Does Participation in a Randomized Clinical Trial Change Outcomes? An Evaluation of Patients Not Enrolled in the SPRINT Trial.

Abstract

To determine the extent to which knowledge from clinical trial protocols is transferred to nonparticipating patients. Retrospective review of prospectively collected data from a large clinical trial. Six level-1 international trauma centers. We compared rates and timing of reoperation in a subset of patients enrolled in the Study to Prospectively evaluate Reamed Intramedullary Nails in Patients with Tibial Fractures (SPRINT) to concurrent patients who were eligible but not enrolled. This was a retrospective review of prospectively collected trial data. The records of 6 of the original SPRINT centers were searched for non-SPRINT patients who underwent intramedullary nailing of a closed tibial fracture. The rate and timing of reoperation were compared. A P < 0.05 was considered significant. One hundred fourteen non-SPRINT patients were compared with 328 patients enrolled in SPRINT from those same sites. There were 7 reoperations (6.1%) in non-SPRINT patients versus 18 (5.2%) in SPRINT patients [odds ratio (OR) 1.19, 95% confidence interval (CI) 0.41 to 3.13; P = 0.811]. There was no difference in the time to reoperation between the SPRINT and non-SPRINT patients (6.2 vs. 6.8 months, 95% CI of the difference -3.8 to 2.6; P = 0.685) or in the proportion of patients who underwent reoperation before 6 months (29% vs. 43%; OR 1.75; 95% CI 0.18 to 15.41; P = 0.647). Patients not enrolled in SPRINT had similarly low rates of reoperation for nonunion, and the average time to reoperation for both groups was longer than 6 months. A 6-month waiting period may have allowed slow-to-heal fractures adequate time to heal, thereby reducing the rate of diagnosis of nonunion. As such, this waiting period could contribute to lower-than-expected reoperation rates for nonunion. It is possible that clinical trials may beneficially influence the care of nonenrolled patients.