Retrospective cohort analysis of 926 tricuspid valve surgeries: Clinical and hemodynamic outcomes with propensity score analysis

Background The objectives were to describe morbidity and mortality after tricuspid valve (TV) surgery, to compare outcomes after repair versus replacement, and to assess risk factors for mortality and tricuspid regurgitation (TR) recurrence. Methods A retrospective cohort study including 926 consecutive cases of TV surgery (792 repairs and 134 replacements) performed at the Montreal Heart Institute was conducted. Median follow-up was 4.3 years (4,657 patient-years). Median age was 62 years (interquartile range 53-69 years), and 72% of patients were women. Results Operative mortality was 14% (128 patients: 1977-1998 20%, 1999-2008 7%, P < .001). Independent risk factors for operative mortality in the 1999 to 2008 period were hypertension (odds ratio [OR] 6.03, P = .02), daily furosemide dose (by 10 mg) (OR 1.06, P = .05), weight (by 10 kg) (OR 0.36, P < .01), and cardiopulmonary bypass time (by 10 minutes) (OR 1.29, P < .001). Ten-year survival was 49% +/- 2% and 38 +/- 5% in the repair and replacement groups, respectively (P = .012). At discharge, severity of TR was >= 3/4 in 13% and 2% of patients in the repair and replacement groups, respectively (P = .01). Propensity score analysis showed that tricuspid repair was associated with higher rates of TR >= 3/4 at follow-up compared with replacement (hazard ratio 2.15, P = .02). Forty-eight reoperations (7% of patients at risk) were performed during follow-up (repair group, 6%; replacement group, 15%; P = .01). At last follow-up, New York Heart Association functional class was improved compared with baseline in both groups (P < .001). Conclusion Tricuspid valve surgery is associated with substantial early and late mortalities but with significant functional improvement. Replacement is more effective in early and late corrections of regurgitation, but it does not translate into better survival outcomes. (Am Heart J 2012;163:851-858.e1.)