Defining centres of expertise for minimally invasive mitral valve surgery: a systematic review and volume–outcome meta-analysis

Background Minimally invasive mitral valve surgery (MIMVS) is increasingly performed, but outcomes such as repair rate, mortality and survival likely depend on expertise. Still, the definition of a high-volume centre varies in the literature and lacks an evidence-based substantiation. Consequently, this study aims to determine the volume–outcome relation in MIMVS in conjunction with a volume threshold, in order to define ‘high-volume centres’, applying a novel statistical concept. Methods The study was preregistered in PROSPERO (CRD42022376293, registered 26 November 2022). A systematic search was applied to three databases, including consecutive patients undergoing MIMVS. Studies describing patients undergoing transcatheter procedures were excluded. Restricted cubic spline analyses were applied and the elbow method was used to retrieve the threshold volume. Long-term outcomes were analysed using reconstructed Kaplan-Meier curves and a novel statistical concept to assess the volume–outcome relation for time-to-event outcomes was applied. The primary outcome was early mortality, secondary outcomes were repair rate, stroke, and long-term survival, freedom from reoperation, and freedom from more than moderate mitral regurgitation. Leave-one-out analyses were performed for sensitivity purposes. Results Data from 68 unique centres were included (n=23 495 patients). Early mortality was 1.3% (95% CI 1.1% to 1.6%), without a statistically significant non-linear relation for this endpoint, nor for stroke. There was a statistically significant volume–outcome relation for mitral valve repair rate (p=0.018). Based on the repair rate, the threshold to define a high-volume centre was 60 cases/year (number needed to treat to prevent a replacement ≤7). A significant volume–outcome relation was observed for long-term outcomes as well, with a threshold of 53 and 54 cases/year for long-term survival and freedom from reoperation, respectively. These results were robust across the sensitivity analyses for the various endpoints. Conclusions The threshold to define a high-volume centre ranges between 53 and 60 cases/year based on repair rate, long-term survival and freedom-from reoperation. These findings have the potential to facilitate centralisation of MIMVS. PROSPERO registration number CRD42022376293.