以多層次分析精神疾病患者之逛醫師行為及其與照護結果之相關性

Using Multilevel Analysis to Examine the Doctor Shopping Behavior of Psychiatric Patients and the Association with Their Treatment Outcomes

陳家榆(Chia-Yu Chen)

指導教授 ： 楊銘欽

國立臺灣大學/公共衛生學院/健康政策與管理研究所/博士(2012年)

https://doi.org/10.6342/NTU.2012.02619

摘要

民眾逛醫師行為是醫療浪費的一大主因，目前國內外已經有許多研究探討逛醫師行為之相關議題，但多數著重於影響逛醫師行為之因素與現況，尚少研究驗證逛醫師行為與其照護結果之相關性。因此，本研究以國內尚無相關研究之精神病患者為對象，嘗試將各種層面可能與其逛醫師行為之相關因素同時納入分析，探討精神疾病患者之逛醫師行為情形，找出相關因素，再檢驗精神疾病患者逛醫師行為與照護結果之相關性，以填補相關研究之不足。
本研究使用國衛院全民健保資料庫「2005年承保抽樣歸人檔」的2006-2008年資料進行研究，選取18歲以上門、急診初次診斷為精神疾病之患者共25,554位為本研究對象，使用多層次分析探討其逛醫師行為之相關因素，接著再使用羅吉斯迴歸、波以松迴歸、負二項式迴歸及複迴歸分析逛醫師行為、其他就醫行為與照護結果之相關性。
本研究發現有18.7%的門、急診初次診斷精神疾病患者有逛醫師行為，平均精神疾病就醫次數為6次，精神疾病患者逛醫師行為的百分比並未比其它疾病患者高；多層次分析結果發現，18-30歲精神疾病患者較多有逛醫師行為，有器質性精神病、精神疾患及情感性精神病等較嚴重之精神疾病患者，比有精神官能症者有較多逛醫師行為，有慢性病者也較多有逛醫師行為；另外，控制病人特質之後，醫師特質也會影響病患逛醫師行為，年齡較輕的醫師其病患較多有逛醫師行為，而精神科醫師的病人有逛醫師行為者最多；而第三層的醫療機構特質則無階層差異。
在醫療利用方面，以男性、有重度精神疾病及慢性病患者的醫療利用較多，而精神疾病患者年齡愈大精神疾病就醫次數愈多，18-30歲患者看診醫師數最多，41-60歲患者就醫科別數最多，有免部分負擔身份者其精神疾病就醫次數較多；另外，病患之主要醫師的特質，以45-54歲醫師及精神科醫師之病患精神疾病就醫次數較多，愈年輕的醫師及急診醫學科醫師之病患看診醫師數較多，而精神科醫師的病患就醫科別數最少；在醫療機構特質方面，中區及高屏地區、私立醫療機構及基層診所的病患精神疾病就醫次數最多，北區及高屏地區、私立醫療機構及地區醫院的病患看診醫師數最多，而中區病患的就醫科別數最多。
另外，探討逛醫師行為與照護結果之相關性，發現有逛醫師行為者會顯著增加精神疾病及所有疾病急、住診醫療利用，則照護結果較差，患者就診醫師數及科別數愈多，照護結果也愈差；然而，精神疾病就醫次數愈多者，照護結果反而較佳。
本研究結論為有逛醫師行為者以18-30歲、有慢性病、有嚴重精神疾病及常看精神科醫師者較多，然而，這些有逛醫師行為的精神疾病患者整體的照護結果愈差，建議政策制定者，可以針對這群疾病嚴重度及醫療利用較高的病患連結支付制度，再對其進行衛教，減少不必要的醫療行為、降低醫療費用，使其有更好的照護連續性。

關鍵字

精神疾病 ； 逛醫師行為 ； 照護結果 ； 多層次分析

並列摘要

Doctor shopping behavior (DSB) is an important issue of medical care utilization. There are already many studies examined issues related to influencing factors of DSB, but studies on the relationship between DSB and treatment outcome are scant. The objectives of this study were to use multilevel analysis model to examine factors associated with DSB among psychiatric patients and the relationship between DSB and treatment outcome, to fill the research gap.
This study used the 2005 National Health Insurance Registry for Beneficiaries Claims Data covering the years 2006 to 2008. A total of 25,554 psychiatric patients who were older than 18 years of age and visited outpatient and emergency departments for the first time were extracted from the data set. Multilevel regression model were carried out to examine influencing factors of DSB. Logistic regression, Poisson regression, negative binomial regression, and multiple regression models were also used to explore the relationship between DSB and treatment outcome.
The prevalence of DSB was 18.7% in this sample. On the average, a patient had 6 mental illness visits in a year. The results suggested that patients who had heavy mental illness, had chronic disease, aged 18-30 were more likely to have DSB. After controlling for patients characteristics, young doctors and psychiatric doctors were more likely to have DSB patients. However, no significant association was found in the hospital level.
In the medical use, the results showed that having heavy mental illness, having chronic disease, or being male patients were more likely to have medical use. Patients aged 18-30 visited more doctors. Patients aged 41-60 visited more specialties. Doctors aged 45-54, with psychiatric specialty had more patients with frequent visits. Young doctors or emergency doctors had more patients with more attending physician. Patients who lived in central and southern Taiwan, visited private hospitals, or clinic had more visits. Patients who lived in northern and southern Taiwan, visited private hospital and local hospital visited more different doctors. Patient lived in central Taiwan visited more different specialties.
This study also found that patients with DSB were more likely to have bad treatment outcome such as emergency use and hospitalization. The more different doctors and specialties visited, the worse the treatment outcomes. However, patients who had more visits were likely to have better outcome.
The results of the study suggest policy makers can build a better payment system and strengthening the health education for DSB patients. By doing so, the unnecessary use of the medical resources may be reduced, and the continuity of the health care will be better.

並列關鍵字

mental illness ； doctor shopping behavior ； treatment outcome ； multilevel analysis

參考文獻

江慧珍. (2006). 台灣民眾的就醫第二意見行為之探討. 碩士, 臺灣大學.

林鉦棽、彭台光. (2006). 多層次管理研究：分析層次的概念、理論和方法. 管理學報, 23(6), 649-675.

張苙雲. (1998). 「逛醫師」的邏輯：求醫歷程的分析. 臺灣社會學刊(21), 59-87.

黃信勳. (2009). 原發性肝癌病人重複確診之醫療利用情形及其相關因素探討. 碩士, 臺灣大學, 台北市.

溫福星, & 邱皓政. (2009). 多層次模型方法論：階層線性模式的關鍵議題與試解. 臺大管理論叢, 19(2), 263-294.