Wei-Heng Kao1, Ji-Hong Hong1,2,3, Lai-Chu See4, Huang-Ping Yu5, Jun-Te Hsu6, I-Jun Chou7, Wen-Chi Chou8, Meng-Jiun Chiou9, Chun-Chieh Wang1,2, Chang-Fu Kuo9,10. 1. Department of Radiation Oncology, Chang Gung Memorial Hospital, Taoyuan, Taiwan. 2. Department of Medical Imaging and Radiological Science, Chang Gung University, Taoyuan, Taiwan. 3. Radiation Biology Research Center, Institute for Radiological Research, Chang Gung University and Chang Gung Memorial Hospital, Taoyuan, Taiwan. 4. Department of Public Health, College of Medicine, Chang Gung University, Taoyuan, Taiwan. 5. Department of Anaesthesiology, Chang Gung Memorial Hospital, Taoyuan, Taiwan. 6. Department of General Surgery, Chang Gung Memorial Hospital, Taoyuan, Taiwan. 7. Department of Paediatrics, Chang Gung Memorial Hospital, Taoyuan, Taiwan. 8. Department of Haemato-oncology, Chang Gung Memorial Hospital, Taoyuan, Taiwan. 9. Division of Rheumatology, Allergy and Immunology, Chang Gung Memorial Hospital, Taoyuan, Taiwan. 10. Division of Rheumatology, Orthopaedics and Dermatology, School of Medicine, University of Nottingham, Nottingham, UK.
Abstract
PURPOSE: We aimed to evaluate the validity of cancer diagnosis in the National Health Insurance (NHI) database, which has routinely collected the health information of almost the entire Taiwanese population since 1995, compared with the Taiwan National Cancer Registry (NCR). METHODS: There were 26,542,445 active participants registered in the NHI database between 2001 and 2012. National Cancer Registry and NHI database records were compared for cancer diagnosis; date of cancer diagnosis; and 1, 2, and 5 year survival. In addition, the 10 leading causes of cancer deaths in Taiwan were analyzed. RESULTS: There were 908,986 cancer diagnoses in NCR and NHI database and 782,775 (86.1%) in both, with 53,192 (5.9%) in the NHI database only and 73,019 (8.0%) in the NCR only. The positive predictive value of the NHI database cancer diagnoses was 94% for all cancers; the positive predictive value of the 10 specific cancers ranged from 95% (lung cancer) to 82% (cervical cancer). The date of diagnosis in the NHI database was generally delayed by a median of 15 days (interquartile range 8-18) compared with the NCR. The 1, 2, and 5 year survival rates were 71.21%, 60.85%, and 47.44% using the NHI database and were 71.18%, 60.17%, and 46.09% using NCR data. CONCLUSIONS: Recording of cancer diagnoses and survival estimates based on these diagnosis codes in the NHI database are generally consistent with the NCR. Studies using NHI database data must pay careful attention to eligibility and record linkage; use of both sources is recommended.
PURPOSE: We aimed to evaluate the validity of cancer diagnosis in the National Health Insurance (NHI) database, which has routinely collected the health information of almost the entire Taiwanese population since 1995, compared with the Taiwan National Cancer Registry (NCR). METHODS: There were 26,542,445 active participants registered in the NHI database between 2001 and 2012. National Cancer Registry and NHI database records were compared for cancer diagnosis; date of cancer diagnosis; and 1, 2, and 5 year survival. In addition, the 10 leading causes of cancer deaths in Taiwan were analyzed. RESULTS: There were 908,986 cancer diagnoses in NCR and NHI database and 782,775 (86.1%) in both, with 53,192 (5.9%) in the NHI database only and 73,019 (8.0%) in the NCR only. The positive predictive value of the NHI database cancer diagnoses was 94% for all cancers; the positive predictive value of the 10 specific cancers ranged from 95% (lung cancer) to 82% (cervical cancer). The date of diagnosis in the NHI database was generally delayed by a median of 15 days (interquartile range 8-18) compared with the NCR. The 1, 2, and 5 year survival rates were 71.21%, 60.85%, and 47.44% using the NHI database and were 71.18%, 60.17%, and 46.09% using NCR data. CONCLUSIONS: Recording of cancer diagnoses and survival estimates based on these diagnosis codes in the NHI database are generally consistent with the NCR. Studies using NHI database data must pay careful attention to eligibility and record linkage; use of both sources is recommended.