糖尿病照護品質測量與預後評估相關性研究: 第一型與第二型糖尿病

Outcome evaluation associated with quality measures of diabetes care: Type 1 and type 2 diabetes

曾芬郁(Fen-Yu Tseng)

指導教授 ： 賴美淑 ； 陳秀熙

國立臺灣大學/公共衛生學院/預防醫學研究所/博士(2006年)

https://doi.org/10.6342/NTU.2006.01486

摘要

背景及研究目的：
台灣地區糖尿病盛行率約為百分之四，估計糖尿病人口超過100萬人。糖尿病是台灣第四大死因，糖尿病之醫療花費每年超過三億美元，不論著眼於民眾健康或是衛生經濟，提升糖尿病照護品質均是重要的課題。本論文對於臺灣的糖尿病照護品質做多面向的分析。

研究對象及方法：
本研究之第一部分以中央健康保險局台北分局2001年的申報資料分析，計算糖尿病人實際接受的檢驗次數並與糖尿病照護準則所建議的檢驗做比較，以評估糖尿病照護的專業盡責度，並且探討病人特徵與醫院特徵對於盡責度的影響。本研究之第二部分以十分之一隨機抽樣2002年於台大醫院就診四次以上的糖尿病人，其中固定於内科、家庭醫學科、內分泌科接受治療的875位糖尿病人為研究對象。以病人之病歷資料作回溯性分析，使用Chi-square、logistic regression、linear regression 等統計方法探討醫師及病人特徵對於相關檢驗執行、以及血糖、血壓、血脂控制之影響。本研究之第三部分以十分之一隨機抽樣2002年在臺大醫院就診四次以上且2003年曾經回診的糖尿病人。以郵寄方式做問卷調查，512位回覆者為研究對象。利用回溯性病歷資料分析以及問卷答覆，以logistic regression、Chi-square等統計方法，分析糖尿病人對於他們自己的檢驗值或臨床狀況的了解程度，並探討臨床狀況認知與自我監測血糖的相關。本研究之第四部份回溯研究自1995年至2005年在台大醫院就診的第一型糖尿病人的病歷，擷取各項臨床資料以及檢驗數據，紀錄糖尿病慢性併發症的發生率。估算自糖尿病發病以至產生視網膜病變、神經病變、末期腎衰竭或死亡之期間，並分析導致腎病變的危險因子。

結果：
以收錄的2259位糖尿病人健保申報資料分析，一年當中約有76.3%、42.7%、40.2%、59.7%、59.2%、53.2%、16.8%的病人曾經接受至少一次血糖、糖化血色素、尿液、腎功能、血脂、肝功能以及視網膜病變的檢查。約有19.2%的糖尿病人一年當中未曾接受上述任何一項檢查。使用降血糖、抗高血壓或降血脂藥物、曾經住院、曾經急診、就診次數多的病人，接受檢驗的比率較高。醫院層級、所在地、權屬、是否為糖尿病照護機構等特徵會影響糖尿病照護專業盡責度。
在台大醫院就診的第二型糖尿病人一年當中曾經接受至少一次血糖、糖化血色素、尿液、腎功能、血脂、肝功能以及視網膜病變檢查的百分比為89.7%、82.5%、48%、53.6%、69%、45.9%、18.0%。內分泌科醫師比內科醫師或家庭醫學科醫師注重對於糖尿病人血糖、糖化血色素、以及尿液檢查的監測。在台大醫院就診的第二型糖尿病人的平均糖化血色素值為7.3 ± 1.5%。與內科或家庭醫學科病人相較，在內分泌科門診追蹤治療的病人有較低的血糖值。
約有73%、17.9%的糖尿病知道自己的血糖值、糖化血色素值及其意義。有33.3%、66.7%、65.5%、70.6%的病人知道自己有蛋白尿、腎功能異常、肝功能異常或是高血脂等臨床狀況。年老、低教育程度、低收入、就診次數較少的病人對於自己的檢驗值或臨床狀況認知較不足。以胰島素治療、醫師年紀較大、或是糖尿病專科醫師的病人臨床認知較足夠。年青、糖尿病得病期間較久、收入較多、使用胰島素治療、有較多併發症或合併症、糖尿病專科醫師的病人比較會執行自我血糖監測。對於血糖值以及糖化血色素的認知與自我血糖監測有正相關。
以病歷上記載的最後一次數據分析，臺大醫院第一型糖尿病患者的平均糖化血色素值為8.9 ± 1.5%。這些病人在長期追蹤後產生視網膜病變、神經病變、腎病變、腦血管疾病、冠狀動脈疾病、周邊血管疾病之比例分別為23.1%、15.1%、25.7%、1.1%、4.0% 以及5.5%。共有5位(1.1%)病人死亡。女性病人、血糖控制不良、血脂異常是微白蛋白尿的風險因子。高血壓、高血脂、抽菸是產生蛋白尿的危險因子。發病年齡大於15歲、糖尿病發病期間長以及高血壓會增加末期腎衰竭的風險。多數合併有高血壓、腎病變、血脂異常的病人並沒有被投予糖尿病照護準則所建議的相關用藥。

結論:
醫院、醫師、以及病人特徵均會影響糖尿病照護品質。臺灣地區對於糖化血色素、尿液、視網膜等檢查之專業盡責度較低，尤其是基層醫療院所。不同專科醫師對於檢驗監測的執行有所不同，他們照護的病人的血糖控制程度也有不同。糖尿病人的臨床認知可以增加他們的自我照護，但並不是所有的糖尿病人都了解他們的檢驗結果或是臨床狀況。與其他國家相較，台灣第一型糖尿病患者的腎病變有較早期、進展較快速的特徵。台灣地區糖尿病患者所接受的檢驗監測以及相關用藥與照護準則所建議的相較，均有所不足。我們希望提昇糖尿病照護品質，可以由定期檢驗追蹤、針對病情需要用藥、加強對於病人的溝通教育等著手，希望可以讓我們的糖尿病人有好的血糖、血壓、血脂控制，並減少糖尿病腎病變等慢性併發症的發生。

關鍵字

第一型糖尿病 ； 第二型糖尿病 ； 照護品質 ； 專業盡責度 ； 病人特徵 ； 醫院特徵 ； 醫師特徵 ； 檢驗結果認知 ； 自我血糖監測 ； 血糖控制 ； 血壓控制 ； 血脂控制 ； 糖尿病腎病變

並列摘要

Aims:
The diabetic population in Taiwan is estimated to be over 1,000,000. Diabetes is the fourth leading cause of death in Taiwan. Improvement in the quality of diabetes care is critical, both for patients and health care costs. The main purpose of this study is to evaluate the quality of diabetes care in Taiwan.
Study design and methods:
Four study populations were recruited for evaluating quality of diabetes care in different domains. Study population I (NHI-DM population) enrolled diabetic patients randomly sampled from year 2001 National Health Insurance Bureau Taipei Branch claims data. Effects of patient factors or hospital factors on adherence to diabetes care measures were analyzed by logistic regression. Study population II (NTUH-T2DM patients) recruited 875 type 2 diabetic patients who were regularly cared for by endocrinologists (EN), internists (IM) and family medicine physicians (FM) at National Taiwan University Hospital (NTUH) in the year 2002. Adherence to laboratory exams and quality improvement measures (glycemic, blood pressure, and lipid control) of diabetic patients were evaluated. Effects of physician specialty on professional performance or patient health status indicators were analyzed by logistic regression or linear regression, accordingly. Study population III (NTUH-T2DM-Q patients) recruited 512 type 2 diabetic patients who were regularly treated at NTUH and responded to our questionnaire. Awareness of test results and self-management in diabetic patients were assessed. Effects of patient characteristics or physician factors on the clinical awareness, and self-monitoring of blood glucose (SMBG) were evaluated by logistic regression. Study population IV (NTUH-T1DM patients) recruited 471 type 1 diabetic patients who were treated at NTUH in the period between 1995 and 2005. Adherence to diabetes care measures, glycemic control, blood pressure control, lipid control, and long-term clinical outcomes of the study cohort were evaluated. Risk factors for developing diabetic nephropathy were analyzed by logistic regression and Cox regression.
Results:
Adherence rates to plasma glucose, glycated hemoglobin A1C (A1C), urinalysis, renal function test (RFT), lipid profile, liver function test (LFT), and eye ground exam (eye) in the NHI-DM population was 76.3%, 42.7%, 40.2%, 59.7%, 59.2%, 53.2%, and 16.8%, respectively. Patients with hypoglycemic medications, anti-hypertensive drugs, lipid-lowering agents, hospitalization, emergency service visit and frequent visits were more likely to receive exams. Hospitals with different levels, ownerships, locales or qualifications as diabetes care institutions presented different accountability for diabetes care measures. After regression, counts of visits and levels of hospitals persistently affected all the measures.
The adherence rates to annual measures in NTUH-T2DM patients for glucose, A1C, urinalysis, RFT, lipid profile, LFT, and eye exam were 89.7%, 82.5%, 48%, 53.6%, 69%, 45.9%, and 18%, respectively. EN patients showed a significantly better adherence to glucose checkup, A1C measure and urinalysis than IM patients. EN patients also had better adherence to glucose checkup and urinalysis than FM patients. The mean A1C for NTUH-T2DM patients was 7.3 ± 1.5%. About 49.3% patients had an A1C level < 7.0%. EN patients had the lowest mean fasting plasma glucose (FPG) and lowest mean postprandial plasma glucose (PPG). The difference in PPG between EN and IM patients, and the difference in FPG between EN and FM patients were persistently significant following adjustment by patient and physician characteristics.
Percentages of patients who knew their test results were: glucose level 73%, A1C level and meaning 17.9%, proteinuria 33.3%, impaired renal function test 66.7%, liver disease 65.5%, dyslipidemia 70.6%. Unawareness of clinical conditions was remarkable in patients with older age, lower education level, lower income, or less frequent visits. Patients treated with insulin, under the care of older physicians or diabetologists had better awareness of test results. About 49.3% patients performed SMBG. Patients with younger age, longer diabetes duration, higher income, insulin therapy, more comorbidities/complications, under the care of diabetologists were more likely to perform SMBG. Clinical awareness of glucose level or A1C meaning had significant correlations with SMBG.
The NTUH-T1DM patients had a mean latest A1C of 8.9 ± 1.5%. The percentages of NTUH-T1DM patients with retinopathy, neuropathy, nephropathy, cerebral vascular disease, coronary artery disease, and peripheral vascular disease were 23.1%, 15.1%, 25.7%, 1.1%, 4.0%, and 5.5%, respectively. Female, poor glycemic control, and dyslipidemia are risk factors for microalbuminuria. Hypertension, dyslipidemia, and smoking are risk factors for proteinuria. Older age at diagnosis, longer diabetes duration, and hypertension significantly increased risks for end stage renal disease (ESRD). The mean interval between DM diagnosis and ESRD was 13.3 ± 7.7 years. Five patients (1.1%) died. Medications for blood pressure control, renal protection, and lipid control were not prescribed as recommended by practice guideline for diabetes care.
Conclusions:
Hospital-, physician-, and patient- characteristics influence quality of diabetes care. Adherence to A1C measure, urinalysis, and eye ground exam is sub-optimal, especially in primary health care units. Physician specialty had significant effects on adherence to diabetes care measures and outcomes of diabetes care. Clinical awareness and self-care of diabetic patients should be further enhanced. Taiwanese type 1 diabetic patients had a tendency of early onset and rapid progress to develop microvascular complications. However, screening and follow-up exams were inadequate. Medications for preventing and slowing diabetic complications were not promptly prescribed. To improve quality of diabetes care in Taiwan, we should establish Taiwanese diabetes database, enhance adherence to diabetes care measures, encourage effective physician-patient communication, and reinforce fulfillment of the practice guideline.

並列關鍵字

Type 1 diabetes ； Type 2 diabetes ； Quality of diabetes care ； adherence to diabetes care measures ； patient case-mix ； hospital characteristics ； physician specialty ； awareness of test results ； self-management ； glycemic control ； blood pressure control ； lipid control ； diabetic nephropathy

參考文獻

13. Lin WC, Chang HJ, Wang PJ, Chou YJ, Lee CH: Cross-region hospitalization behavior and its related factors in Taiwan. Taiwan J Public Health. 2004; 23: 453-461.

11. Chang YC, Liao HE: The effect of the global budget on the western medicine clinics price and utilization - case of western medicine clinics in the northern Taiwan. Taiwan J Public Health 2002; 21: 363-372.

97. Wei JN, Chung LM, Lin RS, Chao CL, Sung FC: Prevalence and hospitalization rates of diabetes mellitus in Taiwan, 1996-2000. Taiwan J Public Health 2002; 21: 173-180.

7. Weng RH, Shiu RS, Hsieh YC: Current status of diabetes shared care in Taiwan. Formosan J Med 2002; 6: 569-573.

121. Lin CC, Lai MS, Syu CY, Chang SC, Tseng FY: The Accuracy of Diabetes Diagnosis in Health Insurance Claims Data in Taiwan. J Formos Med Assoc 2005; 104: 157-163.