Pmd1 Analysis of Decision Models in Total Knee Replacement for End-Stage Knee Osteoarthritis: Initial Lessons for Conducting a Cost-Effectiveness Analysis in the Chilean Context

Abstract

Due to increasing statistics in obesity, sedentarism, population aging in Chile, it becomes necessary to conduct local cost-effectiveness analysis (CEA) for those technologies highly demanded in this scenario. Total Knee Replacement (TKR) for End-Stage Knee Osteoarthritis (ESKO) is not a prioritized procedure and waiting-lists are frequent. Given the demand for TKR and the need of a local CEA, TKR Decision Models (DM) were analyzed. Methods: A Scoping Review was conducted to map DM related to CEA in TKR. Population, Concept and Context strategy was followed. Search was carried out in Medline, ScienceDirect, Cochrane, York University and ISPOR database, Pan-Canadian website, and list of references. A sample of CHEERS parameters was analyzed per each DM, particularly: model type, horizon, sub-groups, comparators, health states, outcomes, and uncertainty. Results: N=2.517 articles were retrieved. Finally, five articles fulfilled inclusion/exclusion criteria. All models were Markovs: Lifetime horizons were more frequent (60%). Three DM included sub-groups starting in 60 years old. Just one of the models considers the Health State of ESKO explicitly. Revision and Re-revision are frequent Health States, mostly when TKR is performed under 65 years old (device-survival related factor). Models apply the same deathprobabilities of general population to the same age groups of ESKO without TKR despite disability deteriorates death rates. One interesting sensitivity analysis found is by hospital volume, indicating more volume improves ICER. Conclusions: Lessons for conducting a local CEA for TKR in ESKO regard the use of Markov Models; Long-terms horizons due to device-survival; Identification of sub-groups above 65 years-old as current local practice does and due to its effect on Revision rates; Identification of specific death rates for ESKO without TKR; And including ESKO Health State explicitly. Conducting sensitivity analysis per hospital volumes is interesting in medical devises CEA as these depend upon the growing experience of providers.