178 | Chapter 7 After each replacement because of leakage, the moment of replacement for the next VP will be calculated as the DL70 on the then last three VPs. When the current VP reaches this moment and hence would be replaced prophylactically, the next moment of replacement will be the DL70 based on the last three VPs increased with 17% to correct for the fact that the actual lifetime of the last VP is longer than its in situ time. The value of 17% of this correction factor has been found by trial and error. Problemsolving VPs are not prophylactically replaced because of their longevity. Monte-Carlo simulations Since VP replacement data is sparse, the possibilities for leakage reduction by simulating device lifetimes through Monte-Carlo (MC) simulations was investigated. The main unknown factor in formulating an optimal device change policy is the probability distribution of the in-vivo device lifetimes. These MC-simulations are used to investigate the relationship between variability in device lifetime (inter- and intrapatients) and the limits of PVPR. The relevant parameter of this unknown probability distribution is the Coefficient of Variation (CV = Standard Deviation/Mean). The MC simulations model the average time between leakage events, relative to the default policy and the associated number of device replacements as a function of the simulated CV. The simulations are repeated for four probability distributions. See supplemental information for details. Ethical considerations This study does not fall under the scope of the Medical Research Involving Human Subjects Act and was approved by the review board of the Netherlands Cancer Institute (IRBd21-092).
RkJQdWJsaXNoZXIy MTk4NDMw