ABSTRACT
Bioprosthetic aortic valves degenerate over time, and differences between brands could be expected. We compared 2 brands implanted in 3 different centers serving 3.3 million people. Between 2000 and 2014, we identified 1241 bioprosthetic aortic valve replacements using Mitroflow (Sorin, Milan, Italy) and 3212 using Perimount (Edwards Lifesciences, Irvine, CA) covering 88% of all aortic valve replacements in the region. Average differences in t-year mortality were derived from Cox regression. The complete case analyses included 881 Mitroflow replacements and 2488 Perimount replacements. The median follow-up time and 25/75 percentiles were 5.0 years (3.3-7.2) and 8.4 years (5.1-10.6) for Perimount and Mitroflow, respectively. Multiple Cox regression analyses demonstrated significantly higher mortality with Mitroflow valves compared with Perimount (hazard ratio 1.27; 95% CI: 1.1-1.5; P < 0.001). Average risk of death within 5 years was 25.0% with Mitroflow and 20.4% with Perimount. Average difference in 5-year mortality based on Cox regression was 4.60% in favor of Perimount (95% CI: 1.02-8.02%; Pâ¯=â¯0.01) and the number needed to harm was 21.9 (95% CI: 12.7-80.5) within 5 years. Propensity matching confirmed 2-year survival differences 4.6% in favor of Perimount (95% CI: 1.2-7.9%; Pâ¯=â¯0.004), and further confirmed in a series of subgroups and a double robust analysis that takes into account both propensity for treatment and covariate relation to outcome. Mitroflow valves were associated with a significantly increased risk of death when compared to Perimount valves.
