OMB approval memo

FDA DOCUMENTATION FOR GENERIC CLEARANCE

“TESTING COMMUNICATIONS ON DRUG PRODUCTS”

(0910-0695)

TITLE OF INFORMATION COLLECTION: Exploring Healthcare Providers’ Practices, Perspectives, and Experiences Prescribing and Co-Prescribing Benzodiazepines and Opioids

DESCRIPTION OF THIS SPECIFIC COLLECTION

1. Statement of Need

The Food and Drug Administration’s Center for Drug Evaluation and Research (CDER) has an ongoing responsibility to communicate about the medical products it approves or authorizes for use in medical emergencies (Sections 564, 564A, and 564B of the Federal Food, Drug, and Cosmetic Act [FD&C Act].

The purpose of FDA’s qualitative social science research study through a contract with Lake Research Partners (LRP) is to enhance understanding regarding healthcare providers’ (HCPs’) prescribing of benzodiazepines and their co-prescription with opioids. The understanding to be gained from this study is necessary to inform FDA’s communication decisions, especially given that benzodiazepine prescribing has continued to be problematic despite FDA warnings about the serious risks associated with these medicines, including abuse, addiction, physical dependence, and withdrawal reactions alone and especially with concomitant use of prescription opioids,¹ and advising healthcare professionals to prescribe both these medications only in patients for whom alternative treatment options are inadequate and then only at the lowest dosages and duration possible.² However, studies have shown that co-prescribing of these two drug classes is still prevalent, putting patients using both at higher risk of visiting an emergency room, being admitted to a hospital, or dying from a drug-related emergency.

As a result, FDA will conduct this study employing online focus groups followed by online individual in-depth interviews (IDIs) with healthcare providers (HCPs) across the United States to better understand their motivations, decision-making, perspectives, experiences, and practices related to prescribing benzodiazepines alone and in conjunction with opioids.

2. Intended Use of Information

This qualitative study is exploratory and will offer the opportunity for HCPs to provide diverse and in-depth input and reactions in their own words about a variety of topics of FDA interest, including how HCPs think and talk about risks and benefits of prescribing and co-prescribing the widely used benzodiazepine and opioid medications, misperceptions they may have, and barriers and challenges they face. The study will also help FDA identify trends and consistencies in participants’ knowledge and decision-making processes and help ensure the Agency has elicited an appropriate range of opinions, beliefs and perceptions about benzodiazepine and opioid prescribing. The information and insights gained from this study will enhance FDA’s understanding, allow it to determine whether and/or what additional areas of research may be needed, and help inform communication decisions in the shorter term. Thus, this work is critical to strengthening FDA’s ability to fulfill its public health mission³ by raising awareness of and better educating HCPs about benzodiazepine and opioid prescribing and the risks associated with co-prescribing these medications.

3. Description of Respondents:

Respondents for both the focus groups and the subsequent IDIs will be HCPs (physicians, nurse practitioners [NPs] and physician assistants [PAs]) who have prescribed benzodiazepines either alone or in conjunction with opioids in the past 90 days and to an average of at least five (5) patients a month. Participants will be recruited from the following medical fields, which were selected based on data showing they had the highest levels of benzodiazepine and/or opioid prescribing levels:

• Primary Care (Family Practice, Internal or General Medicine, Osteopathic Medicine, General Practice, and Geriatric Medicine)

• Specialists in Mental Health (Psychiatry, Geriatric Psychiatry, and Psychology)

• Specialists in Neurology

• Specialists in Emergency Medicine

• Specialists in Pain Medicine

• Specialists in Addiction Medicine

In addition to the medical fields of practice noted above, participants will be recruited to reflect a mix of professional characteristics, including number of years in practice, extent of opioid and benzodiazepine prescribing, and practice location (U.S. Census 9-way region and urban/suburban/rural). We will also recruit participants based on personal demographic characteristics, including gender (goal will be to have at least half of participants be female), race, ethnicity (goal will be to have at least a third of participants who are HCPs of color), and age. HCPs who participate in the focus groups will not be allowed to participate in the IDIs.

Focus Group Participants

In total, we expect 144 participants to participate across the 16 focus groups. We will recruit 12 HCPs to achieve the desired goal of 9 participants per group. The groups will be segmented by type of provider and by specialty, with 12 groups conducted with primary care physicians (PCPs), and two each with PAs and NPs. (See Focus Group Recruitment Screener in Appendix A.)

The 12 physician focus groups will be segmented as follows:

• Five (5) focus groups with PCPs who have NOT prescribed buprenorphine or other opioid medications used to treat opioid use disorder (MOUD).

• One (1) focus group with PCPs who have prescribed MOUD.

• Six (6) focus groups with physician specialists:

  • 1 focus group of Mental Health specialists who have NOT prescribed MOUD.

  • 1 focus group of Psychiatrists who have prescribed MOUD.

  • 1 focus group of Neurology specialists who have NOT prescribed MOUD.

  • 1 focus group of Emergency Medicine specialists who have NOT prescribed MOUD.

  • 1 focus group of Pain Medicine specialists, some of whom may have prescribed MOUD in the past 3 months.

  • 1 focus group of Addiction Medicine specialists who have prescribed MOUD in the past 3 months.

• Two (2) focus groups with PAs and NPs:

  • 1 focus group with PCPs who have NOT prescribed MOUD.

  • 1 focus group with a mix of specialists across Mental Health, Neurology, Emergency Medicine, Pain, and Addiction Medicine specialties who have NOT prescribed MOUD.

In-Depth Interview Participants

Thirty (30) IDIs among HCPs who meet the same recruitment characteristics as the focus groups will be conducted following the groups. The IDIs, which will be informed by the findings gained from the groups, will consist of 24 interviews with PCPs and three (3) each with PAs and NPs. Two (2) additional each of PCPs, PAs and NPs will be recruited to ensure a total of 30 interviews (in anticipation of potential no-shows and schedule changes). (See Interview Recruitment Screener in Appendix B.)

The 24 physician IDIs will consist of:

• 12 with PCPs, three (3) of which will be with those who have prescribed MOUD.

• 12 interviews with specialists:

  • Four (4) with Mental Health specialists, two (2) of which will be with psychiatrists who have prescribed MOUD.

  • Two (2) each with specialists in Neurology, Emergency Medicine, Pain Medicine and Addiction Medicine.

  • Three (3) each with PAs and NPs, including two (2) among those who practice in primary care and one (1) with a Mental Health specialist.

4. Date(s) for Research to be Conducted:

Scheduling of the 16 focus groups will begin within two (2) days after approvals from OMB and FDA’s Office of the Chief Scientist Human Subjects Protection Officer (OCSHSPEO) are obtained and are expected to be completed within eight (8) weeks (4 weeks beyond recruitment). The IDIs will be conducted following the data analysis and reporting of the focus groups.

5. How the Information is Being Collected

Sixteen (16), 90-minute virtual focus groups will be conducted followed by thirty (30), 60-minute virtual IDIs. For each focus group we will recruit 12 HCPs for the desired turnout of nine (9) participants. In total, we expect 144 participants to participate across the 16 focus groups. These groups will be segmented by type of provider and by specialty as noted previously. The 30 IDIs among HCPs nationwide who meet the same characteristics as those in the focus groups (but who did not participate in the groups) will be conducted following the data analysis and reporting of the focus groups. The IDIs will include 24 interviews with PCPs, and three (3) each with PAs and NPs. To fulfill the goal of 30 interviews, we will over-recruit by two physicians, two PAs, and two NPs (to account for schedule changes and no-shows).

The topics explored in the focus groups and IDIs will be similar; however, there will be fewer questions posed to respondents in the IDIs, which will focus on the specific topics about which we want to gather additional information based on the findings and insights generated in the focus groups. Focus group dynamics animate discussion and stimulate creativity and broader responses among participants, while IDIs allow each respondent to discuss topics one-on-one with an interviewer and in much greater detail, typically resulting in increased clarity and greater depth of understanding. Thus, these two methods build on each other when used sequentially, providing a more rigorous methodology and more robust findings to inform FDA understanding and future communications.

The Schlesinger/2020 Research will recruit participants for the focus groups and IDIs. They will first send an initial invitation and then a confirmation email to qualified participants. This correspondence is identical for both the focus groups and IDIs (See Recruitment Outreach for Focus Groups and Interviews in Appendix C). Schlesinger/2020Research has recruited for many health care professional groups and interviews in the past, and their US panel includes more than 39,500 physicians, PAs, and NPs. All participants will be required to sign an electronically administered informed consent form prior to participation in the study. With their consent, we will audio record each session, produce a written transcript of the discussion, and use the transcript to supplement the team’s notes. No participants will be allowed to participate in the study without a signed consent form. Both the focus groups and interviews will begin with an introduction, including a review of some of the key information from the consent form, and ground rules.

Some areas to be explored through the semi-structured moderator and interview guides

include:

• General prescribing practices

• Perceptions about benefits

• Decision-making related to benzodiazepine and opioid prescribing

• Prescribing experiences and practices, beliefs, and concerns

• Consideration of alternatives, tapering, or discontinuation

• Knowledge and use of FDA warnings and other clinical guidelines

• Provider-patient communication

• Resource and training needs

(See Focus Group Moderator Guide in Appendix D and the Individual Interview Guide in Appendix E.)

For each 90-minute focus group, one of the two professional, trained moderators with extensive experience who are working on this project will lead the discussion from her computer and a separate note-taker and logistics coordinator will assist. Once connected, the participants and moderator will be able to see one another on screen. The moderator will lead the discussion using a semi-structured moderator guide that ensures consistency in major topics but allows flexibility in probing each group depending on the discussion. The note-taker, who will not be visible to participants, will observe and document the major themes that surface in each session. FDA staff will be able to observe unobtrusively and will not be visible to or able to interact directly with the participants.

Both the focus groups and the IDIs will be conducted online using Schlesinger/2020Research as the virtual platform. Schlesinger/2020Research is user-friendly for participants and was selected because of the following capabilities:

• It will allow the research team to gather insights in real time using webcam technology and video streaming.

• It provides technical assistance to participants, researchers and observers before the live session begins to ensure participants are prepared. Schlesinger/2020Research will conduct “tech rehearsals” with participants 15-30 minutes prior to each focus group and will remain online and available throughout the duration of the group to address any technical issues by participants, researchers, or observers.

• It has a virtual ‘backroom’ to which members of the FDA and LRP research team can observe in real time and provides a “chat” function on which they can interact in writing with the moderator, e.g., to ask clarifying questions.

• It produces both high-resolution video and audio recordings available immediately after each group and interview ends. Verbatim, time-stamped transcripts of each focus group and interview will be developed and linked to each video recording so it is easy to move back-and-forth between the transcripts and recordings. An electronic copy of the transcripts shall be submitted to the FDA Project Manager within two (2) weeks following completion of the 16 focus groups and two (2) weeks following completion of the 30 interviews.

• It has security measures in place to prevent data breaches, including project-specific links, waiting rooms monitored by an assigned technician to ensure the correct actors (moderator, observers, and participants) have access to the study, and limited screen sharing capabilities.

• It will also act as a video portal to the research team, enabling LRP analysts to view, download files and transcripts, and enable search capabilities to find clips within each video recording.

6. Confidentiality of Respondents:

Each participant will receive a link to an online consent form at the time of recruitment and scheduling for the focus groups (see Focus Group Consent Form in Appendix F) and interviews (see Individual Interview Consent Form in Appendix G). Participants will electronically sign the programmed version of the consent form so that a date/time stamp of consent is collected by Schlesinger/2020Research.

Schlesinger/2020Research will store screening and consent data on a password-protected computer for no more than 90 days to invite respondents and send them reminder emails and phone calls the day before their scheduled focus group. Only Schlesinger/2020Research staff assigned specifically to this project will have access to this information. Neither LRP nor FDA will have access to the consent forms or the full names or any other personally identifiable information (PII) about the participants.

To ensure the required participant diversity, Schlesinger/2020Research will provide LRP with the screening data for the participants via daily updates, which will include first names and last initials only, and answers to the study’s recruitment screening questions. The daily updates will not include contact information or other PII. Only LRP and FDA staff assigned specifically to this project will have access to the daily recruitment grids. PII and contact information will not be provided to the FDA at any point throughout the study. This information will be removed from all documents shared with FDA, including, but not limited to, recruitment information, audio and video recordings, transcripts, and reports and raw data.

At the beginning of each focus group and IDI, the moderators will reiterate the information contained in the informed consent that participants previously signed, and participants will be instructed to use only their first names. Participants will also be told that no full names or any PII will be used in any notes, reports, or subsequent materials created as part of this research study, that only anonymized information reported in aggregate will be provided to the FDA, and that participants’ information will be kept secure to the extent permitted by law. Participants will be asked not to share anything discussed during the group with anyone else.

Schlesinger/2020Research also has security measures in place to prevent data breaches, including project-specific links that are provided only to approved participants, waiting rooms monitored by an assigned technician to ensure the correct actors (moderator, observers, and participants) have access to the study, and limited screen sharing and chatting capabilities during the focus group or IDI.

Audio and video recordings and all other study materials and files will be stored on password-protected servers at LRP and FDA accessible only to the research team assigned to this study. LRP and FDA will retain these files for 10 years and then delete them. LRP will check all transcripts, audio/video files, reports, and other materials for PII and remove it before providing the files and documents to FDA.

7. Amount and Justification for Any Proposed Incentive

The respondents for this national study are physicians, PAs and NPs in a variety of specialties who are difficult to recruit, in part because they have demanding schedules working long and/or irregular schedules, including responding to emergencies, so they may not be able to attend research focus groups, particularly at a scheduled time or may be no-shows even when they do agree to participate. Increasing this difficulty is that this study also very specific recruitment requirements based on such factors as area of specialization, prescribing certain drugs or a combination of drugs in a typical month to at least five patients, further limiting the number of providers who will qualify. In addition, others will be ineligible because they have worked for a pharmaceutical company or have ever consulted with one, or any U.S. Department of Health and Human Services agency or entity.

Little academic research is available that provides guidance on appropriate incentive levels for physicians and/or other HCPs for national online qualitative research involving lengthy and detailed discussions. However, we found one study that is sufficiently rigorous, relevant, and recent to provide guidance for the current project.⁴ The study was conducted online among primary care physicians, cardiologists, neurologists, oncologists, and orthopedists. Of the 625 physicians who clicked on the link via the survey invitation, 425 completed the survey (a 68% completion rate). The survey tested willingness to participate in five different types of research, ranging from a 15-minute quantitative survey to a 60-minute qualitative interview. Fair market value (FMV) was defined using the IRS Revenue Ruling 59-60: “Fair market value is the price, expressed in cash equivalents at which a property would change hands between a hypothetically willing and able buyer and a hypothetically willing and able seller, acting at arm’s length in an open and unrestricted market, where neither is under compulsion to buy or sell and where both have reasonable knowledge of the relevant facts.”⁵ The baseline FMV incentive amount varied by specialty and respondent time commitment based on experience recruiting physicians and other healthcare providers. Exact incentive amounts (Table 1) were determined as a range (±50%, ±25%) around our clients’ average FMV rate for each specialty by study combination; the middle price point was always the average FMV rate. To elicit participation likelihood, a Gabor-Granger exercise was utilized. Developed by Andre Gabor and Clive Granger in the 1960s, Gabor-Granger is a sequential monadic method typically used for estimating price elasticity of a product or service and find the revenue-optimizing price points. In this application, it was used to estimate incentive elasticity of research participation and seeks to find the most cost-effective incentive payment.

Table 1.	Intended Participation in a 60-Minute Qualitative Interview by Amount of Incentive Not Adjusted for Inflation
	Oncologists	Other Specialists (Cardiologists, Orthopedists, Neurologists)	PCPs	Avg. Intended Participation Rate
Lowest	$155	$115	$110	14% (8%,21%)*
Moderate - FMV	$310	$230	$205	66% (58%,71%)*
Highest	$465	$345	$300	95% (93%,98%)*

*The percentages in parentheses represent the lowest and highest specialty.

The Gabor-Granger exercise worked like this: For each study type, respondents were shown a description of the study they’d be participating in, one of the five incentive amounts (randomized), and asked if they would participate. If they responded “yes”, they were shown a lower amount, if they said “no”, they were shown a higher amount. This repeated until the respondent reached either the end of the incentive range or until their reservation amount was reached. The order in which respondents were shown the five market research studies was randomized to avoid bias.

The authors provided an estimate of appropriate per-minute incentives for specialists and PCPs sufficient to assure an average 75% response rate for an individual qualitative interview. However, this must be qualified by the nature of the research since response rates vary by specialty, topic, and many other factors. We have extrapolated these estimates to the 60-minute qualitative interviews and the 90-minute focus groups proposed for the current study based on interview length alone (see Table 2). The study also supported the hypothesis that per-minute incentive required to achieve 75% participation in qualitative research would be significantly higher than that required for survey research. For example, the recommended per-minute incentive amount for PCPs is $3.33 for a 45-minute qualitative interview and $2.43 for a 40-minute quantitative survey. This supports the authors’ hypothesis that respondents require a larger incentive to participate in qualitative research than quantitative surveys even when both take the same amount of time and reinforces that idea that studies that focus on incentives for quantitative research, especially relatively short surveys, should not be compared to those that focus on qualitative research.

Table 2	Estimates of Appropriate Incentives for Specialties Not Adjusted for Inflation
	Cardiologist	Neurologist	Oncologist	Orthopedist	Avg. Specialists	PCPs
Per Minute	$4.25	$4.42	$5.73	$4.63	$4.76	$3.63
60 Minutes	$255.00	$265.20	$343.80	$277.80	$285.60	$217.80
90 Minutes	$382.50	$397.80	$515.70	$416.70	$428.40	$326.70

Thus, the estimates drawn from this research exceed our proposed incentive, which is $275 for physicians and $250 for NPs and PAs for their participation in the 90-minute focus groups and $225 for physicians and $200 for NPs and PAs for their participation in the 60-minute individual depth interviews.

In addition, these incentives are lower than the $300 incentive that OMB approved for the previous FDA research project titled “Testing Communications on Biological Products” approved in 2014 under generic clearance 0910-0687. In addition, the FDA was allowed to increase this incentive an additional $100 if specialists declined to participate based on their stated reason that the $300 incentive is too low. A similar tiered strategy was approved by OMB in 2017 for the Centers for Disease Control and Prevention’s “Formative Research to Develop HIV Social Marketing Campaigns for Healthcare Providers” (OMB Control No. 0920-1182). Similarly, specialists received $250 incentives for participating in a 60-minute focus group as part of FDA’s “Generic Drug Substitution in Special Populations” study (OMB Control No. 0910-0677; 2017) and in 60-minute telephone interviews for “Studies to Enhance FDA Communications Addressing Opioids and Other Potentially Addictive Pain Medications” (OMB Control No. 0910-0695; 2016), an incentive rate that would have amounted to $375 for 90 minutes.

The proposed incentives are based on the principles set forth in OMB’s guidance on factors that may justify provision of an incentive. These factors include the need to: (1) assure data quality by using incentives to increase ability to recruit the diversity of desired subgroups and reduce sample bias; (2) compensate for respondent burden; (3) facilitate respondent recruitment when the target group is difficult to recruit; and (4) improve coverage of specialized respondents or rare groups. All these factors are involved in the current FDA study among HCPs. Thus, the proposed incentives are necessary to increase the likelihood they will agree to participate, actually attend the real time research groups and interviews, and provide their focused attention and active participation.

8. Questions of a Sensitive Nature

Neither the focus group guide nor the IDI moderator’s guide contain any questions of a sensitive nature. The respondents are all HCPs who have experience working with patients and prescribing the medications they will be asked to discuss. The types of questions will probe their regular prescribing practices and are similar to those that might be discussed in an in-house meeting of colleagues or a medical conference or seminar. Nevertheless, the focus group and IDI consent forms state that participants are not required to answer any question they choose not to, and this is reiterated as part of the introduction at the beginning of the groups and IDIs.

9. Description of Statistical Methods

Traditional qualitative analysis, that is thematic coding analysis conducted by humans, will be employed to analyze the findings from this research in addition to analysis facilitated by computer software designed specifically to support qualitative research studies. The two moderators will also be the chief qualitative analysts; thus, they will begin the analytical task with some grounding in common themes and questions that emerge during the focus group session or IDI they conducted. These themes and questions will be addressed in the subsequent analysis after being checked against other supporting sources, such as the video recordings, transcripts, and observer notes.

The analysis will also be facilitated by computer software called NVivo that facilitates the categorization, organization, storage in a variety of formats, and convenient retrieval of qualitative data for analysis. NVivo is a platform where various stages of a project can be housed, and data files can then be efficiently coded by theme, topic, or other case differentiator. Information from transcripts, audio and video recordings, and moderator and observer notes will be coded and categorized for analysis by the moderator/analysts. The software will enable and expedite this process in the following way:

1. Import interview transcripts, recordings (audio and video), moderator notes, and observer notes to a single stand-alone system.

2. Allow the analysts to use coding tools available within the platform to facilitate the organization of clips, quotes, notes, and other research materials gathered into conceptually similar themes.

3. Provide query and charting tools for the analysts to explore the connections between themes.

4. Use an algorithm to independently verify coding categories for no fewer than 20% of the transcripts for the focus groups (not less than 4) and interviews (not less than 6) by each of two coders.

5. Calculate Inter-Rater Reliability (IRR) for the two coders using Cohen’s Kappa Coefficient) for no fewer than 20% of the transcripts for the focus groups (not less than 4) and interviews (not less than 6).

   • The Kappa coefficient will be calculated individually for each combination of node and source. If the two users are in complete agreement, Kappa will be 1. If they are in complete disagreement, Kappa will be ≤ 0. A Kappa value of 0.8 or higher is considered high agreement and will be used for these analyses.

Burden Hour Computation

(Number of responses (X) estimated response or participation time in minutes (/60) = annual burden hours):

Types of Respondents	Number of Respondents	Minutes for Each Respondent	Total Minutes	Burden (Hours)
Focus Group Screening	2,880	10	28,800	480
Focus Groups	144	90	12,960	216
Interview Screening	450	10	4,500	75
Interviews	30	60	1800	30
TOTAL	3,504			801

REQUESTED APPROVAL DATE: September 30, 2021

NAME OF PRA ANALYST & PROGRAM CONTACT:

Ila S. Mizrachi

Paperwork Reduction Act Staff

[email protected]

301-796-7726

Paula Rausch, PhD, RN

CDER Office of Communications

301-325-4662

[email protected]

FDA CENTER: Center for Drug Evaluation and Research

1US Food and Drug Administration. Drug Safety Communication: https://www.fda.gov/drugs/drug-safety-and-availability/fda-requiring-boxed-warning-updated-improve-safe-use-benzodiazepine-drug-class

2US Food and Drug Administration. Drug Safety Communication: FDA warns about serious risks and death when combining opioid pain or cough medicines with benzodiazepines; requires its strongest warning. https://www.fda.gov/drugs/drug-safety-and-availability/fda-drug-safety-communication-fda-warns-about-serious-risks-and-death-when-combining-opioid-pain-or

3FDA Mission: https://www.fda.gov/about-fda/what-we-do#mission

4Clement, Lynn and Claeys, Chris. What’s fair? The fair market value dilemma in health care research. Quirk’s Market Research Review. April-May, 2019.

5This is the definition of FMV set forth in the International Glossary of Business Valuation Terms

9