Investigating Patient Selection in the ay-for-Performance Program for Diabetes

近年來，糖尿病醫療照護廣受國內外重視。因此，一種為引導醫療提供者提供糖尿病患完整性照護服務，進而提昇醫療品質，而以支付額外獎金予醫療提供者為手段，並連結其績效作為支付依據之糖尿病醫療給付改善方案便應運而生，期藉此方案達到疾病控制之目的。然而根據醫師行為理論指出，醫師為追求利潤最大化，可能出現選擇低嚴重度病患之行為，這將嚴重扭曲此方案提昇照護品質與疾病控制之美意，因此，在此方案下參與之醫師是否存在傾向選擇輕症病患之行為，值得探討。 本研究之資料來源為全民健康保險研究資料庫，研究對象為2002~2005年參與糖尿病醫療給付改善方案之醫師，2002年共計220位醫師，2003年共計413位醫師，2004年共計620位醫師，2005年共計810位醫師。本研究以DCSI作為糖尿病疾病嚴重度之評估指標，分析參與糖尿病醫療給付改善方案之醫師下符合收案條件之糖尿病就診病患收案與未收案者其糖尿病疾病嚴重度是否有差異，以推論參與之醫師是否出現選擇病患之行為，期能增添更多攸關資訊，提供有關單位未來制訂政策參考使用。 本研究之主要研究結果如下：、參與糖尿病醫療給付改善方案之醫師下符合收案條件之就診病患收案與未收案DCSI平均分數之平均2002年分別為2.977與6.176分，2003年分別為2.655與6.037分，2004年分別為2.334與5.998分，2005年分別為2.325與6.186分，且均達統計上顯著差異，意即參與糖尿病醫療給付改善方案之醫師存在選擇病患之行為。、參與糖尿病醫療給付改善方案之醫師下符合收案條件之就診病患收案與未收案DCSI平均分數的差值之平均，2002年為3.199分，2003年為3.382分，2004年為3.663分，2005年為3.862分，意即參與糖尿病醫療給付改善方案之醫師存在傾向選擇低嚴重度病患收案之行為。、在控制醫師年齡、性別、專科別、執業機構層級別、執業機構權屬別之變項後，參與糖尿病醫療給付改善方案之個別醫師下未收案組就診病患之DCSI平均分數較收案組高，2002年高出2.62分、2003年高出2.89分、2004年高出2.99分、2005年高出3.03分，且均達統計上顯著差異，意即控制醫師特質與醫療機構特質後，2002~2005參與糖尿病醫療給付改善方案之醫師存在傾向選擇低嚴重度病患收案之行為。另外，2002~2005參與糖尿病醫療給付改善方案之醫師其年齡、性別、專科別、執業機構層級別、執業機構權屬別均未達統計上顯著差異，但在控制其他變項後，執業機構權屬別呈現公立醫療機構之糖尿病患疾病嚴重度高於私立醫療機構之情形，且似乎有越來越顯著的傾向。此結果說明私立醫療機構較傾向收治疾病嚴重度低之糖尿病患之情形。但整體而言，此結果顯示參與此方案之醫師存在傾向選擇低嚴重度病患收案之行為並不顯著受醫師特質與執業機構特質所影響。 本研究之建議如下：一） 應審慎評估推行糖尿病醫療給付改善方案後對整體糖尿病患之成效，非僅關注收案病患之品質改善與資源耗用情形，畢竟仍有大多數糖尿病患未被納入此方案，且未被納入者多屬於疾病嚴重度較高之族群，更應設法提供此族群完善的醫療照護，以提升整提糖尿病患之照護品質。二） 建議糖尿病醫療給付改善方案之支付模式應引入適當之疾病嚴重度校正模式，疾病嚴重度較高之病患應給予醫療提供者較大之支付，鼓勵其提供嚴重度較高之糖尿病患完整之照護服務，進而維護病患健康、控制醫療費用。三） 建議糖尿病醫療給付改善方案應簡化各項行政流程以提高醫師收治嚴重度較高之糖尿病患之意願。四） 加強對嚴重度較高之糖尿病患的衛教服務以提升其回診率進而提高醫師收治嚴重度較高之糖尿病患之意願。In recent years, diabetes care has drawn much attention domestically and all over the world. Therefore, a pay-for-performance program for diabetes was implemented. Goal of the program conducts health care providers to supply the integrity of diabetes care services, thereby enhancing quality of medical care and control the disease. Method of the program pay bonus to health care providers, and link their performance as a basis for payment. However, according to physician behavior theory that the physician for the pursuit of profit maximization, the physician possible select patients of low severity, This situation will seriously distort the intention of this programme for enhancing the quality of care and disease control. Therefore, under this program, whether physicians tend to select patients of low severity needs to be further discussed. The samples selected from National Health Insurance Research Database of claimed data during 2002 to 2005. Accede to pay-for-performance program for diabetes, there were total 220 physicians in 2002, 413 physicians in 2003, 620 physicians in 2004 and 810 physicians in 2005. This study use DCSI to evaluate the severity of patients and understand the difference of severity between patient participating and not patient participating under this program. These results will inference whether physicians tend to select patients of low severity and provide government organization with suggestions when making medical quality policy. The main findings of the study were concluded as below：. The averages that the means of DCSI scores of the patient participating in this program or not are 2.977 and 6.176 in 2002, 2.655 and 6.037 in 2003, 2.334 and 5.998 in 2004, 2.325 and 6.186 in 2005 respectively. This result indicates that patient selection of physicians behaviors. . The difference of average that the means of DCSI scores of the patient participating in this program is 3.199 in 2002, 3.382 in 2003, 3.663 in 2004, 3.862 in 2005. This result indicates that physicians tend to select patients of low severity.. After controlling hospital’ and physician’ characteristics, the averages of DCSI scores of the the patient not participating in this program is significant higher 2.62 in 2002, 2.89 in 2003, 2.99 in 2004 and 3.03 in 2005 than the patient participating.This result indicates that after controlling hospital’ and physician’ characteristics, physicians tend to select patients of low severity. In addition, hospital’ and physician’ characteristics are not significant but patients have lower DCSI scores in private medical institutions than public medical institutions. This result indicates that private medical institutions tend to select patients of low severity. Overall, under this program, hospital’ and physician’ characteristics do not have significant influence on patient selection of physicians behaviors. Based on the results of this study, a few suggestions were proposed as the following:. Effectiveness of pay-for-performance program for diabetes for paitents should be evaluated carefully. This is because the most patient with diabetes were not included in this program and these patients with high severity. Therefore, policy should provide these patients comprehensive medical care to enhance the quality of care. . To suggest this program use a severity of illness adjusted model. Higher severity of patients should be given a larger payment to health care providers to encourage their to supply the integrity of diabetes care to higher severity of the diabetic patients, thereby enhancing quality of medical care and control medical costs.. To suggest this program should simplify administrative processes to improve physicians treated patients of high severity.. To raise the patients’ compliance to improve physicians treated patients of high severity.摘要IbstratIII錄V目錄VI目錄VII一章 緒論 1一節 研究背景與動機 1二節 研究目的 4二章 文獻探討 5一節 醫師行為理論－利潤極大化模式 5二節 醫療提供者選擇病患行為之相關研究 7三節 財務誘因與醫療提供者行為之關係 11四節 論質計酬(pay-for-performance，P4P)支付制度 12五節 糖尿病疾病管理 15三章 研究材料與方法 19一節 研究設計與研究架構 19二節 研究假說 21三節 變項名稱及操作型定義 22四節 資料處理及統計分析方法 24四章 研究結果 30一節 描述性統計結果 30二節 配對T檢定之檢定結果 35三節 迴歸分析之結果 36五章 討論 39一節 參與糖尿病醫療給付改善方案之醫師是否有選擇病患之行為 39二節 研究限制 41六章 結論與建議 42一節 結論 42二節 建議 43考文獻 44文部分 44文部分 46路資料 48錄一49錄二5