CUSUM管制圖應用在類流感症候群監控上

The CUSUM Control Chart for Influenza-Like Illnesses Syndromic Surveillance

趙英琮(Ying-Tsung Zhao)

指導教授 ： 陳慧芬

中原大學/電機資訊學院/工業與系統工程研究所/碩士(2010年)

https://doi.org/10.6840/cycu201001097

摘要

本論文研究目的主要是應用品管手法中的累積和與指數移動加權平均管制圖來偵測類流感症候群的爆發，傳統的傳染病偵測方法時效性上明顯不足，因此主動式的症候群偵測方式被提出用來及早發現傳染疾病疫情異常現象。
研究資料為2003到2008年健保資料庫之健保申報資料，健保申報資料是以16萬人為樣本， 2003至2008年中，平均一年約有五十萬就診人次，高峰期多發生在10月至隔年4月。類流感症候群ICD-9診斷碼則採用美國疾病管制局之症候群分類標準以及台灣專家會議所共同研商訂定類流感症候群疾病分類碼進行分析。
我們利用累積和與指數移動加權平均管制圖來建立疫情監控機制，但是基本假設為觀察值必須互相獨立，然而本研究資料之間卻有相關性，
因此先使用迴歸模式來建立統計預測模式，然後建立其預測誤差的單邊上累積和與單邊指數移動加權平均管制圖，管制上限須滿足管制內平均連串長度等於50 (ARL0 )。
由於資料間具有相關性，所以假設迴歸模式的誤差項服從ARIMA模式。本研究利用2003到2006年的資料建立迴歸模式與95%預測區間，並使用2007年資料來做驗證及監控2008年的類流感症候群的情形。
本研究所建立的類流感症候群統計預測模式與偵測症候群爆發的累積和與指數移動加權平均管制圖，以2007年資料驗證時，類流感症候群流行的高峰期期間，
所建立出的累積和管制圖確實能偵測出疫情異常，但是建立模式的資料有些情形未發生(例如:元旦發生在禮拜一)而產生錯誤的警訊。經由監控2008年的資料中發現每年的1月到4月與12月都是類流感症候群容易爆發的月份。然而,
以實際資料來計算 是相當困難的，因此由CUSUM以及EWMA管制圖的趨勢來解釋兩個管制圖，由結果得到CUSUM管制圖的趨勢類似於EWMA管制圖。

關鍵字

類流感症候群 ； 症候群偵測 ； 累積和管制圖 ； 移動指數加權平均管制圖 ； 時間序列 ； 迴歸分析 ； ARIMA模式

並列摘要

The purpose of this thesis is to apply the cumulative sum (CUSUM) control chart and exponential weighted moving average (EWMA) of the quality improvement tools to detect the outbreak of the influenza-like illness syndrome. The timeliness to identify the outbreak is not good for the traditional infection surveillance. The active detection “Syndromic surveillance” was developed to early detect the aberration of diseases.
Research data are the 2003 to 2008 data on outpatient and emergency department visits obtained from the National Health Insurance Research Database (NHIRD) and the samples of the data are 160,000 patients. The ICD-9-CM codes of the influenza-like illness syndrome are adopted according to the syndromic classification criteria of CDC, USA and Taiwan. The annual average number of the visits is 500,000.
The peak period of the epidemic for influenza-like illness syndrome is from October to April in the next year.The CUSUM charts and EWMA charts are widely used in disease surveillance, but the basic assumption is that the observations are mutually independent. However, the data in this research are correlated. We use the regression model to establish the statistical forecasting model.
We construct the one-sided upper CUSUM chart and one-sided upper EWMA chart based on forecasting errors in which and then undertake the surveillance mechanism of the number of visits for influenza-like illness syndrome. As a result of the correlated data, we assume that the error term of the regression model follows autoregressive integrated moving average (ARIMA) model.
In this study, we use the data in 2003 and 2006 to establish the regression model and 95 percent prediction interval. The data in 2007 are used to validate and then we apply it to monitor the epidemic of the influenza-like illness syndrome in 2008.In this research, we establish the statistical forecasting model for influenza-like illness syndrome and we construct a CUSUM chart and a EWMA chart that detects the outbreak of influenza-like illness syndrome.
When the CUSUM and EWMA charts are applied in 2007, it is indeed able to detect the aberration of influenza-like illness syndrome during the peak period of the epidemic for influenza-like illness syndrome. But some situation not occurred for research data (for example, the New Year is on Monday), the control chart could be affected to cause the false alarms. By monitoring the data in 2008, January, February, March, April and December are the months that the influenza-like illness syndrome would outbreak easily.
However, our data is actual data and it is difficult to compute the ARL. The results showed that trends of the CUSUM charts are similar to EWMA charts in my research.

並列關鍵字

Cumulative sSum (CUSUM) Control Chart ； Exponential Weighted Moving Average (EWMA) Control Chart ； ARIMA model ； Time Series ； Regression Model

參考文獻

[21] Lai, D., 2005. Monitoring the SARS epidemic in China: A time series analysis. Journal of Data Science 3, 279－293.

[4] Centers for Disease Control and Prevention, USA, 2003. Syndrome Definitions for Diseases Associated with Critical Bioterrorism-associated Agents. Available from: http://emergency.cdc.gov/surveillance/syndromedef/.

[2] Box, G. E. P. and Jenkins, G. M., 1976. Time Series Analysis: Forecasting and Control, revised edition, San Francisco: Holedn－Day.

[6] Cowling, B. J., Wong, I. O. L., Ho, L.-M., Riley, S., and Leung, G. M., 2006. Methods for monitoring influenza surveillance data. International Journal of Epidemiology 35, 1314－1321.

[8] Fricker, R. D. Jr., Hegler, B. L., and Dunfee, D. A., 2008. Comparing syndromic surveillance detection methods: EARS’ versus a CUSUM-based methodology. Statistics in Medicine 27, 3407－3429.