糖尿病論質計酬方案與糖尿病合併高血脂與高血壓病人照護相關性研究-以健保資料庫為例

A study of the association between pay-for-performance program and patient care for diabetes patients with hypertension and hyperlipidemia under the Taiwanese National Health Insurance System: a population-based study

辜嵩貿(Song-Mao Gu)

指導教授 ： 謝慧敏

高雄醫學大學/健康科學院/公共衛生學研究所/碩士(2014年)

摘要

糖尿病病患有多重合併慢性病的盛行率逐年增加。面對醫療費用不斷高漲的問題以及提高醫療品質的需求，中央健保署自2001年開始陸續推行糖尿病論質計酬試辦計劃。過去文獻多著重在糖尿病單一疾病的探討，並未有針對是否有多重慢性病情況，在照護品質上是否有差異之相關研究，因此本研究目的主要探討在糖尿病論質計酬的照護下，糖尿病病患有無同時合併三高，對於病人在中長期臨床照護指標是否有影響，以及有無加入論質計酬(P4P versus Non-P4P)的糖尿病病患有無同時合併三高，是否影響其存活狀況與醫療成本效果。
本研究資料來源為健保醫療費用申報資料庫與P4P-VPN資料庫。研究樣本皆是2007年至2008年期間新收案之糖尿病病患。利用收案前一年內是否有三高相關診斷碼或藥品碼來當作分組依據，將病患分成單純糖尿病與三高組病人。依照研究目的將研究方法分為兩部分。第一部分主要比較兩組的P4P病患自收案日往後追蹤三年期間的臨床數值。第二部分則是比較單純糖尿病與三高組的參與方案(P4P)與為參與方案(Non-P4P)病人存活狀況以及醫療成本效果。兩組個案之P4P與Non-P4P以傾向分數配對法來進行配對。統計分析方法有獨立樣本T檢定、卡方檢定、廣義線性迴歸、COX比例風險迴歸以及Bootstrap抽樣法。
本研究結果發現參與論質計酬的病人在過程指標中，三高組病人在主要的檢查指標中與單純糖尿病組呈現顯著差異；三高組中期照護指標中的改善率與單純糖尿病組亦呈現顯著差異。其中，單純糖尿病組在HbA1c與血壓的目標達成率較佳；三高組在LDL的目標達成率較佳。
另外本研究結果亦發現P4P病人的存活人年比Non-P4P病人多，其介入成本較高且在省錢成本中也較省錢。若比較有無合併三高病人族群，研究發現單純糖尿病組的存活率較三高組高，其介入成本較三高組少但在省錢成本相對較為省錢。若以投資報酬率來看單純糖尿病與三高組兩組差異不大。
總結而言，單純糖尿病與三高組在照護過程指標以及中期結果達成率上兩組都呈現不錯的表現。另外，糖尿病患確實會因為其多重合併慢性病狀況而影響到存活狀況以及醫療成本效果。由於三高病人是更需要被照護且從本研究看起來投資報酬率也是很高，建議政府未來可針對三高族群進行論質計酬方案的介入。

關鍵字

糖尿病 ； 多重慢性病 ； 三高 ； 論質計酬品質改善方案

並列摘要

The number of diabetes mellitus patients with multiple chronic disease has been growing worldwide. With respect to the increasing health care expenditures and peopele’ expectations of quality of care, the Taiwan’s National Health Insurance (NHI) Program implemented a diabetic pay for performance program (P4P) in the end of 2001. The majority of literature examining the effect of P4P on diabetes care focused on patients with diabetes in general rather than patients with multiple chronic diseases (e.g., three-hypers, such as hyperglycemia, hypertension, and hyperlipidemia). This study aimed at: (1) examining to what extent the process care and intermediate outcomes were difference between patients with diabetes only and with three-hypers; (2) examining all cause mortality rate as compared P4P with non-P4P patients who were having diabetes only and with three-hypers; (3) evaluating the cost-effectiveness of the P4P program as compared P4P with non-P4P patients having diabetes and three-hypers diseases.
Data sources were derived from the NHI administrative claims and P4P VPN datasets. This study included patients who were diagnosed as diabetes during 2007 and 2008 and used prior year diagnosis or medication prescriptions for hypertension/ hyperlipidemia to group patients into diabetes only and three-hypers groups. The methodology was different based on the study objectives. For the first aim, this study compared the clinical outcomes during three-year follow-ups between diabetes only and three-hypers P4P patients. For the second and third aims, propensity score matching approach was used to find the comparable comparison group for P4P group. Several statistical technologies were used given the aims, such as independent t-test, chi-square test, general linear regression model (GLM), cox proportional hazard model, and bootstrapping method.
The study findings indicated that there were significant different for process outcomes and intermediate outcomes between P4P diabetes only and three-hypers patient groups. Patients with diabetes only had better intermediate outcomes in HbA1c and blood pressures control, while three-hypers groups had better outcomes in LDL control.
In addition, the study also indicated that P4P patients had better survival than non-P4P patients. Compared to non-P4P patients, P4P groups had higher intervention costs (diabetes-related outpatient costs), but better cost-savings under the program. Compared with three-hypers, diabetes only groups had better survival, lower intervention costs and higher cost-savings. Nevertheless, the return on investment of the P4P program between two groups were not significant different.
In conclusion, the process and intermediate outcomes for both diabetes only and three-hypers P4P patients were improved. Multiple chronic diseases (e.g., three-hypers) did affect patients’ survival and quality outcomes. Patients with three-hypers diseases were the groups who have medical needs particularly. Given the study findings, the government may consider to put more emphasis on providing care for this group of patients under the diabetes pay-for-performance program.

並列關鍵字

diabetes mellitus ； multiple chronic diseases ； Pay-for-performance

參考文獻

Doran, T., Fullwood, C., Gravelle, H., Reeves, D., Kontopantelis, E., Hiroeh, U., & Roland, M. (2006). Pay-for-performance programs in family practices in the United Kingdom. N Engl J Med, 355(4), 375-384. doi: 10.1056/NEJMsa055505

英文文獻

Assmann, G., & Schulte, H. (1989). Diabetes mellitus and hypertension in the elderly: concomitant hyperlipidemia and coronary heart disease risk. The American journal of cardiology, 63(16), 33-37.

Association, A. D. (2014a). Diagnosis and classification of diabetes mellitus. Diabetes Care, 37(Supplement 1), S81-S90.

Association, A. D. (2014b). Standards of medical care in diabetes—2014. Diabetes Care, 37(Supplement 1), S14-S80.