Conservative operation for mitral stenosis. Independent determinants of late results.

A retrospective study was designed to define the independent determinants of late results in 282 consecutive patients operated on for mitral stenosis between 1978 and 1985. A total of 25 variables were investigated by multivariate discriminant analysis for their possible influence on symptomatic outcome, subsequent reoperation, postoperative thromboembolism, and death. All the patients underwent a conservative mitral valve operation. In addition, 25 patients required concomitant tricuspid annuloplasty. The surgical mortality rate was 1.4% and the late mortality, reoperation, and thromboembolism rates were, respectively, 0.14%, 0.80%, and 0.95%/pt-yr. Actuarial probability of complication-free survival at 8 years was 83% +/- 3.7%. Left atrial size on the M-mode echocardiograms, tricuspid annuloplasty, mitral annuloplasty, presence of left atrial thrombus, male gender, and separation of subvalvular apparatus portended a significantly higher likelihood of poor postoperative symptomatic status (p less than 0.0005). Preoperative mild mitral regurgitation, E to F slope on the M-mode echocardiogram, and tricuspid annuloplasty had significant influence on the need for late reoperation (p less than 0.05). Postoperative atrial fibrillation, mitral valve amplitude on the M-mode echocardiogram, preoperative embolism, and residual mitral incompetence had an independent predictive power for postoperative thromboembolism (p less than 0.05). Finally, preoperative atrial fibrillation, preoperative cardiothoracic ratio, postoperative left atrial size, and postoperative atrial fibrillation all influenced the probability of long-term complications. These results suggest that earlier operation should be considered in patients with mitral stenosis, to increase the complication-free survival rate.