OBJECTIVES: To assess the feasibility of a probabilistic discrete event simulation (DES) in obesity and to validate the model results against model results from past technology appraisals (TA) from the National Institute for Health and Care Excellence (NICE).

METHODS: Systematic literature review was performed to inform model concept. Relevant disease states and model events were retrieved and implemented in a DES using R. Baseline and modifiable risk factors for patient profiles were based on published literature. Modifiable risk factor trajectories were modelled through risk equations. Development of diabetes was estimated based on the Qdiabetes-2018 Risk Model. Natural history of BMI was estimated based on equations from a study in the UK general practice research database. The Qrisk3 Risk Model and the Framingham Recurring Coronary Heart Disease Risk Model were used to estimate the risk of primary and secondary cardiovascular events, respectively.

RESULTS: 100 patients were simulated for base case feasibility analysis of diet and exercise management in obesity. Model events included onset of T2D, sleep apnoea, cardiovascular disease, knee replacement and all-cause mortality. Clinical events were comparable with TA 494. Modelled QALYs were 16.055 and were comparable to results from TA 664 (QALYs: 15.216) and TA 494 (QALYs: 15.134).

CONCLUSIONS: Probabilistic DES for obesity is feasible with the R framework. Model runtime for 100 patients was less than 1 minute and this modelling approach will allow running large cohorts with sensitivity analysis and acceptable runtime for NICE TA. Simulating larger cohorts and validating economic outcomes against data from published models is warranted.