Enrollment Obstacles in a Randomized Controlled Trial: A Performance Survey of Enrollment in BEST-CLI Sites

Background: Although randomized controlled trials (RCTs) provide the most reliable form of scientific evidence, they are challenging to complete because of a variety of enrollment obstacles. We evaluated obstacles in a large RCT by comparing survey results at high-performing sites (HPS) and low-performing sites (LPS). Methods: The Best Endovascular versus Best Surgical Therapy in Patients with Critical Limb Ischemia (BEST-CLI) trial is a prospective, pragmatic, multicenter, and multispecialty RCT that will compare clinical outcomes, quality of life, and cost in patients with CLI randomized to surgical bypass or endovascular therapy. BEST-CLI aims to enroll 2100 patients at 160 sites in North America, Europe, and New Zealand. We surveyed the 30 HPS and 30 LPS to assess perceptions of enrollment obstacles. HPS were defined by enrollment of 0.5 subjects or more per month or more than 8 total subjects enrolled. LPS were defined by enrollment of 0.1 subjects per month or only 1 subject total. Responses were compared by site performance status. Results: There were 22 of 30 (73%) HPS and 14 of 30 (47%) LPS that answered the survey (P = 0.06), including 17 investigators and 31 coordinators. The mean total enrollment and rate of enrollment at HPS and LPS were 12.5 subjects at 1.5 subjects/month and 1.0 subject at 0.1 subjects/month, respectively. The most common barrier to enrollment at HPS was difficulty convincing patients and their families to participate (36%), whereas at LPS both difficulty convincing patients and difficulty motivating investigators to enroll (29% each) were most frequently cited. At HPS, the most common obstacle to consenting patients for the trial was patient/family having strong preference toward revascularization strategy (32%) and at LPS it was patient/family not wanting to have treatment chosen at random (36%). At 55% of HPS and 43% of LPS, the trial team was reported as extremely collaborative (P = 0.73), whereas 68% of HPS and 64% of LPS reported having identified a trial champion on their team (P = 1). The most restrictive perceived enrollment criterion at HPS was prior index limb stenting with significant restenosis (32%), whereas at LPS it was excessive risk for surgical bypass (43%). Materials to aid enrollment were used equally at HPS and LPS: patient brochures at 59% HPS and 64% LPS (P = 1); investigator talking points at 45% of HPS and 36% of LPS (P = 0.73). Conclusions: Patient perceptions and investigator biases are significant challenges to enrollment in large RCTs. In the BEST-CLI trial, difficulty convincing patients and families to allow treatment randomization and difficulty in motivating investigators were major enrollment obstacles.