Enrollment volume effect on risk factor control and outcomes in the SAMMPRIS trial.

Chiu, David; Klucznik, Richard P; Turan, Tanya N; Lynn, Michael J; McCane, Charles D; Katz, Lawrence B; Nizam, Azhar; Derdeyn, Colin P; Fiorella, David; Lane, Bethany F; Montgomery, Jean; Janis, Scott; Chimowitz, Marc I

OBJECTIVE The role of physician experience and patient volumes on the outcome of surgical or endovascular procedures has been well-studied but there are limited data on how these factors affect the outcome of medical therapy. METHODS In the stenting and medical cohorts of the Stenting and Aggressive Medical Management for the Prevention of Recurrent Ischemic Stroke (SAMMPRIS) trial, we compared Kaplan-Meier (K-M) curves for the primary endpoint (any stroke or death within 30 days of enrollment or ischemic stroke in the territory beyond 30 days) using the log-rank test and the percentages of patients achieving target levels for primary and secondary risk factors during the study using Fisher exact test between patients at high-enrolling (≥12 patients) vs low-enrolling (<12 patients) sites. RESULTS In the stenting group, the K-M curves for the primary endpoint were similar at high-enrolling sites and low-enrolling sites (p = 0.93) with rates of 13.5% vs 14.7% at 30 days and 19.0% vs 20.6% at 2 years. In the medical group, the K-M curves differed between high-enrolling sites and low-enrolling sites (p = 0.0005) with rates of 1.8% vs 9.8% at 30 days and 7.3% vs 20.9% at 2 years. The percentages of patients who achieved targets for low-density lipoprotein cholesterol and systolic blood pressure at high- vs low-enrolling sites in both treatment groups combined were 64% vs 49% (p = 0.003) and 70% vs 59% (p = 0.026), respectively. CONCLUSIONS High-enrolling sites in SAMMPRIS achieved better control of primary risk factors and much lower rates of the primary endpoint than low-enrolling sites in the medical group, suggesting that experience with medical management is an important determinant of patient outcome.