Abstract
BACKGROUND: The Best Endovascular vs Surgical Therapy in Patients with Chronic Limb-Threatening Ischemia (CLTI) (BEST-CLI) trial compared surgical bypass and endovascular treatment in patients with CLTI. Although center-level variation in vascular surgery outcomes is well-documented, its impact within BEST-CLI has not been explored. Moreover, traditional quality metrics often fail to adequately discriminate center-level performance. This study introduces cumulative, probability-based quality metrics-similar to those employed in professional sports (earned outcomes [EO] and wins above average [WAA])-to evaluate center-level performance in both surgical and endovascular treatment of CLTI. We hypothesized that high performance in both modalities conferred the best overall outcomes among centers.
METHODS: Participating BEST-CLI centers were evaluated by composite major adverse limb events (MALE) or death, for all patients treated at a given site (bypass and endovascular, all BEST-CLI cohorts). WAA was calculated as a risk-adjusted, volume-sensitive measure derived from MALE/death using EO methods. Risk adjustment accounted for patient-level differences using a Cox proportional hazards model, excluding patients with incomplete data. Centers were ranked and divided into WAA quartiles from bottom (Q1) to top (Q4). Patient-level demographics and outcomes were compared across quartiles. Centers were further categorized based on WAA performance: above average (WAA >0) or below average (WAA <0) in bypass, endovascular therapy, or both.
RESULTS: Analyses included 1440 patients (79% of randomized patients) across 146 centers. At 2 years, unadjusted MALE/death rates varied significantly by quartile (bottom Q1, 58%; Q2, 43%; Q3, 33%; top Q4: 30%; P < .001). Centers were evenly distributed based on WAA: both modalities above average (27%), bypass above average only (27%), endovascular above average only (21%), and both below average (25%). Among top centers (Q4), 84% achieved above average outcomes in both modalities, whereas 62% of bottom centers (Q1) were below average in both. Centers excelling in only one modality constituted 16% of top centers (3% bypass above average only, 14% endovascular above average only) and 38% of bottom centers (27% bypass above average only, 11% endovascular above average only).
CONCLUSIONS: MALE/death varied considerably among BEST-CLI centers, with a difference of approximately 30% seen at 2 years between the bottom and top quartiles. Top performing centers consistently achieved above-average outcomes in both bypass and endovascular treatment. Conversely, centers exceling in only one modality were less likely to be top performers. These findings suggest that optimal CLTI care demands proficiency in both bypass and endovascular treatment and highlights the need for quality metrics that better differentiate center-level performance.