Import Non-relational Databases to National Health Insurance Research Database to Analyze the Infusion after Bladder Cancer Surgery

• Main Authors: Yu-Yao Chen, 陳昱堯
• Other Authors: Ming-Chin Lin
• Format: Others
• Language: zh-TW
• Published: 2016
• Online Access: http://ndltd.ncl.edu.tw/handle/3buu62

碩士 === 臺北醫學大學 === 醫學資訊研究所 === 104 === According to Taiwan Cancer Registry database, bladder cancer was the 9th most common cancer in males in 2011: 8.7 in 100,000 males had bladder cancer. The standardized cancer incidence rate shows that bladder cancer was a common cancer in males. Moreover, bladder cancer recurrence rates are high. In Cancer Registry Annual Report 2012 conducted by Health Promotion Administration, Ministry of Health and Welfare, it is shown that bladder cancer ranked 12th in cancer mortality rates and 9th in cancer incidence rates. Bladder cancer recurrence rates are high hence the need for boosting local immunity with postoperative bladder irrigation. Irrigation for treating superficial non-invasive bladder consists of multiple drugs. Treatment includes immunotherapy and chemotherapy. In this study, data is accessed from a NoSQL database: MongoDB. MongoDB operates on BSON, whose data format is binary JSON. ManogoDB, the NoSQL database, is superior in output and input to relational database, which can also scale horizontally. One wishes to calculate data from National Health Insurance Research Database (NHIRD) 2001-2011, which contain millions of samples, with Java, in order to discuss the relationship between drugs and recurrence rates. The focus is on comparing related stratifications and recurrence rates in two types of drugs: chemotherapeutic drugs and immunotherapeutic drugs. This study is inclusive of comparisons between: recurrence rates in two types of drugs, recurrence rates in cross-drug samples stratified by sex and age.