Introduction
Recruitment and retention are leading barriers to the success of clinical trials that have been present even prior to the COVID-19 pandemic. Patient recruitment has been found to be the single biggest cause of clinical trial delays [Reference Ali, Zibert and Thomsen1]. Up to 20% of clinical trials either face early termination due to failure to recruit or continue to completion while having failed to meet the original recruitment target [Reference Goodson, Wicks, Morgan, Hashem, Callinan and Reites2]. Challenges that impact recruitment to clinical trials include issues related to study design, physician attitude, participant attitude, accessibility, demographics, and socioeconomic disparities [Reference Chen, Lu and Kummar3]. The methods and logistics through which clinical studies are conducted have traditionally placed the clinic or hospital at the core. Such a model is referred to as site-centric [Reference de Las Heras, Daehnke and Saini4]. Decentralized methods are those that focus clinical study conduct around the patient, using methods such as telemedicine, electronic consent (eConsent), wearable biomarkers, and home visits. Decentralized methods of recruitment include the use of virtual strategies such as social media apps, online campaign advertising targeting specific patient populations, and email blasts to patient databases to improve recruitment and enrollment [Reference Sedrak, Sun and Liu5]. Retention is another issue that clinical investigators face, as trials can have up to 40% of the participants dropout [Reference Ali, Zibert and Thomsen1]. Virtual strategies to improve retention include patient alerts and reminders for visits or completion of questionnaires related to their participation in a clinical study. Varied virtual strategies can not only help ensure accrual but also improve retention, completion, and compliance to the protocol.
Decentralized methods have been incorporated to also improve the diversity represented by the US population in clinical trials. The US Food and Drug Administration (FDA) has drafted guidance regarding how investigators should aim to improve the enrollment of these populations in clinical trials [6]. Offering digital alternatives to increase accessibility to groups who might otherwise have been unable to participate allows for these groups to participate in clinical trials in a more equitable and inclusive manner [Reference Kelsey, Patrick-Lake and Abdulai7]. In addition to these initiatives focused on increasing diversity, the FDA has issued other guidances acknowledging the potential of leveraging new technologies to improve data acquisition [8].
All aspects of clinical research were disrupted by the onset of the COVID-19 pandemic, suspending ongoing trials, delaying the start of new studies, and hindering participant enrollment [Reference Tan, Wu and Day9]. The technology for successful incorporation of digital methods into clinical studies has existed for over 20 years, and although decentralized methods have been used in prior clinical studies, the circumstances brought about by the COVID-19 pandemic created conditions for which their adoption became necessary [Reference Zahradka, Pugmire, Lever Taylor, Wolfberg and Wilkes10]. In order to mitigate the risks of COVID-19 while resuming research activity, the FDA as well as the National Cancer Institute issued guidance providing flexibility to clinical investigators for the adoption of decentralized methods in conducting clinical trials [Reference Pugmire, Lever Taylor, Wilkes, Wolfberg and Zahradka11,12]. These guidances suggest the use of virtual clinic visits, delivery of investigational tools to the participant’s home, and use of alternative laboratories or imaging centers to conduct the collection of trial data outside of the traditional, single-site model requiring in-person visits [Reference Pugmire, Lever Taylor, Wilkes, Wolfberg and Zahradka11,12]. Social media platforms such as Facebook can improve patient recruitment to include rural and other populations typically not represented in a traditional model of clinical studies [Reference Rolnitsky, Kirtsman, Goldberg, Dunn and Bell13]. Third-party allied health providers such as nurses, physical therapists, phlebotomists, and physician assistants can do home visits to provide protocol-related services [Reference Johnson and Marsh14]. The use of telemedicine and eConsent further increases accessibility for clinical study participants. Since the COVID-19 pandemic, national pharmacies like Walgreens and CVS have set up walk-in clinics where patients can be seen by nurse practitioners and physician assistants [15]. The use of these decentralized methods also allows for more widespread access to patients and enables them to participate in clinical studies in an otherwise nontraditional manner.
Although many trials have now adopted guidance from the FDA and have incorporated decentralized methods into clinical studies, there is still a lack of directly comparable data on how these studies have performed with regard to the aspects of recruitment and retention [Reference Rogers, De Paoli and Subbarayan16]. The objective of this review is to address this gap in knowledge by conducting a systematic review of the literature and identifying papers that provide insight on how decentralized methods have impacted recruitment, retention, and participant diversity in clinical studies.
Methods
The objective of this review is to identify peer-reviewed publications reporting on the use of decentralized methods within clinical studies. The following research question was used for primary analysis: How has the implementation of decentralized clinical studies (DCSs) impacted the aspect of recruitment in clinical and translational research?
Search Strategy and Screening
A systematic search of the literature was conducted to identify relevant articles for the use of decentralized methods in clinical studies and their impact on the aspect of recruitment. As subcategories, this review identified articles that covered aspects of retention and participant diversity in addition to recruitment. A preliminary search for “decentralized clinical trials” (DCTs) was conducted in PubMed to identify interchangeable terms and acronyms to eventually be used in the final search equation. Pearl growing methods were used to identify additional search terms for the search strategy.
Articles were identified by searching in databases PubMed, Cochrane, and EMBASE using a variety of keywords and medical subject headings related to the research question. To ensure that a sufficient number of articles were identified, these keywords included terms pertaining to both clinical trials and observational studies. This was also done by having the complete search syntax included truncated terms to maximize potential article identification. Search terms were limited to include only in title and abstract settings for more precise results. The final search was conducted on October 9, 2022, as follows:
-
1. Decentral* clinical trial* OR direct-to-participant trial* OR virtual stud* OR virtual trial*
-
2. enroll* OR recruit*
-
3. 1 AND 2
All the identified articles from the initial database searches were exported to EndNote 20 reference management software. Duplicate studies were removed. The titles and abstracts of the articles were reviewed. Relevant articles were retained for eligibility screening. Full-text articles were requested for the remaining articles. Reports were then screened in accordance with the exclusion and inclusion criteria that were set. Studies that included an outcome related to recruitment were included in the final review. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement was used as a guideline. A PRISMA flow diagram depicting the search process used is illustrated in Fig. 1.
Figure 1. PRISMA flow diagram.
Eligibility Criteria
Following the results of the search conducted in databases PubMed, EMBASE, and Cochrane, articles commenting on the impact of decentralized methods on the aspects of recruitment, retention, and diversity in clinical studies were identified. For the focused review, (1) articles discussing the aspect of participant recruitment in clinical studies implementing decentralized elements were included. Studies including discussion on participant retention and diversity in addition to recruitment were also noted. Other inclusion criteria included the following: (2) study was conducted within the past 5 years, (3) study provides sufficient detail or data with regard to the primary outcome of this review, and (4), study was published in English.
With regard to exclusion criteria, (1) studies conducted prior to 5 years ago were excluded. A time restraint of 5 years was chosen since DCSs largely rely on the current technology available. Therefore, recent articles would provide the most accurate depiction as to how DCSs currently affect clinical and translational research. Other exclusion criteria included the following: (2) studies for which full texts in English were unable to be found, (3) studies that provided insufficient detail with regard to how the study was conducted with decentralized methods, (4) studies that were eventually superseded by another study conducted by the same primary author, and (5) studies that did not focus on the impact of decentralized methods in the context of clinical research.
Data Extraction
The articles for which full texts were obtained were all logged into a spreadsheet using Microsoft Excel. Collected data points included title, authors, publication year, patient population, decentralized methods used, and primary and secondary outcome information.
Outcomes
Outcomes that were logged and included in this review are as follows: decentralized methods implemented during the trial, sample population, study design, reported recruitment outcomes, reported retention outcomes, reported participant diversity outcomes, and participant satisfaction. The primary outcome of this review is recruitment outcome. The secondary outcomes of this review are participant retention, diversity, and participant satisfaction.
Results
The searches initially identified 174 articles. Prior to screening, 36 duplicate articles were removed, leaving 138 left for title and abstract screening. Eighty-seven articles were removed due to lack of relevancy to DCSs or the research question. Full texts were attempted to be obtained for 51 remaining articles, which was successful for 50. The 50 articles were entered into a Microsoft Excel spreadsheet for data extraction and eventual analysis. Fourteen articles were further screened and removed in accordance with the exclusion criteria for this systematic review. Twenty-four articles were included in the review for contextual information and 13 studies provided data to assess the end points of this systematic review. The results of this process are described in a PRISMA flow diagram (Fig. 1). Thirteen studies covering decentralized methods in clinical studies were assessed for outcomes pertaining to recruitment, retention, diversity, and participant satisfaction. These studies are identified in Table 1 and summarized in Table 2, including the type of article and decentralized methods implemented. Table 3 lists the outcomes reported by the investigators pertinent to the focus of this systematic review.
Table 1. Publications included in analysis
Table 2. Description of decentralized methods used and study outcomes
Table 3. Primary and secondary outcomes of systematic review
The articles identified (Table 1) provide detailed data and insight from investigators as to how decentralization has impacted clinical studies. Here, the decentralized methods commonly used are outlined by study (Table 2). Ten of the 13 articles contain information directly from completed or ongoing studies, while 3 articles are surveys or reports evaluating the impact of recruitment on clinical trials. Eight of the 13 studies have been conducted in the USA, while the other studies have been solely conducted in European countries.
Decentralized Methods
The studies identified in this review incorporated methods that leverage technology or other means to deviate from the traditional, single-site model historically employed to conduct clinical trials. Common methods mentioned among these 13 studies include virtual recruitment, electronic patient reported outcomes (ePRO) through wearable biomarkers or smartphone applications, eConsent forms, and virtual visits.
Recruitment
Eleven out of 13 studies report recruitment success implementing social media advertisements or other methods deviating from the traditional approach where all recruitment is conducted in the clinic. In these studies, success was declared through the achieved recruitment rate, the number of participants enrolled, or the ability to meet the recruitment goals of the study. Of the two other studies, one was a survey in which participants expressed likelihood to enroll in trials using decentralized methods to lessen travel burden. The other study only voiced improved recruitment in two of the three trials covered. This review identified studies that directly compared virtual recruitment strategies alongside traditional methods. Seven studies made direct comparisons between the two, where outcomes such as time to reach recruitment goals and number of enrolled participants favored virtual methods instead of in-clinic, traditional methods of recruitment. When commenting about the benefits of virtual recruitment, authors often noted faster recruitment rates and cost-effectiveness, with some studies having to place potential subjects on waitlists due to the dramatic surge in recruitment that was observed.
Retention
Seven out of 13 studies were found to report success with retaining subjects in studies implementing decentralized methods, potentially attributing the study’s success to the ease of participation for subjects. Four studies included direct comparisons between retention in DCSs with traditional methods. These comparisons often occurred by comparing retention of participants recruited prior to the pandemic with retention of participants recruited after the pandemic forced trials to adopt virtual methods. In other articles, decentralized and traditional arms occurred simultaneously in the study, or retention in the decentralized study was compared with the standard retention of other studies for that time.
Diversity
Six out of 13 studies reported some benefits to sample population diversity in DCSs. Diversity was usually observed in the ability to recruit geographically remote subjects, often from rural areas lacking adequate healthcare coverage. Two studies conducting DCSs noted potential limitations in their ability to reach certain populations, with those investigators suggesting that adjustments should be made to reach historically underrepresented racial and ethnic groups. Two other studies set specific goals to reach underrepresented minorities. These studies involved partnering a community hospital with a larger academic center to conduct clinical trials employing decentralized methods, and both studies were successful in achieving their goals pertaining to participant diversity. The authors attribute this success with diversity to the use of telemedicine and uniform access to electronic health records (EHRs) across sites, which they used to overcome geographical boundaries and attract diverse participants representative of the target populations.
Satisfaction
Six studies reported on participant satisfaction, all of which included positive feedback. Satisfaction was expressed in a variety of ways, such as enthusiasm for participation in future DCSs or preference toward the decentralized methods implemented. Participant satisfaction was often measured through feedback questionnaires that took place either during the study or after completion of the study. One study noted high satisfaction with the informed consent process being conducted virtually. Three studies included enthusiasm for participation in future DCSs.
Discussion
The goal of this systematic review was to identify how the implementation of decentralized methods has impacted the ability of clinical studies to overcome long-standing barriers to success such as participant recruitment. The identified papers covered clinical trials and studies that were conducted using methods such as wearable biomarkers, eConsent, and virtual visits, to provide accessibility that may not be present in the traditional, single-site study. This review provides data from studies on the use of these decentralized methods for recruitment, retention, diversity, and participant satisfaction. Overall, the findings of this systematic review provide further support in favor of implementing decentralized methods in clinical studies. The results are consistent with prior literature identifying the advantages associated with decentralized methods [Reference Rogers, De Paoli and Subbarayan16,Reference Moseson, Kumar and Juusola30].
Previous systematic reviews conducted to assess the impact of decentralization on clinical trials have identified the advantages associated with these methods [Reference Rogers, De Paoli and Subbarayan16,Reference Moseson, Kumar and Juusola30]. However, these reviews have stated that there is still a lack of definitive support for whether decentralized methods are beneficial for recruitment, retention, adherence, and cost metrics [Reference Rogers, De Paoli and Subbarayan16]. This may result in hesitation from clinical investigators to adopt these methods in future trials. This review provides data from studies, many of which started implementing decentralized methods as a result of the COVID-19 pandemic, on the aspects of recruitment, retention, diversity, and participant satisfaction.
Participant recruitment is one of the largest barriers to successful clinical trials, and the included articles demonstrate that investigators are willing to implement decentralized methods to access potential participants through nontraditional means [Reference Ali, Zibert and Thomsen1]. The studies identified in this review reported faster recruitment rates and higher amounts of enrolled participants through conducting the enrollment process virtually or by using social media advertisements [Reference Ali, Anderson and Chiriac18–Reference Yiannakou, Howell and Kemppinen29]. The strongest evidence comes from the trials that have conducted recruitment through virtual and traditional methods, comparing the outcomes between the two [Reference Ali, Anderson and Chiriac18–Reference Magnani, Ferry and Swabe21,Reference Ng, Howell and Kemppinen23,Reference Sarraju, Seninger and Parameswaran24,Reference Sommer, Zuccolin and Arnera28,Reference Yiannakou, Howell and Kemppinen29]. In addition, presenting patients with the opportunity to enroll in clinical trials implementing decentralized methods has also been shown to increase self-reported likelihood to enroll [Reference Adams, Long and Fleury17]. As investigators continue to evaluate the benefit to recruitment that DCSs have, it should be noted that parameters such as conversion rate should be considered differently. In virtual recruitment, conversion rate is calculated by dividing the number of individuals signing up for the study by the number of individuals who visited the recruitment website and clicked on the button to participate [Reference Ali, Anderson and Chiriac18]. Although conversion rates with virtual recruitment may be lower compared to conversion rates through conventional recruitment, online recruitment reaches a higher number of people and conversion rates for virtual studies above 5% are considered good [Reference Ali, Anderson and Chiriac18].
Though conducting the initial search for this review focused on participant recruitment in DCSs, it was noted that many articles also commented on how decentralized methods impacted other elements such as retention. Secondary outcomes evaluated in this review included retention, diversity, and participant satisfaction in studies utilizing decentralized methods. Articles reporting on these outcomes in clinical studies have largely been in favor of decentralized methods, reporting positive outcomes such as high retention rates, better inclusion of historically underrepresented populations, and enthusiasm for future participation in DCSs. In order to reduce health disparities in the USA, increase the generalizability of results, and promote equity in healthcare, clinical investigators have an ethical duty to take actionable steps such as leveraging digital tools to decrease the burden of participation for underserved populations. Through implementing telemedicine into clinical trials, geographical boundaries are nearly eliminated for potential participants that may have been previously unable to access trials. DCSs operating through community hospitals partnered with larger academic centers have shown success in reaching enrollment goals focused on participant diversity. The studies assessing this model utilized a satellite hospital serving a geographical area designated as a healthcare shortage area by the Health Resources and Service Administration. This hospital, which serves rural and underserved populations, received institutional oversight from a larger academic medical center. One way this lead institution provided oversight included having the principal investigators and main research coordinators located there and providing training for the other physician investigators located at the community hospital via videoconferencing [Reference Sedhai, Patel and Mihalick25,Reference Sedhai, Sears and Vecchiè26]. Increasing the ease of participation in clinical studies has demonstrated a positive impact on patient retention in DCSs [Reference Ali, Anderson and Chiriac18]. Many of these studies had participants complete study-related tasks from their homes through eDiaries or mobile applications. One study implementing a rewards-based system for task completion had more than 50% of the participants report increased motivation to progress with the trial due to this design [Reference Ali, Anderson and Chiriac18]. Similar methods of engagement such as sending newsletters to patients, easily consumable videos, or providing patients with appropriate data relevant to the trial may show similar success. Participants in DCTs also expressed interest in future clinical trial participation where decentralized methods are being utilized [Reference Sarraju, Seninger and Parameswaran24].
Implementing decentralized methods promotes study design and conduct that moves away from that of the traditional, single-site model. The use of satellite hospitals is one of many examples of the potential for new models of trial oversight. The use of decentralized methods in clinical trials was gaining support from regulatory agencies prior to the onset of the COVID-19 pandemic. In December 2018, the FDA announced a new strategic framework to advance how investigators utilize real-world evidence (RWE) in supporting the development of drugs and biologics [Reference de Las Heras, Daehnke and Saini4]. This framework acknowledges how the healthcare system is finding more effective methods to leverage electronic tools to gather health-related information during routine care of patients, referred to as real-world data. The FDA has already allowed the use of RWE to eliminate the need for post-marketing studies on nine potential safety issues in five products, and the framework includes efforts to utilize RWE to help support the approval of new indications for approved drugs [8]. Following the onset of the pandemic, the advantages of a patient-centric model of conducting clinical trials over the traditional, site-centric model are becoming clearer. In response to the rapid adoption of decentralized methods by investigators, the FDA is requesting the applicants of new drug application/biological license applications (NDA/BLA) to indicate when data points were collected using remote means, looking to determine the potential risks and benefits associated with DCT solutions [Reference de Las Heras, Daehnke and Saini4].
Several studies in this review reported limitations with the use of decentralized methods. In one study utilizing social media recruitment through Facebook, the majority of participants enrolled were young women even though the average age of adults using Facebook is 48.2. These investigators have recommended future studies to consider conducting online recruitment through additional channels, such as the websites of patient associations or pharmacies, in order to address this bias in recruitment [Reference Ali, Anderson and Chiriac18]. Reported outcomes for participant satisfaction are also at risk for bias. High rates of reported satisfaction may be skewed, since questionnaires were often optional and only open to individuals who enrolled in the study, which could indicate precluding interest and enthusiasm in the study.
The results from this systematic review of studies conducted in the last 5 years demonstrated the potential of decentralized methods to enhance the conduct of clinical studies toward a patient-centered model. The inclusion of studies done during the years of the COVID-19 pandemic accelerated the use of these tools as well as strategies to mitigate challenges. Recruitment may be improved in DCSs due to the virtual recruiting methods employed or because of the increased accessibility granted in these trials. The increased accessibility that DCSs demonstrate is often met with participant satisfaction, and therefore, it should be noted that the implementation of patient-centered tools is being positively received. The resulting data from this analysis lessen the gap of documented information needed to justify the integration of decentralized methods when designing clinical studies. Future DCSs should continue to evaluate and report on these methods to better understand the impact on recruitment, retention, diversity, and participant satisfaction.
Acknowledgments
The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.
Funding statement
Research reported in this publication was supported by the National Center for Advancing Translational Sciences (NCATS), a component of the National Institute of Health (NIH) under Award Number UL1TR003017.
Competing interests
The authors have no conflicts of interest to declare.
Open access
