近十年來在臺灣有幾次嚴重疾病的大傳染,像是在2003年的SARS或是更近幾年的H1N1,都是迅速的在全台各地造成恐慌,然而要讓疫情受到控制,除了研發新的疫苗供人們施打外,當然是盡量找出在哪些地方會有嚴重的疫情爆發,讓政府能對當地進行強力管制,減少人與人接觸的機會。而對於慢性疾病的研究,就必須要靠長期的時間去統計、記錄各地的疾病分布情況,再去做分析找出造成慢性疾病是否有其外在的環境因素或者生活習慣不同影響患病率的高低。 偵測疾病群聚現象方法不只一種,本論文主要是介紹scan statistic方法與推導過程,並且應用在研究臺灣慢性阻塞性肺病(COPD)在哪些地方會有疫情爆發的情形,由第三章節的統計結果得知在臺灣東半部和一些鄉下的地區都有COPD群聚現象的發生,調查的結果和最初想要找出哪些地區空氣汙染嚴重導致患病率的增加大不相同,其原因在於調查的疾病是屬於慢性的,病患經過一定時間的累積,才會有明顯得病的徵狀,所以患病者在40歲以上的人口佔有大部分,又因鄉村人口老化的因素,使得調查的結果會偏向海邊、山區及一些鄉下地區有COPD明顯群聚的情況發生。
In recent years,there are several high infectious disease in Taiwan,such as SARS in 2003 and H1N1 in recent years. However, to have the epidemic under control,in addition to developing new vaccines for people,try to identify the areas in which there will be a outbreak.Let the Government can strongly controls to reduce the chance of human contact. For chronic disease research, you must spend a lot of time to record and find the distribution of the disease around, and then doing analysis to identify whether there are chronic diseases caused by the environment factors or lifestyle. This paper is to introduce the scan statistic methods and to apply in the study of Taiwanese patients with chronic obstructive pulmonary disease(COPD). Results from the third chapter, the eastern Taiwan and a few rural areas have COPD clustering phenomenon, the results of the survey is unrelated to the Pollutants Standards Index(PSI) of counties in Taiwan. The reason is that the investigation of the disease is chronic,so it must go through a long time to get sick. Therefore, patients over the age of 40 share most of the population. And because the aging of the rural population such that the result will tend to the beach, mountains and a few rural areas have significant clusters.