[Priority setting in scaled-up cancer screening in China: an systematic review of economic evaluation evidences]

Objective: The existed economic evaluations of cancer screening in Chinese population are almost all single-cancer focused, evidence on parallel comparison among multiple cancers is lacking. Thus, the aim of this study was, from a priority setting perspective, to compare the cost-effectiveness of six common cancers(colorectal cancer, breast cancer, liver cancer, lung cancer, esophageal cancer and stomach cancer) to facilitate policy making in future scaled-up screening in populations in China. Methods: Partially based on our previous single-cancer systematic reviews (colorectal cancer, breast cancer, liver cancer, and lung cancer), evidence of economic evaluations of cancer screening in populations in mainland China were systematically updated and integrated. The main updates include: 1) Stomach cancer and esophageal cancer were newly added to the current analysis. 2) The literature searching was extended to 8 literature databases, including PubMed, EMbase, The Cochrane Library, Web of Science, CBM, CNKI, Wanfang Data, and VIP. 3) The period of publication year was updated to the recent 10 years: January 1, 2009 to December 31, 2018. 4) The study focused on populations in mainland China. Following the standard processes of literature searching, inclusion and exclusion from previous systematic reviews, the basic characteristics, evaluation indicators and main results of the included studies were extracted. All the costs were discounted to 2017 value using the by-year consumer price index of medical and health care residents in China and presented in the Chinese Yuan (CNY). The ratios of incremental cost-effectiveness ratio (ICER) to China's per capita GDP in 2017 were calculated (<1 means very cost-effective, 1-3 means cost-effective, >3 means not cost-effective). Given a specific indicator, the median value among all reported screening strategies for each cancer was calculated, based on which priority ranking was then conducted among all cancers when data available. Results: A total of 45 studies were included, 22 for breast cancer, 12 for colorectal cancer, 6 for stomach cancer, 4 for esophageal cancer (all conducted in high-risk areas), 1 for liver cancer and none for lung cancer (was not then considered for next ranking due to limited numbers of studies). When based on the indicator, the median ratio of cost per life-year saved to China's per capita GDP (reported in 12 studies), the lowest ratio (-0.015) was observed in esophageal cancer among 16 strategies of 2 studies (N=2, n=16), followed by 0.297 for colorectal cancer (N=3, n=12), 0.356 for stomach cancer (N=1, n=4) and 0.896 for breast cancer (N=6, n=52, P(75)=3.602). When based on another commonly used ICER indicator, the median ratio of cost per quality-adjusted life-year gained to China's per capita GDP (reported in 13 studies), the least cost was found in stomach cancer (0.495, N=3, n=8, P(75)=3.126), followed by esophageal cancer (0.960, N=1, n=4, P(75)=1.762) and breast cancer (2.056, N=9, n=64, P(75)=4.217). Data was not found for colorectal cancer. In addition, cost per cancer case detected was the most adopted indicator (32 studies). The median cost among all screening strategies for each cancer was 14 759 CNY for stomach cancer (N=5, n=7), 49 680 CNY for colorectal cancer (N=12, n=25) and 171 930 CNY for breast cancer (N=13, n=24), respectively. Data was not available for esophageal cancer and rare for precancer cases detected. Evidence related to cost per disability-adjusted life-year gained was not available. Conclusions: At China's national level and limited to the six cancers covered by the current study, the preliminary analysis suggests that stomach cancer and colorectal cancer were the most cost-effective target cancers and could be given priority in the future scaled-up screening in general populations. Esophageal cancer screening should be prioritized in high-risk areas. Breast cancer was also cost-effective in general but some of the intensive screening strategies were marginal. Data on liver cancer and lung cancer were too limited to conclude, and more well-designed studies and high-quality research evidence should be required. This priority ranking might be changed if other common cancers were involved analyses.