behavior, (f) time is a central variable in system
theory. It provides a referent for the very idea of dynamics, (g) the “boundary” serves to
delineate the system from the environment and any subsystems from the system as a
whole, and (h) system-environment interactions can be defined as the input and output of
matter, information, and energy. The system can be open, closed, and semipermeable to
the environment (Chen & Stroup, 1993).
However, even with these clarifications, systems theory can be challenging to
understand. Steward and Ayres expressed the view that systems theory is difficult to
specify and Dubrosky argued that principles of the theory have not been formulated
(Dubrovsky, 2004; Stewart & Ayres, 2001). At the abstract level, Kenneth Boulding
defined the theory as “a level of theoretical model-building which lies somewhere
between the highly generalized constructions of pure mathematics and the specific
theories of specialized disciplines” (Boulding, 1956).
As a base for understanding systems theory, it is useful to clarify the use of the
term "system." Laszlo and Krippner defined a system as “a group of interacting
components that conserves some identifiable set of relations with the sum of the
components plus their relations (i.e., the system itself) conserving some identifiable set of
relations to other entities (including other system)” (Laszlo & Krippner, 1997). Russell

73
Ackoff defined a system as a set of two or more interrelated elements, which have the
following properties: (a) each element has an effect on the functioning of the whole, (b)
each element is affected by at least one other element in the system, and (c) all possible
subgroups of elements also have the first two properties (Ackoff, 1981). Cavallo
suggested that a system might not be a physical entity, and that systems might be defined
only in relation to the purposes of the observer (Cavallo, 1979).
The essence of systems theory appears to be its emphasis on interlinkages
between components, that themselves may help to define the components. The
interlinkages might take the form of flows (of money, people, or products), which affect
levels of key policy-relevant variables. The interlinkages can be embedded in the
structure of the system, which might be hierarchical or network-based (Leydesdorff,
1997). As Bertalanffy stated:
It is necessary to study not only parts and processes in isolation, but also to solve
the decisive problems found in organization and order unifying them, resulting
from dynamic interaction of parts, and making the behavior of the parts different
when studied in isolation or within the whole… (International Society for the
Systems Sciences, 2012)
More simply stated, “the whole is more than sum of its parts” (International
Society for the Systems Sciences, 2012). The process of understanding the relationships
between various parts of a system, and how parts influence one another within a whole, is
also known as system thinking (Mathews & Jones, 2008). The system theory and the
process of system thinking has evolved over time and applied in different fields and
contexts.

74
2.9.2 Applications of System Theory to Policy Analysis
Systems theory has been used to only a limited extent for the study of health
related systems. Howley and Chuang and Leischow and his colleagues applied the
systems theory and the process of system thinking to the problem of improving
healthcare (Howley & Chuang, 2011; Leischow et al., 2008). Specifically, Howley and
Chuang hypothesized that the quality management and reporting systems in health care
organizations could influence patient safety and quality of care. They studied the
influence of the systems by constructing a holistic-relationship model of healthcare
systems and a series of adaptive-control studies to explore relationships between
segments within the systems relationship model (Howley & Chuang, 2011). Leischow
and his colleagues examined system thinking as a conceptual framework by focusing on
four areas to promote functional efficiency in the translational and transdisciplinary
aspects of public health: (a) management of systems knowledge, (b) utilization of the
power of transdisciplinary and multidisciplinary systems networks, (c) utilization of
methods for analyzing complex systems, and (d) organization of systems. By
emphasizing interrelationships, they could make policy suggestions directed at
maintaining these interdependencies in a healthy state. For example, they urged
government agencies to foster collaboration, and recognize that public health is the
“culmination of a complex, adaptive federation of systems that no one organization can
or should control.” To encourage collaboration through the lens of system thinking, they
recommended the following necessary steps: (a) to develop and apply systems methods
and processes, (b) to build and maintain network relationships, (c) to build system and

75
knowledge capacity and (d) to encourage transformation to a systems culture (Leischow
et al., 2008).
In an article titled “Health Care Management: The Contribution of Systems
Thinking,” Lebcir applied system thinking to study the causes of policy resistance in
health care systems and demonstrated how the causes might be overcome by using the
System Dynamics methodology. Lebcir described System Dynamics as follows:
System Dynamics assumes that most problematic situations arise from the fact
that systems are dynamic and complex in nature. It conceptualizes systems as
being constructed of complex networks of feedback loops in which time delays
and non-linear relationships are important sources of dynamic complexity and
policy resistance. (Lebcir, 2006)
Simply stated, System Dynamics is an approach to understanding the behavior of
complex systems by analyzing the parts of a system including cause-effect relationships
and their underlying mathematics and logic, time delays, and feedback loops between
elements (Nuhoglu, 2008). The internal feedback loop is a concept worth exploring
further. Lebcir argued that there is no indefinite linear chain of cause-and-effect
relationships; any action taken by one part of the system will upset the equilibrium of the
system and cause other parts to react and hopefully restore the system’s balance. In this
context, the influence is both a cause and an effect, which creates an internal causal
feedback loop. The causal feedback loops can have either reinforcing (positive) or
balancing (negative) effects. In the case of reinforcing loops, a change in one part of the
system will trigger changes that amplify the original change. Conversely, balancing
loops will result in changes that dampen the original change (Lebcir, 2006).

76
Systems theory and thinking, by definition, is large in scope. Thus it might
difficult for policy makers to understand, analyze, influence, or control the various
systems with which they were concerned, as Stewart and Ayres stated in the article
“System Theory and Policy Practice: An Exploration.” They argued that application of
system theory and thinking to complex policy issue might result in little improvement
because “systems theory has been billed as a kind of superpositivism, a way of modeling
and predicting every facet of reality.” Instead, they argued that the application of system
thinking to a policy issue would not provide a “ready-made” answer; instead the system
approaches offer a way of conceptualizing the policy issue. According to Stewart and
Ayres, the nature of problem could not be understood separately from its solution, which
would mean that policy solutions could not be “designed,” but represent a way of
navigating a specific problem. Nonetheless, they argued that system thinking might offer
new perspectives on Linder and Peters’ three constituents of policy-making – causation,
intervention, and evaluation (Linder & Peters, 1984; Stewart & Ayres, 2001). Each of
these constituents is described below and presented in Figure 7.

77
Figure 7: Application of system thinking to three constituents of policy-making

Causation – According to Linder and Peters, a model of causation allowed “association
of certain outcomes in the society with certain initial states” (Linder & Peters, 1984).
From the system thinking perspective, Stewart and Ayres argued that a complex policy
issue, which might have many interlinkages and complex feedback loops, could not
possibly be reduced to a simple, stylized cause-and-effect relationship. They emphasized
the need to acknowledge two-way influences and the role of internal feedback loops to
understand causation (Stewart & Ayres, 2001). Figure 8 illustrates the concept of causal
feedback loop. It uses an example of a model system to control obesity. In this example,
an increase in the number of obese people leads to an increase in the popularity of
weight-loss program, which in turn, reduces the extent of the obesity problem. Such an
outcome would appear to be direct and predictable. The weight-loss program is both a
cause and an effect, which creates an internal casual feedback loop. In this loop, the
perception to view weight-loss programs as the solution reduces efforts to address the

78
underlying causes of obesity related to urbanization and lifestyle. As these factors are
seen to be of less importance, they receive less attention and begin to have a
counterbalancing effect on the system in which their influence increases. Nonintuitively,
the extent of obesity could then increase (Newell, Proust, Dyball, & McManus, 2007).
Figure 8: Causal feedback loop

Note: Adapted from “Seeing Obesity as a System Problem” by B. Newell, K. Proust, R. Dyball, and P. McManus,
2007, NSW Public Health Bulletin, 18, p. 214-218. Copyright 2007 by NSW Public Health Bulletin. Reprinted with
permission.

Intervention – Linder and Peters stated that a model of intervention should guide policy
makers in the “selection of the locus and form of intervention into ongoing social and
economic processes in order to produce a desired outcome” (Linder & Peters, 1984).
Stewart and Ayres argued that by applying system thinking, intervention becomes a more
indirect art that depends for its effectiveness on the system’s capability to understand and

79
change its mode of operation through improved self-awareness and knowledge of other
systems. For example, policy makers may influence business people by manipulating
specific variables, such as regulations or registration requirements, within the business
system. This results in a modified business environment. Business firms respond to
these changes according to their individual decision-making values and practices. This
change in business behavior is therefore linked indirectly to the actions of policy makers.
As stated by Steward and Ayres, “interventions cannot be conceptualized as externally
applied incentives and sanctions. Rather, they will constitute changes in relationships
between (or even within) organizations, based on altered patterns of communication,
influence, and exchange” (Stewart & Ayres, 2001).
Evaluation – From the perspective of evaluation, Linder and Peters stated that a model of
evaluation allowed mapping of the outcomes of the policy process into a set of normative
premises about policy and society such as efficiency or justice (Linder & Peters, 1984).
From system thinking perspective, Steward and Ayres stated that one should focus on
ways of specifying values attached to a particular policy instrument instead of comparing
results with objectives. They implied that the values of policy initiatives should be
compatible with the system being addressed. They illustrated this concept by the
following example. In Australia, policies have been implemented to provide public
housing to elderly residents who do not succeed in obtaining a foothold in the private
housing market. If the policies do not succeed, policy makers might address the failures
by “reinforcing or invigorating existing structures, rather than replacing them with
incentives which assume different values and require a different kind of supporting

80
environment in order to operate.” Furthermore, they concluded that it was the totality of
policies that are relevant, not just the specific areas of interest (Stewart & Ayres, 2001).
Taking the foregoing three constituents into consideration, Steward and Ayres
argued that application of system thinking to a policy issue might allow a policy analyst
to: (a) conceptualize a policy problem as a system, (b) identify stakeholders who might
not play a role through the conventional policy analysis, and (c) construct interlinkages to
communicate across system boundaries and cumulatively, might offer the possibilities of
lateral thinking to a policy issue (Stewart & Ayres, 2001).
2.9.3 Proposed Research Conceptual Framework
A specific conceptual framework for studying the successes and challenges
policies relating to HUDs has not been identified in the literature to the best of my
knowledge. The foregoing work provides a beginning point to frame this study. It
appears possible to collect the issues identified from the literature review under the rubric
of the three constituents of policy-making – causation, intervention, and evaluation as
defined by Linder and Peters, but at the same time blending these elements into a broader
systems-based framework that acknowledges interactions between the two major
regulatory bodies, FDA and CMS. Policies governing patient access to HUDs involve
the interplay of the FDA and CMS that collectively can be viewed as an HUD patient
access “system” sharing the common goal of accelerating the access of patients to HUDs.
Thus, I used a conceptual framework that has been refined specifically to address
interactions within and between these agencies, based upon the application of system
thinking to the three constituents of policy-making as illustrated in Figure 9.

81
Figure 9: Conceptual framework to study patient access of HUDs
Causation FDA
CMS Causation
Intervention
Intervention
Desired
Outcome:
Accelerate
Patient Access
to Products
Evaluation
Evaluation
Legend:
Interlinkages

Within each agency, the policy-making process involves the three constituents
that are nominally considered to be conducted in a sequential order, causation,
intervention, and evaluation, and tied together by a common desirable outcome, that is, to
accelerate products to market. The first constituent, causation, concerned with how well
policies by the agencies are considered to be designed in order to achieve the often stated
goal of the policies. I had queried stakeholders about the specifics of the policies
themselves and used the results to examine whether two-way influences and internal
causal feedback loops appeared to exist that might modify, assist or impede the
development of effective policy. The intervention constituent would typically focus on
whether the actions taken by one agency or the other are sufficient to allow effective
outcomes from policy implementation. However, I explored stakeholder views both on

82
the effectiveness of the current strategies for policy implementation in each agency as
well as whether changes are needed to improve the interconnectedness of the FDA and
CMS as they implement policies. It would seem logical to explore various mechanisms
such as structural relationships, regulatory reforms, or administrative processes that are
currently under consideration or in pilot programs that are designed to improve the
interconnectedness of the FDA and CMS. The evaluation constituent was defined as the
effectiveness with which the outcomes of the policies are measured and then used as
drivers for further change at the policy level. I explored the nature and level of
awareness regarding current evaluative methods, the degree to which these evaluation
methods are seem to be sufficient, and the way in which the two agencies share
evaluations to understand how together their actions be able to improve policy design
(causation) and policy implementation (intervention). Thus, I used the modified triad of
constituents – causation, intervention, and evaluation – as a framework within each
agency, and the theory of systems as a conceptual base to investigate the interactions
between agencies in order to explore the industry’s views on the degree of cohesiveness
and alignment needed between the agencies in achieving a state of regulatory
convergence to support patient access of HUDs.
2.10 Summary and Research Direction
The literature review yielded much information about regulatory hurdles with
regard to the HDE regulatory pathway, but many questions remain answered. According
to the literature review, the most prominent regulatory hurdles are the profitability
restriction on HUDs, perceived investigational status of HUDs, lack of HUD
reimbursement coverage, and undefined “probable benefit” regulatory requirements of

83
the CMS. The extent to which these factors influence the progression of HUD to
approval of HDE is not clear from the literature review. Further, the literature does not
thoroughly evaluate whether additional interventions are needed to improve
interconnectedness of the FDA and CMS. When reviewing the literature critically, it was
not apparent that the policies or concerns have been directed at the system as a whole
rather than as separate parts. It is not clear whether a higher-level approach might
refocus the discussion on these agencies as a collective patient access “system.”
Furthermore, the literature identifies one impediment as a lack of cohesiveness and
alignment between the FDA and CMS but is unclear on the perceived magnitude of this
impediment in relation to other observed challenges. Also unclear is whether the
FDASIA provision will in fact promote easier and faster access of HUDs, as FDASIA is
a “silo” legislative reform involving only the Food, Drug, and Cosmetic Act. Thus, this
exploratory research study contributes to the evolution of thought with respect to the
HUD process by using the modified triad of constituents of policy-making – causation,
intervention, and evaluation – as a framework tied together with a systems theory to
explore the industry’s views on the degree of cohesiveness and alignment needed
between the agencies in achieving a state of regulatory convergence to support patient
access of HUDs.

84

Chapter 3: Methodology
3.1 Introduction
This exploratory study used the mixed methods of literature views, content
analyses, expert reviews of a novel interview tool, and key informant interviews to study
the extent to which the applicable policies of the Food and Drug Administration (FDA)
and Centers for Medicare and Medicaid Services (CMS) impact patient access to
Humanitarian Use Devices (HUDs), and how individuals from outside of these agencies
view the success of their interactions. The first stage was designed to establish a
background regarding the views of diverse stakeholders with respect to the FDA and
CMS parallel review program and the CMS Coverage with Evidence Development
(CED) program by carrying out a content analysis of public comments to the docket
FDA-2010-N-0308 titled “Parallel Review of Medical Products” and the CMS CED
guidance titled “Draft Guidance for the Public, Industry, and CMS Staff Coverage with
Evidence Development in the Context of Coverage Decisions” respectively. To the best
of my knowledge, the parallel review docket was the first of its kind in recent years,
undertaken jointly by the FDA and CMS, and intended to improve quality of patient
health care by ensuring earlier access to innovative medical products for Medicare
beneficiaries (CMS_FRDOC_0001, 2011). The CED draft guidance document was
recently published to solicit public comments to improve the implementation of CED and
to maximize the benefit to Medicare beneficiaries (U.S. Centers for Medicare and
Medicaid Services, 2012b). The content analyses were used as a preliminary method to
identify key concerns relating to FDA and CMS programs and interactions, as a first step

85
in developing areas of focus for the interviews with key informants. In the next stage, a
semi-structured interview tool was developed. The informant interview questions were
guided by the triad of constituents – causation, intervention, and evaluation – embedded
in a framework based on system theory. A small group of experts were identified with
expertise relating to FDA and/or CMS policies and activities, and these individuals were
asked to examine and provide constructive feedback on the interview tool. In particular,
they were asked to consider the content validity and applicability of the interview
questions for the target group of informants from industry. In the third stage of the work,
key informants were interviewed and their insights were evaluated to identify themes and
areas of concern.
3.1.1 Content Analyses
3.1.1.1 Public Docket: FDA-2010-N-0308 Parallel Review of Medical Products
As noted in Chapter 2, the Department of Health and Human Services (HHS)
launched a pilot program in 2011 for parallel review of certain FDA premarket review
submissions and CMS national coverage determinations. A docket (FDA-2010-N-0308)
titled “Parallel Review of Medical Products” was initiated to solicit public comments on
17 specific questions regarding the conduct of the parallel review between the FDA and
CMS. Using the modified triad of constituents – causation, intervention, and evaluation
– as the conceptual framework, the 17 specific questions were assessed and assigned to
one or more of the constituents. Specifically, questions relating to the first constituent,
causation, were concerned with how well policies by the agencies were considered to be
designed in order to achieve the enunciated goals of the policies. Questions relating to
the second constituent, intervention, focused on whether the actions taken by one agency

86
or the other were sufficient to drive effective outcomes from policy implementation, and
whether together their actions were complementary. Questions related to the evaluation
constituent were considered to be related to the effectiveness with which the outcomes of
the policies were measured and then used as drivers for further change at the policy level.
Once questions were assigned to the three constituents of the framework, responses to the
docket were analyzed to identify which of the questions were answered by the respondent
and what was the nature and strength of the opinions that were offered.
It was anticipated that some of the public comment packages might address only
some of the 17 questions because the different respondents might have different interests
or opinions regarding particular issues, some of which might be of greater or lesser
concern to the respondent. Thus, a purposive sampling of N=2 public comment packages
submitted by AdvaMed and Johnson & Johnson were selected first for the content
analysis because their responses addressed all of the 17 questions, and were felt to
represent the collective voice of a broad group of stakeholders in the medical device
industry. To broaden the content analysis beyond comments from the medical device
industry, a supplemental analysis was performed using public comments from the Cystic
Fibrosis Foundation (patient advocacy group) and Blue Cross Blue Shield (payer). These
respondents did not provide specific comments to each of the 17 questions, but they
provided general opinions. Thus, a total of N=4 public comment packages were analyzed
for the content analysis. The public comments analyzed are shown in Table 4.

87
Table 4: Public comment packages used for content analysis
Respondent Type Source
AdvaMed Trade Association http://www.regulations.gov/#!documentDetail;D=FD
A-2010-N-0308-0025
Johnson &
Johnson
Industry http://www.regulations.gov/#!documentDetail;D=FD
A-2010-N-0308-0039
Cystic Fibrosis
Foundation
Patient Advocacy
Group
http://www.regulations.gov/#!documentDetail;D=FD
A-2010-N-0308-0040
Blue Cross
Blue Shield
Payer http://www.regulations.gov/#!documentDetail;D=FD
A-2010-N-0308-0034

3.1.1.2 Draft Guidance for the Public, Industry, and CMS Staff Coverage with
Evidence Development (CED) in the Context of Coverage Decisions
In November 2012, the CMS issued a draft document titled “Draft Guidance for
the Public, Industry, and CMS Staff Coverage with Evidence Development in the Context
of Coverage Decisions.” This document is expected to supersede the guidance document
“National Coverage Determination with Data Collection as a Condition of Coverage:
Coverage with Evidence Development” issued on July 12, 2006. In the draft guidance
document, the CMS sought public comments on the following specific topics: (a)
implementation of CED through the National Coverage Determination (NCD), or other
avenues under Part A and Part B; (b) potential impact of CED on the Medicare program
and its beneficiaries; and (c) suggested approach to CED to maximize benefit to
Medicare beneficiaries (U.S. Centers for Medicare and Medicaid Services, 2012b). The
content analysis methodology of the public comments was similar to that of the parallel

88
review public docket described previously. Specifically, a purposive sampling of N=2
public comment packages submitted by trade associations – AdvaMed and Medical
Device Manufacturer Association (MDMA) – was first selected for the content analysis
because the two documents represented the collective voice of a broad group of
stakeholders in the medical device industry. To broaden the content analysis beyond
comments from the medical device industry, a supplemental analysis was performed
using public comments from the American Medical Association (physician professional
association) and American’s Health Insurance Plan (national insurance trade association).
Thus, a total of N=4 public comment packages were analyzed for the content analysis.
The public comments analyzed are shown in Table 5.

Table 5: Public comment packages used for content analysis
Respondent Type Source
AdvaMed Trade Association
http://www.cms.gov/medicare-
coverage-database/details/mcd-
view-public-
comments.aspx?MCDId=23&Ex
pandComments=n&bc=AAAAA
AAAAIAAAA%3d%3d&

Medical Device
Manufacturer Association
(MDMA)
Trade Association
American Medical
Association (AMA)
Physician Professional
Association
American’s Health
Insurance Plan (AHIP)
National Insurance Trade
Association

To conduct the content analysis from the public comments, I focused on the
complete set of textual data, but did not consider in what order or manner different
elements of the text were introduced. I conducted the inductive content analysis
according to the methodology described by Elo and Kyngas (Elo & Kyngas, 2007). For

89
the open coding step, I wrote categories in the public comments while reading them. I re-
read the public comments to confirm the consistency of the categories and edited them as
necessary. I then created a category table in Microsoft Word where each column
represented the respondent of the public comment and each row represented the specific
questions or areas of interest for the content analysis. I added one additional row to
capture any general comments, if applicable. I transferred the categories to the table
according to the specific questions or areas of interest to which they belonged. The
categories in the table were sorted several times in attempt to group them into a small
number of themes each containing a number of different categories. For example,
categories “understand types of evidence needed” and “designate liaisons” were grouped
under theme “parallel review should encourage three-way collaboration.” Under each
theme, I counted the number of times that the higher-level theme was addressed and the
number of times that a specific category of response was identified. For the purpose of
reporting the representativeness and importance of the same category with respect to
different questions or areas of interest, a specific category was counted more than once, if
applicable, within the same public comment. Finally, each theme and its categories were
related back to the constituents (causation, intervention, and evaluation) of the
conceptual framework.
3.1.2 Expert Review of Interview Tool
A small group of individuals with expertise relating to FDA and/or CMS policies
were asked to examine and provide constructive feedback on the content validity and
overall instrument design of the interview questions. In order to obtain a broad
perspective, the experts included a mixture of individuals who have worked with devices

90
for rare diseases in the regulatory or reimbursement fields. A total of N=4 regulatory
experts and N=3 reimbursement experts were requested to provide feedback with regard
to the interview questions in May 2013. The professional backgrounds of the experts are
illustrated in Table 6.

Table 6: Description of experts’ professional background
Regulatory Professional Background Reimbursement Professional Background
1. Retired Vice President of Regulatory
Affairs for a major medical device company.
25+ years of experience in the
pharmaceutical and medical device industries
including expertise in orphan drugs and
HUDs.
1. Senior Director of Reimbursement for a
major medical device company. 15+ of
experience in the medical device industry,
including 6-9 years of direct work experience
in reimbursement for HUDs.
2. Expert Regulatory Affairs Advisor for a
major medical device company. 20+ years
of experience in the medical device industry
including expertise in HUDs.
2. President of a medical device
reimbursement consulting firm. 20+ years of
experience in coding, coverage, and payment
strategies for medical devices, including
HUDs.
3. Vice President of Regulatory Affairs for a
major medical device company. 15+ years
of experience in the medical device industry
including expertise in HUDs.
3. Program Director and Associate Professor
at a university. 25+ years of experience in
coding, coverage, and payment strategies for
medical products.
4. Vice President for a consulting firm.
Former Associate Director at Office of
Device Evaluation (ODE) of FDA. 25+ years
of experience in the medical device industry
including expertise in HUDs.

Two sets of interview questions were constructed according to triad of
constituents of policy-making – causation, intervention, and evaluation. One set of
interview questions focused on regulatory aspects of HUDs and the other set centered on
reimbursement aspects of HUDs. While two sets of interview questions in some instances

91
overlapped, the questions were intended to explore the degree of cohesiveness and
alignment needed between the FDA and CMS to achieve a state of regulatory
convergence to support patient access to HUDs. The interview questions were
distributed to each expert approximately two weeks in advance to ensure that he or she
had sufficient time to read the questions. The discussion was conducted over the
telephone or e-mail according to the preference and availability of the experts. Two out
of 4 regulatory expert reviews were conducted via e-mail; the others were conducted by
telephone for approximately 60 minutes. For reviews conducted over the telephone, the
discussion started with a brief overview of the purpose and framework of the exploratory
research. Each interview question was then discussed sequentially. For reviews
conducted via e-mail, multiple written correspondences helped to ensure that discussions
were complete and rich. Ideas and suggestions were noted throughout the expert reviews.
Information gained from the reviews was considered and used to update the interview
questions appropriately. The final 17-18 interview questions, most of which used a
variety of “open-ended” and Likert-type questions, were used as an interview guide.
Copies of the regulatory and reimbursement interview questions are provided in
Appendices B and C, respectively.
3.1.3 Deployment and Analysis of Informant Interviews
Using the interview questions as a guide, this exploratory research study involved
semi-structured interviews of key informants who have experienced with regulatory
and/or reimbursement aspects of HUDs. Because the number of individuals experienced
with HUDs is inherently small, extensive search efforts were employed in various public

92
domains to recruit qualified key informants. The individuals identified as potential
interviewees included:
• Owners of HDEs published on the OOPD website.
• Recipients of grants from the Pediatric Device Consortia program published on
the OOPD website.
• Attendees of past FDA meetings such as the OOPD pediatric medical devices
workshop conducted on September 24, 2012, and FDA's public workshop on
"Identifying Unmet Public Health Needs and Facilitating Innovation in Medical
Device Development” on June 24, 2010.
• Individuals found on LinkedIn using the following keywords: humanitarian use
device, humanitarian device exemption, HUD, HDE, rare disease, and orphan
device.
• Individuals viewing information on posted discussion on LinkedIn and Pediatric
Device Consortia (http://www.pediatricdeviceconsortium.org).
• Individuals identified by searches on Google.com using the following keywords:
humanitarian use device, humanitarian device exemption, HUD, HDE, rare
disease, orphan device, 10-(K) filing and humanitarian use designation, HDE
reimbursement, etc.
In addition a request was made to the OOPD to obtain a list of the HUD
designation owners. However, the OOPD was reluctant to provide the information,
indicating confidential nature of the information, and suggested use of FOI pathway to
obtain the information. On September 21, 2012, FOI Services (www.foiservices.com)
filed a FOI request on my behalf to seek information with regard to HUD designation
owners. Unfortunately, to date (January 2014) the FDA has yet to respond to the FOI
request (Soo, personal observation).
This exploratory research study then utilized a snowball technique to recruit key
informants whereby the existing key informants were asked to recommend acquaintances
who might be qualified to participate in the interviews. As a result, a total of N=31

93
regulatory key informants and N=12 reimbursement key informants participated in the
study from July 1, 2013 through November 30, 2013. Examples of invitation e-mails to
informants during this timeframe are provided in Appendix D. The recruitment of key
informants was stopped when I felt that the interviews were reaching the point of
information saturation, defined as the point when collection of new data did not appear to
add to the questions being researched (Mason, 2010).
After key informants agreed to participate in the research study, the background
of the research study and the interview questions were sent for review prior to the
interview. The informants were asked to provide responses to Likert-type questions, and
any other questions that they felt were appropriate according to their areas of expertise
(e.g. HUD designation and/or HDEs) to focus the interview and thus improve its
efficiency. Semi-structured interviews were conducted for approximately 60 minutes
using a Cisco WebEx online meeting tool or by direct phone calls, whichever method
each informant preferred. The interview questions and responses provided by the
informant were used as a starting point from which more detailed and in-depth responses
were solicited from the key informants. Verbal permission was obtained from the key
informants to record the conversation for interview transcription at a later date for
interviews done using the Cisco WebEx tool. During each interview, I started with casual
conversation then described the research framework and questions, explained the
confidential nature of the interview, and clearly stated that questions might be skipped if
they were outside the expertise of the informants. Some key informants preferred to
provide responses in writing instead of being interviewed by phone or in-person. In these

94
instances, multiple correspondences via e-mail were conducted to ensure complete and
rich responses.
Results from the semi-structured interviews were collected and stored
electronically. If available, audio recordings of interviews were replayed multiple times
and transcribed to capture appropriate segments of text. The results from the interviews
were grouped by triad of constituents – causation, intervention, and evaluation – related
to policy-making. Within each constituent, the results were analyzed for content and for
any trends or commonality.

95

Chapter 4: Results
4.1 Research Approach 1: Content Analysis of Public Dockets
4.1.1 Content Analysis: Parallel Review of Medical Products
Results of the content analysis of Food and Drug Administration (FDA) public
docket FDA-2010-N-0308 suggested that the questions asked by the FDA were
concerned primarily with one part of the policy-making that has been utilized as a
framework for this study. As shown in Table 7, all 17 questions from the public docket
focused on the intervention constituent, as might be expected given the fact that the
parallel review process under consideration was a joint implementation of the FDA and
Centers for Medicare and Medicaid Services (CMS) statutory requirements. However,
some questions might have implications for other aspects of policy setting, and they were
therefore assigned the additional identifier of causation.

Table 7: Grouping of questions in public docket using a modified triad of constituents
Questions Constituent
1. Should anyone other than the product sponsor be able to initiate a request for
parallel review (for example, the FDA, CMS, an interested third party)?
Intervention
2. For which classes of products would consumers, payers, or sponsors benefit
most from parallel review? Why?
Intervention/
causation
3. FDA and CMS may propose to limit the number of products concurrently
under parallel review. How should limits be placed on the number and/or type of
products concurrently under parallel review? Should CMS be permitted to
review indications for which the sponsor is not seeking FDA clearance or
approval under parallel review?
Intervention/
causation
4. Are there disadvantages to parallel review? Intervention
5. Are there any barriers (for example, regulatory, legal, scientific) to parallel
review and if so, how might they be overcome?
Intervention/
causation

96
Questions Constituent
6. Should a voluntary process be put in place to encourage the conduct of
clinical trials that are appropriately designed to support both FDA
approval/clearance and CMS national coverage decisions? If so, what process
should be established?
Intervention
7. What criteria should the FDA and CMS use to decide whether to grant a
request for parallel review?
Intervention
8. At what point during FDA premarket review for prescription drugs, biologics,
and medical devices, should parallel review begin in order to reduce the time
between FDA marketing approval or clearance decisions and CMS national
coverage decisions while avoiding the risk that CMS would initiate an NCD for
a product whose premarket application the FDA subsequently does not approve
or clear?
Intervention
9. How should parallel review be implemented? Should the agencies use means
in addition to a guidance document, such as designating agency liaisons, to
educate sponsors about parallel review?
Intervention
10. When, if at all, should the agencies offer joint meetings to interested
sponsors during parallel review? Before parallel review begins? Before a
premarket application is submitted to the FDA?
Intervention
11. Should FDA and CMS have access to the same data and information about
the product during parallel review?
Intervention
12. It is CMS' policy to inform the public when it begins an NCD process for a
particular product. However, under applicable statutes and FDA's regulations,
the existence of a premarket application is considered confidential commercial
information prior to approval or clearance unless the sponsor has publicly
acknowledged the application. With the consent of the sponsor, should CMS
make public that it has begun the NCD process, as part of parallel review, for a
product still undergoing FDA premarket review? As a condition of the agencies'
agreement to initiate parallel review, should a sponsor have to inform the public,
or consent to the agencies informing the public, that the product will be
evaluated under parallel review? If the sponsor declines to consent to disclosure,
should it be permitted to request parallel review anyway, which would prevent
CMS from disclosing the NCD process until after the product is approved by the
FDA? How can the transparency of CMS' NCD process be reconciled with the
need to retain confidentiality of certain commercial information?
Intervention/
causation
13. At present, sponsors whose medical products will undergo both FDA
premarket review and CMS national coverage review submit separate
application packages to FDA and CMS that, in part, contain the same data, and,
in part, contain different data. Keeping in mind the limited resources available to
the agencies, what steps can the agencies take to minimize duplication of data
submissions? Would the use of electronic submissions reduce submission
burdens and facilitate data transfers? Are there other steps the agencies can take
to streamline a parallel review process without modifying the regulatory
standards and evidentiary requirements of both agencies? Would the
transparency of CMS' NCD process subject the FDA to additional public
Intervention

97
Questions Constituent
pressure regarding marketing authorization?
14. Should the agencies convene a joint advisory committee to consider
common issues needing public discussion and advice during the parallel review
process?
Intervention/
Causation
15. What other concerns or considerations should the agencies take into account
when developing a process for parallel review?
Intervention
16. Once FDA and CMS have opened a parallel review should a sponsor be able
to terminate or withdraw the request for parallel review? If this happens, should
that information be made public?
Intervention
17. Sponsors who submit a PMA or 510(k) to the FDA generally must pay a
user fee. One key advantage of parallel review is to streamline the current
process by allowing engagement by a sponsor with both FDA and CMS
concurrently. Earlier engagement could shorten the time between FDA approval
or clearance of the PMA or 510(k) and a coverage decision from CMS. Parallel
review could, however, entail additional costs for the agencies (for example, if
the product ultimately does not receive FDA approval or clearance). Changes to
a user fee would also require legislative changes. Given these factors, should the
current Medical Device User Fee be restructured to support the FDA and CMS
costs of this parallel review and if so, how?
Intervention

Content analysis of the responses in the studied submissions to the docket resulted
in 9 overarching themes associated in some cases with two or more subordinate
categories (Table 8). Comments related to one specific category often recurred
throughout the text of the submission because the point of discussion had relevance to
more than one of the questions for which the FDA had solicited input. To capture the
essence of the parallel review public docket, opinions expressed in the most prominent
themes are described further below.

98
Table 8: Themes and subthemes in the “Parallel Review of Medical Product” public
docket
Constituent Theme Category
Count
Total
Count
Causation Statutory challenge:
CMS’s NCD is a
public process. FDA’s
review is confidential
Difficult to overcome without statutory
change
2 2
Legal framework:
FDA & CMS have
different statutory
authorities and
evidentiary standards

Different statutory standards 11*
13
CMS might influence FDA’s review of
products
2
Intervention Voluntary
participation: Parallel
review should be
voluntary
Voluntary process 4*
14
Initiate by sponsor 5*
Withdraw by sponsor 3*
Slow down FDA and CMS 2*
Confidentiality:
Parallel review should
protect confidentiality
of information
Protect trade secret & proprietary
information
4
7
Share information at consent of
submitting sponsor
2
Maintain confidentiality of existence of
PMA
1
Collaboration:
Parallel review should
encourage three-way
collaboration
Understand types of evidence needed 7*
10
Designate liaisons 2
Define clinical evidence needs through
joint advisory meeting
1
Linkage to coverage:
Parallel review should
not be automatically
linked to NCD
Allow coverage through LCD. Address
payment and coding
3
13
Link to NCD automatically delays
patient access
10
Innovation:
Parallel review is
suitable for innovative
products for which
reimbursement
pathway is not clear
Apply when non-coverage policy exists 4
12
Apply when uncertainty in evidence
exists
1
Apply when impact on patient
outcomes is significant*
1
Apply when unmet medical needs exist
or on life-saving technology
6
Evaluation Measurement:
Metric of parallel
review should be
implemented

Implement metrics
1 1
Analysis:
Benefit and efficiency
of parallel review
should be assessed
Adjust the process after further
comment
1 1
* Occurred in the supplementary content analysis of Cystic Fibrosis Foundation & Blue Cross Blue Shield

99
Parallel Review Should Be Voluntary
All submissions strongly expressed the view that parallel review should be a
voluntary process, in which the product sponsor should retain the right to initiate or
withdraw the parallel review at any time. For example, Blue Cross Blue Shield stated the
following:
We recommend that sponsors be required to consent to the agencies informing the
public of a product’s acceptance for a parallel review and of the outcomes of each
agency’s assessment – whether positie or negative. A sponsor should be able to
withdraw from the parallel review at any time…(FDA-2010-N-0308, 2010b).
FDA & CMS Have Different Statutory Authorities and Evidentiary Standards
All respondents were aware that FDA and CMS had different evidentiary
standards associated with the review of medical technologies; for FDA, reviews were
guided by “safety and effectiveness” but for CMS, review focused on the “reasonable and
necessary” standards. They expressed the view that a parallel review process must not
compromise these different evidentiary standards. The specific comments from
AdvaMed and Cystic Fibrosis Foundation, quoted below respectively, reflect these
consonant views.
FDA and CMS operate under separate and distinct statutory standards with
completely different regulatory mandates and processes. FDA is a regulator with
appropriate emphasis on public health by ensuring the safety and effectiveness of
a new medical technology, whereas CMS is a third party payer of heath care
services for Medicare beneficiaries…AdvaMed is concerned that creating a
parallel review process may result in CMS influencing or providing input into

100
FDA’s review of products, determinations of safety and efficacy, or orders.
(FDA-2010-N-0308, 2010a)
In fact, the agencies are answering different questions through their review
processes…By introducing into the FDA review process questions about the
standards for coverage of a new products as well as its price, the parallel review
approach could slow the review and could effectively put FDA review team
members – hired and trained to focus on issues of safety and efficacy – in the
position of evaluating these additional issues. (FDA-2010-N-0308, 2010c)
Parallel Review Should Not Be Automatically Linked to a National Coverage Decision
The view was expressed that automatic linkage of parallel review to subsequent
review to establish or deny a National Coverage Determination (NCD) could compromise
patient access to life-saving medical technologies. At the time of FDA approval, the type
and level of evidence available to CMS may be insufficient to support a coverage
decision and this could potentially lead to a premature non-coverage determination.
Instead, respondents felt that Local Coverage Determination (LCD) is and should be the
dominant route for determining coverage decisions. According to AdvaMed and Johnson
& Johnson, respectively:
Medicare’s local coverage process is the most appropriate avenue to determine
coverage for advances in medical technology. AdvaMed is concerned that
subjecting new medical technology to the national coverage process (either during
pre-market review or immediately following clearance or approval) could delay
patient access to important advances and have overall negative impact on
innovation…A parallel review process should allow manufacturers to obtain
coverage through Medicare’s local coverage process. (FDA-2010-N-0308, 2010a)

101
A parallel FDA-CMS review should be voluntary, as proposed in the notice, and
that it be de-linked from a national coverage decision or NCD entirely. Currently,
Medicare’s coverage process is decentralized and relies on Local Coverage
Determinations (LCD) by the Medicare Administrative Contractors in most
cases…Parallel review that is not automatically linked to the initiation of an NCD
would be highly valuable in encouraging the conduct of clinical trials that are
designed to support both FDA’s and CMS’s needs. (FDA-2010-N-0308, 2010d)
4.1.2 Content Analysis: CMS 2012 Draft CED Guidance Document
The comments to “Draft Guidance for the Public, Industry, and CMS Staff
Coverage with Evidence Development in the Context of Coverage Decisions” could be
grouped into 5 overarching themes (including a miscellaneous category) and their
subordinate categories as shown in Table 9. The themes were considered to fit more
readily under the intervention constituent of the conceptual framework. The three most
frequently expressed themes are described further below.

102
Table 9: Results of the content analysis of CMS’ draft CED guidance document
Constituent Theme Category Count
Total
Count
Intervention
CED should be
used
infrequently
within the
NCD
Use when non-coverage is the only alternative 1
13
Use when a promising technology lacks
sufficient evidence
1
Promote infrequent use of CED 4
Re-instate the 8 principles from the 2006
guidance
2
Implement CED within NCD. NCD should be
transparent and open
4
Foster transparency to make coverage
decision
1
Intervention
CED study
design should
be prescriptive
and flexible
Define study endpoints, duration, and research
questions
7
21
Require minimally necessary data 2
Consider budget constraints 1
Do not restrict CED study design. Allow
flexibility in evidence that is tailored towards
unique patient population
5
Use for treatment for conditions that occur
infrequently
1
Allow use of intermediate outcomes 1
Do not use superiority as goal 1
Define evidentiary standards in general terms 1
Improve clarity in evidentiary standards 2
Intervention
CED should
not hinder
coverage
through LCDs
Allow data collection via LCD to continue 1
8
Do not allow CED with Medicare
Administrative Contractors (MAC)
2
Don’t hinder local coverage 5
Intervention
CED should
ensure
continued
coverage
following
completion of
study
Continue coverage while CED is conducted or
ended
3
5
Establish a clear pathway of reassessment of
coverage after CED ended
1
Intervention

Miscellaneous

Reverse coverage decision to invoke CED
would prevent access
1
10
Engage stakeholders early 2
Avoid duplication of efforts of other
regulatory agencies
1
Outline anticipated areas of CED 1
Concern if CED intervenes with practice of
medicine
1
Obtain inputs from private payers 1
Define roles of AHRQ 3

103
CED Study Design Should be Prescriptive and Flexible
AdvaMed and MDMA expressed the view that the Coverage with Evidence
Development (CED) study design should be prescriptive. Specifically, the study
objectives, endpoints, patient sample size, and analysis should be defined at the onset of
each CED study to avoid otherwise costly and burdensome changes and execution issues.
They suggested that CMS should provide guidance on the minimum study data necessary
to answer the specific clinical questions needed for the subsequent assessment by CMS.
Furthermore, they commented that the study design should be flexible because some
patient populations might be difficult to study using randomized controlled trials, and
other study designs may be more appropriate. According to AdvaMed and American’s
Health Insurance Plan (AHIP), respectively:
CED-required data collection should have a well-defined endpoint established
before the data collection begins, as determined by the specific protocol, and we
recognize that flexibility will be important in this regard. (U.S. Centers for
Medicare and Medicaid Services, 2013b)
To be effective, CED needs additional clarity in considering health services that
are promising but lack evidentiary threshold for coverage. These services include
treatments that are part of a well-designed clinical trial…and treatments for other
conditions that occur so infrequently that it is unlikely that a body of published
evidence will be developed within a reasonable period of time. (U.S. Centers for
Medicare and Medicaid Services, 2013b).

104
CED Should be Used Infrequently as part of the NCD Process
The respondents believed that the benefit of CED to Medicare beneficiaries would
be maximized by continuing to apply some of the key principles stated in the 2006
guidance document: (a) the CED should be used infrequently and only when the available
evidence does not justify other forms of coverage, and (b) the CED should continue to be
applied within the NCD process, which is transparent and open to the public. As noted
by the MDMA:
In general, MDMA believes that the thresholds for invoking CED should be
established at levels that ensure that CED continues to be used infrequently and
not when other forms of coverage are justified by the available evidence,
consistent with the principles established in CMS’s 2006 guidance document on
CED. (U.S. Centers for Medicare and Medicaid Services, 2013b)
CED Should Not Hinder Coverage through LCDs
The submissions expressed the view that the CED should be executed within the
NCD, and not as part of local coverage decisions. They felt that local Medicare
Administrative Contractors (MAC) should not be given the authority to implement CED
due to the potential of duplicative studies across MACs. AdvaMed expressed the
following concern:
We recommend additional clarification that CED will only be used within the
NCD process, and that CMS does not anticipate allowing Medicare contractors
greater authority to carry out CED locally…noted that local coverage is often a
pathway for beneficiary access in the absence of national coverage. However, we
have strong concerns about any change in policy that would give local contractors
greater authority to restrict or limit coverage through CED. In addition, we

105
continue to have concerns about the potential for multiple, conflicting data
collection requirements that might result from this action. This would be
extremely burdensome and costly to manufacturers, and CMS should ensure that
such circumstances are avoided. (U.S. Centers for Medicare and Medicaid
Services, 2013b).
4.2 Research Approach 2: Expert Review of Interview Questions
In the second phase of the study four regulatory experts and three reimbursement
experts were individually approached to provide constructive feedback on a draft set of
interview questions to be used to extend the analysis of industry views. The focus of the
expert reviews was on assuring that the questions were relevant to the current policies
and would yield clear results. From the reviews, regulatory experts suggested that (a) the
distinction between laws and regulations related to HUDs should be clearly articulated in
the interview questions, (b) the interview guide should be semi-structured to limit the
number of “open-ended” questions, (c) views regarding the “probable benefit” regulatory
requirement should be explored, (d) the way that the regulatory opportunities affect
business strategy should be explored, (e) the impact of the FDASIA 2012 provision
related to HDE should be examined, (f) the IRB review requirement associated with the
use of HUDs should be explored, and (g) the post approval requirements related to HDEs
should be explored. In addition, reimbursement experts indicated that (a) the utilization
of NCD vs. LCD mechanisms to secure reimbursement of HUDs should be examined, (b)
the broader strategy to secure reimbursement of HUDs should be explored, and (c) the
potential for the CED reimbursement pathway to ensure better patient access to HUDs
should be addressed. Some regulatory and reimbursement experts shared a concern that

106
the FDA and CMS parallel review initiative was relatively new and the research might
yield limited information with regard to current practice, but it was nonetheless an area
worth exploring. These suggestions were considered and used to update the interview
questions accordingly. The regulatory and reimbursement interview tools are provided in
Appendices B and C, respectively.
4.3 Research Approach 3: Key Informant Interviews
The analyses of the key informant interviews are presented in four sections.
Section 4.3.1 presents the profiles and demographic information of the key regulatory and
reimbursement informants. The results of the causation, intervention, and evaluation
constituents are described in section 4.3.2, 4.3.3, 4.3.4, respectively.
4.3.1 Profiles of Key Informants
Thirty-one individuals participated in the interviews. Most of these individuals
held positions as consultants (26%) or vice presidents (26%). Six out of the 31 key
regulatory informants were physicians/professors (19%), and the remaining individuals
were directors, senior managers, specialists, and CEOs (Figure 10).

107
Figure 10: What is your current job title?

The key regulatory informants were asked to estimate the annual revenue of their
place of business when they were involved in HUDs (Figure 11). Most worked for mid-
size companies generating up to 10 million dollars annually (39%), or very large
companies generating over 1 billion dollars annually (35%). Five worked for start-up
companies with no revenue (16%), and the others worked for large companies generating
between 10 million and 1 billion dollars annually (10%).

108
Figure 11: What is/was the annual revenue of your place of business?

The informants were also asked to provide the status of the humanitarian use
device (HUD) designation and/or humanitarian device exemption (HDE) held by their
current or former place of business. Most commonly (52%) they worked for companies
with both HUD designations and approved HDEs. In this group, one individual stated
that the approved HDE on which she worked was not commercially available because the
company ran out of funding and therefore was no longer in business. Another individual
indicated that his company was acquired, and the three HUD designations he obtained
have not progressed to HDE applications with that new parent company. A second group
of nine informants (29%) worked with businesses that had obtained HUD designations
successfully and were working on HDE applications. The remaining informants
indicated that their place of business did not have HUD designations or HDEs (19%).
This group of informants consisted of (a) consultants who helped their clients in
obtaining HUD designations and/or HDEs, (b) former FDA employees who were

109
involved in implementing HDE guidance or policies, and (c) individuals who are
currently working to obtain HUD designations.
The majority of the regulatory informants had more than more than 15 years of
medical device experience (71%) (Figure 12(a)). Typically, this group of informants
also had significant experience with regulatory aspects of HUD designations/HDEs
(Figure 12(b)). Most had more than 10 years of regulatory experience in HUD
designations/HDEs (45%).
Figure 12(a): How many years of medical device experience do you have?

Figure 12(b): How many years of HUD regulatory experience do you have?

110
Twelve key reimbursement informants also participated in the study. Most
commonly, the informants were consultants (42%), but job levels of vice president
(25%), director (17%), manager (8%), and specialist (8%) were also sampled (Figure 13).
Figure 13: What is your current job title?

When working with HUDs, most reimbursement informants worked as
consultants for mid-size business generating up to 10 million dollars annually (58%).
Three worked for large businesses generating over 1 billion dollars annually (25%). The
remaining individuals worked for businesses generating between 10 million and 1 billion
dollars annually (17%). All had more than 6 years of experience in the medical device
industry, and most had more than 15 years of medical device experience (77%) (Figure
14(b)). Most had more than 10 years of reimbursement experience in HUDs
(75%)(Figure 14(b)).

111
Figure 14(a): How many years of experience do you have in the medical device
industry?

Figure 14(b): How many years of experience do you have in HUD reimbursement?

4.3.2 Constituent of Policy-Making: Causation
Issues of FDA policy related to HUDs that appeared best related to the element of
causation could be grouped into five areas of concern, including: (a) regulatory standards
related to “probable benefit,” (b) requirements for IRB approval, (c) impact of FDASIA
2012 related to lifting the profitability restriction, (d) incentives necessitating legislative
changes, and (e) factors that contributed to failure of products to progress from HUD

112
designation to HDE application. Issues of CMS policy related to causation included: (a)
the statutory determination of “reasonable and necessary” and (b) factors that might have
contributed to reimbursement challenges of HUDs. Further, the potential for FDA and
CMS to jointly design HUD policy as a “system” was examined.
4.3.2.1 FDA
“Probable Benefit” Regulatory Standard of HUDs
Key regulatory informants were asked whether, in their view, the “probable
benefit” regulatory requirement makes sense and how this requirement should be
established for HUDs. Many informants were not clear about the regulatory definition of
“probable benefit,” and noted that a definition was not specified in the statute and
regulations. Nevertheless, they typically advocated for a flexible regulatory approach to
the interpretation of “probable benefit” requirements. Although the HDE regulations
require that a HDE application contain a description of the suitability of the device for the
rare disease or condition and an explanation of why the probable benefit from the use of
the device outweighs the risk of injury or illness from its use, the informants felt that no
“one size fits all” definition would accommodate all medical devices. When probed
further as to how the “probable benefit” regulatory requirement should be established for
HUDs, informants had specific suggestions for decision-making: (a) consider mode of
action of the device in relationship to relevant physiology/pathology, (b) identify
clinically meaningful parameters, and lower the normal statistical bar for meeting clinical
endpoints, (c) use preclinical data and/or early feasibility clinical data as a basis for some
determinations, (d) permit alternative clinical evidence such as peer-reviewed journal or

113
historical control studies, and (e) use surrogates or measurable improvements, when these
are available and relevant. For example, one of the key informants stated:
In the device world, many populations are so small as to not make it feasible or
even possible to develop such data without inordinately delaying the ability to
supply therapy to such patients who are without other alternatives. That is why the
HDE was developed. It is very important and sensible. Enough safety data is
required to assure an understanding of the risk and no one can guarantee efficacy,
even after large trials, just an expectation of the size of the benefit and the percent
of a population that might benefit. Therefore, probable benefit is fine as an
endpoint for small orphan populations who otherwise have no chance for
improvement. With an HDE they at least have a chance for benefit with
understood risk. Probable benefit comes from there being indicators of
improvement in the affected population. These can be surrogates or measurable
improvements. The probable means it is not statistically confirmed…not that it is
not real. (Consultant)
Further, three of the key informants strongly recommended that the phase
“probable benefit” be revised to “possible effectiveness” or “acceptable level of
effectiveness” for HUDs in light of reimbursement hurdles associated with the
insufficient evidence of clinical effectiveness.
IRB Approval Requirement of HUDs
Most regulatory informants shared a similar view with regard to the effect of the
IRB approval requirement on patient access to HUDs. Of the 31 regulatory informants,
three did not have direct experience in dealing with IRB approvals of HUDs and felt
unable to comment usefully. Most of the remaining informants (23/28, 82%) felt that the
requirement for IRB approval discourages patient access to HUDs because the approval

114
process can be redundant, arduous, and costly at least in part because some IRBs often
were not familiar with the role assigned to them by the FDA. Some informants elaborated
that although some IRBs have become more knowledgeable about the HUD review
requirement, this remained a challenge. For example, one of the informants stated the
following:
The IRB approval requirement for HDEs is a bit arduous for many hospitals.
When I started working with HDEs in 2001 initially, it was difficult for hospitals
to understand why the IRB approval was required for an HDE approved by FDA,
but it is much better understood and accepted today. It is somewhat burdensome
for manufacturers to obtain and track IRB approvals. (Vice President)
A smaller number of informants (5/28, 18%) had a different view. They believed
that the requirement neither discouraged nor encouraged patient access to HUDs. Some
argued that HUDs serve such a small patient population that sponsors are usually savvy
in finding facilities that will deal with such patients and can facilitate the appropriate
approvals. A few saw this as a non-issue since the fact that it is a regulatory requirement
mandated by the Food, Drug, and Cosmetic Act to protect patients should trump the fact
that it might discourage patient access.
When asked if the IRB approval requirement adds value, the views of informants
were split fairly evenly. Some (43%, 12/28) felt that the requirement does add value, and
the others (50%, 14/28) were unconvinced. A couple had no opinion (7%, 2/28). The 26
with opinions were encouraged to explain why they felt the IRB requirement adds or does
not add value. Those who advocated for its value felt that this value came from providing
(a) oversight and protection to the patients because the effectiveness of HUDs has not

115
been demonstrated, (b) safeguards to assure that bad “actors” do not abuse the reduced
regulations, and (c) benefits to stakeholders such as hospitals and physicians other than
patients. As an example, one individual stated that:
The value added by the IRB is actually mostly for the hospitals and doctors to
understand the nature of the data that supported the HDE approval, to assure the
right patients are selected and the products are used properly. (Consultant)
Those who felt that the IRB requirement does not add value had two primary
arguments: (a) in most cases, the patients may not realize the reasons why IRB oversight
is needed, especially when they have no alternative treatment options other than HUDs,
and (b) an HDE constituted a marketing approval from the FDA. The review and
approval of an IRB is not the correct forum to oversee a marketing approval. On this
note, one of the former FDA employees who was involved in implementing HDE
guidance and policies stated the following:
I don’t feel that the IRB requirement adds value…the HDE program was
originally designed to be like an investigational device so it was going to be
governed under the IDE regulations. I think that is how the IRB approval
requirement became associated with HDEs…Before the HDE regulations were
finalized; there was discussion that the HUDs should be governed under the PMA
regulations as commercial products. However, the “strange” IRB approval of
HUDs remained…sometimes the way the regulations are written is inconsistent.
The legislators might not realize the implication of IRB approval of HUDs…I
really think that is why we have the IRB approval of HUDs. The requirement
makes sense for IDEs, but it really confuses IRBs as they have no ideas why they
were asked to review and don’t know what to do. (Former FDA employee)

116
FDASIA 2012: Lifting of HDE Profitability Provision
Twenty nine informants provided opinions on the extent to which the legislative
change to lift of the profitability restriction on HDEs (per FDASIA 2012) would
encourage innovation in development of HUDs and/or progression from HUD to HDE,
using a 5-point scale ranging from high positive impact to high negative impact (Table
10). Overall, a majority of the key informants felt that the legislative change was
directionally positive in spurring innovation and development of HUDs to meet unmet
medical needs.

Table 10: Impact of FDASIA 2012 HDE lifting of profitability restriction provision
Category
High
Positive
Impact
Positive
Impact
No
Impact
Negative
Impact
High
Negative
Impact
Total
Pursue HUD/HDE
regulatory pathway
11
(38%)
14
(48%)
4
(14%)
0
(0%)
0
(0%)
29
Apply for HUD
designation
11
(38%)
13
(45%)
5
(17%)
0
(0%)
0
(0%)
29
Move forward with
HDE marketing
application
13
(45%)
10
(34%)
6
(21%)
0
(0%)
0
(0%)
29
Commercialize HDE
devices
11
(38%)
14
(48%)
4
(14%)
0
(0%)
0
(0%)
29
Note: The shaded cells illustrated the most common selection by the informants.

More than 80% of the informants felt that the legislative change had a positive
impact or highly positive impact on the decisions of companies to “pursue an HUD/HDE
regulatory pathway” (25/29) and “apply for HUD designation” (24/29). Most (79%) also
felt that the legislative change would have a highly positive or positive impact on
decisions to “move forward with HDE marketing application” (23/29). More than 85% of

117
the informants felt that the legislative change had positive impact on decisions to
commercialize HDE devices.
The ability to realize a profit from HDE products was considered an important
part of initiating and sustaining product viability in the market by most informants.
Certain informants indicated that they had a HUD with pediatric indication and so had
been able to make a profit under provision of FDAAA in 2007. Without it their business
could not exist. In addition, informants felt that without profit only some larger
companies could absorb the costs of pursuing the HUD. However, a few informants felt
that the legislative change had little or no impact, and three reasons were provided to
justify this position: (a) an HDE was a “stepping stone” to a PMA, so profitability might
not be a concern for the sponsor at the HDE stage, (b) the legislative change might have
little impact at this point because the legislation was relatively new and many sponsors
may not yet be aware of it, and (c) the business decision to proceed with an HDE was
usually not based on profitability restrictions, but instead on the overall cost and risk of
such an approval. Specifically, the ongoing costs of maintaining the approval, providing
service to the patients and physicians, and monitoring complaints were usually not worth
the incremental gain in revenue, with or without the profitability restriction.
Incentives Necessitating Legislative Change
Informants who were asked about potential legislative incentives that could spur
innovation in the development of HUDs provided a range of responses that could be
grouped into specific themes shown in Figure 15. The most common type of suggestion
(9X) focused on enhancing funding to sponsors in order to facilitate the necessary
development activities and preclinical study. Although FDA now provides a modest

118
funding mechanism to conduct clinical trials, some key informants emphasized that better
early funding for start-ups was key to spur development of HUDs. A second commonly
discussed approach (7X) centered on raising the limit of HUD devices from the current
target population of less than 4000, which was viewed without exception as being too
low. One of the key informants who had successfully obtained HDE approvals stated, “in
my discussion with FDA, they recognize this (4,000 limit) is based on U.S. population in
1990 and does not reflect our current population.” The ability to make a profit was
mentioned several times (6X) despite the fact that legislative change has already been
implemented to satisfy this concern. The remaining legislative incentives suggested by
the key informants are illustrated in Figure 15.

Figure 15: Legislative incentives to promote the development of HUDs
What legislative incentives could be provided to promote the development of HUDs to
fulfill unmet medical needs?

119
Factors that Influence Progression of HUD Designation to HDE Application
Informants were asked to evaluate the factors that might have contributed to
failure of products to progress from HUD designation to HDE application, using a 5-
point scale ranging from extremely influential to not influential (Table 11). Only 20
informants provided complete responses to the question and were therefore included in
the data analysis. Most influential factors were felt to be “lack of funding,” “difficulty to
accrue sufficient number of clinical subjects in a reasonable amount of time,” “high
clinical trial design expectation from FDA,” and “uncertainty in business
environment/strategy.” Viewed as less influential were the “limited reimbursement
codes,” “HDE profitability restriction,” and “lack of clinical evidence.” Further, some
key informants felt that “post-approval commitment” and “business decision post
acquisition” were two additional important factors influencing the progression of HUD
designation to HDE application.

120
Table 11: Factors that contribute to failure of progression of HUD designation to HDE
Factor
Extremely
Influential
Very
Influential
Influential
Moderately
Influential
Not
Influential
Total
Lack of funding
10
(50%)
4
(20%)
2
(10%)
2
(10%)
2
(10%)
20
Lack of clinical
guidance to meet the
“probable benefit”
standard
2
(10%)
4
(20%)
6
(30%)
4
(20%)
4
(20%)
20
Difficulty to accrue
sufficient number of
clinical trial subjects in a
reasonable amount of
time
5
(25%)
7
(35%)
5
(25%)
1
(5%)
2
(10%)
20
Long product
development cycle
4
(20%)
5
(25%)
3
(15%)
3
(15%)
5
(25%)
20
High clinical trial design
expectation (e.g.
randomized controlled
trial) from the FDA
5
(25%)
8
(40%)
1
(5%)
4
(20%)
2
(10%)
20
Uncertainty in business
environment/strategy
3
(15%)
9
(45%)
3
(15%)
2
(10%)
3
(15%)
20
HDE profitability
restriction
1
(5%)
2
(10%)
4
(20%)
7
(35%)
6
(30%)
20
Limited reimbursement
codes
1
(5%)
5
(25%)
3
(15%)
6
(30%)
5
(25%)
20
Lack of clinical evidence
1
(5%)
2
(10%)
4
(20%)
7
(35%)
6
(30%)
20
Note: The shaded cells illustrated the most common selection by the informants.
4.3.2.2 CMS
“Reasonable and Necessary” Statutory Interpretation for HUDs
All twelve reimbursement informants provided views on how the “reasonable and
necessary” statutory determination should be interpreted, and those views were mixed.
About one-third (33%, 4/12) felt that interpretation of “reasonable and necessary” needs
to be flexible and consistent with the regulatory standards imposed by the FDA given that
HUDs are intended to be used on small populations with no alternative treatment
modality. For example, one informant stated that CMS had been reluctant to allow lower
evidentiary bar for HUDs for small patient populations by stating the following:

121
The biggest issue with applying “reasonable and necessary” criteria for HUDs is
not related to the safety aspect, rather the “probable benefit.” They (CMS) do not
feel that they have the authority or ability to allow a lesser standard evidentiary
requirement to HUDs than 510(K)s and PMAs. (Senior Director) (Quotation
added by author)
A couple of the informants (2/12) felt that the “reasonable and necessary”
statutory determination is intentionally designed to be broad, and is therefore sufficient to
allow for coverage of HUDs when needed. Specifically, the informants felt that HUDs
should remain eligible to be covered by Medicare under the “reasonable and necessary”
statutes on a case-by-case basis at the discretion of local CMS contractors. One of the
informants viewed the flexibility of allowing local CMS contractors to make a
“reasonable and necessary” determination for HUDs as important, to ensure that HUDs
were not consistently excluded on the grounds that the HUD was not considered a full
FDA approval or clearance. When probed further, this informant stated the following:
Perhaps confirm that an FDA “HDE approval” equals to final FDA approval or
clearance from a Medicare coverage requirement perspective. For example, some
Medicare contractors might consider HDE approval as not FDA-approval or
clearance and therefore automatically deem HUDs as not reasonable and
necessary. (Consultant)
Another two (2/12) informants felt that the “reasonable and necessary” statutory
interpretation for HUDs should be no different than those for other medical devices or
services for the following reasons: (a) one could argue that HUDs are intended to be used
on patients with no alternative treatment options, which meets the “necessary” part of the
statutory determination. However, from CMS or a payer perspective, unless

122
demonstrated by clinical evidence, it is not known if the HUDs would provide efficacy
benefit to the patients. The informant stated “why would payers want to pay for a
technology that they have no idea if it is going to work for the patients?” and (b) one
could argue that HUDs are intended to be used on a small patient population so that the
budgetary impact for CMS should be small. Further, the informant indicated that CMS is
prohibited by law to consider cost as a factor to determine coverage policy regardless of
HUD or non-HUD status.
The remaining informants (4/12) each provided a unique response; therefore key
themes cannot be summarized.
Factors that Contribute to Lack of Reimbursement of HUDs
Eleven informants evaluated a number of factors that might have contributed to
the lack of reimbursement for FDA-approved HUDs (Table 12). The most influential
factors were felt to be “perceived investigational status of FDA-approved HUDs,”
“difficulty in gathering clinical evidence,” and “difficulty in gathering health outcome
evidence.” For example, one of the informants explained the importance of “perceived
investigational status of FDA-approved HUDs”:
HUDs come to market under different standards, with different data and the
oversight of an IRB. Most, if not all, medical directors equate IRBs with clinical
trials and anything in a clinical trial is considered investigational. Perhaps the
medical directors are being a little lazy here, but often they do not have the time
or the clinical expertise to delve into the nuances of HUDs. They may not even
have a claims history from insured who have the condition that is to be treated by
the HUD, so there is no imperative to deal with the issue. (Consultant)

123
In contrast, viewed as less influential were “IRB not understanding HDE
requirements,” “lack of National Coverage Determination (NCD),” and “conflicting
Local Coverage Determination (LCD).”
Table 12: Factors that contribute to lack of reimbursement of HUDs.
Factor
Extremely
Influential
Very
Influential
Influential
Moderately
Influential
Not
Influential
Total
Difficulty in gathering
clinical evidence
4
(36%)
2
(18%)
2
(18%)
3
(27%)
0
(0%)
11
Difficulty in gathering
health outcome evidence
3
(27%)
2
(18%)
4
(36%)
2
(18%)
0
(0%)
11
Perceived investigational
status of FDA approved
HUDs
7
(64%)
2
(18%)
2
(18%)
0
(0%)
0
(0%)
11
IRB not understanding
HDE requirements
0
(0%)
1
(9%)
2
(18%)
3
(27%)
5
(45%)
11
Lack of National
Coverage Determination
(NCD)
2
(18%)
1
(9%)
0
(0%)
3
(27%)
5
(45%)
11
Conflicting Local
Coverage Determination
(LCD)
1
(9%)
1
(9%)
3
(27%)
1
(9%)
5
(45%)
11
Note: The shaded cells illustrated the most common selection by the informants.
4.3.2.3 FDA and CMS as a “System”
FDA-CMS: Development of HUD Policy
Thirty-one regulatory and eleven reimbursement informants provided their views
on the success of collaboration between the FDA and CMS to design their policies for
HUD development and access. Overall, a majority of the informants felt that FDA and
CMS are two separate entities with different legislative missions and agendas; therefore
they would not be able to develop HUD policy together.
Most key regulatory informants felt that better integration and collaboration
between the two agencies is much needed within the existing statutory framework for
each agency. Suggested options included: (a) early discussion with FDA and CMS during
development to ensure a clinical test strategy that would address the requirements of both

124
agencies and eliminate redundant requirements, (b) streamlined review processes in both
agencies, (c) better alignment of priorities with respect to unmet medical needs such as
rare diseases between the two agencies, (d) creation of a HUD reimbursement pathway
by the CMS that integrates well with the FDA’s HDE regulatory pathway, (e)
implementation of a memorandum of understanding between the FDA and CMS to
ensure a positive coverage decision upon HDE approval, and (f) benchmarking the FDA-
CMS collaboration on the IDE approval process whereby the FDA will place all IDEs it
approves in either Category A (experimental – risk of the device type has not been
established) or Category B (investigational and/or non-experimental – safety and
effectiveness of the device type have been demonstrated or incremental risk is the
primary risk in question). The IDE category is written on the IDE approval letter and
forwarded to the CMS. Under the FDA-CMS IDE collaboration, the CMS does not cover
Category A devices but may provide coverage for Category B devices. One of they
informants who is a physician and inventor emphasized the importance of the
collaboration between the two agencies by stating the following:
If FDA and CMS coordinate to support the inclusion of cost-utility factors in the
decisions of medical device developers, that would be advantageous in pushing
developers toward devoting resources to a project only when its final outcome is
likely to be covered by CMS. However, if FDA/CMS coordination just makes the
already cumbersome process even more difficult, then that would be a significant
disadvantage, as it would discourage device development. (Physician and
Inventor)

125
A majority of the reimbursement informants felt that the FDA and CMS are two
separate entities with different legislative missions and agendas; therefore it is important
to maintain the division between controls over market entry and decisions on coverage
and payment. Some of the major concerns and issues highlighted by the informants were:
(a) the possibility of “crossover” of agencies in a narrow field like HUDs that might
allow the FDA and CMS to stray from their respective evidentiary standards, (b) the lack
of compatibility between FDA and CMS policies on HUDs in that FDA does not require
demonstration of effectiveness, (c) the role of Congress to mandate FDA and CMS to
design HUD policies jointly is not clear, and (d) the need for open communications,
enhanced information-sharing, and selective coordination between the two agencies
would certainly be directly positive, but is not well developed.

4.3.3 Constituent of Policy-Making: Intervention
Examination of the intervention constituent focused on a particularly wide range
of elements as identified in the earlier document analysis: (a) the application of the
“medically plausible subset” regulatory definition when making decisions, (b) the
implementation of FDA guidance and regulations, (c) the strategic business decisions to
pursue the HDE regulatory pathway, (d) the efforts to spur innovation within the existing
FDA regulatory framework, (e) the feasibility of Coverage with Evidence Development
(CED) for HUDs, (f) the feasibility of using a “NCD with CED” approach for HUDs, (g)
the assessment of LCDs and NCDs for HUDs, (h) the strategy to optimize dispersed LCD
for HUDs, (i) the nature of intervention to assist patients/hospitals to obtain
reimbursement, (j) the effect that lack of codes could have on the opportunity for

126
reimbursement, and (k) the issue of lack of reimbursement on device selection by
physicians. Further, the feasibility of FDA-CMS parallel review of HUDs as a “system”
was examined in detail.
4.3.3.1 FDA
Regulatory Definition of “Medically Plausible Subset”
The Office of Orphan Product Development (OOPD) defines the “medically
plausible subset” as the “subset of individuals with a non-rare disease or condition on
whom use of a device is appropriate, where use of the device on remaining individuals
with that disease or condition would be inappropriate given some intrinsic feature of the
device” (U.S. Food and Drug Administration, 2013a). When asked if the regulatory
definition of “medically plausible subset” is appropriate and how it should be interpreted,
the responses of the informants were mixed. Some (42%, 13/31) who were not familiar
with the regulatory definition of “medically plausible subset” or did not have direct
experience working on a “medically plausible subset” HUDs felt that they could not
provide a useful opinion. Some informants (31%, 10/31) felt that the regulatory
definition of “medically plausible subset” is appropriate, either because it (a) provides
useful “insurance” against abuse or off-label use devices with larger markets, (b) helps to
identify a “functional” orphan group (i.e. a group that is defined by its response to the
proposed technology) rather than an “intrinsic” orphan group (i.e. a group defined by the
clinical manifestation of the disease), and (c) acknowledges that individuals often vary
substantially with respect to the severity and stage of their disease, and some subsets are
better treated in other ways. However, a few informants (13%, 4/31) expressed
dissatisfaction with the definition by using the words such as “very confusing,” “narrow,”

127
“needs improvement,” and “unclear.” The remaining informants (13%, 4/31) felt that the
interpretation of the “medically plausible subset” was not consistent with the intent of the
HUD statute and regulations. Specifically, one of the informants indicated that the
“medically plausible subset” should not be based on device features that preclude its use
in a broader patient population, by stating the following:
In my experience, it is interpreted by the Agency that the device is not “able” to
be utilized in any other population. For example, in obtaining the HUD/HDE, we
had to demonstrate that due to a design feature, the device absolutely could not be
used in populations outside of the indicated population. I do not believe this the
correct interpretation of the intent of medically plausible subset. I believe the
correct interpretation should be that there is no other device/treatment available to
the subset population, or that due to some feature of the device it is only safe in
this subset population. (Vice President)
Additionally, a former FDA director who was involved in the implementation of
HDE policies and guidance described the interpretation of the “medically plausible
subset” as a significant problem that limits the potential of HUDs by stating the
following:
A lot of OOPD reviewers came from CDER where they were familiar with how
orphan drugs are designated, which based on the principle that some properties of
the drug preclude its use outside of an “orphan subset” of a non-rare disease or
condition. They wanted the “orphan subset” principle to be consistent between
orphan drugs and devices. Unfortunately, I think this is the case where ex-CDER
reviewers dominated the HUD guidance document during its development. They
did not understand that we needed one standard for drug and another for device...
For example, you design a device for an orphan indication and send in a HUD

128
designation request, however OOPD reviewer denies the designation because the
device could be potentially used in a broader patient population via off-label.
They wanted HUD designation discussion based on device features…some
sponsors appeared to use HDE as a mechanism to commercialize products and
promote off label. FDA then started to think that they needed to limit
HUDs/HDEs to prevent off label use. However, FDA reviewers are not supposed
to concern about off label use during the review process…the interpretation of
“medically plausible subset” is not consistent with the HDE statue and
regulations. (Consultant)
Implementation of FDA Guidance/Regulations/Policies
Twenty informants provided a complete response when asked about the impact of
specific FDA guidance/policies on the pursuit of HDEs (Table 13). Their responses
suggested that “funding to complete the necessary testing expected by the FDA,” “accrue
sufficient number of clinical trial subjects expected by FDA,” and “IRB approval
requirement” had negative impact on the pursuit of HDEs. On the other hand, a majority
of informants felt that “interaction with OOPD,” “transparency of the grant program,”
“accuracy and usefulness of FDA website and documentation,” and “explanation from
FDA that an HDE constitutes a marketing approval” had a positive impact on the pursuit
of HDEs.

129

Table 13: Extent to which pursue of HDEs impacted by FDA policies/requirements
Factor
High
Positive
Impact
Positive
Impact
No Impact
Negative
Impact
High
Negative
Impact
Total
Funding to complete
necessary testing expected
by FDA
3
(15%)
1
(5%)
4
(20%)
9
(45%)
3
(15%)
20
Accrue sufficient number
of clinical trial subjects
expected by FDA
2
(10%)
0
(0%)
11
(55%)
5
(25%)
2
(10%)
20
Product development
cycle
1
(5%)
0
(0%)
12
(60%)
6
(30%)
1
(5%)
20
Standard for clinical trial
3
(15%)
1
(5%)
11
(55%)
3
(15%)
2
(10%)
20
IRB requirement
2
(10%)
3
(15%)
5
(25%)
10
(5%)
0
(0%)
20
Interactions with OOPD
3
(15%)
8
(40%)
7
(35%)
2
(10%)
0
(0%)
20
Transparency of the grant
program
3
(15%)
6
(30%)
9
(45%)
2
(10%)
0
(0%)
20
Accuracy and usefulness
of FDA website and
documentation
3
(15%)
9
(45%)
4
(20%)
4
(20%)
0
(0%)
20
Explanation from FDA
that an HDE constitutes a
marketing approval
8
(40%)
5
(25%)
6
(30%)
1
(5%)
0
(0%)
20
Note: The shaded cells illustrated the most common selection by the informants.
HDE Regulatory Pathway as a Strategic Business Decision
Twenty-six (84%, 26/31) informants provided their views as to why the HDE
regulatory pathway was pursued; others did not respond to protect confidential strategies
of their businesses. Content analysis of the responses revealed that the overarching reason
for pursing the HDE regulatory pathway was to fulfill unmet medical needs (Figure 16).
A sense of social responsibility was apparent amongst informants who frequently used
phrases such as “serve patients,” “patient’s needs,” “patients with no alternative
treatment,” and “unmet medical needs” during the interviews. The most common specific
reason (9X) focused on using HDE as a “stepping stone” regulatory pathway for the
following reasons: (a) slow clinical enrollment for PMA pathway, (b) proof of device

130
concept first before future device refinements, (c) need for revenue to fund further device
development, and (d) intent to covert to PMA in the future. A second commonly
discussed reason (8X) centered on using the HUD designation or HDE as a basis for a
business value proposition that could fuel business investment or attract a potential buyer
(7X).

Figure 16: Can you tell more about your firm’s decision to pursue HDE pathway?

Informants were asked to describe if the decision to commit to a HDE regulatory
pathway was made before or after the product had been substantially developed.
Relatively more informants (62%, 16/26) indicated that the decision was made before
product development, but 38% (10/26) stated that the decision was made after it had been
developed.

131
Proposals to Spur Innovation in HUDs within Existing Regulatory Framework
Informants were asked to describe what FDA could do to spur innovation of
HUDs within the existing regulatory framework (Table 14). Content analysis categorized
the responses into three themes: (a) improve collaboration with stakeholders, (b) improve
regulatory cycle/speed to market, and (c) resolve regulatory hurdles. Most comments
seemed to emphasize the need to “resolve regulatory hurdles” (13X), which informants
felt that FDA could clarify the “probable benefit” requirement (4X) and reduce post
market reporting and GMP requirements of HDEs (3X). A second very common theme
was to “improve collaboration with stakeholders” (10X) where FDA would collaborate
with CMS to issue a Memorandum of Understanding (MOU) to cover HUDs (3X) or
publish market adoption information relevant to HUDs (2X) (e.g. number of patients with
access to HUDs). Further, some informants expressed in strongly stated language that
HDE review could be more consistent between various branches of the Office of Device
Evaluation (ODE) branches (4X).
Table 14: What could FDA do to spur innovation in HUDs?
Theme Category Sub-Category
Improve
collaboration with
stakeholders
(12)
CMS (5)
Create MOU to cover HUDs. Improve processes (2)
Publish HUD market adoption information (2)
FDA – Office of
Device Evaluation
(5)
Ensure consistency of HDE regulatory bar within ODE
branches (4)
Designate a HUD champion within ODE (1)
Physician
community (2)
Improve understanding of HUDs (2)
Improve regulatory
cycle/speed to
market (6)
Create HDE performance goal (3)
Ensure timely review and approval (2)
Improve communication and interactive review (1)
Resolve regulatory
hurdles (13)
Create expedited pathway for pediatric indications (2)
Reduce post approval reporting and GMP requirements (3)
Lower regulatory bar (2)
Clarify “probable benefit” (4)
Clarify HDEs are marketing approval (2)
Others (4)
Conduct workshop/training (2)
Create HUD specific funding/grant (2)

132
4.3.3.2 CMS
Coverage with Evidence Development for HUDs
Coverage with Evidence Development (CED) is a reimbursement scheme that
allows payers to offer conditional coverage for potentially beneficial, yet unproven,
medical technologies while requiring the development of clinical evidence (Kary et al.,
2009). Only five informants appeared to be sufficiently knowledgeable to rank the
importance of several features that might characterize a CED (Table 15). Judged as
“important” or “highly important” or “extremely important” included views that: (a) CED
should ensure continued coverage after a study was over, (b) CED should be used when
the only alternative is denial of coverage, (c) CED should not duplicate efforts of FDA,
(d) CED should not intervene with practice of medicine, and (e) funding mechanism of
CED should be considered.

133
Table 15: Importance of CED characteristics
If CED were to be implemented for HUDs, how important the following characteristics
should be considered?
Characteristics
Extremely
Important
Highly
Important
Important
Moderate
Important
Not
Important
Total
CED study design should
be pre-defined and
flexible
2
(40%)
1
(20%)
1
(20%)
1
(20%)
0
(0%)
5
CED should be used
infrequently within the
NCD process
1
(20%)
1
(20%)
1
(20%)
1
(20%)
1
(20%)
5
CED should not hinder
coverage through LCs
2
(40%)
1
(20%)
1
(20%)
0
(0%)
1
(20%)
5
CED should ensure
continued coverage after a
study was over
3
(60%)
2
(40%)
0
(0%)
0
(0%)
0
(0%)
5
CED should be used when
the only alternative is
denial of coverage
1
(20%)
2
(40%)
2
(40%)
0
(0%)
0
(0%)
5
CED should not duplicate
efforts of FDA
1
(20%)
3
(60%)
1
(20%)
0
(0%)
0
(0%)
5
CED should not intervene
with practice of medicine
2
(40%)
2
(40%)
1
(20%)
0
(0%)
0
(0%)
5
CED should allow input
from private payers
2
(40%)
0
(0%)
2
(40%)
0
(0%)
1
(20%)
5
Roles of AHRQ should be
defined in the
implementation of CED
2
(40%)
0
(0%)
2
(40%)
0
(0%)
1
(20%)
5
Funding mechanism of
CED should be considered
1
(20%)
2
(40%)
2
(40%)
0
(0%)
0
(0%)
5
Note: The shaded cells illustrated the most common selection by the informants.
Feasibility of National Coverage Determination with CED for HUDs
A National Coverage Determination (NCD) conditional on CED is a
reimbursement scheme by which CMS would issue a NCD, but would require additional
evidence collection to supplement the clinical evidence (U.S. Centers for Medicare and
Medicaid Services, 2007, 2012a). When asked if an NCD with CED was a reasonable
reimbursement scheme for HUDs, the views of informants were split quite evenly. Half
(50%, 6/12) of the informants who gave comments believed that an NCD with CED was
a reasonable approach that would allow further evidence collection until the objective
benefit of the device could be demonstrated. Some informants within this subgroup

134
suggested that an NCD with CED might be appropriate in some circumstances but should
not be a universal requirement. Some believed that the CED scheme would provide a
centralized mechanism to collect and analyze data to facilitate reimbursement of HUDs.
The other half (50%, 6/12) of the informants felt that an NCD with CED was not a
reasonable expectation because: (a) it was often impractical to enroll sufficient patients in
a reasonable amount of time due to the small size of the affected population, (b) it would
defeat the purpose of a HUD because it would delay access to the needed device, and (c)
it could impose a significant financial burden to the sponsor. Some informants (4/12)
suggested that an alternative reimbursement scheme for HUDs would be to tie coverage
to additional data collection such as observational studies or a simple registry rather than
a more demanding CED scheme.
Is a local Coverage Policy more assured than an NCD?
When informants were asked if a local coverage policy (LCD) is more appropriate
or more assured than an NCD to secure reimbursement, most (7/12) responded in the
affirmative for the following reasons: (a) NCDs are difficult to obtain and carry higher
burden for evidence that is hard to produce in small populations, (b) local Medicare
contractors have the discretion to approve HDEs for appropriate patients, so that more
opportunity is presented to justify the treatment for specific patients, and (c) a negative
NCD will have block reimbursement of HUDs locally. In contrast, a minority of
informants (3/12) preferred the NCD pathway because it is a predictable and uniform
process to secure reimbursement. They expressed the view that the LCD pathway is less
transparent and more subjective than the NCD pathway, with fewer procedural
safeguards for sponsors. Thus, LCD policies and decisions can be inconsistent and

135
conflicting, so that the process of obtaining coverage is difficult for HUD sponsors to
manage. The remaining (2/12) informants indicated that neither the LCD nor NCD
pathway was satisfactory and neither would be successful in securing reimbursement of
HUDs until statutory language directs CMS and/or local Medicare contractors to review
HUDs differently than other medical devices. They suggested that any NCD for an HUD
would likely be negative or be granted only with restrictive CED requirements. The
alternative of LCDs with various local Medical contractors would be unsatisfactory due
to the burden of managing requirements of many different contractors.
Strategy to Optimize Dispersed Local Coverage Determination (LCD) for HUDs
When asked to suggest strategies to optimize dispersed LCDs for HUDs, the
informants suggested that a sponsor should: (a) convene various Medicare medical
directors with product expertise to create a “model” LCD. Once the LCD is established,
it could be used by other local Medicare contractors, who would rather use the “model”
LCD than invest time and energy to craft a new LCD, (b) determine which local
Medicare contractors are the key opinion leaders (KOL) for a given HUD. Engage the
KOLs to discuss the intended use and the medical needs of the patient population before
FDA approval of HDEs, and continue to keep those KOLs well informed throughout the
FDA review process, (c) develop KOL messaging and applicable publications to
represent the opportunity to address unmet medical needs, (d) perform advocacy with
local KOLs, physicians, patient advocacy groups, and specialty society, to exert pressure
on local Medicare contractors, (d) work with largest local Medicare contractors to obtain
coverage, and then engage smaller contractors on a case-by-case basis, and (e) meet with
the local Medicare contractor and include a treating physician and a patient candidate.

136
The participants can then better explain the therapy, the associated potential risks and
benefits, and the unmet medical needs.
Nature of Intervention to assist Patients/Hospitals to obtain Reimbursement
When asked if informants have intervened or anticipated intervention to assist
patients/hospitals in getting reimbursement of HUDs, most informants (9/12) responded
that they had, a couple (2/12) indicated that they had not, and a single (1/12) informant
had no comment. Most informants suggested that sponsors would have a reimbursement
assistance program to assist hospitals and patients in getting reimbursement of HUDs.
This assistance program might have some of the following features: (a) a reimbursement
support dossier that includes clinical rationale, evidence, statements of support from
KOLs, (b) procedural guidance that includes information on the appeal process and pre-
authorization process, sample letters, payer information, for example, and (c) telephone
assistance to offer and schedule participation in critical meeting with payers. The
informants repeatedly stated that these types of interventions are burdensome and costly.
For example, one of the informants noted:
In my experience we have provided hands on assistance to providers and hospitals
given the burden associated with getting these devices approved for
reimbursement. Some technologies are easier than others, but the burden is very
significant…almost 1-2 FTEs [full-time equivalent staff members] in some cases.
(Senior Director)
Reimbursement Codes of HUDs
Most informants (83%, 10/12) were able to discuss the extent to which a lack of
billing codes hindered the reimbursement of HUDs. All informants appeared to believe

137
that the absence of a code could hinder reimbursement, but most (70%, 7/10) also felt
that the bigger underlying challenge relates to coverage rather than coding. Their
explanations regarding the impact of coding were diverse. A few (30%, 3/10) felt that
payers who agree to cover a HUD will typically designate a temporary code until a
permanent code is applicable. Some (40%, 4/10) felt that codes are not critical to have
because miscellaneous/unlisted codes have been sufficient to allow successful
reimbursement of many previous technologies. Others (20%, 2/10) felt that miscellaneous
codes for CPT and HCPCS could be used with additional documentation. Only one
informant (10%, 1/10) felt that CPT codes were particularly problematic because all
physicians use them for outpatient hospital billing. This informant stated “new or revising
existing CPT codes can take several years while market adoption suffers as a result.”
Reimbursement of HUDs and Device Selection by Physicians
When asked if the challenges of obtaining reimbursement have been a factor in
the device selection by physicians, most informants (8/12) indicated that it had been a
factor. These informants felt that physicians might not use the device if they cannot be
paid promptly, or at a sufficient level. For example, one of the informants stated the
following:
I have witnessed providers switching their choice of device because one device
has a discrete code associated with its procedure while the other does not.
Hospital makes them (physician) switch to the alternative for reimbursement
reasons. (Consultant)
In contrast, one informant stated that lack of reimbursement has not been a factor
in device selection because few options or alternative treatments were available for

138
patients. One informant felt that the impact might vary depending on the situation
because in some situations physicians might be willing to take a loss given a small patient
population, or they might be able and willing to advocate for reimbursement. Two did not
respond to the question.
4.3.3.3 FDA-CMS as a “System”
Parallel Review of HDE Regulatory Pathway
Both the regulatory and reimbursement informants were asked if they believe that
the parallel review between the FDA and CMS would benefit patient access to HUDs. Of
this combined group, 19 regulatory and 7 reimbursement informants responded (Figure
17). The most common answer of reimbursement informants was that the parallel review
was “ineffective” (43%, 3/7) whereas that of most regulatory informants was “somewhat
effective” (37%, 7/19).

139
Figure 17: How effectively the parallel review benefits patient access to HUDs?

When informants were asked to rank several features of the parallel review
process that they believed to be important when implementing parallel review (Table 16),
all but one were viewed as “extremely important” or “highly important.” The single
exception was the idea that “parallel review should not be automatically linked to an
NCD” where data suggested that reimbursement and regulatory experts might have
different views about the importance of the linkage. Specifically, reimbursement
informants (72%, 5/7) most commonly characterized this idea as “extremely important,”
in contrast most regulatory informants (52%, 10/19) characterized it as only “important.”

140
Table 16: Importance of parallel review characteristics

Characteristics
Extremely Important Highly Important Important Moderate Important Not Important
Reg.
(N=19)
Reim.
(N-7)
Reg.
(N=19)
Reim.
(N-7)
Reg.
(N=19)
Reim.
(N-7)
Reg.
(N=19)
Reim.
(N-7)
Reg.
(N=19)
Reim.
(N-7)

Parallel review should be
voluntary

5
(26%)
3
(43%)
6
(32%)
1
(14%)
5
(26%)
1
(14%)
2
(11%)
0
(0%)
1
(5%)
2
(29%)

FDA and CMS need to
adhere to their respective
statutory requirements and
evidentiary standards

4
(21%)
7
(100%)
9
(47%)
0
(0%)
3
(16%)
0
(0%)
2
(11%)
0
(0%)
1
(5%)
0
(0%)

Parallel review should not
be automatically linked to a
National Coverage
Determination (NCD)

4
(21%)
5
(72%)
3
(16%)
1
(14%)
10
(52%)
0
(0%)
0
(0%)
0
(0%)
2
(11%)
1
(14%)

Parallel review should
protect confidentiality of
information

8
(43%)
5
(71%)
5
(26%)
0
(0%)
5
(26%)
2
(29%)
0
(0%)
0
(0%)
1
(5%)
0
(0%)

Parallel review should
encourage three-way (FDA,
CMS, and sponsor)
collaboration

11
(58%)
5
(71%)
7
(37%)
2
(29%)
1
(5%)
0
(0%)
0
(0%)
0
(0%)
0
(0%)
0
(0%)
Note: “Reg” denotes “Regulatory” informants. “Reim” denotes “Reimbursement” informants. The shaded cells illustrated the most common selection
by the informants

141

4.3.4 Constituent of Policy-Making: Evaluation
4.4.4.1 FDA
Potential Evaluation Methods of HUD/HDE Regulations/Guidance
Informants were asked to identify potential ways to measure the effectiveness of
FDA’s HUD designation and HDE pathway regulations/guidance (Figure 18). The most
commonly identified methods were the “number of patients with access to HUDs” (5X)
and the “ratio of HDE approvals versus rejections” (4X). Some informants stressed
further that it was important to know why some HDEs were rejected so that others in the
industry could learn from that experience. Some informants did not offer specific
evaluation methods, but nevertheless felt that “industry has little insight into how FDA
measures effectiveness where HUDs are concerned…there should be some level of
accountability from FDA” and the fact that “there have been so few HDEs per year
should tell you something about the effectiveness” (Vice President).

142
Figure 18: Potential FDA evaluation methods
What are the potential evaluation methods to measure the effectiveness of FDA’s HDE
Regulations/Guidance?

4.3.4.2 CMS
Potential Evaluation Methods of HUD/HDE Reimbursement
Informants were asked to provide the potential evaluation methods that could be used
to measure the effectiveness of practices by CMS practice to assure reimbursement of
HUDs. Less than half (42%, 5/12) of the informants provided responses to the question.
The evaluation methods suggested by these informants were:
• “Analysis of their claim files to evaluate the number submitted and paid”
• “Return on Investment analysis applied to administrative activity, in this case,
how much does the payer spend to save treatment costs and/or improve patient
outcomes”
• “Number of NCDs issued (out of total considered)”

143
• “Number and trend of LCD issued (out of total considered)”
Some informants found the question difficult to answer because “effectiveness can
only be measured against specified objectives, and the objectives are unstated”
(Consultant).
4.3.4.3 FDA-CMS as a “System”
Potential Evaluation Methods of FDA-CMS Regulations/Policies as a “System”
When asked how the informants would evaluate the effectiveness of FDA and
CMS regulations/policies collectively as a “system” to govern patient access to HUDs,
responses from informants suggested that some evaluation methods were possible, but
statutory changes might be needed to allow meaningful measurements of patient access to
HUDs.
Many regulatory informants responding to this question (71%, 22/31) felt that the
effectiveness of FDA and CMS policies could be evaluated best by measuring whether or
not the needs of HUD patients are being addressed. They suggested to measure: (a) the
number of patients treated each year with HUDs out of the total number of HUD patients,
(b) the number of HUD designations converted to HDEs, (c) the number of days from the
HDE submission to HDE approval and a reimbursement decision, (d) the percentage of
HDEs that have obtained reimbursement successfully, and (e) the FDA and CMS review
timelines. However, most felt that the current FDA and CMS regulations and policies are
insufficient to allow for meaningful measurements as a “system” because CMS had given
relatively low priority to HUDs compared to the FDA.

144
Reimbursement informants (83%, 10/12) appeared to be particularly sensitive to
the fact that FDA and CMS are two distinct agencies operating under separate statutory
standards with different regulatory mandates and processes. Therefore, U.S. regulatory
approval and payment processes are not designed to function as a “system.” Without
statutory changes, any attempt to measure effectiveness of policies governing patient
access to HUDs would not give a meaningful outcome. Reimbursement informants felt
that more oversight from Department of Health and Human Services (DHHS) and
involvement from Congress were much needed. Specifically, DHHS could provide
impartial review of policies driving patient access to HUDs and make changes as needed
as both FDA and CMS fall under DHHS. Further, reimbursement informants felt that
Congress needed to understand the underlying issues that hinder patient access to HUDs
and ensure the implementation of the right solution to the issue. For example, one
regulatory informant stated the following:
Congress needs to be better educated on the problems, needs, and solutions so it
can either change the law to make it better or prod the agencies to implement the
law in a way that better promotes approval and payment for HUDs and increases
patient access. (Consultant)

145

Chapter 5: Discussion
As regulatory requirements of the Food and Drug Administration (FDA) and
Centers for Medicare and Medicaid Services (CMS) become more demanding, concerted
efforts between these agencies appear important to ensure that patients with rare
conditions obtain access to devices that will address their unmet medical needs. The
purpose of this exploratory research using content analyses and informant interviews was
to explore the views of industry regarding the advantages and challenges of FDA and
CMS policies related to HUDs individually, and the degree of alignment between the
agencies to achieve a state of policy convergence. To my best knowledge, this
exploratory research offers the first systematic examination of FDA and CMS policies
governing patient access to HUDs as a “system.” The results from this exploratory
research suggest that HUD policies specific to FDA and CMS remain a challenge, and
statutory mandates impede the alignment between the agencies governing patient access
to HUDs.
5.1 Methodological Considerations
This exploratory study used the mixed qualitative research methods of literature
review, content analysis, expert reviews of a novel interview tool, and key informant
interviews. This range of methods was intentionally arranged in stages to gain
incremental knowledge relevant to the next stage of the research. The use of mixed
qualitative research methods has been described as a valuable approach that offers
complementary insights and understanding that may be difficult to access to achieve
using a single method of data collection (Darbyshire, Macgougall, & Schiller, 2005). The

146
mixed method approach has been described as “taking out an insurance policy” on the
accuracy of the answer to the research question to reduce the risks associated with
investing all qualitative methodological “eggs in one basket” (Hall & Rist, 1999).
However, the strengths and weaknesses of each qualitative research method must be
examined critically to challenge the validity of the research outcomes.
Content analysis as a qualitative research method has strengths and weaknesses.
In its favor, content analysis has the ability to condense a multiplicity of observations into
a smaller number of content related categories within which the elements in a single
category share a common characteristic, connotation, or meaning (Elo & Kyngas, 2007).
Other strengths of content analysis are its “flexible research design and naïve technique
that results in a simplistic description of data” (Cavanagh, 1997). Despite these strengths,
content analysis is often criticized for being too simple to allow detailed quantitative
analysis, particularly statistical analysis (Morgan, 1993). Another limitation of the
approach can be its “biased selectivity,” because the method often relies on “an
incomplete collection of documents” (Yin, 1994) and this can limit external validity. In
the content analyses of “Parallel Review of Medical Products” (section 3.1.1.1) and
“Draft Guidance for the Public, Industry, and CMS Staff Coverage with Evidence
Development in the Context of Coverage Decisions” (section 3.1.1.2), only four
responses from a diverse group of stakeholders were analyzed, so that care must be taken
to assume the external validity of the results. However, the intent of the content analysis
was to inform the construction of a guide for later interviews of key informants rather
than to provide systematic information on the views of industry as a whole. Further the
chosen responses were the considered and collective views of a group of contributors.

147
In addition to external validity, internal validity is also of concern when using
social science methods such as the interviews conducted here. An important aspect of
internal validity is construct validity, that depends on the assumption that the
observations or measurement tools actually represent or measure the construct being
investigated (Polit & Beck, 2012). The construct is the initial concept, notion, question or
hypothesis that determines which data is to be gathered and how it is to be gathered
(Golafshani, 2003). Two different approaches were used to reduce threats to construct
validity. First, content analyses of the FDA public docket “Parallel Review of Medical
Products” and CMS “Draft Guidance for the Public, Industry, and CMS Staff Coverage
with Evidence Development in the Context of Coverage Decisions” were performed to
derive key themes and issues relevant to the research questions. The results of the content
analyses were used to establish the constructs and to inform the interview tool. Second, a
group of subject matter experts was convened as a panel to review the interview tool. The
backgrounds and qualifications of the subject matter experts were carefully considered to
ensure that they could provide relevant feedback to improve the interview questions. All
the subject matter experts had at least 15 years of work experience and direct experience
in regulatory or reimbursement of HUDs or orphan products.
The number of regulatory and reimbursement informants who have worked on
HUDs is inherently small in the United States. Identifying the informants with sufficient
experience and obtaining their agreements to participate in this exploratory research was
an important limitation to this work. Snowball techniques were employed extensively to
recruit key informants to participate in the research study. The snowball technique has
been recognized as a useful methodology in qualitative research, especially in studies in

148
which potential respondents are few in number or hard to reach (Balter & Brunet, 2011).
However the snowball technique has also been criticized for its inability to draw a diverse
group of informants because the “sample composition is influenced by the choice of
initial seeds [individuals]. Those samples tend to be biased towards more cooperative
individuals” (Magnana, Sabin, Saidel, & Hackathorn, 2005). The interview results
suggested that key informants recruited through the snowball technique had significant
work experience in HUDs; and some also had significant experience in advocating and
shaping policy landscape of HUDs. This is considered important to assure some level of
generalizability to the overall study population. In addition to the snowball technique,
use of other channels to recruit potential key informants was attempted to make the
selection process more random. In particular, a request was made to the Office of Orphan
Product Development (OOPD) to obtain a list of the HUD designation holders.
However, the OOPD was reluctant to provide the information, because they considered
that this information should be held confidential. They suggested using the FOI pathway
to obtain the information. On September 21, 2012, FOI Services (www.foiservices.com)
filed a FOI request on my behalf to seek information with regard to HUD designation
holders. Unfortunately, to date (January 2014) the FDA has yet to respond to the FOI
request (Soo, personal observation).
The ability to obtain a sufficient number of key informants was less problematic
for regulatory experts than reimbursement experts. The recruitment of sufficient
reimbursement informants was difficult because the pool of experts with expertise in
HUDs is very small. A small sample size used in qualitative research can threaten the
external validity of the results; the sample must be sufficiently large to reduce the

149
probability that the sample has different characteristics from the population from which it
was drawn (Bracht & Glass, 1968). Out of the twelve reimbursement informants who
participated in the exploratory research, 75% had more than 10 years of reimbursement
experience in HUDs. Thus, they represented some of the most experienced
reimbursement professionals in the medical device industry. Further, according to Guest
and coworkers, twelve interviews should be sufficient for most research enterprises if the
aim is to understand common perceptions and experiences among a group of relatively
homogeneous individuals (Guest, Bunce, & Johnson, 2006). The twelve reimbursement
informants interviewed seemed to provide in-depth understanding of HUD
reimbursement issues. I felt that the information that was derived was reaching saturation
by the end of the interview sessions.
The semi-structured interviews were conducted by requesting that the informants
provide their written responses to questions composing the interview tool prior to one-on-
one interviews that added richness to the responses. Interviews have the advantage that
informants may be more spontaneous in responding to questions (Hall & Rist, 1999).
Open-ended questions and probing encourages participants to respond in their own
words, rather than forcing them to choose from fixed responses (Stuckey, 2013).
However, there are limitations associated with interviews that must be acknowledged.
Specifically, the interviews rely on the willingness of the informants to provide accurate
and complete answers. The informants may be reluctant to respond to some questions
truthfully or fully if they feel embarrassed by their lack knowledge of the topic (Oatey,
1999). To reduce the likelihood of such defensive behavior, on each interview I started
the conversation with small talk, described the research framework and questions,

150
explained the confidential nature of the interview, and clearly stated that questions might
be skipped if they were outside of the expertise of the informants. By doing this I hoped
that the informants would be more comfortable and willing to provide accurate responses.
Further, the interview questions were asked in a non-leading and open-ended manner to
avoid introducing confirmation bias. Nickerson describes confirmation bias as “the
seeking or interpreting of evidence in ways that are partial to existing beliefs,
expectations, or a hypothesis in hand” (Nickerson, 1998).
A large portion of the interview process required me to use judgment when
designing, analyzing, interpreting, and reporting the results. I have worked in the medical
device quality and regulatory areas for about 12 years, two of which involved policy
advocacy with respect to the HDE regulatory pathway. It is possible that I might
introduce some bias although I attempted to minimize the potential bias by having other
experts review the interview tool and asking open-ended and non-leading interview
questions to the best of my ability. Further, my expertise related to HUDs did not extend
to reimbursement aspects, so that I had no fixed views related to this area of HUD
oversight.
5.2 Considerations of the Results
The central aim of this exploratory research was to gain insight into the views of
industry regarding the advantages and challenges associated with the two independent
sets of policies developed by FDA and CMS respectively, and the degree of alignment
and interconnectedness between the agencies in achieving a state of policy convergence.
Knowledge of such advantages, challenges, and misalignment can be instrumental to
guide effective policy development, implementation, and evaluation in ensuring patient

151
access to HUDs. A number of observations suggested areas that deserve further
consideration. These areas and the constituents of research framework to which they
belong are:
• Causation (FDA): The implications of well-intentioned FDA regulatory
requirements and their unintended reimbursement consequences,
• Intervention (FDA): The gaps of knowledge amongst the HUD designation
holders,
• Intervention (CMS): The way that coverage, coding, and payment decisions are
made for HUDs,
• Intervention (FDA-CMS): The effectiveness of FDA-CMS parallel review to
improve patient access,
• Evaluation (FDA-CMS): The evaluation methods used to assess effectiveness of
policies with respect to the goal of ensuring patient access to HUDs.
FDA Requirements: Challenges and unintended reimbursement consequences
Results of this study suggest at least two concerns with FDA policy as it now
stands. First are those associated with the current requirements for IRB oversight of
HUDs. The view was commonly expressed that this requirement discourages patient
access to HUDs because the IRB approval process can be redundant, arduous, and costly.
In the view of some informants, the hurdle of IRB approval was seen to be magnified
because some IRBs often were not familiar with the role assigned to them by the FDA.
The IRB requirement also created confusion among some payers who assumed that
HUDs were investigational in nature and therefore not eligible for reimbursement.
The views of the informants in this study add to concerns of other stakeholders,
captured in previous studies or anecdotal reports, who also seem uncomfortable with the

152
involvement of the IRB in HUD management. Even the IRBs themselves appear to have
a low level of comfort with the responsibilities that have been handed to them. A survey
conducted by Gordon and Philportt suggested that many IRB chairs were confused about
the nature of HUDs, the way that they should be reviewed and the reasons why an IRB
was involved. They stressed that guidance from the FDA was needed (Gordon &
Philpott, 2008).
As discussed earlier (section 4.3.2.1), some individuals associated with the FDA
appear unconvinced that IRB oversight adds value to the process. Further, in a consensus
report titled “Rare Diseases and Orphan Products: Accelerating Research and
Development,” the Institute of Medicine (IOM) went so far as to suggest that the IRB’s
responsibility of approving and monitoring use of HUDs is “an anomaly” by stating the
following:
The primary responsibility of IRB is to protect human research participants
through review of proposed research. Their role in approving the use of an HDE
device is an anomaly and a potential source of confusion since the purpose in
question is not research on device but use of the device in clinical care…the
requirement for IRB approval thus is potentially another factor that may
discourage companies from developing products for small markets under HDE
procedures. (Institute of Medicine, 2011)
Does the IRB approval add value? The views of industry experts captured in this
study suggest that value may be added for clinicians by improving the information that
they receive concerning the HUD to ensure that they are used only for suitable patients.
Because HUDs are exempted from the effectiveness requirements of Section 514 and 515

153
of the Food, Drug, and Cosmetics Act, it may be prudent for clinicians to understand the
nature of the data on which HUD approval was based on so that a more informed
decision can be made. The information reviewed by the IRB may help to ensure that risks
and benefits are assessed more rigorously. However, those same experts did not seem to
view the IRB stage as helpful for sponsors and patients. Instead, they expressed concern
about the delays incurred by IRB review. By definition, an HUD is intended to diagnose
or treat small patient populations with no alternative treatment modality (AdvaMed,
2007). IRB approval could impede the ability of those patients to access the much-
needed treatment as quickly as possible. Further, it is costly to develop documentation
for the IRBs and to s interact with them. These costs must be rolled into the price of the
device. It would be interesting to conduct a follow-on study of IRBs to assess the number
of HUDs approved versus disapproved and the reason(s) for the disapproval(s). This
might help to identify whether the step of obtaining IRB approval is a pro forma process
that adds little value. It would be also be interesting to explore whether the IRBs
themselves think that their role as an oversight body for HUDs adds value by enhancing
patient protection. Given the feedback here and elsewhere from the stakeholders,
regulators might want to consider whether the intent of the requirement for IRB approval
could be eliminated or achieved through in a less burdensome approach.
The second issue of some concern to the informants was related to the
interpretation of “probable benefit.” Informants appeared to advocate a flexible
regulatory model in which the “probable benefit” requirements are based on an
evaluation of product-specific risks and benefits. Concerns around the “probable benefit”
standard have also been apparent in previous literature. For example, AdvaMed

154
criticized FDA for its failure to provide general guidance to manufacturers regarding the
type or level of evidence that must be developed to demonstrate that an HDE meets the
“probable benefit” standard (Institute of Medicine, 2009). Notwithstanding the
preference for a flexible approach with regard to probable benefits for specific products,
informants would like different FDA device branches to apply a consistent “probable
benefit” regulatory bar throughout. This observation suggests that it is important to align
and to interconnect different units of the FDA as a “system” in applying the “probable
benefit” standard. How can such alignment and interconnectedness be achieved? This
exploratory research did not have the ability to answer the question definitively. Perhaps
it can be accomplished by centralizing the regulatory coordination and review as well as
designation of HDEs under a single office within the FDA or issuing better regulatory
guidance on “probable benefit” requirements (U.S. Food and Drug Administration,
2010e). One informant also suggested that FDA should designate a HUD-specific
ombudsman who could serve as an advocate for patients within CDRH. Discussions of
ways to facilitate this process would be important to streamline the process and level the
playing field for different product types.
Rueter states that in most areas of public policy, interventions designed to achieve
one goal might affect other goals as well; these unintended consequences can be desirable
or undesirable (Rueter, 2009). The well-intentioned statutory regulatory requirements for
HUDs, including the need for IRB oversight and demonstration of “probable benefit,”
seem to have had unintended reimbursement consequences. Results here, like reports
elsewhere, suggest that the statutory requirement for IRB oversight creates confusion
among some payers who assume that HDEs are investigational and therefore unready for

155
reimbursement (FDA-2011-D-0847, 2012; National Organization for Rare Diseases,
2011). With regard to the “probable benefit” provision, CMS, and presumably other
insurers, struggle with the fact that the more modest collection of clinical evidence,
sufficient for FDA approval, is insufficient to demonstrate the effectiveness of HUDs
under the “reasonable and necessary” statutory determination (Bentivoglio et al., 2006).
The unintended reimbursement consequences of the FDA statutory requirement also
seem to reinforce the view of that CMS/FDA cannot help but function as a “system;”
specifically “each element has an effect on the functioning of the whole and each element
is affected by at least one other element in the system” (Ackoff, 1981). It is hoped that
policy makers would conceptualize the various players responsible for patient access to
HUDs as a “system” and engage a variety of stakeholders to voice their perspectives and
foster the development of HUD policies that are better aligned and more effective.
When approaching any system, Stewart and Ayres argued that a complex policy
issue cannot be reduced to a simple cause-and-effect relationship. They emphasized the
need to acknowledge the role of internal feedback loops to understand causation (Stewart
& Ayres, 2001). Using the results from several interview questions (e.g. what is the
extent to which the following factors influence the progress of your HUD designation to
HDE? What is the extent to which the following factors influence lack of reimbursement
of HUDs?), this study attempted to construct a summative “causal loop” diagram
focusing on patient access to HUDs (Figure 19).
The causal loop diagram attempts to capture the following interrelated elements:
Growing research on pathophysiology identifies new rare diseases and subsequently
increases the demand for development of appropriate medical technology. This leads to

156
an increase in the number of HUD requests/designations that in turn increases the number
of HDE applications. At the same time, various developmental burdens such as lack of
funding, small patient population, uncertainty in business environment, rigorous clinical
evidence requirement, undefined probable benefit requirement, and rigorous or
unreasonable expectations for clinical trial design increases the time from HUD
designation to HDE application and reduces the number of HDE applications. If the
number of HUD applications goes up, the number of approved applications should also
increase, as long as the ratio of approvals to disapprovals can be assumed to stay
constant. Increases in the numbers of approved HDEs will increase the need for coverage
decisions. However, the perceived investigational status and lack of clinical effectiveness
of HUDs will impede the coverage decisions (FDA-2011-D-0847, 2012). Further, the
lack of coverage decisions creates an environment hostile to investment and leads to
decrease in funding available for medical technologies such as HUDs (Ackerly,
Valverde, Diener, Dossary, & Schulman, 2009). The absence of a coverage decision
reduces patient access to HUDs, which ultimately increases the continuing demand for
rare disease therapy options. This is known as a “reinforcing feedback loop” as the
demand for rare disease therapy options reinforces even more demands within the
feedback loop. The reinforcing feedback loop can therefore lead to growth of demand
(Kirkwood, 1998).

157
Figure 19: Causal loop diagram governing patient access to HUDs

The Gaps of Knowledge amongst HUD Designation Holders
Most HUD designation holders comprising the relatively small group of
informants currently pursing HDE applications (78%, 7/9) in this study were found to be
working in academic or hospital settings as physicians and professors. It is not clear
whether the high proportion of designation holders in academia reflects accurately the
proportion of academic HUD designation holders in the population more generally.
However, other statistics also suggest that many, if not most, HUD holders are in
academic settings or in small companies headed by individuals with academic or hospital
affiliations. For example, in a recent small business report to Congress, as required by
FDASIA 2012, the FDA stated that “an important area for small device firms is the

158
development of a HUD” and also indicated that since 2008, OOPD had granted 30 of its
92 grants for clinical trials to small businesses (U.S. Food and Drug Administration,
2013d). Many such small businesses are formed to satisfy requirements of the Small
Business Innovation Research (SBIR) granting scheme administered by NIH that limits
funding for products with commercial potential to small businesses rather than academic
laboratories, thus driving academic faculty to become small business entrepreneurs.
Research has suggested that having academic faculty connected to an SBIR project
increased the likelihood of its successful commercialization (Siegel & Wessner, 2012).
Further, the university is the setting where most innovation tends to occur and where an
understanding of unmet medical needs is most easily appreciated. A survey conducted by
Easley and Miller suggested that if all companies rooted in the university were to form a
nation, it would comprise the world’s tenth largest economy (Eesley & Miller, 2012). It is
difficult, however, to analyze the distribution of HUD designation holders further
because the identities of these holders, unlike holders of orphan drug product
designations, are not publicly disclosed by the FDA.
We might expect that many individuals or groups in universities or their spin-off
companies have not had previous experience with the development of a commercializable
product and may be holding their first HDE. Thus it might not be surprising to find that
these individuals are not fully aware of the rules and challenges associated with HUD
development, and these gaps in knowledge could impede patient access to HUDs. Gaps
of particular concern include an incomplete understanding of reimbursement challenges,
post-approval HDE change control commitments, and quality system requirements. The
extent of the gap varies but lack of knowledge related to reimbursement challenges of

159
HUDs seem most problematic, because access for patients is so clearly reliant on whether
the HUD can be reimbursed by a patient’s health care coverage. However small
development teams are often driven to focus on the immediate hurdles associated with
FDA approval, and do not appreciate the reimbursement challenges that will become
central after the FDA approval is obtained. Such a short-term focus might be reinforced
by the need of the HUD designation holders to attract funding from outside investors,
who often view progress toward FDA approval as a key milestone. Reinforcing this
view, one of the regulatory informants indicated that HUD designation from the FDA
gave the device “credibility.” The results from this exploratory study highlight the need
to better educate HUD designation holders on dealing with reimbursement and other
hurdles that could impede access to HUDs so that plans to collect needed data can be put
into place earlier in the development cycle.
Reimbursement of HUDs
Areas of concern highlighted by this study included not only the volume of
evidence needed for reimbursement decisions but also the approach taken by CMS to
make those decisions. In the past CMS has not typically engaged in much dialogue about
how best to secure reimbursement of HUDs. Perhaps the clearest insight has come from
its 2006 review of policies regarding national coverage determination (NCDs) for clinical
research. There, CMS stated the belief that HUDs would be covered most appropriately
by granting an NCD in conjunction with requirements for further clinical studies under
the coverage with evidence development (CED) program (U.S. Centers for Medicare and
Medicaid Services, 2007). Many informants in this study appeared to be concerned about
such an approach. One reason for this concern may be the long implementation timeline

160
of an NCD with CED requirements. CMS implements an NCD by: (a) initiating the NCD
review and allowing a 30-day public comment period, (b) engaging in a formal evidence
review, (c) issuing a proposed decision within 6 months of opening the review for NCD
not requiring an external technology assessment or Medicare Evidence Development and
Coverage Advisory Committee (MEDCAC) review. If these additional reviews are
required, CMS will issue a proposed decision within 9 months of opening the NCD
review, (d) allowing another 30-day public comment period on the proposal, and (e)
issuing a final NCD and decision memorandum no later than 60 days after the end of the
30-day public comment period (Lundy et al., 2013). In fiscal year 2012, CMS reported
that it took an average of 170 days to implement an NCD fully (U.S. Centers for
Medicare and Medicaid Services, 2013a). This timeframe could be even longer for HUDs
because they serve such small patient populations. It could be difficult or impossible to
accrue a sufficient number of patients for the necessary studies, and such accrual would
likely take a long time (Institute of Medicine, 2009). Since HUDs involve treatments for
small populations with no alternative treatment, the process would delay access to
technology for an unreasonable length of time. On the other hand, some informants
appeared to support the NCD with CED approach because it allowed the sponsor to
collect further evidence that would help to demonstrate objective benefit. As an NCD
with CED is a relatively new submission approach for HUDs, it would be very interesting
evaluate case studies of HUDs such as the AbioCor Implantable Replacement Heart
(HDE 040006) that are covered under this reimbursement scheme to gain better insight
into the timelines, challenges, and impact on patient access to HUDs.

161
Typically the process of obtaining reimbursement coverage from Medicare or
other insurers has required decisions to be made on three elements – coverage, coding,
and payment – as outlined in Chapter 2. In previous literature, AdvaMed appeared to
highlight the importance of coding as a hindrance to reimbursement. AdvaMed reported
that it could take more than14 to 26 months to obtain a new or substantially revised CPT
code from the CMS subsequent to FDA approval or clearance. AdvaMed urged CMS to
develop a faster process by which temporary reimbursement codes could be assigned for
HUDs, until such time as American Medical Association (AMA) assigned a permanent
CPT code (FDA-2010-N-0237, 2010). However, the views of the experts interviewed
here seem to suggest that access is hindered more by the coverage of HUDs than by their
coding. Some informants indicated that the coding problem was less important because
payers could designate a code or use miscellaneous, unlisted, or temporary codes once
the coverage decision has been determined. However, sponsors could face tremendous
challenge to navigate these codes as they work with various local Medicare contractors. It
would be interesting in future to assess the degree to which the use of temporary codes
solves the problem identified by Advamed and mitigates the limitations of having no
coding determination.
In addition to the elements of coverage and coding is the “payment” element of
reimbursement. As noted in Chapter 2, most approved HUDs to date (January 2014)
have been implantable, high risk medical devices. Thus, they are likely to be used in
hospitals where an in-patient reimbursement scheme would be applicable. Most
commonly hospital procedures are reimbursed through a Diagnostic Related Group
(DRG) approach, in which patients undergoing similar procedures are classified into

162
clinically cohesive groups with similar costs related to such factors as length of stay and
resource consumption are similar (Hallas, 2012). Although the amount reimbursed under
the DRG approach is reviewed and recalibrated annually, the data used by CMS to
recalibrate the payment rates could be several years old (Lewis, 2008). Thus, DRG
payments to hospitals might not necessarily be adjusted to reflect the real costs of HUDs
and their related interventions. To address this concern, CMS offers temporary add-on
payments to DRGs if the technology under consideration meets three prerequisites: (a)
the technology must be new, (b) the technology must not be covered adequately under the
current DRG schedule, and (c) the technology must offer clinical benefits over existing
treatment options. CMS has granted an add-on payment to a few HUDs, including, for
example, the AbioMed artificial implantable heart, Spiration device to control post-
operative prolonged air leaks or significant air leaks of the lung, and Argus II retinal
prosthesis system to restore some functional vision for people suffering from blindness
(Institute of Medicine, 2011; Second Sight Medical Products, 2013). However, out of the
twelve reimbursement informants interviewed in this study, only one individual reported
success in obtaining the technology add-on payment for HUDs. Perhaps this observation
was not surprising because it could be difficult for many manufacturers to produce the
comparative data needed to prove that their technology offers significant benefit, since
proof of efficacy is difficult to gather in small patient populations.

163
The effectiveness of FDA-CMS parallel review to improve patient access
The parallel review pilot program is believed to facilitate the development of
innovative medical devices, reduce regulatory burden, and shorten the time it takes to
bring these devices to patients (U.S. Food and Drug Administration, 2011b). By
comparing the views of regulatory and reimbursement experts, it appeared that the
regulatory informants were more optimistic than reimbursement informants about the
potential ability of parallel review to accelerate patient access to HUDs. The sample of
experts in each category is relatively small, so it is not clear whether the more negative
views expressed by reimbursement experts are typical for the wider population. Further,
the underlying reasons for the difference in views between the informants are not clear.
However, previous reports may offer some insight into the greater negativity of
reimbursement experts. Richer suggested that FDA and CMS processes are incompatible
for at least two key reasons: (a) it is very difficult to align the timing of CMS processes
for coding and payment decisions with the faster FDA approval process, and (b) the
disclosure of technological and clinical evidence required when CMS coverage is sought
would be problematic before a PMA is granted because the disclosure might release
proprietary information to competitors who are developing a similar technologies
(Academic OneFile, 2010).
More experience with the parallel review process will be needed before its
problems and benefits can be appreciated fully. The parallel review program is a
relatively new initiative; only five innovative devices can enter this pathway each year
and most of these devices are probably not going to be HUDs (U.S. Food and Drug
Administration, 2011c). It is therefore likely that the responses from most of the

164
informants were based on their perceptions or expectations of the parallel review
program, rather than their direct experience with the program. It will be important to
assess the usefulness of the parallel review program with respect to HUDs in the future
after the program has become more mature. It may prove that HUDs are not even
deemed to be good candidates for the program because the regulators may recognize the
difficulties of working with dossiers in which relatively little clinical data is available.
The use of both content analysis of submitted dossiers and informant interviews
offered the opportunity to compare the results of the findings from each approach.
Analysis of content in the reviewed dossiers identified a number of industry preferences
with regard to parallel review: (a) that parallel review should be voluntary, (b) that FDA
and CMS should adhere to their respective statutory requirements and evidentiary
standards, (c) that parallel review should not be automatically linked to a NCD, (d) that
both sets of regulators should protect confidentiality of information, and (e) that parallel
review should foster three-way discussion between FDA, CMS, and sponsor. It is
important to remember that the content of the reviewed dossiers was not product specific
or HUD specific. Nonetheless, the interviews seemed to affirm that the comments made
in those dossiers were relevant for HUDs. Further, regulatory and reimbursement experts
appeared to share largely similar views with regard to the importance of most features of
the parallel review process. The only feature that elicited differing opinions from the
regulatory and reimbursement informants was that “parallel review should not be
automatically linked to a NCD.” In this instance reimbursement informants all ranked
this characteristic as “extremely/highly important,” whereas only about a third of
regulatory experts rated this at such a high level of importance. Since the majority of

165
reimbursement informants also expressed the view that a local coverage determination
(LCD) was more assured than an NCD, perhaps they were more aware of the
reimbursement dynamics and tended to appreciate better why a dissociation of HUDs
from NCDs would be useful.
On December 19, 2013, the FDA announced a program appears to respond to
some of the concerns that are captured by the interviews with industry informants
documented here. The “CDRH Reimbursement Program” was designed to strengthen the
collaboration with CMS and payers by developing plans to (a) establish a formal process
under which a sponsor can voluntarily request a payer to participate in a FDA pre-
submission meeting to foster early collaboration between the FDA, payer, and sponsor,
(b) modify its current IDE checklist, as needed, and provide it to CMS upon request to
support the CMS to make coverage decision, and (c) collaborate with payers and health
technology assessment organizations in that FDA can provide a summary of safety and
effectiveness information to support coverage determinations (Kern, 2013). This is not a
program that addresses HUD devices specifically. However, it might be particularly
valuable for such products. Importantly, any form of collaboration or “intervention”
between the FDA and CMS such as parallel review or the “CDRH Reimbursement
Program” is an opportunity for the agencies to align expectations and requirements. By
applying system thinking, these “interventions” seem to fit with Stewart and Ayres’ view
that “interventions cannot be conceptualized as externally applied incentives and
sanctions. Rather, they will constitute changes in relationships between (or even within)
organizations, based on altered patterns of communication, influence, and exchange”
(Stewart & Ayres, 2001).

166
Evaluation methods driving patient access to HUDs
Very little has previously been written about approaches and methods by which
the effectiveness of the HUD policies could be measured. Nevertheless, it is a central
tenet of good policy management that the outcomes of policies should be evaluated in
order to measure the success of interventions and modify policies to better accomplish
their original goals. In its publication entitled “System thinking for health systems
strengthening,” the Alliance for Health Policy and Systems Research of the World Health
Organization (WHO) states that “a system thinking approach can connect intervention
design to evaluation more explicitly, both to each other and to the health system
framework – thought it should be added that not all interventions require evaluation or
evaluation with a systems thinking lens” (World Health Organization, 2009). This
exploratory study attempted to assess whether informants knew of any methods used by
the FDA and CMS to evaluate their policies as a “system” governing patient access to
HUDs. Results suggested that evaluative methods either were not in place, or if in place
were closely held activities within the agencies that were not communicated effectively to
the informants interviewed here. The lack of apparent evaluation might be attributed to
at least two factors. First, the perceived importance of the HUD pathway and policy
initiatives for public good have not been recognized widely until recently, so policy
initiatives have been modest at best. It is difficult to evaluate new or immature policies.
Second, the differing statutory mandates (causation) and relatively low priority of HUDs
from CMS (causation/intervention) will make such evaluations difficult to structure. For
example, Randel Richer, Vice President of Government Affairs of Boston Scientific,

167
responded to the public docket (2004S-0233) on innovation in medical technologies by
stating the following:
Congress and FDA have taken explicit actions to promote the development of
these technologies through the Humanitarian Device Exemption (HDE) process,
no corresponding mechanism has been established to promote Medicare and
Medicaid coverage and payment for these technologies. (U.S. Food and Drug
Administration, 2004)
How then might HUD policies be evaluated in the current environment?
Responses from informants included not only suggestions of quantitative metrics but also
more qualitative approaches. One particular “system” evaluation method suggested by
the informants was the number of patients treated each year with HUDs out of the total
number of HUD patients. It is a meaningful evaluation method to measure patient access
to HUDs because it targets at the effectiveness of FDA and CMS policies and practices
collectively as a “system.” However, it is not clear what the correct group to comprise
the denominator would be, given that many rare diseases are at the moment beyond our
technological capabilities to treat. Nevertheless, this particular evaluation method seems
to align with Stewart and Ayres’ system thinking for “evaluation,” which stated that it is
the totality of policies that are relevant, not just the specific areas of interest such as
regulatory or reimbursement alone (Stewart & Ayres, 2001).
Another way to evaluate the effectiveness of policies might be to examine how
well they are taken up in the market, especially if a number of barriers, such as IRB
oversight and reimbursement challenges, have a major effect on uptake. In an article
related to market access of orphan drugs, Matthews and Glass described such an

168
evaluation method as “adoption,” which can be calculated using the following
methodology (Matthews & Glass, 2013):

Adoption = (annual sales)/(calculated sales)*100%

Where:

Calculated sales = (annual cost)*(prevalence of disease/total annual population)

However, application of the above “adoption” evaluation method to HUDs could
be challenging. Variables such as annual sales and annual costs may be difficult to obtain
due to confidential nature of the information. Further, there might be many reasons why
some patients who have the disease would not be candidates for the treatment. Thus care
must be taken with respect to the number of patients suffering from the rare disease
(prevalence/total annual population) to develop a more realistic estimate of potential
patients. This is an area where a rare disease patient advocacy organization such as
NORD could assist the regulators in gathering and sharing such information without
divulging proprietary or sponsor-specific details.
5.3 Conclusion and Recommendation
“Rare diseases are not rare, at least in aggregate” (Institute of Medicine, 2011).
Although each individual rare disease affects no more than 200,000 people, and some
affect only a few hundred or even a few dozen, all of the rare diseases together affect an
estimated 25 million to 30 million Americans. Importantly, approximately 50% of people
affected by rare disease are children (National Organization for Rare Diseases).

169
Concerted efforts between the FDA and CMS are critical if society aspires to provide
treatments for people suffering from rare diseases. However, it is evident from this
exploratory study that HUD policy convergence specific to FDA and CMS remains a
challenge, and the differing statutory mandates impede the alignment between the
agencies governing patient access to HUDs as a “system.” To remedy the challenge, the
issue of patient access to HUDs shall be viewed from the lens of system thinking, as
WHO stated:
The responses from many health systems so far have been generally considered
inadequate and naïve…a system’s failure requires a system’s solution – not a
temporary remedy. (World Health Organization, 2009)
This insight further underlines the need for statutory changes that can enhance
compatibility between the agencies, and thus encourage a “system” solution to the
problems of patient access. The changes will be difficult and time-consuming, but they
are central to improving the lives of patients with rare diseases.

170

References
Academic OneFile. (2010). Washingon roundup: FDA, CMS try again to drum up interest in
parallel review. Medical Device Daily. Retrieved December 20, 2013, from
http://go.galegroup.com/ps/i.do?id=GALE%7CA238061794&v=2.1&u=usocal_main
&it=r&p=AONE&sw=w&asid=621a73eb03b851689b985113194f93cb

Ackerly, D. C., Valverde, A. M., Diener, L. W., Dossary, K. L., & Schulman, K. A. (2009).
Trends: Fueling innovation in medical devices (and beyond): Venture captical in health care.
Health Affairs, 28(1), 68-75. doi: 10.1377/hlthaff.28.1.w68

Ackoff, R. L. (1981). Creating the corporate future. New York: John Wiley & Sons.

AdvaMed. (2007). Proposed decision memo for clinical trial policy (CAG-00071R). Retrieved
August 18, 2012, from
http://www.elsevierbi.com/~/media/Images/Publications/Archive/The Gray
Sheet/33/021/01330210009/052107_advamed_clinical_research_policy_comments.p
df

Amagoh, F. (2008). Perspectives on organizational change: Systems and complexity theories.
The Public Sector Innovation Journal, 13(3), 1-14.

American Academy of Pediatrics. (2008). American Pediatric Society Association: Public policy
council legislative report March 2008. Retrieved October 23, 2012, from http://www.aps-
spr.org/files/get-involved/advocacy-public-policy/legislative-reports/2008-03-Mar-
PPCLegislativeReport.pdf

American Medical Association. About CPT®. Retrieved October 8, 2012, from
http://www.ama-assn.org/ama/pub/physician-resources/solutions-managing-your-
practice/coding-billing-insurance/cpt/about-cpt.page?

171
American Medical Association. CPT - Current Procedural Terminology. Retrieved August 21,
2012, from http://www.ama-assn.org/ama/pub/physician-resources/solutions-
managing-your-practice/coding-billing-insurance/cpt.page?

Anonymous. (2010). The needs of the few. Nature, 466(7303), 160.

Arnold and Porter LLP. (2012). A reference guide to Food and Drug Administration Safety and
Innovation Act. Retrieved October 25, 2012, from
http://www.arnoldporter.com/resources/documents/A Reference Guide to the FDA
Safety Innovation Act (Updated 7-20-12 at 4 15PM).pdf

Balter, F., & Brunet, I. (2011). Social research 2.0: Virtual snowball sampling method using
Facebook. Retrieved December 7, 2013, from http://www.emeraldinsight.com/1066-
2243.htm

Bass, C. F. (2012). Bass legislation to spur innovations in medical device technology passes
committee. Retrieved August 30, 2012, from
http://bass.house.gov/index.php?option=com_content&task=view&id=3463&Itemid=
400158

Bentivoglio, J., Lorell, B., Scanlan, K., Loeb, L., & Weinstein, A. (2006). CMS reconsiders
clinical trial policy significant changes in play for manufacturers, providers, and patients at
MCAC hearing. Retrieved August 18, 2012, from
http://www.kslaw.com/library/publication/ca122006.pdf

Bernad, D. M. (2009). Humanitarian Use Device and Humanitarian Device Exemption regulatory
programs: pros and cons. Expert Review of Medical Devices, 6(2), 137-145.

Bleicher, E. W. B. (2009). Encouraging research and development of pediatric medical devices
through legislative and regulatory action: The Pediatric Medical Device Safety and
Improvement Act of 2007 in context. Food and Drug Law Journal, 64, 531-564.

172
Blue Cross Blue Shield of North Carolina. (2012). Blue Cross Blue Shield of North Carolina
corporate medical policy: Gastric electrical stimulation. Retrieved August 29, 2012, from
http://www.bcbsnc.com/assets/services/public/pdfs/medicalpolicy/gastric_electrical_s
timulation.pdf

Boulding, K. E. (1956). General system theory: The skeleton of science. Management Science, 2,
197-208.

Bracht, G. H., & Glass, G. V. (1968). The external validity of experiments. American
Educational Research Journal, 5(4), 437-474.

Cavallo, R. E. (1979). The role of systems methodology in social science research. Boston:
Martinus Nijhoff.

Cavanagh, S. (1997). Content analysis: concepts, methods and applications. Nurse Researcher, 4,
5-16.

Center for Medical Technology Policy. (2011). Coverage with evidence development key issues.
Retrieved April 20, 20113, from http://www.cmtpnet.org/wp-
content/uploads/downloads/2012/03/CED-Key-Issues.pdf

Chen, D., & Stroup, W. (1993). General system theory: Toward a conceptual framework for
science and technology education for all. Journal of Science Education and Technology, 2(3),
447-459.

Cigna. (2011a). Cigna medical coverage policy: Deep brain and motor cortex stimulation.
Retrieved August 29, 2012, from http://www.cigna.com/assets/docs/health-care-
professionals/coverage_positions/mm_0184_coveragepositioncriteria_deep_brain_sti
mulation_for_pd_and_et.pdf

Cigna. (2011b). Cigna medical coverage policy: Gastric pacing/gastric electrical stimulation.
Retrieved August 13, 2012, from http://www.cigna.com/assets/docs/health-care-

173
professionals/coverage_positions/mm_0103_coveragepositioncriteria_enterra_therap
y_gastric_pacing.pdf

CMS_FRDOC_0001. (2011). Pilot program for parallel review of medical products.
Regulations.gov: Retrieved from http://www.regulations.gov/ -
!documentDetail;D=CMS_FRDOC_0001-0831.

Covington and Burlington LLP. (2012). Food and Drug Administration Safety and Innovation
Act of 2012. Retrieved 2012, October 25, from
http://www.cov.com/files/Publication/0c4b516e-d83a-4f48-9d34-
1144bc0e3f90/Presentation/PublicationAttachment/773cafe3-f4c8-48fa-be2b-
1fd0a9fcbd11/FDA_Safety%26Innovation_Act_2012.pdf

Darbyshire, P., Macgougall, C., & Schiller, W. (2005). Multiple methods in qualitative research
with children: more insight or just more?. Qualitative research, 5(4), 417-436.

Dolcimascolo, F. (2009). Pediatric cardiac devices—An FDA pediatrician’s perspective of the
challenges and potential solutions. Journal of Cardiovascular Translational Research, 2(2),
147-149.

Dubrovsky, V. (2004). Toward system principles: General system theory and the alternative
approach. Systems Research and Behavioral Science, 21(2), 109-122.

Eeden, M. V. (2011). Every drug an orphan drug: A stakeholders analysis. Stakeholder
Involvement in Health Economic Evaluation of Innovative Technologies Retrieved
September 8, 2012, from http://ihea2011.abstractsubmit.org/sessions/117/

Eesley, C. E., & Miller, W. F. (2012). Impact: Standard University's economic impact via
innovation and entrepreneurship. Retrieved December 16, 2013, from
http://engineering.stanford.edu/sites/default/files/Stanford_Alumni_Innovation_Surve
y_Report_102412_1.pdf

174
Elo, S., & Kyngas, H. (2007). The qualitative content analysis process. Journal of Advanced
Nursing, 62(1), 107-115.

Erwin, L. (1972). Introduction to systems philosophy: Toward a new paradigm of contemporary
thought. Gordon & Breach Science.

Eydelman, M. B., & Chen, E. A. (2011). The FDA's humanitarian device exemption program.
Health Aff (Millwood), 30(6), 1210-1212. doi: 10.1377/hlthaff.2011.0550

FDA-2010-N-0218. (2010a). FDA review and regulation of articles for the treatment of rare
diseases Advanced MedicalTechnology Association (AdvaMed) comment. Regulations.gov:
Retrieved from http://www.regulations.gov/ - !documentDetail;D=FDA-2010-N-0218-
0038.

FDA-2010-N-0218. (2010b). FDA review and regulation of articles for the treatment of rare
diseases Amgen comment. Regulations.gov Retrieved from http://www.regulations.gov/ -
!documentDetail;D=FDA-2010-N-0218-0040.

FDA-2010-N-0218. (2010c). FDA review and regulation of articles for the treatment of rare
diseases Biotehnology Industry Organization's comment. Regulations.gov: Retrieved from
http://www.regulations.gov/ - !documentDetail;D=FDA-2010-N-0218-0004.

FDA-2010-N-0218. (2010d). FDA review and regulation of articles for the treatment of rare
diseases Medtronic comment. Regulations.gov: Retrieved from
http://www.regulations.gov/ - !documentDetail;D=FDA-2010-N-0218-0042.

FDA-2010-N-0237. (2010). FDA review and regulation of articles for the treatment of rare
diseases Advanced MedicalTechnology Association (AdvaMed) comment. Regulations.gov:
Retrieved from http://www.regulations.gov/ - !documentDetail;D=FDA-2010-N-0237-
0009.

175
FDA-2010-N-0308. (2010a). Parallel review of medical products by the FDA and CMS agencies
Advanced MedicalTechnology Association (AdvaMed) comment. Regulations.gov: Retrieved
from http://www.regulations.gov/ - !documentDetail;D=FDA-2010-N-0308-0025.

FDA-2010-N-0308. (2010b). Parallel review of medical products by the FDA and CMS agencies
Blue Cross and Blue Shield Association (BCBSA) comment. Regulations.gov: Retrieved
from http://www.regulations.gov/ - !documentDetail;D=FDA-2010-N-0308-0034.

FDA-2010-N-0308. (2010c). Parallel review of medical products by the FDA and CMS agencies
Cystic Fibrosis Foundation comment. Regulations.gov: Retrieved from
http://www.regulations.gov/ - !documentDetail;D=FDA-2010-N-0308-0040.

FDA-2010-N-0308. (2010d). Parallel review of medical products by the FDA and CMS agencies
Johnson & Johnson comment. Regulations.gov: Retrieved from
http://www.regulations.gov/ - !documentDetail;D=FDA-2010-N-0308-0039.

FDA-2010-N-0308. (2010e). Parallel review of medical products. (Federal Register, Vol. 75, No.
180,). Retrieved from http://www.gpo.gov/fdsys/pkg/FR-2010-09-17/pdf/2010-
23252.pdf.

FDA-2011-D-0847. (2012). Draft guidance for industry and staff humanitarian use device
designations Advanced Medical Technology Association (AdvaMed) comment.:
Regulations.gov Retrieved from http://www.regulations.gov/ -
!documentDetail;D=FDA-2011-D-0847-0006.

FDAVoice. (2012). Advocating for patients with rare diseases. Retrieved August 30, 2012, from
http://blogs.fda.gov/fdavoice/index.php/2012/03/advocating-for-patients-with-rare-
diseases/

Fins, J. J., Mayberg, H. S., Nuttin, B., Kubu, C. S., Galert, T., Sturm, V., . . . Schlaepfer, T. E.
(2011). Misuse of the FDA's humanitarian device exemption in deep brain stimulation for

176
obsessive-compulsive disorder. Health Aff (Millwood), 30(2), 302-311. doi:
10.1377/hlthaff.2010.0157

Golafshani, N. (2003). Understanding reliability and validity in qualitative research. The
Qualitative Report, 8(4), 597-607.

Gordon, E. J., & Philpott, S. (2008). The convergence of research and clinical practice:
Institutional Review Board review of humanitarian use device applications. Journal of
Empirical Research on Human Research Ethics, 3(4), 81-98. doi: 10.1525/jer.2008.3.4.81

Guest, G., Bunce, A., & Johnson, L. (2006). How many interviews are enough? An experiment
with data saturation and variability. Field Methods, 18(59).

H.R. 3211. (2011). Humanitarian Device Reform Act of 2011. (112th Congress). Retrieved from
http://www.govtrack.us/congress/bills/112/hr3211/text.

Hall, A. L., & Rist, R. C. (1999). Integrating multiple qualitative research methods (or avoiding
the precariousness of a one-legged stool). Phychology and Marketing, 16(4), 291-304.

Hallas, K. (2012). Fundamentals of reimbursement. Retrieved December 20, 2013, from
http://www.emersonconsultants.com/_assets/document/Fundamentals_of_Reimburse
ment_Chapter.pdf

Harvard Pilgrim HealthCare. (2010). Medical policy: Gastric Electrical Stimulation (GES) for
gastroparesis and obesity. Retrieved August 13, 2012, from
https://http://www.harvardpilgrim.org/pls/portal/docs/PAGE/PROVIDERS/MEDMG
MT/STATEMENTS/GASTRIC_ELECTRICAL_STIM_0512.PDF

Health Plan of Nevada. (2011). Metacarpophalangeal and proximal interphalangeal joint implant.
Retrieved August 13, 2012, from
http://www.healthplanofnevada.com/documents/provider files/Metacapophalangeal
and Proximal Interphalangeal Joint Implant SUR045.pdf

177
Howley, P., & Chuang, S. (2011). System theory and improving healthcare. Paper presented at
the Applied Statistics Education and Research Collaboration (ASEARC) Conference,
University of Western Sydney, Paramatta, Australia.

Institute of Medicine. About the Institute of Medicine. Retrieved September 5, 2012, from
http://www.iom.edu/About-IOM.aspx

Institute of Medicine. (2009). Committee on accelerating rare diseases research and orphan
product development: AdvaMed written testimony. Retrieved August 19, 2012, from
http://www.iom.edu/~/media/Files/Activity
Files/Research/OrphanProductResearch/23-NOV-09/Alpert.pdf

Institute of Medicine. (2011). Rare diseases and orphan products : Accelerating research and
development. Washington, DC, USA: National Academies Press.

International Society for the Systems Sciences. (2012). Ludwig von bertalanffy: General systems
theory. Retrieved 2012, September 14, from http://www.isss.org/quotelvb.htm

Kaplan, A. V., Harvey, E. D., Kuntz, R. E., Shiran, H., Robb, J. F., & Fitzgerald, P. (2005).
Humanitarian Use Devices/Humanitarian Device Exemptions in cardiovascular medicine.
Circulation, 112(18), 2883-2886. doi: 10.1161/CIRCULATIONAHA.105.553701

Kary, W., Bergthold, L., Talbott, M., & Aubry, W. (2009). Coverage for evidence development:
A conceptual framework. Retrieved April 20, 2013, from http://www.cmtpnet.org/wp-
content/uploads/downloads/2012/03/CED-Issue-Brief.pdf

Kern, R. (2013). CDRH preps new program to streamline approval-to-reimbursement path.
Retrieved December 20, 2013, from http://www.elsevierbi.com/publications/the-gray-
sheet/39/51/cdrh-preps-new-program-to-streamline-approvaltoreimbursement-
path?elsca2=rss&elsca1=gray

178
Kirkwood, C. W. (1998). Chatper 1: System behavior and causal loop diagrams. Retrieved
December 30, 2013, from http://www.public.asu.edu/~kirkwood/sysdyn/SDIntro/ch-
1.pdf

Laszlo, A., & Krippner, S. (1997). System Theories: Their origins, foundations, and development.
Retrieved September 14, 2012, from
http://archive.syntonyquest.org/elcTree/resourcesPDFs/SystemsTheory.pdf

Lebcir, R. M. (2006). Health care management: The contribution of systems thinking. Retrieved
September 15, 2012, from https://uhra.herts.ac.uk/dspace/bitstream/2299/683/1/S65.pdf

Leischow, S. J., Best, A., Trochim, W., Clark, P. I., Gallagher, R., Marcus, S. E., & Matthews, E.
(2008). System thinking to improve the public's health. American Journal of Preventive
Medicine, 35(2S), S196-S203.

Lewis, M. (2008). Medicare reimbursement for drugs and devices (chapter 18) emerging life
sciences companies deskbook. Retrieved December 21, 2013, from
http://www.morganlewis.com/documents/erh/ERH_MedicareReimbursementForDrug
sAndDevices_ELSCDeskbook.pdf

Leydesdorff, L. (1997). Sustainable technological developments and second-order cybernetics. .
Technology Analysis and Strategic Management, 9(3), 329-341.

Linder, S. H., & Peters, B. G. (1984). From social theory to policy design. Journal of Public
Policy, 4(3), 237-259.

Lundy, S. H., Barry, D. M., Thiel, D. K., Polston, M. D., Farber, D. J., Pinto, P. N., & Lorell, B.
(2013). CMS updates policies and procedures for National Coverage Determinations and
annoucnes a new expedited process for removing certain older NCDs. Retrieved December
27, 2013, from
http://www.kslaw.com/imageserver/KSPublic/library/publication/ca081213.pdf

179
Lundy, S. H., Barry, D. M., Thiel, D. K., Polston, M. D., Pinto, P. N., Lorell, B. H., & Adam, L.
M. (2012). CMS issues new draft guidance on coverage with evidence development policy
for national coverage determinations. Retrieved April 20, 2013, from
http://www.kslaw.com/imageserver/KSPublic/library/publication/ca120612.pdf

Magnana, R., Sabin, K., Saidel, T., & Hackathorn, D. (2005). Sampling hard to reach and hidden
populations for HIV surveillance. AIDS, 19(2), 67-72.

Mason, M. (2010). Sample size in saturatoin in PhD studies using qualitative interviews.
Retrieved October 28, 2012, from http://www.qualitative-
research.net/index.php/fqs/article/view/1428/3027

Mathews, L. G., & Jones, A. (2008). Using systems thinking to improve interdisciplinary learning
outcomes: Reflections on a pilot study in land economics. Integrative Studies, 26, 73-104.

Matthews, J. T., & Glass, L. (2013). The effect of market-based economic factors on the adoption
of orphan drugs across multiple countries. Therapeutic Innovation and Regulatory Science,
47(226).

Medica. (2011). Medica coverage policy: Gastric electrical stimulation. Retrieved August 29,
2012, from http://provider.medica.com/c15/policyindex/Document
Library/GastricElectricalStim_CP.pdf

Morgan, D. L. (1993). Qualitative content analysis: A guide to paths not taken. Qualitative
Health Research, 1, 112-121.

National Institutes of Health. The Office of Rare Diseases Research (ORDR): Frequently ask
questions. Retrieved December 24, 2013, from http://rarediseases.info.nih.gov/about-
ordr/pages/30/about-ordr

National Organization for Rare Diseases. NORD physician guides: Rare disease resources for
medical professionals. Retrieved August 23, 2012, from http://nordphysicianguides.org/

180
National Organization for Rare Diseases. (2007). Ensuring safe medicines and medical devices
for children. Retrieved September 9, 2012, from
http://www.rarediseases.org/docs/policy/TestimonyofRobertM.pdf

National Organization for Rare Diseases. (2011). Treating rare and neglected pediatric diseases:
Promoting the development of new treatments and cures. Testimony of Diane Edquist
Dorman before the U.S. Senate Committee on Health, Education, Labor, and Pensions.
Retrieved August 14, 2012, from
http://www.rarediseases.org/docs/policy/copy_of_StatementDorman710.pdf

National Organization for Rare Diseases. (2012). Humanitarian use devices. Retrieved September
5, 2012, from http://www.rarediseases.org/advocacy/initiatives-updates/huds

Neumann, P. J., & Chambers, J. (2013). Medicare's reset on "coverage with evidence
development". Retrieved April 20, 2013, from
http://healthaffairs.org/blog/2013/04/01/medicares-reset-on-coverage-with-evidence-
development/

Newell, B., Proust, K., Dyball, R., & McManus, P. (2007). Seeing obesity as a systems problem.
NSW Public Health Bulletin, 18, 214-218.

Nickerson, R. S. (1998). Confirmation bias: A ubiquitous phenomenon in many guises. Review of
General Phychology, 2, 175-200.

Nuhoglu, H. (2008). Modeling spring mass system with system dynamics apporach in middle
school education. The Turkey Online Journal of Educational Technology, 7(3), Article 3.

Oatey, A. (1999). The strengths and limitations of interviews as a reserch technique for studying
television viewers. Retrieved Dec 8, 2013, from
http://www.aber.ac.uk/media/Students/aeo9702.html

Premarket approval of medical devices., 21 CFR Part 814 C.F.R. § (n) (2011e).

181

Polit, D. F., & Beck, C. T. (2012). Nursing research: Generating and assessing evidence for
nursing practice. (9th ed.): Lippincott Williams & Wilkins.

Raubicheck, C. J., & Weigel, W. F. (1976). How to comply with the new medical device law.
Food and Drug Law Journal, 31, 312-326.

Redington, L. (2009). The Orphan Drug Act of 1983: A case study of issue framing and the
failure to effect policy change from 1990 - 1994. (Doctor of Public Health), University of
North Carolina.

Reuters, T. (2012). The economic power of orphan drugs. Retrieved September 8, 2012, from
http://img.en25.com/Web/ThomsonReutersScience/1001450.pdf

Rowberg, R., Randall, B., Porter, D., Reyes-Akinbileje, B., Vogt, D., & Duffy, D. (1998). Food
and Drug Administration Modernization Act of 1997: The provisions: Science, Technology,
and Medicine Division, CRS; American Law Division, CRS.

Rueter, P. H. (2009). The unintended consequences of drug policies. Retrieved December 14,
2013, from
http://www.rand.org/content/dam/rand/pubs/technical_reports/2009/RAND_TR706.p
df

Food and Drug Administration Safety and Innovation Act (FDASIA). C.F.R. (2012).

Samaras, E. A., & Samaras, G. M. (2010). Using human-centered systems engineering to reduce
nurse stakeholder dissonance. Biomedical Instrumentation & Technology: Human Factors,
44(s1), 25-32.

Second Sight Medical Products. (2013). Medicare designates Argus II retinal prosthesis system as
new technology and provides payment. Retrieved December 30, 2013, from http://www.2-
sight.eu/en/us-medicare-en

182
Sherman, M., Wentworth, S., & Whitaker, L. (2002). Humanitarian use devices. Food and Drug
Law Journal, 57(1), 95.

Siegel, D. S., & Wessner, C. (2012). Universities and the success of entrepreneurial ventures:
Evidence from the small business innovation research program. Journal of Technology
Transfer, 37(4), 404-415. doi: http://dx.doi.org/10.1007/s10961-010-9186-3

Smith, J. J., & Henderson, J. A. (2003). Bringing new medical technology to market:
Understanding CMS coverage and payment determinations. Retrieved August 18, 2012,
from http://www.cimit.org/news/regulatory/cms_coverage.pdf

Steven, A. (2007). Could the work system method embrace systems concepts more fully?
Information Resource Management Journal, 20(2), 33-43.

Stewart, J., & Ayres, R. (2001). Systems theory and policy practice: An exploration. Policy
Sciences, 34(1), 79-79.

Stuckey, H. (2013). Three types of interviews: Qualitative research methods in social health.
Journal of Social Health and Diabetes (pp. 56).

Sullivan, T. (2010). Senator Al Franken calls on additional incentives for medical device
companies to tackle rare diseases. Retrieved August 19, 2012, from
http://www.policymed.com/2010/08/page/4/

Trueman, P., Grainger, D., & Downs, K. (2010). Coverage with evidence development:
Applications and issues. International Journal of Technology Assessment in Health Care,
26(1), 79-85.

Tunis, S. R., Berenson, R. A., Phurrough, S. E., & Mohr, P. E. (2011). Improving the quality and
efficiency of the Medicare program through coverage policy. Retrieved April 20, 2013, from
http://www.urban.org/UploadedPDF/412392-Improving-the-Medicare-Program-
Through-Coverage-Policy.pdf

183
Tunis, S. R., & Kang, J. L. (2001). Improvements in medicare coverage of new technology.
Health Affairs, 20, 83-84.

Tunis, S. R., & Pearson, S. D. (2006). Coverage options for promising technologies: Medicare's
coverage with evidence development. Health Affairs, 25(5), 1218-1230.

U.S. Centers for Medicare and Medicaid Services. (2007). Proposed decision memo for clinical
trial policy (CAG-0071R). Retrieved September 10, 2012, from
http://www.cms.gov/medicare-coverage-database/details/nca-proposed-decision-
memo.aspx?NCAId=186&ver=26&NcaName=Clinical+Trial+Policy+(1st+Recon)&
bc=BEAAAAAAEAAA&&fromdb=true

U.S. Centers for Medicare and Medicaid Services. (2008a). Decision memo for artificial hearts
(CAG-00322N). Retrieved September 12, 2012, from http://cms.hhs.gov/medicare-
coverage-database/details/nca-decision-
memo.aspx?NCAId=211&NcaName=Artificial+Hearts&bc=gCAAAAAAIAAA&

U.S. Centers for Medicare and Medicaid Services. (2008b). Decision memo for intracranial
stenting and angioplasty (CAG-00085R5). Retrieved August 18, 2012, from
http://www.cms.gov/medicare-coverage-database/details/nca-decision-
memo.aspx?NCAId=214&fromdb=true

U.S. Centers for Medicare and Medicaid Services. (2010). Medicare National Coverage
Determinations manual. Chapter 1, Part 1 (Section 10 - 80.12) coverage determinations.
Retrieved August 18, 2012, from http://www.cms.gov/Regulations-and-
Guidance/Guidance/Manuals/downloads/ncd103c1_Part1.pdf

U.S. Centers for Medicare and Medicaid Services. (2012a). Coverage with evidence
development. Retrieved September 12, 2012, from
https://http://www.cms.gov/Medicare/Coverage/Coverage-with-Evidence-
Development/index.html

184
U.S. Centers for Medicare and Medicaid Services. (2012b). Draft guidance for the public,
industry, and CMS staff coverage with evidence development in the context of coverage
decisions. Retrieved April 20, 2013, from http://www.cms.gov/medicare-coverage-
database/details/medicare-coverage-document-details.aspx?MCDId=23

U.S. Centers for Medicare and Medicaid Services. (2013a). Report to congress on National
Coverage Determinations for fiscal year 2012. Retrieved December 14, 2013, from
http://www.cms.gov/Medicare/Coverage/InfoExchange/Downloads/2012-Report-To-
Congress.pdf

U.S. Centers for Medicare and Medicaid Services. (2013b). View public comments for draft
guidance for the public, industry, and CMS staff coverage with evidence development in the
context of coverage decisions. Retrieved April 23, 2013, from
http://www.cms.gov/medicare-coverage-database/details/mcd-view-public-
comments.aspx?MCDId=23&ExpandComments=n&bc=AAAAAAAAAIAAAA%3d
%3d&

U.S. Department of Health and Human Services. (2005). Moving medical innovations forward:
New initiatives fromHHS. Retrieved August 25, 2012, from
http://www.hhs.gov/ash/bloodsafety/archives/medicalinnovations.shtml

U.S. Food and Drug Administration. (2004). Solicitation of comments on stimulating innovation
in medical technologies. Retrieved August 18, 2012, from
http://www.fda.gov/ohrms/dockets/dockets/04s0233/04S-0233-EC-14-Attach-1.pdf

U.S. Food and Drug Administration. (2009a). Food and Drug Administration Modernization Act
of 1997. Retrieved August 12, 2012, from
http://www.fda.gov/RegulatoryInformation/Legislation/FederalFoodDrugandCosmeti
cActFDCAct/SignificantAmendmentstotheFDCAct/FDAMA/FullTextofFDAMAlaw/
default.htm

185
U.S. Food and Drug Administration. (2009b). IDE definitions and acronyms. Retrieved
September 8, 2012, from
http://www.fda.gov/MedicalDevices/DeviceRegulationandGuidance/HowtoMarketYo
urDevice/InvestigationalDeviceExemptionIDE/ucm046698.htm

U.S. Food and Drug Administration. (2009c). Office of Orphan Products Development budgetary
report FY2009. Retrieved August 18, 2012, from
http://www.fda.gov/downloads/AboutFDA/ReportsManualsForms/Reports/BudgetRe
ports/2009FDABudgetSummary/ucm116144.pdf

U.S. Food and Drug Administration. (2010a). Considerations regarding Food and Drug
Administration review and regulation of articles for the treatment of rare diseases; public
hearing. Federal docket number: FDA–2010–N–0218. Retrieved September 5, 2012, from
http://www.fda.gov/downloads/ScienceResearch/SpecialTopics/CriticalPathInitiative/
UCM243857.pdf

U.S. Food and Drug Administration. (2010b). Guidance for HDE holders, Institutional Review
Boards (IRBs), clinical investigators, and FDA staff - Humanitarian Device Exemption
(HDE) regulation: Questions and answers. Retrieved August 14, 2012, from
http://www.fda.gov/medicaldevices/deviceregulationandguidance/guidancedocuments
/ucm110194.htm

U.S. Food and Drug Administration. (2010c). Medical device regulations: Overview,
opportunities, and challenges. Retrieved August 14, 2012, from
http://www.fda.gov/downloads/MedicalDevices/NewsEvents/WorkshopsConferences
/UCM233492.pdf

U.S. Food and Drug Administration. (2010d). Medical devices; Humanitarian Use Devices Part
V. 21 CFR Part 20 and 814. Docket No.91N-0404. 61 FR 33232. Retrieved August 15, 2012,
from
http://www.fda.gov/ScienceResearch/SpecialTopics/RunningClinicalTrials/ucm1205
49.htm

186
U.S. Food and Drug Administration. (2010e). Public workshop: Identifying unmet public health
needs and facilitating innovation in medical device development. AdvaMed testimony.
Retrieved September 4, 2012, from
http://www.fda.gov/downloads/MedicalDevices/NewsEvents/WorkshopsConferences
/UCM221035.pdf

U.S. Food and Drug Administration. (2010f). Treating rare and neglected pediatric diseases:
Promoting the development of new treatments and cures. Retrieved August 19, 2012, from
http://www.fda.gov/NewsEvents/Testimony/ucm219765.htm

U.S. Food and Drug Administration. (2011a). Draft guidance for industry and Food and Drug
Administration staff: Humanitarian Use Device (HUD) designations. Retrieved August 14,
2012, from
http://www.fda.gov/downloads/ForIndustry/DevelopingProductsforRareDiseasesCon
ditions/DesignatingHumanitarianUseDevicesHUDS/UCM283
504.pdf

U.S. Food and Drug Administration. (2011b). FDA-CMS parallel review. Retrieved December
20, 2013, from
http://www.fda.gov/MedicalDevices/DeviceRegulationandGuidance/HowtoMarketYo
urDevice/PremarketSubmissions/ucm255678.htm

U.S. Food and Drug Administration. (2011c). FDA-CMS parallel review. Retrieved August 25,
2012, from
http://www.fda.gov/MedicalDevices/DeviceRegulationandGuidance/HowtoMarketYo
urDevice/PremarketSubmissions/ucm255678.htm

U.S. Food and Drug Administration. (2011d). Institutional Review Boards frequently asked
questions: Information sheet. Retrieved September 8, 2012, from
http://www.fda.gov/RegulatoryInformation/Guidances/ucm126420.htm

187
U.S. Food and Drug Administration. (2011f). Strategic prioritiries 2011 - 2015: Responding to
public health challenges of the 21st century. Retrieved August 14, 2012, from
http://www.fda.gov/downloads/AboutFDA/ReportsManualsForms/Reports/UCM252
092.pdf

U.S. Food and Drug Administration. (2012a). Listing of CDRH Humanitarian Device Exemption.
Retrieved August 8, 2012, from
http://www.fda.gov/MedicalDevices/ProductsandMedicalProcedures/DeviceApproval
sandClearances/HDEApprovals/ucm161827.htm

U.S. Food and Drug Administration. (2012b). Office of Orphan Product Development budgetary
report FY2013. Retrieved August 18, 2012, from
http://www.fda.gov/downloads/AboutFDA/ReportsManualsForms/Reports/BudgetRe
ports/UCM298358.pdf

U.S. Food and Drug Administration. (2012c). Premarket Approval (PMA). Retrieved September
8, 2012, from
http://www.fda.gov/MedicalDevices/DeviceRegulationandGuidance/HowtoMarketYo
urDevice/PremarketSubmissions/PremarketApprovalPMA/default.htm

U.S. Food and Drug Administration. (2013a). Guidance for industry and Food and Drug
Administration staff: Humanitarian Use Device (HUD) designations. Retrieved March 10,
2013, from
http://www.fda.gov/ForIndustry/DevelopingProductsforRareDiseasesConditions/Desi
gnatingHumanitarianUseDevicesHUDS/LegislationRelatingtoHUDsHDEs/ucm28351
7.htm

U.S. Food and Drug Administration. (2013b). Office of Orphan Product Development budgetary
report FY12 through FY14. Retrieved December 21, 2013, from
http://www.fda.gov/downloads/AboutFDA/ReportsManualsForms/Reports/BudgetRe
ports/UCM349751.pdf

188
U.S. Food and Drug Administration. (2013c). Office of Orphan Product Development designates
200th Humanitarian Use Device. Retrieved January 10, 2014, from
http://www.fda.gov/ForIndustry/DevelopingProductsforRareDiseasesConditions/OOP
DNewsArchive/ucm351296.htm

U.S. Food and Drug Administration. (2013d). Section 1128 of the Food and Drug Administration
Safety and Innovation Act (FDASIA): Small business report to congress. Retrieved
December 19, 2013, from
http://www.fda.gov/downloads/RegulatoryInformation/Legislation/FederalFoodDrug
andCosmeticActFDCAct/SignificantAmendmentstotheFDCAct/FDASIA/UCM36005
8.pdf

U.S. Government Accountability Office. (2011). Report to congressional committees: Pediatric
medical devices - provisions support development, but better data needed for required
reporting. Retrieved August 22, 2012, from http://www.gao.gov/assets/590/587164.pdf

U.S. Government Accountability Office. (2012). About GAO. Retrieved October 12, 2012, from
http://www.gao.gov/about/index.html

UPMC Health Plan. (2011). UPMC health plan: Selective internal radiation therapy. Retrieved
August 13, 2012, from http://www.upmchealthplan.com/pdf/PandP/MP.037.pdf

World Health Organization. (2009). System thinking for health systems strengthening. Retrieved
December 21, 2013, from
http://whqlibdoc.who.int/publications/2009/9789241563895_eng.pdf

Yin, R. K. (1994). Case study research: Design and methods. (2nd ed.): Sage Publications.

Yoon, S. W., & Kuchinke, K. P. (2005). Systems theory and technology. Performance
Improvement, 44(4), 15-20.

189
Appendix A: Analysis of Approved HDEs
HDE
Designati
on
HDE
Approval
Progressi
on of
HUD 
HDE
Device Description Sponsor Implant Pediatric
Indication
Type of Devices
Vas
1
Ortho
2
Uro
3
/
Gastro
4

ENT
5
/
Opth
6

Neuro
7
/
Other
Nov 1996 Feb 1997 4 months Harrison Fetal
Bladder Stent
Set
Fetal Urologic
Obstruction
Cook OB/GYN X X X
June
1997
Sep 1997 4 months King’s College
Fetal Bladder
Drainage
Catheter
Fetal Urologic
Obstruction
Rocket
Medical PLC
X X X
Sep 1997 Dec 1997 4 months Urostim Neurostimulation
– Bladder Control
WE Kaplan &
I Richards
X X X
Nov 1996 Apr 1998 18 months Excorim
Immunoadsorpt
ion System
Coagulopathy
Treatment
Cobe BCT X
Aug 1997 Apr 1998 9 months Perma-Flow
Coronary Stent
Prosthetic
Conduit for
CABG
Posis Medical,
Inc
X X
Dec 1997 Sep 1998 10 months Interphalangeal
Finger
Prosthesis
Orthopedic
Impant – Joint
Avanta
Orthopaedics,
Inc
X X
Apr 1998 Dec 1998 9 months Vocare Bladder
System
Neurostimulation
- Bladder
NeuroControl
Corp
X X
Apr 1998 Feb 1999 11 months Vocare Bladder
System
Neurostimulation
- Bowel
NeuroControl
Corp
X X
Jan 1999 Sep 1999 9 months CardioSeal
Septal
Septal Occluder NMT Medical X X X

190
Occlusion
System
Dec 1998 Sep 1999 10 months Action
Neosphincter
Artificial Rectal
Sphincter
AMS X X X
Feb 1999 Sep 1999 8 months CardioSeal
Septal
Occlusion
System
Septal Occluder NMT Medical X X X
Aug 1998 Sep 1999 14 months Shelhigh
Pulmonic
Valave Conduit
Pulmonic Valave Shelhigh,Inc. X X X
Mar 1999 Dec 1999 10 months BioGlue
Surgical
Adhesive
8

Bioadhesive Cryolife, Inc X
Dec 1997 Dec 1999 25 months TheraSphere
MDS
Brachy Tx
Implant-Hepatic
Nordion, Inc X X
Aug 1999 Feb 2000 7 months CardioSeal
Septal
Occlusion
System
9

Septal Occluder NMT Medical X X
June
1999
Mar 2000 10 months Telescopic
Plate Spacer
Spinal System
9

Orthopedic
Implant – Disk
InterporeCross
Internat
X X
Sep 1999 Mar 2000 7 months Enterra
Therapy
System
Gastric Electric
Stimulation
Medtronic, Inc X X
Sep 1999 May 2000 9 months TAS Ecarin
Clotting Time
Test
Coagulation
Assay
Cardiovascular
Diagnostics
X
Oct 1999 Jan 2001 16 months Jomed Jostent
Coronary Stent
Coronary Stent
Graft
Jomed AB X X

191
Graft
Apr 1998 Feb 2001 35 months Composite
Cultured Skin
Artificial Skin Ortec
International
Inc.
X
Jul 1997 Mar 2001 45 months Prostalac Hip
Temporary
Prosthesis
Orthopedic
Implant – Hip
DePuy X X
Apr 1998 Aug 2001 41 months Metatarsophala
ngeal (MCP)
Joint Implant
Orthopedic
Implant – Joint
Avanta
Orthopedics,
Inc.
X X
May 2001 Oct 2001 6 months OP-1 Implant Bioadhesive Stryker Corp X
Mar 2000 Dec 2001 22 months VISX Custom
Contoured
Ablation
Pattern
Ophthalmologic
Laser Pattern
VISX, Inc. X
Dec 2001 Mar 2002 4 months Ascension PIP Orthopedic
Implant – Joint
Ascension
Orthopedics
X X
10
See
note
below
Apr 2002
10
See
note
below
Amplatzer PFO
Occluder
9

PFO Occluder AGA Medical X X
Sep 2001 Aug 2002 12 months Neurolink Stent
& Delivery
Catheter
Neurovascular
Stent
Guidant Corp X X
Aug 2000 Sep 2002 26 months Neuroform
Microdelivery
Stent System
Neurovascular
Stent
Smart
Therapeutics,
Inc.
X X
Nov 2001 Apr 2003 18 months Activa
Dystonia
Therapy
Neurostimulation Medtronic, Inc. X X X
Jan 2002 Jul 19 months Dermagraft
9
Artificial Skin Smith & X

192
2003 Nephew
Apr 2002 Nov 2003 20 months Contegra
Pulmonary
Valved Conduit
Pulmonic Valve Medtronic, Inc X X X
Oct 2002 Feb 2004 17 months Heartsbreath
Cardiac
Injection
Cardiac Rejection
Assay
Menssana
Research, Inc
X
Jun 2003 Feb 2004 9 months DeBakey –
Child Left
Ventricular
Assist System
Pediatric LVAD MicroMed
Technology
Inc.
X X X
Jan 2003 Apr 2004 16 months OP-1 Putty Bioadhesive Stryker X
Mar 2004 Jul 2004 5 months INTACS
Prescription
Inserts for
Keratoconus
Ophthalmologic
Insert
Addition
Technology
X X
Oct 1997 Aug 2004 84 months Vertical
Expandable
Prosthetic
Titanium Rib
Orthopedic
Implant – Rib
Synthes, Inc X X
Aug 2002 Mar 2005 32 months CoAxia
NeuroFlo
Catheter
Catheter –
Treatment of
Cerebral
Ischemia
CoAxia, Inc X
Jan 2004 Aug 2005 20 months Wingspan Stent
System
PTA Balloon
Catheter Stent
System
Boston
Scientific
X X
Oct 2003 Mar 2006 30 months Karl Storz
Fetoscopy
Instrument
Transabdominal
Amnioscope
(Fetoscope)
Karl Storz
Endoscopy –
America, Inc.
X X
Sep 2003 Sep 2006 37 months Abiocor®
Implantable
Artificial Heart Abiomed, Inc. X X

193
Replacement
Heart
Sep 2006 Jan 2007 5 months Fujirebio
Mesomark
Assay
Enzyme-linked
Immunosorbent
Assay
Fujirebio
Diagnostics,
Inc.
X
Nov 2003 Apr 2007 42 months Onyx HD-500,
Model 105-
8101-500
Artificial
Embolization
Device
EV3
Neurovascular
X
May 2005 May 2007 25 months ENTERPRISE
Vascular
Reconstruction
Device and
Delivery
System
Intracranial Stent Cordis
Neurovascular,
Inc
X X
Nov 1998 Oct 2007 109
months
Epicel Cultured
Epidermal
Autografts
Genzyme
Biosurgery
X X
Mar 2006 June 2008 28 months NeuRx RA/4 Diaphragmatic
Pacing System
Synapse
Biomedical,
Inc.
X X
Jan 2007 Oct 2008 22 months Levitronix
Centrimag
Right
Ventricular
Assist System
(RVAS)
Ventricular
Bypass (Assist)
Device
Levitronix,
LLC
X X
Mar
2003
Oct 2008 69 months INFUSE/Maste
rgraft
Posterolateral
Revision
Device
9

Bone Graft Medtronic Inc. X
Apr 2004 Oct 2008 55 months IBV Valve
System
Bronchial Valve Spiration, Inc. X X

194
Mar 2005 Feb 2009 48 months Reclaim Deep
Brain
Stimulation –
Obsessive
Compulsive
Disorder
Deep Brain
Stimulation
Medtronic, Inc. X X
Jul 2007 Jan 2010 31 months Melody
Transcatheter
Pulmonary
Valve
Pulmonary Valve Medtronic Inc. X X X
Sep 2003 Mar
2011
92 months Elana Surgical
Kit
Arteriotomy
System
Elana, Inc. X
Feb 2010 Apr 2011 15 months cPAX
Aneurysm
Treatment
System
Neurovascular
Embolization
Device
NeuroVasx,
Inc.
X X
Sep 2010 Sep 2011 13 months NeuroRX DPS
Diaphragm
Pacing System
Diapgragm
Pacing Stimulator
Synapse
Biomedical
X X
Feb 2005 Nov 2011 83 months Ovation
Abdominal
Stent Graft
System
Endovascular
Graft
TriVascular,
Inc.
X X
May 2009 Nov 2011 31 months BSD-2000
Hyperthermia
System
Hyperthermia
System
BSD Medical
Corporation
X
Jan 2001 Dec 2011 133
months
Berlin Heart
EXCOR
Pediatric
Ventricular
Assist Device
(VAD)
Ventricular Assist
Device
Berlin Heart,
Inc.
X X X

195
1
Cardiovasular Devices.
2
Orthopedics Devices.
3
Urology Devices.
4
Gastroenterology Devices.
5
Ear, Nose, and Throat Devices.
6
Opathalamic Devices.
7
Neurological
Devices.
8
HDE converted to PMA.
9
HDE withdrawn.
10
“Summary of Safety and Probable Benefit” was removed from the FDA website. No data available

May 2009 Feb 2013 47 months Argus II Retina
Prosthesis
System
Retina Prosthesis
System
Second Sight
Medical
Products
X X
Jan 2010 Oct 2013 47 months Kaneka
Liposorber®
LA-15 System
Extracorporeal
blood processing
system
Kaneka
Pharma
American LLC
X X
Percentage % (out of 56 HDEs) 71 24 33 16 12 3 35

196

Appendix B: Regulatory Interview Guide

197

198

199

200

201

202

203

204
Appendix C: Reimbursement Interview Guide

205

206

207

208

209

210

211

212
Appendix D: Examples of Invitation E-Mails to Informants

Note: Names of informants have been redacted to preserve confidentiality

213

Note: Names of informants have been redacted to preserve confidentiality

Abstract (if available)

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

FDA influence on advisory committees through documentation: a content analysis and survey of industry views

PDF

Current practices in pharmaceutical container closure development

PDF

Promotion of regulated products using social media: an industry view

PDF

The impact of FDA Patient-Focused Drug Development (PFDD) on US drug development strategies: a survey of views from pharmaceutical product companies

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

Incentivizing quality in the manufacture of pharmaceuticals: manufacturers' views on quality ratings

PDF

Reprocessing of single-use medical devices: a survey investigation comparing the views of three unheard stakeholders

PDF

Implementation of unique device identification in the medical device industry: a survey of the change management experience

PDF

Risk management and recalls: a survey of medical device manufacturers

PDF

Regulatory dissonance in the global development of drug therapies: a case study of drug development in postmenopausal osteoporosis

PDF

Examining the regulatory framework for drug compounding: industry views and experiences

PDF

Regulatory programs to foster medical product development: user experience in the United States and Japan

PDF

Organizational communication of regulatory intelligence: a survey of the medical device industry

PDF

Current practices in biocompatibility assessment of medical devices

PDF

Computerized simulation in clinical trials: a survey analysis of industry progress

PDF

Implementation of tobacco regulatory science competencies in the tobacco centers of regulatory science (TCORS): stakeholder views

PDF

Risk approaches and standards used in hospitals: a survey of industry views

PDF

The impact of agglomeration policy on CO₂ emissions: an empirical study using China’s manufacturing data

PDF

Using telemetry to ensure safe and reliable medical device operation: experience with defibrillators and infusion pumps

Asset Metadata

Creator Soo, Chin-Wei (author)

Core Title Convergence of United States regulatory and reimbursement policies impacting patient access to humanitarian use devices (HUD)

School School of Pharmacy

Degree Doctor of Regulatory Science

Degree Program Regulatory Science

Publication Date 02/19/2014

Defense Date 01/24/2014

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

Tag humanitarian device exemption,humanitarian use device,OAI-PMH Harvest,orphan disease,patient access

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Richmond, Frances J. (committee chair), Bar-Cohen, Yaniv (committee member), Lawless, Grant (committee member), Loeb, Jerry (committee member)

Creator Email sooc@usc.edu,weisusu@yahoo.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-363055

Unique identifier UC11296652

Identifier etd-SooChinWei-2256.pdf (filename),usctheses-c3-363055 (legacy record id)

Legacy Identifier etd-SooChinWei-2256.pdf

Dmrecord 363055

Document Type Dissertation

Format application/pdf (imt)

Rights Soo, Chin-Wei

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