Takekuma Yoh, Imai Shungo, Sugawara Mitsuru
Department of Pharmacy, Hokkaido University Hospital.
Faculty of Pharmaceutical Sciences, Hokkaido University.
Yakugaku Zasshi. 2022;142(4):331-336. doi: 10.1248/yakushi.21-00178-3.
The JMDC Claims Database contains completely anonymized receipt information on the insured members of health insurance associations. The number of registered users is approximately 9.6 million (6% of the population) as of May 2020. In this database, it is possible to track even outpatient treatment, even if the patient changes the medical facility, as long as the insurer of the subscriber's health insurance does not change, so that long-term medical treatment could be targeted as a research theme. However, as the data do not contain medical record information, it is not possible to obtain laboratory values, although it is possible to know whether clinical tests have been performed. For pharmaceutics-related research, the most suitable use of the receipt database like JMDC Claims Database seems to be the investigation of actual prescriptions. However, the research topics that pharmacists are interested in are probably comparisons of drug effects, drug-drug interactions, or causal analysis of drugs and side effects. However, laboratory data for evaluating drug efficacy is not available in the receipt database, and the accuracy of the disease name in the database becomes problematic when using the disease name as information indicating the occurrence of side effects. In this review, we introduce our studies performed by using JMDC Claims Database and how to manage the above-described problems. We hope that this study will be helpful to those who are going to engage in research using medical big data.
JMDC理赔数据库包含健康保险协会参保成员完全匿名的报销信息。截至2020年5月，注册用户数量约为960万（占人口的6%）。在这个数据库中，即使患者更换了医疗机构，只要参保人的健康保险承保公司不变，就可以追踪门诊治疗情况，因此长期治疗可以作为一个研究主题。然而，由于数据不包含病历信息，虽然可以知道是否进行了临床检查，但无法获得实验室检查值。对于与药学相关的研究，像JMDC理赔数据库这样的报销数据库最适合用于实际处方的调查。然而，药剂师感兴趣的研究主题可能是药物疗效比较、药物相互作用或药物与副作用的因果分析。然而，报销数据库中没有用于评估药物疗效的实验室数据，并且当将疾病名称用作表明副作用发生的信息时，数据库中疾病名称的准确性就会出现问题。在本综述中，我们介绍了使用JMDC理赔数据库进行的研究以及如何处理上述问题。我们希望这项研究对那些打算利用医疗大数据进行研究的人有所帮助。
