天氣對於皮膚病之相關性研究

The Influence of Weather Factors on Dermatologic Diseases: A Population Based Study

10.6342/NTU201700522

楊天怡

全民健保資料庫 ； 空氣品質監測網 ； 酒糟性皮膚炎 ； 複回歸分析 ； 天氣 ； National Health Insurance Research Database ； air quality ； Rosacea ； multiple regression analysis ； weather factors

國立臺灣大學資訊管理學系學位論文

2017年

碩士

曹承礎

繁體中文

皮膚病雖然大多數為非急性疾病，但卻對於患者的生活有很大的影響，因此若能及早發現或預防，將可以較有效避免或減緩皮膚病的帶來的傷害。隨著資訊科技的發展，健保局的資料庫也逐漸完整記錄台灣民眾的診療明細，本研究想要透過資料分析方法，結合健保資料的大量數據，找出從前多數僅在臨床醫學觀察，卻缺乏大量數據證實與罹患皮膚病有關聯的相關因子。皮膚為接觸外界首當其衝的器官，因此外界天氣與天氣的變化被認為是影響許多皮膚病生成與惡化的因素。其中，多數的醫生認為酒糟性皮膚炎的生成可能受到天氣因子影響，例如溫度變化、紫外線照射等。因此，本研究的目的為針對台灣的患者探討天氣對於酒糟性皮膚炎的影響。 本研究資料包含台灣健保資料庫與行政院環保署的空氣汙染監測資料，在2000-2010年間共有263675筆病例，102053名病患。統計分析了解病患特徵: 罹患酒糟性皮膚炎機率最高的族群為21-30歲的女性；春季(3-5月)罹患此疾病的病例數最多。 研究的結果發現天氣與天氣的變化確實與酒糟性皮膚炎有所關聯，複回歸分析所建立的天氣因子回歸模型對於酒糟性皮膚炎病例數有達到近七成的解釋力(R^2=0.68)，其中最重要天氣因子為一氧化碳濃度、濕度以及紫外線指數。而天氣變化因子的複回歸模型最高的r-square可達0.66，影響力最大的指標為二氧化硫濃度變化、溫度變化以及紫外線指數變化。 此結果除了印證先前文獻探討所說的紫外線、陽光曝曬以及濕度大易導致酒糟性皮膚炎外，同時也發現了先前臨床醫療未曾注意的天氣因子：一氧化碳的濃度以及二氧化硫濃度變化。 另外，研究發現二氧化硫濃度變化、溫度、溫度變化、一氧化碳濃度這四個天氣因子與病例數呈負相關，這個發現值得未來研究再做更深入的探討。

Recently, dermatologic diseases have been common to everyone. Although most of the dermatologic diseases are not fatal, they still have huge influences on patient’s life. So Prediction or early diagnosis of these diseases can be a great contribution to the patients. With the rapid improvement in data analysis techniques and the gradually completed healthcare databases, this study tried to analyze the large volume of patient treatment data to investigate the correlation between specific factors and dermatologic diseases. According to previous clinical studies, many doctors suggested that weather factors might be the triggering factors of Rosacea. However, few studied had really found the correlation with massive data. Therefore, the purpose of this study is to excavate the relation between weather factors and Rosacea with more integral data and improved analysis techniques. The patient treatment data were released by the National Health Insurance Administration Ministry of Health Welfare (NHIA) and the weather data was established by the Environmental Protection Administration. A total of 263,675 cases and 102,053 patients of Rosacea were reported between 2000 and 2010. The gender ratio of these cases are: male vs. female = 24% vs. 76%. The typical high-risk group of the disease was female between 21 to 30 years old and the seasonal pattern of Rosacea with a peak occurred in March to May. The result of this study found that weather and the change of weather did correlate with Rosacea. The multiple regression model showed that all the weather factors had almost 70% predicting ability of Rosacea (R^2=0.68). Among the 10 independent variables, CO concentration, humidity, and ultraviolet index were the best predicting variables for the Rosacea. And the weather change regression model showed a similar result (R^2=0.66). And the three best predicting variables for Rosacea were: changes in SO2 concentration, temperature changes, and ultraviolet changes. This finding supported the previous literature on the cause of Rosacea, i.e., the ultraviolet, the exposure to the sunlight, and the increase of humidity would result in more Rosacea. Meanwhile, in this research, we further found out 2 factors neglected by the previous clinical studies: CO concentration and Changes in SO2 concentration. Besides, this research also found out that changes in SO2 concentration, temperature, change in temperature, CO concentration showed negative correlation with Rosacea. Future studies should be done to dig into these interesting phenomena which are against some of our common senses.

1. [2] C.C. Huang, H.J. Wen, P.C. Chen, T.L. Chiang, S.J. Lin and Y.L. Guo. (2015). Prenatal air pollutant exposure and occurrence of atopic dermatitis. British Journal of Dermatology.
   連結：
2. [4] E. Barati, M. Saraee, A. Mohammadi, N. Adibi and M. R. Ahamadzadeh. (2011). A Survey on Utilization of Data Mining Approaches for Dermatological (Skin) Diseases Prediction. Cyber Journals: Multidisciplinary Journals in Science and Technology.
   連結：
3. [5] HK Kim, JH Park, R Lee, I Sohn, BH Kim, KS Lee. (1995). World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects.
   連結：
4. [6] J., Tickle, K.S., Nahar. (2008). Significant Cancer Risk Factor Extraction: An Association Rule Discovery Approach. International Wrokshop on Data Mining and Artificial Intelligence, (頁 108-114). Khulna.
   連結：
5. [8] Noah Scheinfeld, Thomas Berk . (2010年1月). A Review of the Diagnosis and Treatment of Rosacea. Postgraduate Medicine, 頁 139-143.
   連結：
6. [9] R., Imielinski, T., Swami, A.N., Agrawal, . ( 1993). Mining Association Rules between Sets of Items in Large Databases. ACM SIGMOD International Conference on Management of Data , (頁 207-216). Washington, D.C.,.
   連結：
7. [10] Rajani Katta and Yanqing Zhang. (2002). Medical Data Mining. Data Mining and Knowledge Discovery: Theory, Tools, and Technology IV, 頁 305-308.
   連結：
8. [11] Y.-L. Lee, C.-W. Liw, F.-C. Sungz, and Y.L. Guo. (2007). Increasing prevalence of atopic eczema in Taiwanese adolescents from 1995 to 2001. Clinical and Experimental Allergy, 頁 543-551.
   連結：
9. [14] 林慶和. (2015). 巨量醫療資料之時序事件追蹤與分析.
   連結：
10. [16] 郭純伶. (2012). 異位性皮膚炎與節氣相關探討.
    連結：
11. [17] 傅世靜. (2012). 酒糟性皮膚炎治療心得報告. 北市中醫會刊67期, 頁 100-106.
    連結：
12. [18] 廖桂聲. (2010). r 酒糟鼻」的醫學美容淺析及其臨床. 用藥. 中醫藥研究論叢, 頁 121-131.
    連結：
13. [22] 賴昂廷,林益卿,楊鈺雯,吳美鳳. (2012). 氣候變遷與人類健康. 內科學誌, 343-350.
    連結：
14. [23] 謝邦昌, 張瓈文. (2013). 資料採礦技術應用於全民健保資料庫分析腦中風病患死亡相關因. 頁 165-194.
    連結：
15. [24] 謝富森, 李玉雲, 許明隆. (2004). 酒渣(玫瑰斑): 一個常被忽略或誤診的疾病. dermatol sinica.
    連結：
16. [1] B. WAYNE BLOUNT, M.D., M.P.H., and ALLEN L. PELLETIER, M.D. (2002年8月1日). Rosacea: A Common,Yet Commonly Overlooked, Condition. AMERICAN FAMILY PHYSICIAN, 頁 435-440.
17. [3] Chen, M. S., Han, J., and Yu, P. S. (1996). Data Mining: An Overview From a Database Perspective. IEEE Transactions on Knowledge and Data Engineering (8:6), 頁 866-883.
18. [7] Jiawei Han and Micheline Kambe. (2006). Data Mining Concepts and Techniques. San Francisc: CA: Elsevier Inc.
19. [12] 全民健康保險研究資料庫. (2011).
20. [13] 林信成. (2005). 應用全民健保資料庫分析台灣地區糖尿病患者之醫療服務利用.
21. [15] 問題肌膚保養對策-酒糟性皮膚. 擷取自 台灣雷射美容藥妝館: http://www.taiwancosm.com/skincare/rosacea01.htm
22. [19] 蔡長祐. (2010). 徹底解決臉紅紅的困擾--談酒糟紅臉的治療. 擷取自 尹書田醫療財團法人書田泌尿科眼科診所: http://www.shutien.org.tw/dr_doc_detail.aspx?bookid=214
23. [20] 鄭惠文. 元氣網. 擷取自 皮膚病: http://health.udn.com/disease/sole/1
24. [21] 歷年監測資料下載. 擷取自 行政院環境保護署空氣品質網: http://taqm.epa.gov.tw/taqm/tw/YearlyDataDownload.aspx