Content uploaded by Diana Brixner
Author content
All content in this area was uploaded by Diana Brixner on May 24, 2020
Content may be subject to copyright.
164 Jour nal of Managed Care & Specialty Pharmacy JMCP February 2 019 Vol. 25, No. 2 www.jmcp.org
Preapproval Information Exchange: Perspectives of
U.S. Population Health Decision Makers on Preferences
for Early Engagement with Investigational Therapies
Diana Brixner, PhD, RPh; Tatia C. Woodward, MPH, MS; Nik Seifter, PharmD; Joseph Biskupiak, PhD, MBA;
Martin Marciniak, PhD; Peter Neumann, ScD; and Gary Oderda, PharmD, MPH
ABSTRACT
BACKGROUND: Preapproval information exchange (PIE) is the communication
of clinical and health care economic information (HCEI) on therapies in devel-
opment between U.S. population health decision makers (PHDMs) and drug
manufacturers before regulatory approval. Early access to HCEI can help
PHDMs plan budgets, inform formulary coverage decisions, and accelerate
policy development to improve patient access to innovative health technolo-
gies. While recent FDA guidelines and proposed legislation aim to clarify defi-
nitions and execution of PIE, the level of U.S. PHDMs’ awareness and prefer-
ences for early engagement with investigational therapies is unclear.
OBJECTIVES: To (a) assess U.S. PHDMs’ current knowledge and percep-
tions of PIE and (b) identify their preferences for PIE, in order to shape
future development of related guidelines and policy.
METHODS: An expert panel of 5 U.S. PHDMs representing national and
regional payers from integrated health plans, pharmacy benefit manage-
ment, and specialty pharmacy organizations par ticipated in a 2-round
modified Delphi process. A targeted literature review of PIE was used to
develop a web-based survey administered to the panel. Survey responses
were grouped by consensus, with ≥ 80% agreement or disagreement as
the threshold in round 1. In round 2, content experts moderated an in-
person meeting where panelists deliberated and then revoted on round 1
nonconsensus topics.
RESULTS : In the round 1 survey, the panelists reached consensus on 35 of
54 (65%) multiple-choice questions. In the round 2 face-to-face discussion,
19 nonconsensus questions were debated. One question was removed due
to duplication, and consensus was achieved on 16 additional questions,
with 2 items of nonconsensus remaining. Overall, consensus was achieved
on 51 of 53 topics (96 %). There was full consensus by the panelists that
PIE should encompass new molecular entities and new indications of mar-
keted therapies. Panelists completely agreed on the need for a legislative
“safe harbor” for PIE. Four of five panelists reported that the value of PIE
was high to PHDMs, and they expressed a strong preference for peer-to-
peer conversations with manufacturers’ medical or outcomes liaisons for
PIE. The main topic of nonconsensus was the optimal timing of PIE.
CONCLUSIONS: This panel of U.S. PHDMs achieved consensus on the value
of PIE to proactively budget, make informed formulary decisions, and
develop pharmaceutical policy to facilitate patient access to new therapies.
The PHDM panel’s preferences for PIE should be considered in legislative
discussions and planning for future PIE by PHDMs and manufacturers.
The full contribution of PIE to improving the U.S. health system can best
be realized under a safe harbor that allows U.S. PHDM and manufacturer
experts to engage in robust scientific and economic discourse. Additional
research and broad stakeholder engagement is needed to advance the
development of formal U.S. PIE guidelines.
J Manag Care Spec Pharm. 2 019 ; 25 (2) :164-73
Copyright © 2019, Academy of Managed Care Pharmacy. All rights reserved.
RESEARCH
• In the current environment, the majority of information exchange
between U.S. population health decision makers (PHDMs) and
manufacturers on new therapies occurs after approval, which
delays the formulary decision process and affects patient access
to drugs.
• The AMCP Partnership Forum has established that early access
and exchange of clinical and health care economic information
between PHDMs and industry is supportive of accelerating the
formulary decision process.
• Existing legislation and regulations permit the exchange of some
types of clinical information and health care economic informa-
tion between manufacturers and PHDMs before FDA approval;
however, ambiguity has increased uncertainty and inconsistent
information-sharing practices across stakeholders.
What is already known about this subject
• A panel of experts from the PHDM community agreed that PIE
has value and is needed to support efficiency and timeliness
in planning budgets and formulary design when discussions
include (a) new technologies and new indications for approved
technologies, (b) peer-to-peer conversations with manufacturers’
medical or outcomes liaisons, and (c) proposed marketing and
financial information from the manufacturer account manage-
ment team.
• The PHDM panelists held high expectations for financial infor-
mation as a critical component of PIE, including disclosure of
the projected wholesale acquisition cost price and price shift
assumptions (listed and justified) that could be affected by a set
of defined variables.
• PHDMs cited 3 key drivers for prioritizing PIE discussions: (a) size
of target patient population, (b) magnitude of the new budget
impact, and (c) expected speed of regulatory approval.
What this study adds
As the U.S. health care system pivots from volume
to value, U.S. population health decision makers
(PHDMs), such as payers and health plans, have
increasingly expressed a need for clinical and health economic
information on pharmaceuticals before U.S. Food and Drug
Administration (FDA) approval. In 1997, Congress passed the
www.jmcp.org Vol. 25, No. 2 Febr uary 2019 JMCP Journal of Managed Care & Specialt y Pharmacy 165
Preapproval Information Exchange: Perspectives of U.S. Population Health
Decision Makers on Preferences for Early Engagement with Investigational Therapies
opposition to HR 2026.8 These organizations expressed con-
cerns that HR 2026 may promote off-label use of drugs and
may disincentivize manufacturers from producing clinical data
after launch if PHDMs accept the early outcomes and modeling
information to make coverage decisions.
As new legislation and policies are under consideration,
there is an urgent need to assess their potential effects and
to recommend ways to operationalize PIE with appropriate
safeguards to ensure that such information reaches intended
entities, namely PHDMs.
There is a lack of research exploring U.S. PHDM perspectives
regarding PIE to further its dialogue, development, and refine-
ment. To this end, an expert panel of PHDMs convened to clarify
definitions of PIE, assess current PIE processes between PHDMs
and manufacturers, and identify U.S. PHDM preferences for
future preapproval communication with manufacturers.
■■ Methods
A study plan was developed to outline key steps in the research
protocol. A list of 31 U.S. PHDM thought leaders was assem-
bled, based on previous participation in health policy panels,
publications, and projects. From that list, 5 PHDMs were
recruited for an expert panel to ensure representation from
different types of payer organizations (e.g., integrated health
plans, pharmacy benefit management, and specialty pharmacy
organizations); geographic regions; and availability. A targeted
literature search identified relevant publications describing
PIE, policy consensus development, and the modified Delphi
method (an organized method for collecting views and infor-
mation).4, 9-11
A modified Delphi method consisting of 2 rounds was used
to gather and analyze the data for this project. Round 1 was an
online survey developed by investigators and content experts.
The survey collected anonymous opinions from the 5 PHDM
experts in November 2017. The literature suggests that a suit-
able minimum sample size is 7, but sample sizes as small as
4 have been used; that the decision regarding the number of
panelists is empirical and pragmatic and takes into consid-
eration time and expense; and that representation should be
assessed more by the quality of the participants rather than
the number.12-16
The questions in the survey were organized into 4 core
domains: (1) definition of PIE, (2) value of PIE, (3) current utility
of PIE, and (4) future improvement of PIE. The survey consisted
of 54 multiple-choice and 7 open-ended questions (Table 1).
The possible responses for 51 multiple-choice questions were
strongly agree, agree, disagree, and strongly disagree. The
remaining 3 multiple-choice questions had answers based on
a time frame. Each multiple-choice question contained a free-
text box for respondents to provide rationales for their answers.
Results from the survey were tabulated, and consensus
was evaluated by grouping the responses into 2 dichotomous
Food and Drug Administration Modernization Act (FDAMA),
which included Section 114 authorizing the communication of
“health care economic information” (HCEI) generally between
manufacturers and “a formulary committee or similar entity,”
as long as the information “directly related to an approved
indication” and was based on “competent and reliable scien-
tific evidence.”1 The law was intended to offer manufacturers
and PHDMs a “safe harbor” under which manufacturers and
PHDMs could proactively discuss HCEI without fear of legal
repercussions, provided that they do not violate the rules set
out in the law.2 However, the wording in Section 114 is vague
and open to interpretation, leaving manufacturers and PHDMs
uncertain about the types of information that could be legally
shared, the timing of information sharing, and the appropriate
participants.
This need for information exchange is especially acute in
the context of accelerated, targeted, or innovative therapies
for which the FDA has created expedited approval pathways.3
Early access to clinical and health economic information before
FDA approval can allow PHDMs to incorporate information
on drugs in development or expanded indications of existing
drugs into budgeting, formulary design, and patient access
decisions.4 In addition, PHDMs can anticipate the incorpora-
tion of such information into guidelines versus current stan-
dard of care, thereby ensuring patient access at the time of
regulatory approval.
In 2016, an Academy of Managed Care Pharmacy (AMCP)
Partnership Forum was held to address these issues and make
recommendations to Congress and the FDA. The forum’s rec-
ommendations included clarification that preapproval infor-
mation exchange (PIE) is intended as a bidirectional exchange
between manufacturers and PHDMs, a request that the FDA
and Congress provide a safe harbor allowing discussion of
HCEI with PHDMs 12-18 months in advance of product
approval to allow for budgeting and planning, and the provi-
sion of information to patients to support personal health care
decisions.4
At the end of 2016, section 3037 of the 21st Century Cures
Act amended FDAMA Section 114, seeking to clarify language
and add flexibility on how to interpret the legality of HCEI
exchanges. Section 3037 defined HCEI and sought to increase
the transparency of health economic methodology and expand
the audience for HCEI beyond formulary committees.5 In early
2017, the FDA issued a highly anticipated draft guidance clari-
fying some aspects but again leaving room for interpretation
of others.6
On April 6, 2017, HR 2026, the Pharmaceutical Information
Exchange Act was introduced into Congress by Representative
Bret Guthrie.7 However, support for HR 2026 has not been
universal. In a letter dated June 5, 2017, to the Energy and
Commerce Committee of the U.S. House of Representatives,
a group of organizations, including Public Citizen, voiced
166 Jour nal of Managed Care & Specialty Pharmacy JMCP February 2 019 Vol. 25, No. 2 www.jmcp.org
Preapproval Information Exchange: Perspectives of U.S. Population Health
Decision Makers on Preferences for Early Engagement with Investigational Therapies
Question
Number Question
Q1 PIE stands for “preapproval infor mation exchange.”
Q2 On balance, PIE is most helpful to manufacturers. PIE would be most advantageous to manufacturers.
Q3 PIE is intended to provide information to PHDMs on new molecular entities approaching NDA submission, or new indications being submitted
for approval.
Q4 PIE can provide health economic information, enabling value-based decision mak ing by PHDM s.
Q5 There is a need for a “safe harbor” for PIE dissemination.
Q6 PIE may include budget impact and comparative effectiveness models as evidence.
Q7 PIE allows for 2-way dialog ue between PHDMs and manufacturers, including relevant clinical and economic outcomes for their populations.
Q8 Our organization feels PIE benefits manufacturers more than it helps PHDMs.
Q9 A PIE discussion between PHDMs and industry is important.
Q10 Recognizing that NDA filing m ay occur up to a year before FDA approval, the most appropriate time for PIE before filing is: 3 months, 6 months,
12 months, 18 months.
Q11 The most appropriate time frame for PIE before PDUFA date is: 3 months, 6 months, 12 month s, 18 months.
Q12 The primar y audience for PIE is PHDMs at r isk for the cost of providing access to health care for populations.
Q13 On balance, among PIE recipients, PIE would be most advantageous to PHDMs.
Q14 By helping PHDMs optimize formulary forecasting and rates, PIE provides improved patient management.
Q15 PIE prov ides the ability to more accurately forecast health plan budgets to adjust premiums.
Q16 PIE should be disseminated only to PHDMs.
Q17 PIE should be dissem inated to practicing physicians.
Q18 PIE evidence requires a statistical comparison.a
Q19 PIE would offer information on disease prevalence and member impact on health plans for formulary planning.
Q20 PIE is nothing new. We already follow drugs in the pipeline and use the data for budget forecasting.
Q21 PIE is needed for orphan drugs where there is limited data available at launch.
Q22 PIE is NOT intended to promote off-label use of currently approved indication s.
Q23 The cur rent research (i.e., hori zon scanning ) conducted by our organization provides the PIE we need for decision making.
Q24 PIE is important for PHDMs for planning benefit design in the near term.a
Q25 PIE on pipeline products is supportive for anticipating Medicare Part D coverage decisions.
Q26 The value of PIE to our organi zation is high.
Q27 Currently, our organization conducts PIE on the following schedule: monthly, quarterly, semiannually, annually.
Q28 Information presented in PIE is perceived by my plans as reliable, similar to expectations for information provided under FDAMA Section 114.
Q29 PIE should be restricted to new molecular entities only, with no prior approvals.
Q30 Unsolicited requests for preapproval information from our organi zation prov ides sufficient informat ion we need from m anufacturers.
Q31 PIE on a new indication of an approved product r isks being considered off-label promotion.
Q32 There are gaps between what preapproval inform ation is currently available via preapproval unsolicited requests and what PHDM s need.
Q33 In the future, I believe PIE will help improve PHDMs’ formulary decision making.
Q34 PIE will address unmet informational needs for PHDMs.
Q35 I anticipate that the future value of PIE as proposed will be high for our organization.
Q36 PIE could anticipate f uture adherence/compliance considerations for investigational products.
Q37 PIE can provide burden of illness data on investigational products.
Q38 PIE can inform budget impact models on investigational products.
Q39 Including prescr ibing physicians as recipients of PIE will promote off-label use and will increase the misperceptions of PIE.
Q40 Manufacturer communication on products before approval including product information, approval timing, pricing, phasing, and
targeting/marketing strateg ies should be within the scope of PIE.
Q41 It i s important to my plan that data on new uses for approved products is available as PIE.
Q42 The appropr iate choice of who presents PIE information would depend on what data are being presented.
Q43 I prefer to receive clinical PIE information from the manufacturer’s medical organization (e.g., medical affairs representatives, health outcomes
liaisons, managed care liaisons).
Q44 I prefer to receive clinical PIE information from the manufacturer account management team.
Q45 I prefer to receive clinical PIE information from the organization’s brand marketing/market access representatives.
Q46 I prefer to receive health economic PIE information from the manufacturer’s medical organization (e.g., medical affairs representatives, health
outcomes liaisons, managed care liaisons).
Q47 I prefer to receive health economic PIE information from the manufacturer’s account management team.
Q48 I prefer to receive health economic PIE information from the m anufacturer’s brand marketing/market acce ss representatives.
Q49 I prefer to receive forecasting and marketing materials of PIE information from the manufacturer’s medical organization.
TABLE 1 Round 1 Online Survey Questions
continued on next page
www.jmcp.org Vol. 25, No. 2 Febr uary 2019 JMCP Journal of Managed Care & Specialt y Pharmacy 167
Preapproval Information Exchange: Perspectives of U.S. Population Health
Decision Makers on Preferences for Early Engagement with Investigational Therapies
organizations that provided medical and /or pharmaceutical
benefit coverage.
All panelists completed the round 1 web-based survey.
Consensus was reached on 35 of 54 multiple-choice questions
(65%). Across the 4 core survey domains, PHDMs expressed
strongest consensus in core domain 3 (current utility of PIE)
and core domain 4 (future improvement of PIE). The 19 non-
consensus questions identified in round 1 were predominately
from core domain 1 (definition of PIE) and core domain 2
(value of PIE). These were carried forward as questions for the
round 2 in-person workshop phase of the modified Delphi
process. In-person deliberation of nonconsensus questions
resulted in the removal of 1 question due to duplication. In
addition, PHDM discussion resulted in the revision of the
wording in question 24, changing “benefit design” to “formu-
lary design” for clarity. Further consensus was achieved on
16 questions in round 2. Overall, after both rounds of the process,
consensus was achieved on 51 of 53 questions (96%; Figure 1).
Figure 2 shows key points that panelists highlighted regard-
ing the definition and utility of PIE. All panelists reported
knowledge of PIE as an exchange of HCEI between PHDMs
and manufacturers and reported various PIE experiences to
date. There was full agreement that PIE is needed in the United
States and that it should apply to new molecular entities and
new indications of marketed therapies. All panelists agreed that
PIE should encompass HCEI and detailed pricing projections,
and panelists were unanimous in their call for a legislative
safe harbor to facilitate transparent PIE discussions. Panelists
reported broad consensus for peer-to-peer PIE conversations to
ensure that appropriately skilled participants from PHDMs and
manufacturers are engaged in the discussion. Most panelists
categories: agree and disagree. Consensus on agreement was
defined as at least 4 answers of agree or strongly agree, and
disagreement consisted of at least 4 answers of disagree or
strongly disagree. Questions with ≥ 80% agreement or dis-
agreement were considered to have reached consensus and
were complete. Questions with < 80% agreement or disagree-
ment were considered not in consensus. A summary of the
anonymous responses to the survey was provided to all par-
ticipants before round 2.
The second round consisted of an in-person panel discussion
with respondents in December 2017. The in-person discussion
reviewed the survey responses, focusing deliberations on ques-
tions for which consensus was not met. The 7 open-ended ques-
tions helped guide the discussion during the in-person meeting,
which focused on clarification of the definitions and panelists’
interpretation of the wording used in some of the questions.
In this meeting, respondents were required to resubmit their
answers to the multiple-choice questions that had not achieved
consensus in the first round, with their new responses based on
the facilitated discussions with the other panel members. Voting
in this round was done by paper ballots. Again, ≥ 80% was
required to achieve consensus on agreement or disagreement.
After voting for each domain was completed, the content expert
(author Neumann) provided commentary and summarized the
discussion. Following the meeting, an anonymized summary
report of the panel discussions was produced.
■■ Results
Five U.S. PHDMs participated in this study. They repre-
sented national and regional PHDMs from integrated health
plans, pharmacy benefit management, and specialty pharmacy
Question
Number Question
Q50 PIE is of greater value for drugs that receive F DA breakthrough designation.
Q51 PIE as proposed includes scientific and health economic information. I anticipate PIE as proposed will help address my organization’s needs.
Q52 PIE should in form comparative effectiveness models on investigational products.
Q53 Product information, approval timing, pr icing, and targeting/marketing strategies are components of PIE information that would be useful to my
plan.
Q54 I prefer to receive forecasting and marketing materials of PIE information from the manufacturer’s account management or marketing team.
Q55 Open-ended: PIE has been described as “preapproval information exchange” or “pharmaceutical information exchange.” Which do you think best
describes PIE and why?
Q56 Open-ended: From your perspective, what is your definition of PIE?
Q57 Open-ended: What are the gaps between what is allowed by PIE and what payers need?
Q58 Open-ended: What are your current unmet needs and challenges to effective forecasting?
Q59 Open-ended: What is most valuable to you and your organi zation from PIE?
Q60 Open-ended: In your view, why does your organi zation need to conduct PIE?
Q61 Open-ended: If your organization is not conducting PIE, what discussions are you currently having that contribute to forecasting?
aIn round 2, question #18 was omitted due to duplication, and question #24 was revised from “benefit design” to “formular y design” for clarity.
FDA = U.S. Food and Drug Administration; FDAMA = Food and Drug Administration Modernization Act; NDA = new drug application; PDUFA = Prescription Drug User
Fee Act; PHDM = population health decision maker; PIE = preapproval information exchange.
TABLE 1 Round 1 Online Survey Questions
(continued)
168 Jour nal of Managed Care & Specialty Pharmacy JMCP February 2 019 Vol. 25, No. 2 www.jmcp.org
Preapproval Information Exchange: Perspectives of U.S. Population Health
Decision Makers on Preferences for Early Engagement with Investigational Therapies
indicated that PIE would not be most advantageous or helpful
for manufacturers, but rather most advantageous and benefi-
cial for PHDMs. There was 100% agreement that PIE was not
intended for prescribing practitioners and that PIE should not
be disseminated to practicing physicians.
Figure 3 highlights key points from panelists regarding
what they perceived as the value of and need for PIE. For the
majority of panelists, the perceived value of PIE was high and
thought to be important in helping PHDMs optimize formulary
forecasting and rates, thereby improving patient management.
Key value drivers of PIE were data maturity and the trans-
parency of the investigational products’ HCEI, especially for
therapies designed to treat conditions with high unmet need
and high prevalence. One dissenting opinion on the value of
PIE expressed skepticism that it would ultimately help reduce
health care costs. All expert panelists agreed that PIE can
address unmet informational needs for PHDMs. The majority
of panelists agreed that PIE was not intended to promote off-
label use of medications for indications not currently approved.
Areas of nonconsensus included the optimal timing of PIE
(Table 2). Other areas of concern raised by the expert panelists
were the potential misperceptions of off-label promotion if PIE
involved a new indication for a marketed product, the logistics
of enforcing PIE policy, and the need for wider public education
regarding appropriate context and scope of PIE discussions.
■■ Discussion
The reality for most U.S. health plans is that budgets, formular-
ies, and patient access decisions are typically being made 12-18
months before a new plan year begins. These forecasting and
policy-setting activities are largely driven by Medicare Part D
regulations. But the effect of these decisions may be broad
because PHDMs are often responsible for public and private
plan subscribers. In most cases, PHDMs are handicapped
by information asymmetries when making outcomes and
economic projections regarding drugs in development or
drugs with expanded indications. Information gaps include
those related to epidemiological and disease-specific
FIGURE 1 Distribution of PHDM Responses to Survey Questions After Both Rounds of Voting
Question Number
Cumulative Response to QuestionsStrong Disagree Disagree Agree Strong Agree
Domain 4
Future Improvement
of PIE
Domain 3
Current Utility of PIE
Domain 2
Value of PIE
Domain 1
Definition of PIE
¤ 45
¤ 48
¤ 44
¤ 49
¤ 47
¤ 30 ¤ 29
¤ 20
¤ 24
¤ 23
¤ 17
¤ 8
¤ 2
¤ 4
¤ 16
¤ 9
¤ 5
¤ 14
¤ 15
¤ 13
¤ 12
¤ 12
¤ 7
¤ 6
¤ 3
¤ 1
¤ 31
¤ 28
¤ 22
¤ 19
¤ 21
¤ 25
¤ 26
¤ 35
¤ 36
¤ 38
¤ 42
¤ 40 ¤ 39
¤ 46
¤ 41
¤ 37
¤ 34 ¤ 33
¤ 43
¤ 32
¤ 50
¤ 54
¤ 51
¤ 53
¤ 52
Note: Numbers represent question numbers. Questions 10, 11, and 27 are absent because they require time frame-based responses. Question 18 was eliminated due to
redundancy. X-axis scoring was derived from a cumulative score averaged from a sum of scores assigned to the 5 responses per question. The scores assigned to each are as
follows: strong di sagree = -2; disagree = -1; agree = 1; strong ag ree = 2.
PHDM = population health decision maker; PIE = preapproval information exchange.
Distribution of PHDM Preferences by Domain
www.jmcp.org Vol. 25, No. 2 Febr uary 2019 JMCP Journal of Managed Care & Specialt y Pharmacy 169
Preapproval Information Exchange: Perspectives of U.S. Population Health
Decision Makers on Preferences for Early Engagement with Investigational Therapies
characteristics, financial information, and rates of clinical and
economic outcomes. The increased focus on value-based con-
tracts also requires PIE to develop contracting scenarios that
are ready to implement upon product launch. A lack of preap-
proval information can impede the proactive development of
pharmaceutical policies, hindering expedient patient access
to new innovations. Early information can also help assess
potential product placement in future updates to guidelines
versus current standards of care, as supported by published
literature. The unpublished data presented through PIE will
largely be related to pharmacoeconomics, modeling, and cost
information.
Although these assumptions have been largely promulgated
through proactive dialogue,4 consensus on the value of preap-
proval information across payer types and how such infor-
mation exchange would be operationalized remains largely
unknown. Our panel explored these critical areas through a
modified Delphi process, which allowed for an initial reaction
via an online survey and then the opportunity for in-person
dialogue and exchange regarding areas of nonconsensus.
Previous independent surveys have evaluated the use and
value of FDAMA Section 114. For example, in a 2013 study,
U.S. health economics and outcomes research (HEOR) leaders
for drug and device companies were surveyed to examine their
views on the state of the field.17 Approximately 62% agreed
that AMCP dossiers were useful to U.S. health plans, and 55%
stated that FDAMA Section 114 was useful. These findings
suggest strong recognition of the value of HEOR capabilities
and communication at senior management levels and optimism
about the field.17
We saw a similar sentiment in our initial survey results,
with great interest in PIE and support of a framework to facili-
tate dissemination. Respondents stated that PIE engagement
around clin ical data and HCE I was specif ically desired f rom man-
ufacturer medical and scientific representatives (i.e., medical or
outcomes liaisons) as opposed to commercial representatives.
FIGURE 2 Views of Key Panelists on Definition and Scope of PIE
NDA = new drug application; PHDM = population health decision maker; PIE = preapproval information exchange.
020 40 60 80 100
Percentage
Manufacturer communication on products preapproval, including product
information, approval timing, pricing, phasing, and targeting/marketing strategies,
should be within the scope of PIE. (Q40)
PIE allows for 2-way dialogue between PHDMs and manufacturers, including
relevant clinical and economic outcomes for their populations. (Q7)
PIE should be disseminated to practicing physicians. (Q17)
The primar y audience for PIE is PDHMs at risk for the cost of
providing access to health care for populations. (Q12)
There is a need for a safe harbor for PIE dissemination. (Q5)
PIE is intended to provide information to PHDMs on new molecular entities
approaching NDA submission or new indications being submitted for approval. (Q3)
What is PIE?
Strongly Agree Agree Disagree Strongly Disagree
170 Journal of Managed Care & Specialty Pharmacy JMC P February 2 019 Vol. 25, No. 2 www.jmcp.org
Preapproval Information Exchange: Perspectives of U.S. Population Health
Decision Makers on Preferences for Early Engagement with Investigational Therapies
Previous surveys queried respondents about the interpreta-
tion of FDAMA Section 114, exploring the FDA’s role in the
information exchange.18 Similarly, our survey revealed a strong
interest in a safe harbor that would allow for exchange to occur,
as well as the need for oversight of these exchanges.
Finally, a panel of experts, similar to our Delphi panel
round 2, was convened to debate the quality assessment of
real-world data for decision making by health plans. The need
for standardization of information exchange processes across
health plans and decision makers was a significant point of
consensus.19
In our combined approach, the value of PIE and the need
for a legislative safe harbor for PIE was clearly identified.
Considerations of the current PIE process and the future
optimization of PIE benefited from the in-person discussion.
The PHDM panelists held high expectations for financial
information as a critical component of PIE, including disclosure
of the projected wholesale acquisition cost price and pricing
assumptions (listed and justified) that could be affected by a
set of defined variables. PHDMs articulated a strong preference
for simple sensitivity analyses, such as price bands, with full
acknowledgement of inherent uncertainties before approval
and commercialization.
An important lesson from the panel discussion was that not
all new drugs will merit PIE discussions. The value of PIE will
likely vary among drugs, depending on the size of the affected
population, the drug’s cost and economic impact, and whether
the new therapy will create new costs or simply shift costs in a
crowded market. The value of PIE increases as a function of the
prospective new budget impact. The budget impact of a new
therapy for a large population may not be significant if payers
are already providing multiple therapy options or if they have
few of the affected population in their plans. Conversely, intro-
ducing a novel therapy for a small population at a high cost,
or a new adjunctive therapy, could introduce a greater budget
impact than the previous example increasing the value of PIE.
FIGURE 3 Perceptions of Key Panelists on Value of and Need for PIE
020 40 60 80 100
Percentage
PIE is NOT intended to promote off-label use of currently
approved indications. (Q22)
PIE will address unmet informational needs for PHDMs. (Q34)
The value of PIE to our organization is high. (Q26)
PIE is needed for orphan drugs where there is limited data available at launch. (Q21)
By helping PHDMs optimize formulary forecasting and rates, PIE
provided improved patient management. (Q14)
PIE can provide health economic information, enabling value-based
decision making by PHDMs. (Q4)
Value/Need for PIE
Strongly Agree Agree Disagree Strongly Disagree
PHDM = population health decision maker; PIE = preapproval information exchange.
www.jmcp.org Vol. 25, No. 2 Febr uary 2019 JMCP Journal of Managed Care & Specialt y Pharmacy 171
Preapproval Information Exchange: Perspectives of U.S. Population Health
Decision Makers on Preferences for Early Engagement with Investigational Therapies
The issue of safe harbor was not addressed by the recent
FDA guidance, which reaffirms the need for legislative solu-
tions such as HR 2026. Notably, a similar concern regarding
off-label promotion was highlighted by the 2017 Public Citizen
et al. letter to Congress in opposition to HR 2026.8 A key dif-
ference, however, is recognition by the PHDM panelists that
PIE is communication between manufacturers and PHDMs, not
between manufacturers and prescribers. This confusion among
public policy stakeholders about the intended PIE audience
and participants—PHDMs and manufacturers—highlights the
need for a formal framework to standardize PIE.
A final point of legislative consideration was enforcement of
PIE. Presumably the FDA would be responsible for codifying
HR 2026 with respect to monitoring and enforcing adherence.
We recommend broad stakeholder input (patients, PHDMs,
and manufacturers) for the development of PIE regulations to
ensure full transparency.
Limitations
While making important steps towards defining and preparing
for new regulations on PIE, this study has limitations. First,
the panel consisted of a small number of PHDMs who may
not adequately represent the broader U.S. PHDM population.
Second, while the combination of an anonymous web-based
survey and a face-to-face meeting to discuss areas with less
than 80% consensus was beneficial, a limitation was that a
2-step modified Delphi process was used. Multiple iterations
of data gathering and syntheses have been previously recom-
mended for full Delphi processes.20
■■ Conclusions
A panel of 5 national and regional U.S. PHDMs agreed that PIE
is valuable, needed, and should apply to new molecular entities
and new indications of marketed therapies. This sample of U.S.
PHDMs desired PIE with manufacturers as peer-to-peer discus-
sions between appropriate experts encompassing clinical, sci-
entific, health economic, and financial information. PIE aligns
with PHDM goals of reducing uncertainty and risk exposure
while improving forecasting. PHDMs were uncertain whether
PIE would reduce overall costs; however, there was agreement
that PIE supports ongoing efforts to manage health care costs.
When manufacturers plan a PIE discussion with PHDMs, they
should consider the drug’s potential effect on relevant popula-
tions, overall new budget impact, price range projection, and
the unmet needs in the indication as they decide on the level of
PIE provided and the communication strategy of this informa-
tion to the most relevant health plans.
One area of nonconsensus pertained to the optimal timing
of PIE, which ranged from 3-12 months before FDA submis-
sion. While no consensus was achieved, panelists agreed that
manufacturers should engage with PHDMs regarding inves-
tigational products with potential “high PIE value” earlier in
this window. A related discussion focused on whether a new
indication for an existing drug may be considered off-label pro-
motion before launch. Without the enactment of a legislative
safe harbor to discuss preapproval clinical and HCEI without
fear of legal repercussions, panelists voiced concerns about the
full utility of PIE.
On June 12, 2018, the FDA released a final guidance titled
“Drug and Device Manufacturer Communications with Payors,
Formulary Committees, and Similar Entities—Questions and
Answ ers.”21 This guidance encompasses the agency’s guide-
lines on PIE requirements for manufacturers to discuss HCEI
for unapproved products, as well as unapproved indications
for existing products.21 The new guidance does not include a
requirement for an intent to file but instead requires a state-
ment that the product or indication has not been approved by
the FDA and that the safety and effectiveness of the product or
indication has not been established by the FDA. In addition,
companies will be required to provide information on the stage
of filing and any additional planned studies. While this guid-
ance does not directly remedy the nonconsensus observed in
this study pertaining to the optimal timing of PIE, compulsory
disclosure of the product’s stage of filing and any additional
planned studies may help PHDM audiences to plan and predict
the expected product launch. Interestingly, the FDA guidance
also covers health care economic communications related to
medical devices, an area that was not discussed or considered
in our Delphi panel approach.
Question
Round 1 Round 2
3 Months 6 Months 12 Months 18 Mon ths 3 Months 6 Months 12 Months 18 Mont hs
Recognizing that an NDA filing may occur up
to a year before FDA approval, the most
appropriate time for PIE before filing is:
60% 0% 40% 0% 40% 20% 40% 0%
The most appropriate time frame for PIE
before PDUFA date is:
0% 40% 40% 20% 0% 20% 40% 40%
aReflects questions in which consensus was not achieved after both rounds of the Delphi process.
FDA = U.S. Food and Drug Administration; NDA = new drug application; PDUFA = Prescription Drug User Fee Act; PIE = preapproval information exchange.
TABLE 2 Questions of Nonconsensusa
172 Journal of Managed Car e & Specialty Pharmacy JMC P February 2019 Vol. 25, No. 2 www.jmcp.org
Preapproval Information Exchange: Perspectives of U.S. Population Health
Decision Makers on Preferences for Early Engagement with Investigational Therapies
DISCLOSURES
This study was funded by GlaxoSmithK line (GSK). Bri xner, Oderda, and
Biskupiak are principals of Millcreek Outcomes Group, a consultancy that
received funding from GSK to conduct this study. Marciniak and Woodward
are employees of GSK and own stock in GSK. Seifter was employed by GSK at
the time of this study. Neumann served as external health policy advisor for
this study and has consulted or served on advisory boards with Merck, Bayer,
Pacira, Novo Nordisk, Amgen, Abbvie, Boston Health Economics, Vertex,
Precision Health Economics, the Congressional Budget Office, CEA Registry
Sponsors, Axovant, Veritech, Janssen, Parateck, Avexis, GSK, Celegene,
Bluebird, Roche, Sage, Sarepta, Biogen, and Ipsen. Neum ann al so reports grants
from Amgen, Lundbeck, Gates, NPC, Alzheimer’s Association, and NIH.
ACKNOWLEDGMENTS
The authors thank the following panelists for their participation in this proj-
ect: Steven G. Avey, MS, RPh, FAMCP, Vice President, Enterprise Specialty
Clinical Solutions, Medimpact; Dougla s S. Burgoyne, BSPhar m, Phar mD,
FAMCP, President, V Rx Pharmacy Services; H. Er ic Cannon, PharmD,
FAMCP, Assistant Vice President, Ph armacy Benefits, Select Health; Stanley E.
Ferrell, RPh, BSPh arm, MBA, Senior Director, Clinical Account Management,
Express Scripts; and John Leonard Fox, MD, Associate Chief Medical Officer,
Priority Health. The authors also thank medical writer Kelley J. P. Lindberg,
of Blue Raven Services, for her editing and formatting assistance w ith this
article. Funding for her services was provided by Millcreek Outcomes Group.
REFERENCES
1. Food and Drug Administration Modernization Act of 1997. Pub L 105-115,
111 Stat 2296. November 21, 1997. Available at: https://w ww.gpo.gov/fdsys/pkg/
PLAW-105publ115/pdf/PL AW-105publ115.pdf. Accessed November 19, 2018.
2. AMCP Partnership Forum: F DAMA Section 114—improv ing the
exchange of health care economic data. J Manag Care Spec Pharm.
2016;22(7):826-31. Available at: https://www.jmcp.org/doi/10.18553/
jmcp.2016.22.7.826.
3. U.S. Food and Drug Administration. Fast track, breakthrough therapy,
accelerated approval, priority review. November 28, 2017. Available at:
https://www.fda.gov/forpatients/approvals/fast /ucm20041766.htm. Accessed
November 19, 2018.
4. AMCP Partnership Forum: Enabling t he exchange of clinical and economic
information pre-FDA approval. J Manag Care Spec Pharm. 2017;23(1):105-12.
Available at: https://www.jmcp.org/doi/10.18553/jmcp.2016.16366.
5. 21st Century Cures Act. Pub L 114-255,130 Stat 1033. December 13,
2016. Available at: https://www.congress.gov/114/plaws/publ255/PLAW-
114p ubl255.p df. Accessed November 19, 2018.
6. U.S. Food and Drug Administration. Drug and dev ice manufacturer com-
munications wit h payors, formulary committees, and sim ilar entities —ques-
tions and answers; draft g uidance for industr y and review staff. Notice of
availability. Docket No. FDA-2016-D-1307. January 19, 2017. Available at:
https://w w w.federalregi st er.gov/do cu me nt s/2017/01/19/2017-01011 /dru g-
and-dev ice-manufacturer-communications-with-payors-for mular y-comm it-
tees-and-similar. Accessed December 14, 2018.
7. Pharmaceutical Information Exchange Act. HR 2016, 115th Cong, 1st
Sess. (2017). Available at: https://ww w.congress.gov/115/bills/hr2026/BILLS -
115hr2026ih.pdf. Accessed November 19, 2018.
8. Public Citizen, Annie Appleseed Project, Association for Medical Ethics,
et al. Letter to members of the U.S. House of Representatives, Energy and
Commerce Committee. June 5, 2017. Available at: https://www.citizen.org/
sites/default/files/2371.pdf. Accessed November 19, 2018.
9. Eubank BH, Mohtadi NG, Lafave MR, et al. Using the modified Delphi
method to establish clinical consensus for the diagnosis and treatment of
patients with rotator cuff pathology. BMC Med Res Methodol. 2016;16:56.
10. Gustafson DH, Shukla R K, Delbecq A, Walster GW. A comparative
study of differences in subjective likelihood estimates made by indiv iduals,
interacting groups, Delphi groups, and nominal groups. Organ Behav Hum
Perform. 1973;9(2):280 -91.
11. Lynn MR. Determination and quantification of content validity. Nurs Res.
1986;35(6):382-85.
12. Boulkedid R, Abdoul H, Loustau M, Sibony O, Alber ti C. Using and
reporting the Delphi method for selecting healthcare quality indicators: a
systematic review. PLoS One. 2011;6(6):e20 476.
13. Thangaratinam S, Redman CW E. The Delphi technique. Obstet Gynecol.
2005;7:120-25.
14. Linstone HA. The Delphi technique. In Fowles J, ed. Handbook of Futures
Research. Westport, CT: Greenwood; 1978:271-300.
Incorporation of these PHDM perspectives on PIE into
ongoing legislative discussions and planning for future PIE by
PHDMs and manufacturers is warranted. U.S. PHDMs unani-
mously agree that a legislation solution for PIE is required,
since the full contribution of PIE to improving the U.S. health
system can best be realized under a safe harbor that allows U.S.
PHDMs and manufacturer experts to engage in robust scientific
and economic discourse. Legislative action must allow for the
proactive exchange of HECI before FDA filing and between
manufacturers and PHDMs of payer organizations, formulary
or technology review committees, and other such entities with
expertise and responsibility for formulary coverage decisions
or population health care management. Enacted legislation
must also require that information exchange be truthful, not
misleading, and based on competent and reliable scientific
evidence related to the investigational product or investiga-
tional use of a drug or device. The appropriate allowance of
PIE between professional peers based on reliable evidence
is intended to address unmet informational needs, support
efficient timing of PHDM review, and decrease uncertainty,
aligning towards a common goal—improving our health care
delivery system in the United States.
DIANA BRIXNER, PhD, RPh; JOSEPH BISKUPIAK, PhD, MBA;
and GARY ODERDA, PharmD, MPH, University of Utah College of
Pharmacy, Salt Lake City, and Millcreek Outcomes Group, Salt Lake
City, Utah. TATIA C. WOODWARD, MPH, MS; NIK SEIFTER,
PharmD; and MARTIN MARCINIAK, PhD, Gla xoSmithKline,
Research Triangle Park, North Carolina. PETER NEUM ANN, ScD,
Tufts Medical Center, Boston Massachusetts.
AUTHOR CORR ESPONDENCE: Gary Oderda, PharmD, MPH,
University of Utah College of Pharmacy, 30 South 2000 East,
Salt Lake City, UT 84112. Tel.: 801.581.6257;
E-mail: gary.oderda@pharm.utah.edu.
Authors
www.jmcp.org Vol. 25, No. 2 Febr uary 2019 JMCP Journal of Managed Care & Specialt y Pharmacy 173
Preapproval Information Exchange: Perspectives of U.S. Population Health
Decision Makers on Preferences for Early Engagement with Investigational Therapies
15. Hassa n F, Keeney S, McKenn a H. Research guidelines for the Delphi sur-
vey technique. J Adv Nurs. 2000;32(4):1008-15.
16. Powell C. Myths and realities of the Delphi technique. J Adv Nurs.
2003;41(4):376-82.
17. Neumann PJ, Saret CJ. A sur vey of individuals in U.S.-based pharma-
ceutical industr y HEOR departments: attitudes on policy topics. Expert Rev
Pharmacoecon Outcomes Res. 2013;13(5):657-61.
18. Neumann PJ, Lin PJ, Hughes TE. US FDA Modernization Act, section
114: uses, opportunities and implications for comparative effectiveness
research. Pharmacoeconomics. 2011;29(8):687-92.
19. Brixner DI, Holtorf AP, Neumann PJ, Malone DC, Watkins JB.
Standardizing quality assessment of observational studies for decision mak-
ing in health care. J Manag Care Pharm. 2009;15(3):275-83. Available at:
https://www.jmcp.org/doi/10.18553/jmcp.2009.15.3.275.
20. Yousuf M. Using experts’ opinions through Delphi technique. Practical
Assessment, Research & Evaluation. 2007;12(4):1-8. Available at: https://pare-
online.net/getvn.asp?v=12&n=4. Accessed November 19, 2018.
21. U.S. Food and Drug Administration. Drug and device m anufac-
turer communications with payors, formulary committees, and simi-
lar entities—questions and answers: g uidance for industr y and rev iew
staff. June 2018. Available at: https://www.fda.gov/down loads/Dr ugs/
GuidanceComplianceRegulatoryInformation/Guidances/UCM537347.pdf.
Accessed November 19, 2018.
