Background: Mitral valve repair (MVR) has been shown to achieve good long-term results. However, this procedure is associated with relevant immediate postoperative mortality. The aim of this study is to identify those preoperative variables associated with an increased risk of 30-d postoperative death.
Methods: One hundred and sixty-four patients underwent MVR at our institution from January 1993 to December 2000.
Results: Eleven patients (6.7%) died during the immediate postoperative outcome, a median of 14 d after surgery (range, 1-29 d). One patient (1.3%) out of 80 who underwent MVR as lone procedure died on postoperative day 14 of cardiac tamponade. The mortality rate in those who underwent MVR associated with other procedures was 11.9%. Multivariable analysis (154 patients included in the analysis) showed that patients' age (p = 0.006, for an increase of 10 units: OR 4.33, 95% CI 1.53-12.27), history of prior cardiac surgery (p = 0.006, OR 118.56, 95% CI 4.03-3491.14) and NYHA functional class (p = 0.011, OR 5.66, 95% CI 1.49-21.49) were significantly associated with an increased risk of postoperative death. The receiver operating characteristics (ROC) curve showed that patients' age had an area under the curve of 0.762 (95% CI 0.622-0.901, p = 0.004), its best cut-off value being 65 years (mortality, 13.4% vs 2.1%, p = 0.008, sensitivity 81.8%, specificity 62.1%, accuracy 63.4%). None of the patients older than 65 and with a history of prior cardiac surgery survived the operation.
Conclusions: MVR is associated with a relevant 30-d mortality risk in patients older than 65 years, with advanced NYHA functional class and a history of prior cardiac surgery.