Decentralized Community-Integrated Research Sites Drive Higher Randomization Rates: Insights from a Large-Scale Neurodegenerative Disease Trial

Seyed-Mahdi Khaligh-Razavi^1*Tyler E Miller^1,2Ralph Passarella¹Ahmad Namvargolian^1*

• ¹Care Access Research LLC, Boston, United States
• ²Department of Pathology, Case Western Reserve University School of Medicine and Case Comprehensive Cancer Center, Cleveland, United States

Introduction:
Recruitment and retention remain critical challenges in clinical trials, particularly in neurodegenerative diseases, which require large participant populations, rigorous screening, and prolonged follow-up periods. Care Access is a global research site management organization that operates clinical trial sites employing various operational models. This study evaluates the operational performance of Care Access site models—including traditional, hub-and-spoke, and decentralized community-integrated research (DCIR) sites—within a Phase 3 neurodegenerative disease trial, focusing on their relative efficiency in recruitment, randomization, and retention. The inclusion of multiple site models within a single trial provides a unique opportunity for direct comparison, enabling identification of effective strategies to enhance trial efficiency. Methods:
The trial involved 32 Care Access sites, each employing one of these distinct operational models. Key performance metrics, such as participant screening rates, randomization rates, screen failure rates, and post-randomization discontinuation rates, were analyzed across traditional, hub-and-spoke, and DCIR site models. Enrollment performance of Care Access was compared with 196 non-Care Access sites using publicly available data.Results: DCIR sites demonstrated the highest recruitment efficiency, screening 20.61 participants per site per month and randomizing 0.79 participants per site per month, compared to 11.78 and 0.50 for traditional sites, and 12.20 and 0.45 for hub-and-spoke sites, respectively. Despite being newly established and decentralized, DCIR sites achieved post-randomization discontinuation rates (28.17%) comparable to traditional sites (26.28%), highlighting their effectiveness in maintaining participant engagement. All site models encountered high screen failure rates (~95%), consistent with Phase 3 neurodegenerative trials. A community-engaged, research-only facility achieved the lowest discontinuation rate (17.65%), underscoring the potential of local engagement. Care Access sites had an average randomization rate of 15.6 participants per site, outperforming non-Care Access sites (8.7 participants per site). Data quality, monitoring practices, and data integrity were consistent across all site models.Discussion:
DCIR sites exhibited superior screening and randomization efficiency while maintaining comparable discontinuation rates to traditional sites. Success was driven by decentralized community-based outreach, expanding access to research among populations previously lacking clinical trial opportunities, enhancing geographic coverage, representation, and trial performance. Findings highlight a flexible, multi-model strategy to address recruitment and retention challenges, potentially applicable across therapeutic areas.