A Comparison of Markov and Discrete-Time Microsimulation Approaches: ă Simulating the Avoidance of Alcohol-Attributable Harmful Events from ă Reduction of Alcohol Consumption Through Treatment of Alcohol Dependence

When modelling the pathophysiology of a disease, it is important to ă select a modelling approach that can adequately replicate its course. ă The objective of this paper was to compare the outcomes obtained by the ă Markov and discrete-time microsimulation modelling approaches using ă nalmefene clinical trial data. ă Markov and microsimulation modelling approaches assessing alcohol ă dependence treatment with psychosocial support with or without nalmefene ă were compared in terms of the modelled evolution of patients' alcohol ă consumption and the resulting occurrence of alcohol-attributable harmful ă events over 1 year. ă Comparison of the proportion of the modelled population at different ă levels of alcohol consumption over time revealed systematic differences ă arising from the different modelling techniques: a lower number of ă patients reaching abstinence, a higher number of patients at higher ă drinking levels, and, overall, a smoother evolution of alcohol ă consumption in the microsimulation. Reasons are discussed in the paper. ă While the models produced similar occurrences of alcohol-attributable ă harmful events as a whole, distinct results for the individual events ă were observed, explained by the specific pathophysiology of occurrence ă of these events and how their implementation was adapted to fit the ă limitations of the compared modelling approaches; however, these ă differences were only statistically significant for one of the eight ă events. ă For a general public health or health economic assessment of alcohol use ă disorders, it is possible to achieve similar results with the compared ă approaches. To assess a patients' disease course, taking into ă consideration alcohol-attributable harmful events, the microsimulation ă approach may provide more precise results. However, further external ă validation of the models is needed and this additional precision may be ă outweighed by the greater computational burden of a microsimulation ă approach.