The ultimate prognosis after valve replacement: an assessment at twenty years.

Our experience over a 20-year period consists of 2,135 patients with initial caged-ball valve replacement: 52% aortic, 34% mitral, 12% double, and 2% triple-valve replacements, with 59.2, 39.8, 10.3, and 2.7 patient-centuries of follow-up, respectively. Fifteen-year actuarial survival (+/- standard error) was 43 +/- 2% for aortic and 44 +/- 3% for mitral valve replacement, and 27 +/- 5% for double-valve and 23 +/- 7% for triple-valve replacement. Restricting attention to patients operated on since 1973 divides the series almost in half and does not dramatically improve the 5-year actuarial survival (from 66 +/- 2% to 71 +/- 3% and from 70 +/- 2% to 78 +/- 3% for aortic valve replacement and mitral valve replacement, respectively). There was some alteration in the causes of late death: the largest percentage of deaths in both the earlier and current groups, 52%, was cardiac related whereas only 24% and 13%, respectively, were valve related. Over the past two decades operative mortality has declined and, to a lesser extent, late survival after mitral valve replacement has improved. The incidence of embolism has decreased significantly, most notably with the Silastic ball valves. Dramatic improvements in late results will occur primarily by modifying the cardiac-related death rate through earlier operation and improvements in the medical management of postoperative arrhythmias and congestive heart failure.