A tri-institutional comparison of tissue and mechanical values using a patient-oriented definition of "treatment failure".

Selection of valve type for predominant usage is obscured by limiting the analysis to prosthesis-related rather than patient-oriented failure modes. In this report, "treatment failure" is defined as a valve-related death or permanent patient disability; successful reoperations are excluded, and emboli with permanent residua are included. Results with the Starr-Edwards Silastic ball valve (Oregon) and the Hancock (Stanford) and Carpentier-Edwards (Vancouver) porcine valves are compared using this new definition of treatment failure. Evaluated according to structural failure, the mechanical valve is superior to the tissue valve, and using the Stanford definition of valve failure, it becomes so between 5 and 10 years. Using treatment failure, tissue valves are superior at 5 years; at 10 years in the aortic position, the results are comparable; and in the mitral position at 8 to 10 years, tissue valves show a continued but small advantage.