Mitral valve surgery in the US Veterans Administration health system: 10-year outcomes and trends

Objective: To compare mitral valve repair (MVRepair) and mitral valve replacement (MVReplace) trends in the Veterans Affairs (VA) Surgical Quality Improvement Program. Methods: Trends were compared by bivariate analyses, followed by backward stepwise selection and multivariable logistic modeling to determine the effect of preoperative comorbidities and facility-level factors on MVRepair (vs MVReplace) rate. A subgroup analysis focused on patients who underwent elective surgery for isolated primary degenerative mitral regurgitation. Propensity matching was done in the overall and primary degenerative cohorts. Results: From October 2000 to October 2013, 4165 veterans underwent MVRepair (n = 2408) or MVReplace (n = 1757) for MV disease of any cause at 40 VA medical centers (procedural volume, 0-29/y; median 7/y). The MVRepair percentage increased from 48% in 2001 to 63% in 2013 (P < .001). MVRepair rates varied widely among centers; center volume explained only 19% of this variation after adjustment for case mix (R-2 = 0.19, P = .005). Unadjusted 30-day and 1-year mortality rates were lower after MVRepair than after MVReplace (3.5% vs 4.8%, P = .04; 9.8% vs 12.1%, P = .02). Among the propensity-matched patients (n = 2520), 30-day and 1-year mortality were similar after MVRepair and MVReplace. In the propensity-matched primary degenerative subgroup (n = 664), unadjusted long-term mortality for up to 10 years postoperatively was lower after MVRepair (28% vs 37%, P = .003), as was risk-adjusted long-term mortality (hazard ratio, 0.78; 95% confidence interval, 0.61-1.01). Conclusions: In the VA Health System, mortality after MV operations is low. Despite the survival advantage associated with MV repair in primary mitral regurgitation, repair is infrequent at some centers, representing an opportunity for quality improvement.