頁籤選單縮合
題名 | 臺灣地區境外移入登革熱病例之時間數列分析=A Time Series Analysis of Imported Dengue Fever in Taiwan |
---|---|
作者 | 古桂菊; 許承先; 劉介宇; | 書刊名 | 輔仁醫學期刊 |
卷期 | 8:4 2010.12[民99.12] |
頁次 | 頁199-208 |
分類號 | 415.239 |
關鍵詞 | 境外移入登革熱; 時間數列分析; 自我迴歸整合移動平均模式; Imported Dengue Fever; Time series analysis; ARIMA model; |
語文 | 中文(Chinese) |
中文摘要 | 背景和目的:登革熱(Dengue fever)疫情之監測對於台灣以及東南亞國家而言,一直是高度關注之議題,對於境外移入登革熱之防堵,我國於國際港埠設有發燒篩檢站之檢疫措施,約有40%的境外移入病例藉由檢疫作業及早發現。近年來陸續有泰國、緬甸等國家採用時間數列分析方法(time series analysis),進行登革熱疫情之建模與預測,相關文獻研究顯示此分析方法是一個有效的方法。因此,為有效探討台灣地區境外移入登革熱疫情之特性,本研究之目的在於採用時間數列分析法,探討境外移入登革熱之時間數列特性。方法:資料來源為行政院衛生署疾病管制局,2004至2009年間境外移入登革熱確定病例資料庫,研究變數為每月之境外移入登革熱病例數。使用2004至2008年之資料當作訓練集(training set)以建立ARIMA模式,並使用2009年每月實際值當作測試集(testing set)驗證該模式之預測值。結果:台灣地區境外移入登革熱病例數所建立的模式以ARIMA(1,0,0)(1,0,0)12模式為最佳模式,其具有最高之R-squared、最小之RMSE及normalized BIC。結論:本研究針對境外移入登革熱所建立之季節性ARIMA模式,可運用於預測未來之病例數,建立登革熱傳染病防疫監控機制,作為政府、旅遊業及公衛界決策之參考。 |
英文摘要 | Background and purpose: The monitoring of Dengue fever epidemics has been a top priority in Taiwan and South East Asian countries. To prevent the importation of Dengue fever, there are quarantine stations for fever screening in the international ports and harbors in Taiwan. About 40 percent of imported Dengue fever cases were detected early by those quarantine stations. In addition, Thailand, Myanmar and other countries have recently begun using time series analysis for modeling and prediction of Dengue fever, and literature shows that time series analysis is an effective method to analyze Dengue fever. In order to investigate the characteristics of imported Dengue fever in Taiwan more effectively, this study aims to investigate the characteristics of the time series of imported cases of Dengue fever by means of time series analysis. Methods: The monthly numbers of confirmed imported Dengue fever cases in Taiwan between 2004 and 2009 were obtained from Taiwan CDC, Department of Health. We developed seasonal autoregressive integrated moving average (ARIMA) models using the dataset from 2004 to 2008, and the results were subsequently used to forecast monthly cases in 2009, which were then validated using the 2009 data as testing dataset. Results: The results showed that the ARIMA (1,0,0) (1,0,0)12 model was the best fitted model of the cases of imported dengue fever in Taiwan with the maximal R-squared and the minimal RMSE and normalized BIC. Conclusion: This seasonal ARIMA model was used to forecast the number of imported Dengue fever cases and establish a mechanism of the surveillance of travel-related infectious diseases. The findings should provide a reference for policy-making by the government, tourist industry and public health institutions. |
本系統之摘要資訊系依該期刊論文摘要之資訊為主。