全民健保門診高利用者就醫型態分析

Abstract

研究背景 全民健康保險自1995年開辦以來，立基於增進全體國民健康，為政府辦理之社會保險，希望透過自助、互助制度，將經濟弱勢族群納入健康保障。然而，隨著整體環境與社會人口結構等影響，醫療支出增加速度遠快於保費收入成長速度，因此，醫療浪費的問題又受到大家的關注，其中之一便是有部分高醫療利用保險對象過度的使用醫療資源，造成醫療浪費。而高醫療利用保險對象的頻繁就醫不僅會造成健康保險財務上的嚴重負擔，同時也會使高醫療利用保險對象本身的生活品質受到影響，進而造成自身健康的危害。 研究目的 本研究目的為探討門診高利用者人口特徵及就醫背景因素，可分為下列二項： 一、利用 Andersen 第一階段之醫療服務利用行為模式，瞭解門診高利用保險對象之人口特徵及就醫背景因素。 二、探討持續門診高利用保險對象與非持續門診高利用保險對象之就醫行為差異及影響因素。 研究方法 本研究為縱貫性次級資料分析，使用全民健康保險資料庫，分析期間為2011年至2015年，分析對象為針對2012年至2014年一年門診就醫超過90次(含)之保險對象，為研究母群體，並觀察其前一年及後一年之醫療利用狀況。以SAS 9.4版進行統計分析。描述性統計分析各變項之平均數、標準差、次數分配及百分比，了解樣本分布情形。推論性統計，使用卡方檢定及t-檢定來檢驗各自變項在高利用期後1年對是否持續為門診高利用者呈現統計上的顯著差異，再使用羅吉斯迴歸計算OR值及95%信賴區間，分析與高利用後期持續為門者高利用者情形之相關因素。 研究結果 一、門診高利用者之人口特徵及就醫背景因素 研究結果顯示在傾向因素中65(含)歲以上、女性、第一類保險對象、被保險人易成為門診高利用者；使能因素高可近性、低免自付額對象易成為門診高利用者；需求因素中無殘障、無重大傷病、無精神病、有3類(含)以上慢性病、共病指數3分(含)以上，易成為門診高利用者。 二、易成為持續門診高利用者之人口特徵及就醫背景因素 研究結果顯示在傾向因素65(含)歲以上、男性、第五類保險對象、被保險人易成為持續門診高利用者；使能因素中、高可近性、高免自付額對象易成為持續門診高利用者；需求因素中殘障、有重大傷病、有精神病、有慢性病、共病指數2及3分(含)以上，易成為持續門診高利用者。 三、門診高利用者社經、使能、需求程度變化，對後續醫療利用之影響 研究結果發現社經程度降低，並不會成為持續門診高利用者，使能程度上升較易成為持續門診高利用者，需求程度上升較易成為持續門診高利用者，惟均未達統計上顯著差異。 四、門診高利用者醫療資源利用型態 以照護連續性測量指標(COCI)研究結果發現，院內逛醫者較易成為持續門診高利用者，且達到統計上顯著差異；到處就醫者較不易成為持續門診高利用者，惟未達統計上顯著差異。 研究結論 本研究結果顯示高利用前期為門診高利用者，則持續為門診高利用者之勝算比高達5.15倍，故建立門診高利用者之即時監控機制，應可有效減少門診高利用者就醫次數。主管機關對於門診高利用之輔導計畫，不應僅單純從就醫次數上去訂定不同的輔導措施，必須針對醫療行為因素之相關變項，設定就診次數的上限值，導入有效的輔導措施。

Background Since the National Health Insurance system was implemented in 1995, all the citizens in Taiwan has been covered in health insurance. It hopes to include the economically disadvantaged groups through self-help and mutual assistance. However, the growth rate of medical expenditure was much faster than that of the premium. Therefore, the problem of medical waste has drawn everyone's attention. One of the focuses is that some frequent users of outpatient service overused medical care, resulting in medical waste. However, the oveuse of medical care will not only cause a serious financial burden on health insurance, but also affect the patients’ quality of life and may cause damage to their own health. Purposes This study has three study purposes, first, is to use the Andersen healthcare service utilization model to understand the demographic and medical behavior characteristics of the frequent users of outpatient service. Second, is to explore the different behavior and influencing factors between the frequent and infrequent users of outpatient service. Third, is to use the concept of continuous care to analyze the medical utilization patterns of frequent users of outpatient service. Method This study identified those frequent users of outpatient service with more than 90 times from 2012 to 2014 of the national health insurance database, also with the data in the previous year and the following year. This study uses SAS 9.4 to analyze if there is a statistically significant difference between the frequent and infrequent users. We also use logistic regression to identify relevant factors associated with frequent users of outpatient service. Results Factors associated with those who were prone to become the frequent users of outpatient service include three variables in predisposing factors, that is category 1 insured, 65 years old and over, and being female; one variable in enabling factor, that is low copayment in enabling factors; and four variables in need factors, that is non-disabled, without catastrophic illnesses, without serious mental illness, suffered r 3 or more chronic diseases, CCI score equals or higher than 3. Next, the factors associated with those who were prone to become the continuous frequent users of outpatient service include: 3 variables in predisposing facrors, i.e. category5 insured, 65 years old and over, being male; 2 variables in enabling facrors, i.e. paied high copayment, had high accessibility; and 5 varibles in need facrors, i.e. being disabled, had catastrophic illnesses, had serious mental illness, had chronic diseases, CCI score equals to 2 or higher. In addition, the lower socio-economic status was not significantly associated with prone to become the frequent users of outpatient service. Patiens with increasing levels in enabling and need factors tended to become the frequent users of outpatient service, but not statistically significant. Finally, outpatient doctor shopping in the same hospital/clinic is prone to become the continuous frequent users of outpatient service, but less likely in different hospitals or clinics. Conclusions An instant monitoring and responding mechanism can effectively decrease the visit times of the frequent users of outpatient service. The competent administration could look into the factors that are associated with prone to become the frequent users of outpatient service, so that it can not only limit the number of visit times but also provide more effective healthcare.