Introduction
Over two decades ago, the National Academy of Medicines (formerly the Institute of Medicine) set a goal that by the year 2020, “90 percent of clinical decisions will be supported by accurate, timely, and up-to-date clinical information [Reference Olsen, Aisner and McGinnis1].” Traditionally, researchers have considered randomized clinical trials as the gold standard for determining the safety and efficacy of medications or other interventions. However, randomized trials are frequently criticized for their lack of generalizability to patient care in routine clinical practice settings [Reference Kennedy-Martin, Curtis, Faries, Robinson and Johnston2,Reference Rothwell3].
To enable evidence generation to inform patient care, researchers and policymakers have begun to appreciate the potential of studies in which elements of clinical trials, such as patient eligibility and identification, randomization, and data acquisition, are embedded into clinical practice settings when patients are seeking routine care from their healthacare providers [Reference Califf, Cavazzoni and Woodcock4,Reference Wieseler, Neyt, Kaiser, Hulstaert and Windeler5]. These studies align with clinical workflows and leverage clinical care data sources for research purposes. They can increase clinical trial access to representative populations and have the potential to increase trial efficiencies by reducing duplication of trial and care activities, such as data collection.
The concept of embedding clinical trials into clinical practice is not new [Reference Ramsberg and Platt6,Reference Garcia, Haynes and Pokorney7]. The National Institutes of Health’s Health Care Systems Research Collaboratory initiative was created in 2012 with a mission to “strengthen the national capacity to implement cost-effective large-scale research studies that engage healthcare delivery organizations as research partners [Reference Weinfurt, Hernandez and Coronado8,9].” However, implementation remains an issue. In addition, the terminology around embedding trials into clinical practice has not reached a state of consensus [Reference Nicholls, Carroll and Hey10,Reference Pawson11]. The literature often associates embedding trials with being pragmatic as viewed through the Pragmatic-Explanatory Continuum Indicator Summary (PRECIS)-2 scale, yet the threshold of what is considered pragmatic is not consistent [Reference Nicholls, Carroll and Hey10–Reference Loudon, Treweek, Sullivan, Donnan, Thorpe and Zwarenstein13].
The Clinical Trials Transformation Initiative (CTTI), a public–private partnership, conducted a multi-partner project to appreciate: (1) the rationale for integrating trials into clinical practice, (2) the optimal methodological and operational approaches for embedding trials, and (3) the infrastructure needed to facilitate system-wide integration of trials at the point of care.
The intention of this project was not to define how pragmatic a trial is or whether it is considered a point-of-care trial. Rather, recognizing that cultural, administrative, financial, and data barriers exist and that operational direction is needed to assist with embedding trials, we aimed to develop evidence-based recommendations on how to design and conduct embedded trials, especially those intended for regulatory review of a medical product. Here, we present the qualitative research findings from CTTI’s evidence-gathering phase and offer recommendations informed by the findings to facilitate the integration of clinical trials into clinical practice [14,15].
Materials and methods
CTTI projects follow an evidence-based methodology that includes stating an efficiency and quality impediment to clinical trials, convening a multi-partner project team, gathering evidence to understand barriers, and translating the findings into actionable recommendations and tools [Reference Corneli, Hallinan and Hamre16]. The CTTI Trials in Clinical Practice Project Team consisted of partners representing academia, industry, government agencies, institutional review boards, professional societies, patient representatives, and patient advocacy organizations [14].
Study design and participants
As part of our evidence gathering, we conducted a qualitative descriptive study [Reference Sandelowski17,Reference Sandelowski18] using in-depth interviews (IDIs). Study participants were trial designers (those responsible for designing and making decisions about the trial) and implementers (those carrying out day-to-day operations for the trial) of embedded interventional clinical trials with at least one site in the USA. We did not seek to interview representatives from all possible embedded US-based trials but rather we purposively selected [Reference Patton and Ritzer19] designers and implementers who were engaged in US-based trials to ensure that they could comment specifically on challenges and solutions to embedding trials in the context of the US regulatory environment. We also selected designers and implementers who were engaged in registrational trials, or in non-registrational trials intended to be submitted for regulatory review, and whose trials we considered to be embedded into clinical practice because they were integrated into healthcare delivery, closely aligned with clinical workflows, and leveraged existing infrastructure and clinical care data for research purposes, such as using electronic health records (EHRs) to collect research data. Additionally, designers could participate if they took part in the trial decision-making and design process; implementers could participate if they were engaged in the day-to-day operations of an embedded interventional trial.
We drew upon CTTI’s multi-partner project team and other expert contacts to identify potentially eligible trials and then representatives of those trials (i.e., designers and implementers). We also conducted informal searches on ClinicalTrials.gov [20] for interventional studies from January 2011 to April 2021 using the search terms “embed,” “integrate,” “pragmatic,” “practical,” “large simple trial,” “real world,” “learning health care,” and “point of care,” and filtering for interventional trials with a location within the USA. We identified approximately 20 trials that met our selection criteria, and introductory emails were sent to contacts either listed in ClinicalTrials.gov, identified by CTTI’s multi-partner project team, or recognized as authors of publications about the study. Upon further screening, six of those individuals did not meet the criteria, as their studies were not leveraging existing data infrastructure, such as the EHR, or were not US based, which was outside of the scope of this research. Additionally, three individuals did not respond; one was unable to meet our timeframe to conduct an interview, and one was not interested in participating. The final interview sample provided sufficient information power [Reference Malterud, Siersma and Guassora21], which occurs when a qualitative dataset provides rich, descriptive evidence that is useful for understanding the concept under investigation – and in our case, information that is helpful for developing recommendations that are grounded in the experiences of developers and implementers.
Data collection
We first identified conceptual categories to investigate in the interviews based on the study objectives and CTTI project team members’ knowledge of the type of experiential information needed to design and operationalize clinical trials in healthcare settings. Next, we developed interview questions for each category and tailored them based on the participants’ role and trial type: (1) trial designers, registrational; (2) trial designers, non-registrational; (3) implementers, registrational; and (4) implementers, non-registrational. Interview questions for trial designers focused on the rationale for conducting an embedded trial versus a conventional trial; how healthcare settings were chosen; details about how elements of clinical trials were integrated into the healthcare settings, including any modifications that were necessary and how data were captured and harmonized; perceived benefits, barriers, and risks to using an embedded trials approach; and lessons learned. Trial designers of registrational trials were also asked to describe any conversations they had with regulators during the trial design process. All implementers were asked the same questions. Their interview questions primarily focused on identifying the details of integrating clinical trial elements into healthcare settings, including hiring, recruitment, consent, randomization, scheduling, and data capture and entry; implementer interviews also covered how trial processes were woven into standard of care and any modifications that were necessary for either the trial team or the healthcare setting, as well as benefits, barriers, and drawbacks of the embedded trials approach and lessons learned.
Two trained qualitative interviewers conducted telephone interviews from April 23 to December 17, 2021. Either individual or group interviews (i.e., two people from the same trial) were conducted, depending on participant preference. Demographic information was collected from each participant.
Data analysis
Interviews were audio-recorded with participants’ permission, and verbatim transcripts were created using a transcription protocol [Reference McLellan, MacQueen and Neidig22]. Participant demographic characteristics were summarized using descriptive statistics, and applied thematic analysis [Reference Guest, MacQueen and Namey23] was used to analyze participant narratives. NVivo version 12 (QSR International) [24] qualitative data analysis software was used to organize the data and apply codes [Reference Saldana, O’Connor and Joffe25] to the transcripts. Two trained analysts first independently applied structural (a priori) codes to segment participant narratives into conceptual categories related to the study objectives (e.g., motivations for conducting an embedded interventional trial). Next, the analysts identified and applied content-driven (emergent) codes to participant narratives in each conceptual category, reflecting specific details of designer and implementer experiences with the design and conduct of embedded trials. Inter-coder reliability [26] was assessed on approximately 15% of transcripts during each phase of analysis, and where necessary, discrepancies in code application were resolved through discussion, and agreed-upon revisions to the codebook and coding were made.
Following the completion of coding, analysts reviewed the content coding frequencies to identify common perceptions and experiences. Perceptions and experiences varied greatly. The analysts therefore primarily focused on identifying perceptions and experiences that were shared by three or more trials or participants. However, acknowledging the valuable expertise of all IDI participants, we did occasionally report on experiences or suggestions that were only noted once or twice, where these appeared particularly salient for informing actionable recommendations for successfully embedding interventional trials. Of note, when participants described the same procedures for their specific trial, we combined their narratives and described findings at the trial level. When participants described their own perspectives or opinions, or when they provided information on a topic specific to their role, we described the findings at the individual level. Findings were described in analytical summaries, including illustrative quotes, to convey participant experiences with integrating interventional trials into clinical practice.
Ethics
The Duke University Health System Institutional Review Board (IRB) determined that this research was exempt from further IRB review and waived documentation of informed consent. During the recruitment process, participants were provided with an informational sheet that described the purpose of the interviews and related information (e.g., potential risks). The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on research with human participants.
Results
Study participants
We conducted 16 interviews with 18 participants (14 individual interviews and two group interviews), representing 9 embedded trials (4 registrational trials and 5 non-registrational trials). Of the 16 interviews, 9 were with trial designers (representing 4 registrational trials and 5 non-registrational trials), and 7 interviews were with implementers (representing 3 registrational trials and 4 non-registrational trials). For seven of the trials, we interviewed both a designer and an implementer; for two of the trials (one registrational and one non-registrational), we interviewed a designer only.
Participants represented a diversity of trial designs and disease areas. Study designs were adaptive platform, Phase 2 open-label, Phase 3 placebo-controlled double-blind, virtual decentralized, and label extension. Disease areas were COVID-19, cardiovascular disease, multiple sclerosis, fibromyalgia, and Crohn’s disease. A majority of participants represented and conducted trials in academic settings. Tables 1 and 2 provide descriptive characteristics of the participants interviewed and their organizations.
Table 1. Participant characteristics
a Data are missing from one participant.
Table 2. Characteristics of participant organizations, institutions, and companies
a Two participants were from the same academic institution.
b One participant had joint appointments at both a public and a private institution.
c Two participants were from the same “other” organization. NIH, National Institutes of Health.
Motivations for conducting an embedded interventional trial
Rationale for embedding interventional trials
Trial designers listed three primary reasons for choosing to integrate interventional trials into clinical care, versus conducting conventional clinical trials primarily within clinical research facilities or in healthcare facilities but outside of the clinical care process. First, designers explained that they implemented a learning health system perspective, where they aimed to narrow the gap between clinical care and clinical research to improve knowledge generation and its translation back into clinical care. Second, designers explained that they designed trials to evaluate treatment approaches in real-world clinical practice settings under the premise that conducting pragmatic or naturalistic studies would enable them to determine whether the intervention would prove effective under routine practice conditions. Third, designers said they considered the potential cost savings associated with leveraging existing health networks, informatics infrastructure, and previously curated EHR data. Some designers noted that the high costs of conducting conventional clinical trials can serve as a deterrent to research.
Prospective patient benefits
Participants shared several anticipated benefits to patients that served as motivation to embed trials, including that the results of such trials are likely to be more generalizable than those of conventional trials. Participants also described that these types of trials are powered to detect small differences that matter to patients and may be clinically relevant, and they may facilitate clinician engagement in evidence-based practice.
Prospective healthcare system benefits
Participants postulated that embedding trials holds potential benefits for healthcare systems as well. Participants described that health centers at all levels could increase their visibility by participating in trials, stating that becoming known as a place where cutting-edge clinical research is performed may serve to draw more patients to the healthcare setting. They expressed that this could also improve the retention of existing patients within the healthcare setting.
Prospective sponsor benefits
Participants also pointed out potential benefits to sponsors, suggesting that the possibility of increased efficiency and cost savings across clinical care, research, and discovery could motivate sponsors to become involved with embedding interventional trials.
Perceived and actual benefits after conducting embedded trials
Reflecting on their experiences with integrating interventional clinical trials into clinical care, participants described perceived and actual benefits in two main categories: operational benefits and patient benefits. Operationally, both designers and implementers expressed that the embedded trials methodology enabled a larger scale of operation and larger trials, which in turn allowed for cheaper and more efficient trial conduct. One participant described that their trial had been able to enroll 15,000 patients using only 40 sites. Patient-related benefits of embedded trials included the ability for patients to take advantage of evidence-based care. Participants described that embedded trials can provide a scientific basis for improvement in health care and can serve as a fair test of whether new interventions are effective.
Utilizing existing healthcare infrastructure was also perceived to potentially increase diversity and representation by making it easier for patients from traditionally underrepresented populations to participate in trials. Table 3, Section 1 includes illustrative participant quotes related to motivation for conducting an embedded interventional trial.
Table 3. Select participant quotes
EHR, electronic health record; NIDDK, National Institute of Diabetes and Digestive and Kidney Diseases; NIH, National Institutes of Health.
Barriers to conducting embedded trials
Site staff time, availability, and perceived burden of taking part in the trial
Participants explained that clinicians have limited time and fewer incentives to participate in research if it takes time away from their numerous clinical tasks. Finding time to familiarize clinicians with the study concept and conduct training sessions could also be challenging. Training and study start-up were particularly time-consuming with research-naïve sites, where the research team needed to educate site personnel about all aspects of conducting a trial, and where conducting screening and recruitment activities was not always top-of-mind for clinical practice staff.
Lack of buy-in at the health system level
Participants noted that it would be much more difficult to conduct embedded trials without supportive and engaged healthcare system personnel. While buy-in at all levels was deemed important, support from top-level leadership, particularly IT leadership, was seen as critical; participants explained that a lack of IT collaboration could impede trial conduct.
Difficulties encountered accessing and using EHR data
Participants noted that in addition to challenges with obtaining approval to export and use EHR data outside of the healthcare system, interoperability of systems was sometimes an issue. While many healthcare systems use the same EHR programs, some participants noted that they still needed to develop separate templates or platforms to extract and harmonize EHR data across systems. Table 3, Section 2 includes illustrative participant quotes on the barriers to conducting embedded trials.
Solutions: overcoming barriers to conducting embedded trials
Participants offered several recommendations for overcoming barriers to embedding trials into clinical practice. The most commonly mentioned improvements are discussed below and listed in Figure 1.
Figure 1. Improvements for overcoming barriers to conducting embedded trials.
Create culture change/paradigm shift
Both trial designers and implementers spoke about the importance of changing culture at the institutional and/or designer level in order for embedded interventional trials to gain acceptance as a viable research model. Participants described the need for a change in perspective regarding the relationship between clinical research and clinical care, noting that ideally, research would come to be viewed as a normal part of clinical care, with the clinical team also serving as the research team. Closer alignment between clinical care and clinical research requires changing the way that clinical care and research are conceptualized, integrated, and supported; therefore, participants noted the importance of having engaged leaders who support changes to the traditional research approach. Some participants also called for funding agencies and sponsors to take a greater interest in alternate study designs, such as embedded trials. In particular, participants voiced that it would be helpful for the US Food and Drug Administration (FDA) to be open to the embedded trial model and that the National Institutes of Health, FDA, and other agencies should learn from adjustments made during the COVID-19 pandemic to encourage more adaptation and innovation in the way trials are run.
Participants described concrete ways in which embedded trials represent a paradigm shift from conventional clinical trials and require new ways of thinking about research processes. For example, safety reporting may be different in embedded trials. A participant explained that rather than aiming for drug approval, their non-registrational trial looked at the bigger picture of whether overall treatment paradigms affect patient disability outcomes; this meant that unless an event caused the patient to change therapies, there was no need to report it. Another described that their study medication processes involved much less administrative burden than those of conventional trials, as their study medications were already FDA-approved, and the trial was only tracking outcomes after patients were randomized to receive one of the medications. Participants additionally noted that a change in research culture could encompass an increased acceptance of more parsimonious data collection or changes to institutional cold call policies.
Obtain buy-in/engagement
Participants expressed that successfully conducting embedded trials often requires staff in the healthcare setting to change elements of their usual procedures. For example, clinicians could be asked to screen and introduce the trial to potentially eligible patients, or they may need to adapt the way they present treatment options to their patients to accommodate randomization. IT or informatics personnel may need to add study-specific programming to the EHR to enable accessing, sorting, and extracting EHR data while pharmacy personnel may be asked to deviate from their normal processes for handling study medications. Thus, obtaining buy-in from healthcare setting staff at all levels was viewed as important for engaging site staff with the study and setting up an effective and collaborative partnership.
Some participants also noted the importance of both provider and patient engagement for successful trial conduct, commenting that providers who are interested and invested in the study are more likely to sign on to the trial themselves and encourage others to participate. Patient buy-in was described as useful for both recruitment and retention and could also be helpful during earlier stages, when engaged patients may inform aspects of study design. Additionally, participants expressed a need for buy-in at the sponsor level, noting that funding for embedded interventional trials needs to be expanded.
Reduce burden/minimize impact on the healthcare system
Participants described that sites and healthcare system staff may be more likely to participate if embedded trials did not impose much additional burden. For example, regulatory reforms around embedded interventional trials could help to alleviate the administrative burden on healthcare sites. Eliminating perceptions of clinician and staff burden, and making the trial seem more approachable, could be accomplished by demonstrating that the trial does not need to impede clinical workflow and will only require minimal effort from clinicians. Participants also noted that providing research support to clinical staff would reduce the burden and make it more likely that they would participate by minimizing the number of tasks staff have to perform in addition to their normal clinical duties. A participant specifically advocated for reducing the burden of redundant data entry, explaining that it would take less time and increase efficiency if data could be entered only once and then transferred into other systems that need it.
Invest in research infrastructure
Designers and implementers expressed that investments in research infrastructure could also serve to minimize burden on site personnel, for example, by having research staff available to assist with regulatory issues. Many participants described that research coordinators played a key role in embedded trials, with duties that included enrolling patients and obtaining informed consent, tracking and scheduling the collection of various data elements, performing data entry, and assisting with data extraction. A few participants explained that their research team included individuals versed in data management, analytics, and statistics who dealt with searching, abstracting, and analyzing the EHR data while other teams included research clinicians, such as physicians or nurses whose responsibilities could include overseeing the study personnel, ensuring proper study conduct, conducting chart reviews, confirming patient eligibility, and conducting study assessments that were outside of routine care. In some settings, these research clinicians were members of a larger research unit that was embedded in the healthcare system and that routinely assisted with the conduct of clinical trials across the enterprise.
Manage interoperability of EHR systems
Participants most commonly reported managing data collection across EHR systems using conventional electronic data capture platforms. To address the challenge of interoperability issues across healthcare EHR systems, participants developed a variety of solutions, many of which involved creating templates to extract data from EHRs or abstracting data manually. A trial designer mentioned specifically partnering with health networks that used the PCORNet Common Data Model, which includes curated EHR data, to address the issue of interoperability [27]. To facilitate embedded trials, participants suggested that changes are needed to EHR systems to make it easier to obtain enrolled patient data from any healthcare system and leverage EHR data in a more consistent way across healthcare systems. Table 3, Section 3 includes illustrative participant quotes on overcoming barriers to implementing embedded trials.
Discussion
Embedding randomized clinical trials into routine clinical practice is a noteworthy goal, yet experience remains limited. Indeed, for such trials to be utilized more, it is important to leverage the learnings from those who have previously conducted embedded trials.
This research describes experiences from trial designers and trial implementers and highlights a number of key suggestions, specifically, minimize the impact on healthcare settings and patients; obtain buy-in from healthcare settings and staff; have sufficient data and research infrastructure in place; and create a culture change facilitated by tailored messages to partner groups and education of partners who are part of “the usual care process” (e.g., patients, providers, leadership, and pharmacists). Understanding these barriers and proposed solutions is necessary to develop evidence-based, actionable recommendations to implement embedded trials.
The results of these qualitative interviews along with input from two CTTI-hosted Expert Meetings, informed a set of actionable recommendations developed by the multi-partner project team to facilitate the integration of randomized, interventional trial elements into clinical care [14]. These recommendations provide study design considerations, operational approaches, and suggestions on the cultural shifts needed to enable widespread integration (Figure 2). The recommendations emphasize that embedding elements of a trial into clinical practice is not “all or none.” Benefits can be gained regardless of the number of elements embedded. The recommendations also note that (1) the use of healthcare data sources for research purposes should be fit for purpose; (2) the trial design should aim to align with clinical workflows; (3) healthcare settings and sponsors should ensure site readiness to embed trial elements; and (4) leaders at the regulatory, funding, and health system level need to recognize and advance the message that embedding trials can improve evidence generation. In order to appreciate site readiness, CTTI developed the Embedding Trials Feasibility Survey for sponsors and researchers to assess the capacity and feasibility of sites to embed elements of a clinical trial into clinical practice [14,Reference Pessoa-Amorim, Campbell and Fletcher28]. Five case examples accompany the recommendations, illustrating individual trials that have embedded trial elements into care [14]. The case examples feature experiences within and outside the USA, review challenges encountered, and provide words of wisdom to those who may consider integrating a trial into clinical practice. The RECOVERY study is one such example in which patients in intensive care units in the United Kingdom were randomized to different investigational and approved medical products to assess appropriate interventions for COVID-19, and data collection was facilitated with linkage to national healthcare datasets [Reference Angus, Gordon and Bauchner29]. The Randomized, Embedded, Multifactorial Adaptive Platform Trial for Community-Acquired Pneumonia (REMAP-CAP) is yet another example, with hundreds of sites throughout the United States and globally, which used the EHR to identify patients and capture outcomes and endpoints that aligned with routine care [30]. Both examples aimed to make relevant results available in close to real time.
Figure 2. Summary of CTTI’s embedding clinical trials into clinical practice recommendations. CTTI = Clinical Trials Transformation Initiative.
In addition to CTTI’s work, other initiatives are focused on advancing the ability of healthcare systems to integrate trials [31]. The Duke-Margolis Center Health Policy Center conducted meetings on the subject of point-of-care trials, released a white paper on the topic, and created the Coalition for Advancing Clinical Trials at the Point of Care (ACT@POC) to understand and implement well-designed, large-scale point of care trials [32,33]. FDA leadership continues to emphasize the importance of exploring the potential for embedding trials, acknowledging the need to integrate clinical trials directly into clinical care to avoid a separate infrastructure for clinical research [Reference Califf, Cavazzoni and Woodcock4]. The Advanced Research Projects Agency for Health recently launched the Advancing Clinical Trial Readiness (ACTR) initiative to establish a robust clinical trial infrastructure to enable 90% of eligible Americans to take part in a clinical trial within a half hour of their home [34]. ACTR aims to demonstrate the trial design and infrastructure needed to operate trials at the point of care [34].
This work acknowledges the benefits of embedding trials into clinical practice, while also appreciating the barriers, and provides operational recommendations to facilitate integration. Ultimately, this work aims to draw attention to areas where researchers and policymakers can rethink the design and conduct of clinical trials to ensure appropriate protection and respect of participants, allow for the collection of quality data to answer meaningful research questions, and encourage the development of a learning health system through improved clinical evidence generation. Additional work is needed to fully appreciate the implementation of these trials in various contexts and how the recommendations provided here support successful implementation.
Limitations
A limited number of trials met our inclusion criteria for an embedded clinical trial. Although we were able to group narratives and identify commonalities within some of the topics investigated, some of the information we provided was mentioned by only one trial or one participant. Additionally, a different group of participants may have described different or additional experiences to those documented here. However, particularly given the challenge of identifying the specific design and methodology that reflect this type of trial, recognizing that we viewed the term “pragmatic” in the definitional, practical sense rather than based on a PRECIS-2 score [32], we believe our findings are broadly reflective of the motivations, barriers, and facilitators to conducting these types of trials. It is also important to acknowledge that barriers and facilitators to integrating research into practice may vary according to the trial’s context, including the setting (e.g., academic medical center vs. community setting), approach (pragmatic vs. traditional), question of interest (e.g., efficacy vs. effectiveness), and intervention (medication vs. behavior vs. healthcare delivery). Lastly, we note the focused scope of the study and the accompanying CTTI recommendations. The recommendations focus on operational and design considerations of embedding elements of clinical trials into clinical practice. Topics, such as financial and ethical implications, were outside the scope of this work.
Conclusions
Embedding elements of clinical trials into clinical practice can enhance knowledge generation and promote the translation of that knowledge into improved patient care. It also has the potential to increase trial quality and efficiency by reducing duplication of trial and care activities and lessening patient burden by allowing patients to participate in research in their usual routine care setting. The research and recommendations outlined in this article recognize the barriers to embedding trials into clinical practice, provide operational recommendations to facilitate integration, and draw attention to areas where we can rethink the design and conduct of clinical trials to ultimately improve access to research and care.
Acknowledgments
CTTI thanks all participants for sharing their perspectives and experiences with us. CTTI also thanks the CTTI Embedding Trials in Clinical Practice Team for their contributions to the project, Heather Stone, FDA, for her review of the manuscript, and Brooke Walker, Duke Clinical Research Institute, for editorial assistance. CTTI and Duke staff who contributed to the implementation of the research or manuscript preparation were compensated as part of their salaries.
Author contributions
Conception and design of the research: CD, TS, LK, KK, MT, MS, and AC. Data collection and analysis: CD, TS, and AC. Drafting of the manuscript: CD, TS, LK, and AC. Critical review and responsibility for the manuscript as a whole: CD, TS, LK, KK, MT, MS, and AC.
Funding statement
The research was supported by the FDA of the US Department of Health and Human Services (HHS) as part of an award totaling $3,778,241.33 with 15% financed with non-governmental sources. The contents are those of the author(s) and do not necessarily represent the official views of, nor an endorsement, by FDA, HHS, or the US Government. For more information, please visit FDA.gov. Partial funding was also provided by pooled membership fees from the CTTI’s member organizations (https://ctti-clinicaltrials.org/who_we_are/funding/).
Competing interests
The authors declare that they have no competing interests.
Open access
