運用資料探勘技術於六大死因慢性疾病之研究

The Study of the Six Major Causes of Death Chronic Diseases by Data Mining

吳瑞堯(Rei-Yao Wu) ； 周駿賢(Chun-Hsien Chou)

《資訊管理學報》 18卷1期 (2011/01) Pp. 187-211

https://doi.org/10.6382/JIM.201101.0187

摘要

心臟疾病、腦血管疾病、糖尿病、肝臟病變、腎臟病變、高血壓性疾病等慢性疾病一直名列我國國民重大死因，衛生署公布的「2008年國人十大死因」疾病中，慢性疾病佔總死亡人數的六成以上，顯示慢性疾病已成為當前危害國人身體健康的致命殺手。國家衛生研究院的全民健康保險研究資料庫包含了大量的病患門診就醫資料，這些資料中可能隱藏許多未知而有用的知識，因此本研究以該資料庫為基礎，經資料淨化與篩選後，進行統計分析，找出上述六大死因慢性疾病在不同生活圈、年齡層及性別上之盛行率，並利用資料探勘技術找出不同生活圈、性別、年齡之慢性疾病間的關聯性。
研究結果發現，高血壓性疾病盛行率男女性分別為11.37%、9.24%，均居6種慢性疾病之冠，65歲以上族群盛行率更高達37.61%，且高血壓性疾病不分生活圈、性別、年齡，均與其他慢性疾病密切關聯；男性在6種慢性疾病的盛行率均高於女性，但離島生活圈女性在高血壓性疾病、心臟疾病、糖尿病之盛行率皆高於男性；一般生活圈65歲以上高血壓性疾病及心臟疾病盛行率高達47.79%及30.11%，明顯高於都會生活圈；離島生活圈65歲以上高血壓性疾病及腦血管疾病盛行率達59.37%及21.87%，明顯高於另兩個生活圈；肝臟病變在40~64歲之盛行率高達10.88%，反而高於65歲以上的8.03%；心臟疾病在65歲以上族群中與其它慢性疾病間都有高度關聯；在40~64歲中患有腎臟病變及肝臟病變者很可能同時會有高血壓性疾病；女性心臟疾病與高血壓性疾病、糖尿病間的關聯較男性明顯。

關鍵字

慢性疾病 ； 全民健康保險研究資料庫 ； 資料探勘

並列摘要

Chronic diseases including Heart Disease, Cerebrovascular Disease, Diabetes, Liver Disease, Renal Disease, and Hypertension Disease are always listed in the top ten leading causes of death. Chronic diseases account for more than 60% of death toll according to the report of ten leading causes of death in 2008 published by the Department of Health. This fact shows that the chronic diseases have become the deadly health killer currently. The National Health Insurance Research Database that is maintained by the National Institutes of Health has accumulated a large volume of patient treatment information, and maybe hidden useful knowledge there. Applying data cleaning, analysis and mining, this research investigates the population rate of these chronic diseases and the association rules among them in each life circles, sexes, and ages.
The research reveals that the Hypertension Disease is the most popular disease among the six chronic diseases. The population rates are 11.37% in male and 9.24% in female, and 37.61% in the above 65 years old group. Despite of life circle, sex, and age, it is tightly associated with other chronic diseases. The population rates of male are higher than that of female for all the six chronic diseases in general. But the population rates of females in off-shore island life circle are greater than that of males there in Hypertension Disease, Heart Disease and Diabetes. The population rates of the above 65 years old group in the general life circle are 47.79% for Hypertension Disease and 30.11% for Heart Disease, highly beyond that in the metro life circle. The population rates of the above 65 years old group in the off-shore island life circle are 59.37% for Hypertension Disease and 21.87% for Cerebrovascular Disease, respectively, also highly beyond that in the other two life circles. The population rate of the 40~64 years old group for Hepatitis and Cirrhosis is 10.88%, higher than 8.03% of the above 65 years old group. In the above 65 years old group, Heart Disease is highly associated with the other chronic diseases. In the 40~64 years old group, Renal Disease and Liver Disease are associated with Hypertension Disease. And, the association of Heart Disease with Hypertension Disease and Diabetes in female is more significant than that in male.

並列關鍵字

Chronic Disease ； National Health Insurance Research Database ； Data Mining

參考文獻

連賢明(2008)。如何使用健保資料進行經濟研究。經濟論文叢刊。36(1)，115-143。

內政部統計處，2008，老人狀況調查摘要分析 (available online at http://sowf.moi.gov.tw/stat/Survey/94old.doc)。

中華民國糖尿病衛教學會網站，2009，糖尿病與高血壓 (available online at http://www.tade.org.tw/05/web/糖尿病與高血壓.pdf)。

全民健康保險研究資料庫網站，2008a，資料庫內容說明 (available online at http://www.nhri.org.tw/nhird/date_01.htm)。

全民健康保險研究資料庫網站，2008b，承保抽樣歸人檔 (available online at http://www.nhri.org.tw/nhird/date_cohort.htm)。