A Comparison Of Markov Cohort And Discrete Event Simulation Models In Cost-Effectiveness Analysis Of Sorafenib And Everolimus In 3rd Line Metastatic Renal-Cell Carcinoma In The Czech Republic

OBJECTIVES: Markov cohort (MC) and discrete-event simulation (DES) models are inherently different. Therefore, they are rarely used in economic evaluation of the same disease. Sorafenib and everolimus are cornerstones of current metastatic renal-cell carcinoma (mRCC) treatment in 2nd line, however, there are limited effectiveness/efficacy and cost-effectiveness studies in 3rd line. The first objective was to compare MC and DES models in oncology, and the second to assess the cost-effectiveness of sorafenib versus everolimus in 3rd line mRCC. METHODS: We developed two mirror life-time cost-utility models using a) Markovian and b) DES approaches, which projected quality-adjusted life-years (QALYs) and costs from healthcare payers' perspective. In MC, we used weekly cycle length and three states, i.e. progression-free, progression, death. In DES, there were progression/death events instead of states. Transition probabilities, utilities and costs were derived from published literature/sources. Costs and outcomes were discounted by 3%. Probabilistic sensitivity analysis (PSA; 10,000 simulations) was performed with Czech willingness-to-pay threshold (WTP) equal to 645,000. RESULTS: Over a lifetime horizon, sorafenib is less costly but also slightly less effective than everolimus. In MC, sorafenib is less costly by €2,045 (€11,558 vs. €13,603) and slightly less effective by 0.0028 QALYs (0.7815 vs. 0.7843). In DES, sorafenib is less costly by €2,320 (€11,326 vs. €13,646) and slightly less effective by 0.0027 QALYs (0.7927 vs. 0.7954). The ICERs, expressed as savings per QALY lost, are equal to €792,646 (MC) and €874,585 (DES). The results of PSA showed that sorafenib is cost-effective with probability of 95% (MC) and 85% (DES) at the WTR CONCLUSIONS: Despite their differences, MC and DES models yields almost identical results in simple oncologic model. The slight disparity might be due to computational differences, half-cycle correction or cycle length. Finally, sorafenib clearly proved that it is a cost-effective intervention in 3rd line therapy of mRCC.