Chao Shi1, Mengfei Liu1, Zhen Liu1, Chuanhai Guo1, Fenglei Li2, Ruiping Xu3, Fangfang Liu1, Ying Liu1, Jingjing Li1, Hong Cai1, Zhonghu He4, Yang Ke5. 1. Key Laboratory of Carcinogenesis and Translational Research (Ministry of Education/Beijing), Laboratory of Genetics, Peking University Cancer Hospital & Institute, #52 Fucheng Road, Beijing, People's Republic of China. 2. Hua County People's Hospital, Wenming Road, Hua County, Henan Province, People's Republic of China. 3. Anyang Cancer Hospital, #1 Hengbin North Road, Anyang City, Henan Province, People's Republic of China. 4. Key Laboratory of Carcinogenesis and Translational Research (Ministry of Education/Beijing), Laboratory of Genetics, Peking University Cancer Hospital & Institute, #52 Fucheng Road, Beijing, People's Republic of China. Electronic address: zhonghuhe@foxmail.com. 5. Key Laboratory of Carcinogenesis and Translational Research (Ministry of Education/Beijing), Laboratory of Genetics, Peking University Cancer Hospital & Institute, #52 Fucheng Road, Beijing, People's Republic of China. Electronic address: keyang@bjmu.edu.cn.
Abstract
OBJECTIVES: The objective of this study was to establish an optimal population-level follow-up strategy for identifying incident cancers using health insurance reimbursement data in rural China. STUDY DESIGN AND SETTING: We compared active follow-up and passive linkage with claims data for identification of incident cancer cases. Claims data were derived from the New Rural Cooperative Medical Scheme (NCMS). Follow-up data from subject enrollment to December 31, 2016, regarding 33,948 subjects in a large-scale randomized controlled trial were used in this study. RESULTS: The overall sensitivity of passive linkage with NCMS claims data was significantly higher than that of active follow-up (95.6% vs. 54.9%, P < 0.001). Of 12 cases missed by the NCMS data set, seven were treated on an outpatient basis and there were therefore no records in the NCMS system, and five were diagnosed at primary (township-level) health facilities and excluded from the quality control process. Of the 123 cases missed by active follow-up, 54 were reported as negative, 69 were reported as positive but had inaccurate information regarding the site of cancer, or exceeded the 6-month limitation from the date of diagnosis. CONCLUSION: Passive linkage with NCMS claims data is an efficient approach for identifying incident cancers in areas without cancer registries in rural China.
OBJECTIVES: The objective of this study was to establish an optimal population-level follow-up strategy for identifying incident cancers using health insurance reimbursement data in rural China. STUDY DESIGN AND SETTING: We compared active follow-up and passive linkage with claims data for identification of incident cancer cases. Claims data were derived from the New Rural Cooperative Medical Scheme (NCMS). Follow-up data from subject enrollment to December 31, 2016, regarding 33,948 subjects in a large-scale randomized controlled trial were used in this study. RESULTS: The overall sensitivity of passive linkage with NCMS claims data was significantly higher than that of active follow-up (95.6% vs. 54.9%, P < 0.001). Of 12 cases missed by the NCMS data set, seven were treated on an outpatient basis and there were therefore no records in the NCMS system, and five were diagnosed at primary (township-level) health facilities and excluded from the quality control process. Of the 123 cases missed by active follow-up, 54 were reported as negative, 69 were reported as positive but had inaccurate information regarding the site of cancer, or exceeded the 6-month limitation from the date of diagnosis. CONCLUSION: Passive linkage with NCMS claims data is an efficient approach for identifying incident cancers in areas without cancer registries in rural China.