Big data provided by National Health Insurance Service(NHIS) and Health Insurance Review & Assessment Service (HIRA) for academic research have limitations because of the lack of detailed medical information and other reasons, an expert says.
Professor Shin Soo-yong 신수용 of Kyung Hee University made these and other allegations in his paper titled “Current state of healthcare big data in Korea and the new direction of research,” which was published in the recent issue of Electronic & Information Research Information Center (EIRIC)전자정보연구정보센터.
|Professor Shin Soo-yong of Kyung Hee University|
“There are much data opened for research in Korea, but most of them are medical checkup results or those used for making insurance claims,” Shin said. “They show limitations of lacking in detailed healthcare information, including the results of test results, for the research of specific diseases.”
NHIS makes public 1 million sample cohort DB and other data for researchers after removing personal information. The data includes sample people’s social and economic variables, such as residences, dates of death, reasons for death, and income levels, as well as their treatment records and medical checkup data.
HIRA is also providing part of its data for researchers through healthcare big data opening system.
Professor Shin said there are limitations in the data provided by NHIS and HIRA, the two healthcare big data which represent the government’s information opening policy.
“It is true the medical community has released many research papers using the sample cohort DB,” Professor Shin said. “More often than not, however, the data for insurance claims have wrong information, including the slight variation of patients’ disease names from real ones.”
Hospitals do not open their detailed treatment information to the outside citing privacy protection and other reasons. This forces external researchers wanting to obtain study data to conduct joint research with inside researchers of the related institutions, he added.
However, there’s a positive change that some medical institutions tend to disclose detailed data to outsiders through anonymizing them, Shin noted.
As the areas that require further research and improvement, he suggested “storage-analysis platform of healthcare data” and “secure zone of personal information.”
In the case of platform research, Shin explained, it is necessary to convert hospital information system into cloud form. He stressed in security area, too, new technology to protect individuals’ privacy is needed.
<© Korea Biomedical Review, All rights reserved.>