腎臟與膀胱癌與其他癌症的關聯性分析 – 健保資料庫研究
盧誌明1、高以信2、江瑞坤3
佛教慈濟醫療財團法人大林慈濟醫院1泌尿科、3家醫科；2台南市立醫院家醫科
An association analysis to test the classification for cancers of kidney and bladder using National Health Insurance Research Database
Chih-Ming Lu1, Yee-Hsin Kao2, Jui-Kun Chiang3
Departments of 1Urology and 3Family Medicine, Dalin Tzu Chi Hospital, Buddhist Tzu Chi Medical Foundation , Chiayi, Taiwan; 2Department of Family Medicine, Tainan Municipal Hospital, Tainan, Taiwan
Purpose: The aim of this study was to explore the association among different classifications for patients with kidney and bladder cancer using Taiwan’s NHIRD.
Materials and Methods: All patients with diagnosis of cancers were stratified into 12 classifications searched from NHIRD. We tested the association between different classifications with and without kidney and urinary bladder cancer by Apriori test.
Results: A total of 12923 persons were enrolled for analyses. All persons with cancers were classified into 12 strategies (11 unique cancer groups and the other group). The most 3 unique cancers were liver cancer (2836, 21.9%), lung cancer (2808, 21.7%) and colon cancer (1741, 13.5%). A total of 2901(22.4%) patients composed the other group. A positive correlation (0.412) between kidney and urinary bladder cancer group and the other group was found. By association analysis, there was a higher lift (4.37) from the kidney and urinary bladder cancer group to the other group and vice versa. It indicated that these two cancers groups had occurred simultaneously on the same individuals more than expected. There was no association after we reclassify kidney and urinary bladder cancer group.
Conclusion: Complex repeated classification originated from big data such as Taiwan’s NHIRD could be detected by association analysis. This technique enhanced more correct grouping for researchers focusing on all cancer-related studies.