Actual failure rates: A method of assessing tissue valve reoperation rates

Background Patients older than age 70 with coronary disease undergoing mitral valve replacement have a limited survival rate, suggesting that few of these patients will require reoperation if they receive a tissue mitral valve. However, traditional actuarial analysis of valve durability censors patients who die before valve failure and doesn’t reflect the lower frequency of reoperation in patient subgroups with limited survival rates. Methods Actual or cumulative reoperation rates represent the rates of reoperation in patients without censoring for death and estimate the risk that the valve will fail before the patient dies. The actual rates of valve rereplacement were determined in 255 recipients of tissue mitral valves, categorizing patients by age and presence of coronary artery disease at 8 years after surgery and compared with standard actuarial estimates of valve durability. Results Operative mortality rates were 17.5% for patients with coronary artery disease and 6.0% for those without (P < .001). At 8 years, the actual reoperation rates for patients younger than age 70 with and without coronary artery disease were 9.2% (n = 76) and 10% (n = 90), respectively. In patients older than age 70 without coronary artery disease the reoperation rate was 9.4% (n = 32). In contrast, the actual reoperation rate was only 1.7% (n = 58) in patients older than age 70 with coronary artery disease. This rate was significantly lower (P = .05) than the other groups. The difference in reoperation rates was not significant if standard actuarial analysis was used. Conclusions Actual reoperation rates are significantly lower in patients older than 70 years with CAD receiving mitral tissue valves than in younger patients or in patients without coronary artery disease. Calculation of actual reoperation rates provides a unique and clinically useful perspective in evaluating data on prosthetic valve reoperation rates. (Am Heart J 1999;138:108-13.)