照護連續性與民眾逛醫師行為之相關性探討

Abstract

過去雖有許多研究探討過民眾逛醫師行為的相關影響因子，卻未曾有研究提出照護連續性與逛醫師行為間的相關性，故本研究之目的為探討目前台灣民眾逛醫師行為與照護連續性(COC)間的相關性，並檢視照護連續性指標是否可做為預測未來逛醫師行為的指標。資料來源以全民健保資料庫2005年承保抽樣歸人檔第一組至第五組共20萬人為基礎，排除全年就醫三次以下的民眾後計算在2005年有因上呼吸道感染(共129,076人)或腸胃炎(共53,029人)就醫的民眾其照護連續性的分數與逛醫師、逛醫院的行為。照護連續性的指標則依照Jee和Cabana在2003年的五種分類中選擇三種不同類型的照護連續性指標作為代表，密度型指標選擇UPC，離散型指標選擇COCI，次序型指標則選擇SECON作為計算。 分析結果發現上呼吸道感染的民眾有53759人有逛醫師行為、44087人有逛醫院行為，盛行率分別為41.65%、34.16%；腸胃炎的民眾有7412人有逛醫師行為、4780人有逛醫院行為，盛行率分別為13.98%、9.01%。平均逛醫師與逛醫院次數，上呼吸道感染分別共3.38次與2.53次；腸胃炎分別共1.98次與1.81次。上呼吸道感染的民眾其照護連續性的分數UPC為0.45、COCI為0.27、SECON為0.42；而曾患腸胃炎的民眾其醫師的照護連續性分數UPC為0.42、COCI為0.242、SECON為0.39。 利用負二項式迴歸分析照護連續性分數的分組與是否逛醫師、逛醫院間的相關性，以UPC為例，上呼吸道感染的民眾其照護連續性低者，其逛醫師的情形為高COC者的3.66倍 (OR=3.66;p<0.001)，而逛醫院情形則是高COC組的3.42倍(OR=3.42; p<0.001)，而低COC者逛醫師的次數較高COC者多出217% ( eβ=3.17; p<0.001)、逛醫院的次數也較高COC者多出199%( eβ=2.99; p<0.001)；腸胃炎的民眾，在低COC分組者逛醫師情況為高COC者的2.14倍(OR=2.14; p<0.001)，逛醫院的情況較高COC者多出2.06倍(OR=2.06; p<0.001)；在逛醫師的次數，低COC者較高COC者多出109%( eβ=2.09; p<0.001)、而逛醫院的次數則較高COC者多出108%(eβ=2.08; p<0.001)。 若使用2005年的COC預測2006年的逛醫師、逛醫院行為則發現：上呼吸道感染的部分，低COC的民眾其逛醫師的情況為高COC者的1.99倍(OR=1.99; p<0.001)、逛醫院的情況為高COC者的1.89倍(OR=1.89; p<0.001)，低COC者逛醫師的次數較高COC者多出92%(eβ=1.92; p<0.001)、逛醫院的次數則多出83%(eβ=1.83; p<0.001)；腸胃炎的部分，低COC的民眾其逛醫師情形較高COC者多出1.54倍(OR=1.54; p=0.001)、逛醫院的情況則多出1.47倍(OR=1.47; p<0.001)，低COC者逛醫師的次數則較高COC者多出51%(eβ=1.51; p<0.001)、逛醫院的次數則多出45%(eβ=1.45; p<0.001)。 結論：照護連續性與逛醫師、逛醫院行為都有顯著負相關，不論是在上呼吸道感染或是腸胃炎的部分，且與次年的逛醫師、逛醫院行為也有統計上顯著負相關。後續研究者可以再將逛醫師、逛醫院行為細分為第二意見行為與高忠誠行為進行分析，或可針對慢性疾病患者調查其逛醫師、逛醫院的行為。也期待可藉由提高民眾的照護連續性來降低台灣民眾逛醫師、逛醫院行為的盛行率。

The purpose of this study was to understand the association between continuity of care (COC) and the doctor-shopping behavior (DSB) in Taiwan and if the COC indicies could be use as a predicator of the DSH in the future. This study’s data was using the 2005 National Health Insurance Registry for Beneficiaries Claims Data files, the medical service utilization data of 200,000 persons. Only collected the people who have been visited doctors because of URI (upper respiratory tract infection) and AGE (acute gastroenteritis), and those people should visited doctors for more than 3 times in whole year. So that there are 129,076 people involve because of URI and 53,029 people involved because of AGE. We found out people who have DSB and hospital-shopping behavior (HSB) because of URI were 53759 and 44087 people, respectivel, and the prevalence of DSB and HSB were 41.65% and 34.16%, respectively. About the AGE part, who have DSB and HSB were 7412 and 4780 people, respectively, and the prevalence of DSB and HSB were 13.98% and 9.01%, respectively. The average times of DSB and HSB were 3.38 and 2.53 times in URI group and 1.98 and 1.81 times in AGE group, respectively. Negative binominal regressions were performed to examine the effect of three different COC indices on the numbers of DSB and HSB, logistic regressions were used to exam the COC indices with do or do not have the DSB and HSB. The data suggest that lower COC was associated with increased risk of DSB and HSB. Compare with the high COC groups, subjects in the low and medieum COC groups had 2.10-3.66 times and 1.45-2.19 times of DSB, respectively, as well as 2.02-3.42 times and 1.37-2.05 times of HSB, respectively. Compare with the high COC groups, subjects in the low and medieum COC groups had 87%-217% and 55%-141% more DSB, respectively, as well as 101%-199% and 58%-122% more HSB, respectively. While using the 2006’s DSB and HSB as an dependent variable, compare with the high COC groups, subjects in the low and medieum COC groups had 1.43-1.99 times and 1.29-1.59 times of DSB, respectively, as well as 1.38-1.89 times and 1.23-1.50 times of HSB, respectively. Compare with the high COC groups, subjects in the low and medieum COC groups had 35%-92% and 31%-55% more DSB, respectively, as well as 32%-89% and 28%-51% more HSB, respectively. This study indicates that lower COC indices were associated with increased DSB and HSB, also with higher risk of DSB and HSB. The data even shows the same results in 2006’s DSB and HSB, no matteer in the URI group or AGE group. This study suggests that improving the COC might decrease the prevalence of DSB an HSB.