HIRA

새로운 환자표본자료 표본 추출 및 대표성 검증

Metadata Downloads
Type
Article
Author(s)
윤철영안재준이경민최용석김리현하대우유기봉
Issued Date
2021-11-30
Keyword
National patient sampleSampling studiesValidation study
Abstract
Background: The sampling framework of the National Patient Sample of Health Insurance Review & Assessment Service is needed to be improved due to the current demographic structure. We proposed a sampling method and additional strata for extracting the National Patient Sample data due to the current demographic structure, such as low birth rate and aged population.
Methods: A total of 36 strata were set by adding four strata compared to the existing one. The maximum rate of minimal sample number was defined among the entire strata. Based on the rate, we extracted a small-scale sample dataset consisting of about 400,000 people and a large-scale sample dataset of more than 700,000 people.
Results: The representativeness of the high-frequency disease and the low-frequency disease was confirmed. For health expenditure, the representativeness of samples was confirmed in large-scale samples. However, the representativeness of small-scale samples was not confirmed in five strata.
Conclusion: Using the maximum rate of minimal sample number can reflect the demographic structure changes and diverse medical utilization. Although lack of representativeness in the five strata of the small-scale sample, both the small-scale sample and the large-scale sample are necessary to improve data accessibility and a sustainable data provision system. It will be helpful in establishing health policies and conducting medical research.
Publisher
심사평가연구소
DOI
https://doi.org/10.52937/hira.21.1.2.166
URI
https://repository.hira.or.kr/handle/2019.oak/2593
Alternative Title
Developing the New National Patient Sample and Evaluating Representations
Publisher
심사평가연구소
Location
KOR
Citation
윤철영. (2021-11-30). 새로운 환자표본자료 표본 추출 및 대표성 검증. HIRA Research, 1(2), 166-178. doi: https://doi.org/10.52937/hira.21.1.2.166
ISSN
2765--6764
Appears in Collections:
HIRA 발간 > 1. HIRA Research
공개 및 라이선스
  • 공개 구분공개
Files in This Item:

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.