台灣中老年糖尿病患健康行為對住院服務利用影響之探討

The Impact of Self-care Behaviors on Hospitalization among Diabetes Patients in Taiwan

邱于華(Yu-Whi Chiou)

指導教授 ： 吳淑瓊

國立臺灣大學/公共衛生學院/衛生政策與管理研究所/碩士(2007年)

https://doi.org/10.6342/NTU.2007.00501

摘要

中文摘要
目的：
本研究目的為探討中老年糖尿病患的健康行為和保健行為和住院服務利用的情形，並分析影響住院服務利用的重要因子，提供糖尿病照護策略制定之參考。

方法：
研究資料來自2001年國民健康訪問調查，從中選取四十歲以上自訴有糖尿病的中老年個案之調查資料，連結其次年一年中全民健保住院服務使用申報檔，作為資料分析使用。採用SPSS 12.0電腦套裝軟體，進行適合度檢驗、描述性統計和邏輯斯複迴歸分析。

結果：
本研究發現糖尿病樣本772人當中，在次年一年中有因糖尿病相關疾病使用住院服務者共160人，佔20.7%。個案自我健康管理包含用藥情形、健康行為和保健行為。樣本當中有用藥者佔73.3%，在健康行為方面：有運動習慣者佔39.9%、有飲食控制者佔65.8%、有控制體重者佔23.5%、不抽菸者佔71.2%、生活規律者佔39.4%。上述五項健康行為全部執行者佔8.8%；執行四項、三項、二項、一項者分別佔16.5%、21.5%、22.3%、20.6%；不執行任何健康行為者佔10.2%。保健行為方面，有做視網膜檢查者佔39.4%、有做腎功能尿液微量白蛋白篩檢者佔43.1%、有量血壓者佔92.5%、有驗血糖者佔81.9%、有做血酯肪檢驗者佔57.5%、有接受流行性感冒預防注射者佔23.3%。上述六項保健行為全部執行者佔 5.4%；執行五項、四項、三項、二項、一項者分別佔17.9%、24.6%、26.7%、14.1%、7.8%、不執行任何保健行為者佔3.7%。
双變項分析發現年齡、籍貫、居住城鄉別、收入、共存疾病數目和共存疾病有心臟病、高血壓、腎臟病、中風、糖尿病罹病年數、運動習慣、飲食控制、流行性感冒疫苗注射、累計保健行為總分和前一年住院，對住院服務利用有顯著影響。邏輯斯複迴歸分析發現，在控制其他重要變項後，累計健康行為分數顯著影響住院服務利用，即執行越多項健康行為，其住院服務利用越少（95%信賴區間0.732-0.987）。年齡55-64歲（勝算比2.212；95%信賴區間1.194-4.098）和大於75歲個案（勝算比3.193；95%信賴區間1.503-6.780），其住院服務利用顯著高於40-55歲個案。外省籍（勝算比3.345；95%信賴區間0.999-11.203）和其他包括原住民和金馬籍（勝算比3.574；95%信賴區間1.056-12.095）的個案，其住院服務利用顯著高於客家籍。收入顯著影響住院服務利用，收入越高其住院服務利用越少（95%信賴區間0.772-0.961）。前一年住院個案其住院服務利用顯著高於前一年未住院的個案（勝算比2.950；95%信賴區間1.947-4.471）。

結論：
本研究發現樣本因糖尿病相關相關原因住院率為20.7%，大約每五位糖尿病患就有一位使用住院服務，由此可知糖尿病對有限醫療資源的沉重負擔。在控制其他重要變項後，累計健康行為分數顯著影響住院服務利用，執行越多項健康行為，其住院服務利用越少。此外年齡大、低收入、外省籍、原住民和金馬籍的糖尿病個案，為利用較多住院服務的族群，為糖尿病的照護政策需關注的對象。

關鍵字

糖尿病 ； 住院服務利用 ； 健康行為 ； 保健行為

並列摘要

Objectives:
This study aims to understand health behaviors, preventive care, and inpatient services utilization of middle-aged and elderly diabetic patients; and to identify factors associated with their utilization of inpatient services.The findings are to provide implications for diabetes care strategies.

Methods:
Subjects were from the 2001 National Health Interview Survey, and included those aged 40 years and over who self reported having diabetes.Information on inpatient services utilization was collected by linking with of the National Health Insurance Claim database in the entire year of 2002. Analytical methods used were goodness of fit test, descriptive statistics, and logistic multiple regression with the application of SPSS12.0.

Results:
Among 772 diabetic subjects, 160 (20.7%) utilized diabetes-related inpatient services in the following year. (Self care management included drug use, health behaviors, and health care behaviors.) and 73.3% took medications. With respect to health behaviors, 39.9% had regular physical activities, 65.8% had dietary control, 23.5% had weight control, 71.2% were non-smokers, and 39.4% were with a regular regime. The percentages of taking one up to five of the aforementioned behaviors were 20.6%, 22.3%, 21.5%, 16.5%, and 8.8%, respectively, while 10.2% had none. In terms of preventive care, 39.4% took retina exams, 43.1% had urine micro-albumin exams, 92.5% tested blood pressure, 81.9% had blood sugar tests, 57.5% had blood lipid tests, and 23.3% had flue immunization injections. Diabetic patients who took none of the above preventive care accounted for 3.7%,while those for one up to six items were 7.8%, 14.1%, 26.7%, 24.6%, 17.9%, and 5.4%.
The results of bivariable analyses showed that age, birthplace (ethnic origin), region of residence, income, level of co-morbidity, diagnosis of heart disease, hypertension, renal disease, or stroke,years of diagnosed with diabetes,physical activities, dietary control, flu immunization injection,the score of taking preventive care, and hospitalization in the previous year were significantly associated with inpatient services utilization. After controlling for other important variables, the logistic multiple regressions revealed that the scores of health behaviors had a significantly negative relationship with inpatient services utilization. (95% confidence interval 0.732-0.987). Patients aged between 55-64 (odd ratio 2.212, 95% confidence interval 1.194-4.098) and those aged over 75 (odd ratio 3.193, 95% confidence interval 1.503-6.780) had significantly higher inpatient services utilization than those aged between 40-55. Chinese Mainlanders (odd ratio 3.345, 95% confidence interval 0.999-11.203) aboriginal patients and those from Kinmen and Matsu (odd ratio 3.574, 95% confidence interval 1.056-12.095) had significantly higher inpatient services utilization than Hakka patients. Besides, income was significantly associated with inpatient services utilization, with those of higher income levels having a lower rate of inpatient services utilization (95% confidence interval 0.772-0.961). Patients who had received hospitalization in the previous year had significantly higher inpatient services utilization than those who had not (odd ratio 2.950, 95% confidence interval 1.947-4.471).

Conclusion:
This study reveals that the rate of hospitalization among diabetic patients was 20.7%, indicating that about one out of five diabetic patients would need inpatient services. Thus, diabetes would cause a heavy burden on the limited medical resources. The multivariate analyses showed that taking more health behaviors would lead to less inpatient services utilization. In addition, patients who were older, of lower income, Mainlanders, aborigines or born in Kinmen and Matsu utilized more inpatient services .Health care policy for diabetic management should pay more attention on these high-risk groups.

並列關鍵字

diabetes ； inpatient services utilization ； health behaviors ； preventive behaviors

參考文獻

111.侯佳雯：「糖尿病醫療服務改善方案」試辦計畫對第二型糖尿病人醫療資源利用之影響。臺灣大學醫療機構管理研究所，2005。

114.康翠秀：台北市北投區居民對傳統醫療的知識、信念行為意向與其醫療服務利用之相關研究。國立陽明大學公共衛生學研究所，1995。

96.吳淑瓊、粱浙西、張明正、林惠生、孟蘿拉：台灣老人醫療服務之使用。中華衛誌 1994；13（2）: 168-82。

125.中華民國內分泌暨糖尿病協會www.endo-dm.org.tw/

105.張順全、賴美淑、徐豈庸：台灣地理區域社經差異與糖尿病患罹病死亡的關聯性探討。醫護科技學刊 2005；7（2）：140-8。