Propensity-score matched comparison between minimally invasive and conventional aortic valve replacement

Abstract

Aim To evaluate the impact of minimally invasive aortic valve replacement (mini-AVR) on clinical outcomes in comparison with the gold standard. Methods We retrospectively reviewed the records of all patients who underwent isolated AVR at the University Hospital Center Zagreb from 2010 to 2020. Patients undergoing mini-AVR were compared with patients undergoing conventional AVR (fs-AVR). The primary outcome measure was blood product consumption. Propensity score matching was used to create a balanced covariate distribution across treatment groups. Additionally, we compared the contemporary outcomes with a historical control. Results The final sample consisted of 1088 patients. In the unmatched cohorts, mini-AVR patients were younger (65 ± 12 vs 68 ± 10 years, P < 0.001) and had lower risk profiles (EuroSCORE2 2.8 ± 2.0 vs 3.5 ± 3.1, P = 0.003). After matching, mini-AVR patients required less blood transfusion than fs-AVR patients (270 [0-790] vs 510 [0-970] mL, P = 0.029). The incidences of stroke, dialysis, new AV block, and mortality were comparable. Cross-clamp times were longer in the mini-AVR group (71 [60-87] vs 66 [53-83] minutes, P = 0.013). Outcomes were improved in the contemporary mini-AVR era compared with our early mini-AVR experience across multiple metrics. Blood product consumption was reduced in the latter tercile of experience (0 [0-520] vs 500 [0-1018] mL, P < 0.001), and the operation was performed more expeditiously (cross-clamp times: 63 [54,80] vs 74 [62,88] minutes, P < 0.001) in comparison with earlier periods. Conclusions We showed that mini-AVR was associated with less blood product requirement than conventional surgery. Our data supports wider adoption of minimally invasive techniques in dedicated centers of excellence.