引用
就醫行為的潛在類型與健康測量：全民健康保險資料的應用
Latent Healthcare Utilization Classification and Objective Health Measures
作者：張苙雲（Ly-Yun CHANG）、楊孟麗（Meng-Li YANG）、趙景雲（Jing-Yun JHAO） | 首次發表於 2020-06-04 | 第 46 期 June 2011
DOI：https://dx.doi.org/10.6786/TJS.201106_(46).0005
研究紀要（Research Notes）
DOI：https://dx.doi.org/10.6786/TJS.201106_(46).0005
研究紀要（Research Notes）
論文資訊 | Article information
摘要 Abstract
健康測量是研究健康不平等課題的核心，長久以來，學界對自陳評量的效度，保持某種程度的批判，並用以反省健康不平等的研究發現。 然而，因著就醫紀錄的不易取得和複雜性，學界仍以受訪者自陳資料，作為衡量個人健康的主要根據，甚少使用就醫紀錄作為健康測量指標。本文建議運用潛在側圖分析法，分析就醫紀錄，從行為面推測健康情況，以解決就醫紀錄過於複雜的挑戰。本研究使用的資料有二：具全國代表性的「2000年全民健康保險承保抽樣歸人檔」及「2001年國民健康調查」。「承保抽樣歸人檔」近93萬樣本於2001年全年的就醫紀錄之側圖分析，結果顯示四組的分類是最適模型，依其就醫特徵與疾病診斷，其健康等級依序為「零就醫組」、「門診使用組」、「短期住院組」和 「重度使用組」。文中除了討論分組結果的信度外，並將此分組與疾病診斷、就醫文獻相比對，以檢驗其效度。最後，受惠於全民健保資料和調查訪問資料的串連，將分組結果與調查所得之自評健康進行比較，彰顯本分析結果在理論和政策應用上的潛力。
關鍵詞：健康測量、醫療利用、健康不平等、「承保抽樣歸人檔」、潛在側圖分析
關鍵詞：健康測量、醫療利用、健康不平等、「承保抽樣歸人檔」、潛在側圖分析
Health measures sit at the core of health inequality studies. Although objective, medical care utilization is seldom used as a measure of health status because of the unavailability of data and the obscure relationship between medical care utilization and health. A latent profile analysis technique was used to classify medical care utilization, and in turn to investigate the validity of different utilization types as ranked measures of general health. Data were taken from Large Health Insurance Dataset of 2000” (LHID 2000) and the “National Health Interview Survey for 2000 (n=923,892). Four membership groups were identified: “zero utilization,” “outpatient visits only,” “short term hospitalized,” and “heavy users.” Discrepancies between latent types and self-reported health data from a National Health Interview Survey (n=19,732) were noted in terms of age, gender, education, employment status, and income. The validity of classification as a rank order of health status and potential applications are discussed.
Keywords: Health measures, health care utilization, health disparities, latent profile analysis
Keywords: Health measures, health care utilization, health disparities, latent profile analysis
