PMS19 Validation of a Cost-Effectiveness MODEL in Rheumatoid Arthritis

Abstract

Rheumatoid arthritis (RA) has multiple previous health technology assessment (HTA) appraisals for advanced therapies. Cost-effectiveness models (CEMs) in RA are complex, especially with the increasing use of patient-level simulation, therefore, replicating them is an important step prior to HTA submission. Validation can ensure that models provide consistent representation of the disease and intervention, and a solid basis for decision-making. The objective of this study was to validate our patient-level CEM against a published model that was developed by academic experts in previous NICE appraisals. A patient-level, discrete-event simulation model was developed to evaluate the cost-effectiveness of new treatments for moderate-to-severe RA, following an approach developed by Stevenson et al.1 which estimated cost and health benefits using long-term, non-linear health-assessment questionnaire disability index trajectories. Our model was validated against a cost-effectiveness analysis of sequences of biologics (abatacept, adalimumab, certolizumab pegol, etanercept, golimumab, infliximab, tocilizumab) for RA patients in England published by Stevenson et al. in 2017.2 It was possible to replicate the analysis of Stevenson et al.2 2017 in most respects. Stevenson et al.2 reported lifetime incremental QALY gains of 1.5-2.0 and incremental costs of £60,000-£100,000 for the seven drugs. Our model predicted incremental QALYs of 1.54-1.71 and incremental costs of £69,000-£102,000. Estimated ICERs were £48,800-£52,300 and £44,700-£66,700 for Stevenson et al.2 and our model, respectively. The upper ICERs from our model were considered overestimates due to using list prices where patient access scheme prices were confidential. The results of the model validation demonstrate that it is possible to successfully replicate a complex, patient-level model. This illustrates the importance of validation when developing models for HTA to ensure consistency with previous assessments and provide a reliable basis for decision-making. REFERENCES 1. Stevenson M et al. Health Technology Assessment. 2016;20(35):1-610. 2. Stevenson M et al. J Rheumatol. 2017;44(7):973-80.