Low risk of thrombosis and serious embolic events despite low-intensity anticoagulation. Experience with 1,004 Medtronic Hall valves.

: One thousand four Medtronic Hall valves (601 mitral, 398 aortic, and five tricuspid) were implanted in 847 patients between December 1979 and June 1987. Total experience at the end of June 1987 was 2,640 patient-years of follow-up. Prothrombin time ratios were reviewed for all patients (16,866 observations), and these ratios were found to be therapeutically low (median international normalized ratio, 2.6) and highly variable (lower 10th percentile, 1.6; upper 10th percentile, 3.9). During the follow-up period, there were no valvular thromboses; the 95% confidence limit for the risk of thrombosis (0.14 per 100 patient-years) is below that reported for other mechanical prostheses. Sixty percent of all thromboembolic events left no residual deficit, and 75% of all bleeding events did not require treatment. Only 11% of thromboembolic events and 7% of bleeding events were fatal. The linearized rate of fatal bleeding was 0.2% per year, and the linearized rate of moderate to severe bleeding was 0.5% per year. Five-year actuarial embolic-free rates were 92%, 84%, and 83% for aortic, mitral, and double valves, respectively. The low risk of valvular thrombosis and of serious thromboembolic events in the Medtronic Hall valve, regardless of the range and variability of anticoagulation, offers greater patient safety than other mechanical prostheses, provides a credible alternative to bioprostheses, and may be particularly relevant to third-world populations.