Replacing Anecdotal With Empirical Evidence On Impact of DCTs

Ken Getz, MBA, Executive Director and Research Professor, Tufts Center for the Study of Drug Development, Tufts University School of Medicine

The low incidence of, and wide variation in, the use of decentralized clinical trial (DCT) components has made it very difficult historically to gather hard evidence and measure impact.Hard data on experiences with—and the return on investment of—virtual and remote visits, direct-to-patient clinical trial supplies, local labs, wearables, and sensors would be invaluable in informing ways to optimize protocol design and execution.

Fortunately, we are making some progress as organizations such as the Partnership for Advancing Clinical Trials (PACT) consortium are helping to replace anecdotal reports with empirical evidence. One area in particular where we are observing a measurable impact on DCT component use is in patient recruitment and retention.

The results of recently-gathered data by the PACT consortium shows that, although clinical trials using DCT components typically encounter lower site activation rates, the enrollment rate per site is significantly higher with actual enrollment timelines considerably shorter than planned. PACT data and the results of other studies are also showing a strong association between the types of DCT solutions used and the proportional representation of patients enrolled by race and ethnicity.

A brief word about PACT: The consortium was launched in 2024 with start-up funding from Medable and the Reagan-Udall Foundation for the FDA. Facilitated and housed within the Tufts Center for the Study of Drug Development (Tufts CSDD), PACT focuses on gathering hard metrics from actual experiences with virtual and remote solution deployments in clinical trials. Thirty-four member companies, including pharmaceutical and biotechnology firms (e.g., AstraZeneca, CSL Behring, GSK, Eli Lilly, Novartis, Regeneron, Roche, Sanofi, Takeda) and contract research organizations (CROs) (e.g., Fortrea, Icon, Parexel, Veristat) contributed data and shared insights on 69 clinical trials in 2024.

Impact on enrollment timelines

Most of the data collected comes from Phase II and III clinical trials, the majority (75%) of which had protocol approval dates since 2021. A broad mix of therapeutic areas are represented in the dataset, with the highest proportions for investigational treatments in neurology and psychiatry (20%), immunology (19%), oncology (12%), and endocrinology (6%). Data provided specifically by CROs comes from studies sponsored by mid-sized and smaller pharmaceutical and biotechnology companies that are not members of the PACT consortium.

For the 69 clinical trials analyzed, a wide variety of virtual and remote solutions were deployed primarily to support study visit activities. The most prevalent solutions used included virtual visits and telehealth, data collection apps and reminders on mobile devices, and home visits.

DCT component use contributed to faster-than-planned enrollment timelines. Lower screen failure rates and similar completion rates were observed when comparing PACT’s DCT-supported clinical trials to traditional clinical studies that used no DCT components (i.e., benchmark data). Within the PACT dataset, faster enrollment cycle times were significantly correlated with higher proportional use of remotely performed procedures (p<.01). And although the correlation only approached statistical significance, faster enrollment cycle times were associated with higher proportional use of remote visits.

Compared to Tufts CSDD benchmarked practices, it is clear that sponsors and CROs expect their clinical trials supported by virtual and remote solutions to take considerably longer.Controlling for phase and therapeutic area, when DCT components are included, planned cycle times are substantially longer. Actual enrollment cycle times of the 69 clinical trials in the PACT dataset were much more likely to meet or beat their corresponding planned timelines compared to clinical trials that used no DCT components.

Actual timelines from first patient first visit to last patient first visit were, on average, 22.3 days shorter than the plan for DCT solution-supported clinical trials. And mean actual timelines from protocol approval to database lock were 102.7 days shorter than planned timelines.

Impact on enrollment diversity

Analysis of the PACT year-one dataset found that the use of virtual and remote solutions was also associated with higher proportional representation of select demographic subgroups. A much higher proportion (20.9%) of participants who identify as Asian was observed when DCT components were deployed compared to the benchmark 14.2% Asian participants when no solution was used.

Proportional representation of indigenous communities (e.g.,American Indian and Alaska Native) was nearly quadruple (1.9% vs 0.5%) the benchmark level. No significant differences were observed in the proportional representation of patients who identified as Black or of African descent in clinical trials with or without DCT solutions support in the aggregate.

These results are consistent with findings from another Tufts CSDD study that was featured in Applied Clinical Trials last year. This study evaluated select virtual and remote solutions and provides insight into the impact that individual DCT components play in improving patient access and enrollment convenience.

The use of four broad DCT solutions categories was assessed in 194 Phase II and III protocols: (1) Virtual visits (defined as a scheduled participant evaluation in which procedures are performed outside of the clinic (e.g., telephone/mobile telephone, video conference app, secured electronic mail, social media platforms); (2) Home visits (A home healthcare professional visits the participant’s home to perform evaluations or procedures); (3) Devices and wearables (where the study participant is provided with a wearable or other device (i.e., iPad or a tablet) to use at home for data collection and recording, or where participants were asked to download an app to a personal device which will be used to complete surveys or collect other data); and (4) Local locations (the study volunteer travels to a local lab or clinic instead of the investigative site to have procedures or tests performed).

The results showed an association between the DCT solutions categories used and changes in the proportional representation of racial and ethnic diversity among patients enrolled (see Table 1 below). The use of local locations was associated with a 38% increase in the percentage of patients who identify as Black or of African descent and a modestly lower percentage of White patients enrolled. The use of virtual and of home visits in clinical trials were associated with much lower percentages of enrolled patients who identified as Black. The use of virtual visits was associated with a 20% increase in the proportion of patients who identify as Asian. Only modest changes in proportional representation were observed in clinical trials that used mobile and wearable devices.

Patient preferences and protocol customization

In all, the evidence gathered suggests that DCT component use will become increasingly fit-for-purpose, customized to each clinical trial based on patient preferences and needs. Protocol execution complexity and burden associated with DCT component use will fall squarely on investigative sites to navigate and manage.

No doubt the corresponding positive impact on recruitment and retention performance will motivate sites to become ever more adept at incorporating DCT components into their operating activity.It will encourage sites that have achieved sufficient scale and infrastructure (e.g., site networks, academic and community-based health systems) to leverage their own integrated solutions, including virtual communications, remote patient visits, and direct to patient clinical supplies.

Through ongoing data contributions in 2025, the PACT dataset will continue to grow.New members with actual experience using DCT components are joining the consortium, including sponsors, CROs and investigative sites. Members are adding more usage and performance variables to measure. With this growing body of empirical evidence, more valuable insights into specific DCT components associated with, and predictive of, clinical trial performance will be incorporated into protocol customization strategies.

Ken Getz, MBA, Executive Director and Research Professor, Tufts Center for the Study of Drug Development, Tufts University School of Medicine