臺灣「醫療利用歸人檔資料庫」之建置介紹

目的：過往研究全民健保資料庫（以下簡稱健保資料）多以投保地分析地區差異，但投保地不必然等於居住地，故本研究建置「醫療利用歸人資料檔」（以下簡稱本資料庫）提供多種個人區位資料，期能有助於正確分析健康的地區差異。方法：本資料庫以健保承保檔為主檔，彙整門急診及住院資料，建立多種不同的就醫地點（鄉鎮市區），並結合旅運時間，經與國民健康訪問調查(NHIS)受訪者自答之現居地比對，提出適用於全人口的推估居住地；另亦結合健保和死因資料以進行健康資料加值。結果：本資料庫（年代為2001-2013）提供多種不同的個人區位資料、醫療利用及死因資料，且經驗證後發現鄉鎮市區層級全樣本之推估居住地與個人現住地相同的比例高達81.2%。結論：相較於過去研究多使用健保投保地來代表個人實際居住地的作法，本資料庫預期可以提供更為正確的居住區域與健康的研究。

關鍵字

健康資料庫；居住地；投保地；健康不平等

並列摘要

Purpose: Previous studies usually used the insured location to analyze health disparity by geographic location, but the insured location may not be the actual residence. By providing information on various personal locations, this Healthcare Utilization Database can help one to analyze health disparity by geographic location more accurately. Methods: We used multiple databases including Registry for beneficiaries database, Ambulatory care expenditures by visits database, Inpatient expenditures by admissions database, and the commuter time matrix to create different locations of access to healthcare at the district-township level. By comparing location proximity using the commuter time matrix, we proposed an estimated residence which applied to individuals who have seen a doctor within a year. In addition, we also provided annual outpatient and inpatient utilization records and mortality data to add value to this database. Results: This database (2001-2013) provides various personal locations, healthcare utilization information and mortality data. After a validation study using the self-reported current residence from National Health Interview Survey (NHIS) database, we found the equivalence between our estimated residence and the current residence is 81.2% at the district-township level. Conclusions: While past studies have used insured location as a proxy, this database provides more accurate data for studies on geographic location and health.