Bevacizumab plus erlotinib prolonged patients＇ progression-free survive （PFS） versus bevacizumab alone for the maintenance treatment of none-small cell lung cancer （NSCLC） in phase III clinical trial ATLAS （ClinicalTrials. gov identifier NCT00257608）, which repealed a benefit outcome and acceptable side-effects, but whether its cost performance would be accepted by patients is blurry. The aim of our research is to figure out which strategy is the best option in clinic and would spread broadly. Markov Model was used to calculate incremental cost-utility radios （ICURs） and 10-year quality-adjusted life years （QALY） of both strategies. The clinical data were collected from phase III clinical trial ATLAS （ClinicalTrials. gov identifier NCT00257608）. The cost data were obtained from Chinese health care system. In the research, one-way sensitivity analysis, probabilistic sensitivity analysis （PSA） and Monte-Carlo analysis were performed to test the stability of the results. The better strategy was bevacizumab alone strategy, and the cumulative costs of both strategies were $178 648.47 and $46 445.28, respectively, and the QALY was 12.506 and 10.643, respectively. The ICUR of combined application was $70 962.53/QALY, which was much higher than 3 times of mean gross domestic product （GDP） in China, suggesting that this strategy was no economical at all. In one-way analysis, the change of willingness-to-pay could not influence the consequence. In addition, in Monte-Carlo analysis, the probability distribution of cost, effectiveness and ICUR was in normal distribution. Taken together, bevacizumab alone strategy was the better strategy in terms of cost-effectiveness.
