Prophylactic Replacement of Voice Prosthesis | 187 7 Supplemental information Monte-Carlo simulations of optimal time-to-change for voice prostheses Monte-Carlo simulations of prophylactic voice prosthesis changes were performed with variable expected device lifetimes and variability in lifetimes. Simulations were done for different prophylactic replacement policies, comparing no-prophylactic device change (wait-to-leakage), to theoretical optimal replacement, adaptive replacement, initial best estimate, and fixed time replacement policies. Each simulated condition was run for 5000 simulated patients and 32 device changes. To investigate the influence of the Coefficient of Variation on expected time-to-leakage events and number of device changes in the ideal case of known probability distributions, time-to-leakage was simulated for 1 million device changes per data point. Results show that for the high variability in lifetimes found in the investigated patient population, Coefficientof-Variation (Standard Dev. / Mean) > 0.8, no prophylactic device change policy will be effective in reducing the number of leakage events, even in the ideal case where the expected device lifetimes and Coefficient of Variations of individual patients are known. The supplemental information contains the R and Rmarkdown code used in the simulations and the generation of the figures and text.
RkJQdWJsaXNoZXIy MTk4NDMw