Please use this identifier to cite or link to this item:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/48237
Title: | 建立臺灣傳染病偵測與流行病學調查整合性疫訊系統及其評估 Establishment and Evaluation of Infectious Disease Surveillance Systems Integrating Epidemiologic Investigation in Taiwan |
Authors: | Jiunn-Shyan Wu 吳俊賢 |
Advisor: | 何美鄉(Mei-Shang Ho) |
Co-Advisor: | 金傳春(Chwan-Chuen King) |
Keyword: | 新興傳染病,傳染病偵測,流行病學,疫情調查,資訊系統,大型群眾集會,生物恐怖應變, Emerging infectious disease,Communicable disease,Surveillance,Epidemiologic investigation,Information system,Mass gathering,Bioterrorism preparedness, |
Publication Year : | 2011 |
Degree: | 博士 |
Abstract: | 流行病學的科學精髓在於預防疾病的應用性,傳染病的防制著重於藉助疫情調查與接觸者健康管理期達及早診斷感染源、治療,繼而以提早阻斷疫情傳播。國際級的大型群眾集會是現代化生活的世界潮流,但在主辦國(地)具有境外移入的新興傳染病與生物恐怖攻擊之潛在危機,建立及時偵測蠢動的疫情,是預防嚴重危害的必要條件。本研究探研台灣公共衛生體系防疫措施的兩個實例經驗,並以統計方法加以評估:(一)建立可彈性調整的傳染病疫情調查資訊系統,並評估疫情調查程序應變時間、主動發現不顯症狀感染者率、群聚疫情延續時間及其對公共衛生之影響;(二)利用2009年在高雄市舉辦的世界運動會,嘗試強化現有傳染病偵測體系與作為,並評估此異常偵測模式在不同季節、不同活動群集性質與不同期的大型群眾集會之適用性。
本研究的目標有三:(一)建立臺灣傳染病偵測與流行病學調查之整合性疫情資訊系統;(二)評估此疫情調查資訊系統對於防疫應變時間、疫情防治與公共衛生的影響;及(三)建立整備臺灣因應大型群眾集會傳染病偵測與防治能力。 做法上,首先在2006年二月完成建置動態彈性調整題庫與問卷的整合性傳染病疫情調查資訊系統時,經由文獻回顧與專家諮商,將疫情調查問題依照詢問內容屬性歸類為風險因子與流行情境模組,提供未來不同疫情需調查(疫調)所需使用的問卷,可依照傳染途徑、風險因子或當發生流行情境時,並設計不同的資料庫元件,如堆積木式可組合成因流行狀況而制宜的模組問題與問卷;並解決通報兩種以上疑似疾病的重覆問題。為評估本系統建置之公共衛生衝擊,同時分析系統建置前與完成前三年(2006至2008年)疫情調查中與問卷設計產生有關的人機介面互動時間(personnel-system interface, PSI)、系統作業時間(system response time, SRT)與人員應變時間(personnel response time, PRT)、疫情調查所能主動發現桿菌性痢疾與德國麻疹的無症狀個案比率於系統建置前後之差異,以及分析這些應變時間對於德國麻疹群聚疫情規模與延續時間之相關性。 為建立大型群眾集會之傳染病偵測與防治模式,運用2009年世運籌辦規劃,以強化現有的傳染病偵測體系,以兩階段(世運前與世運中)進行策略性風險評估,由機動防疫隊進行定點(比賽場點、選手村、旅館與醫院)的症狀/缺勤及蟲媒指數二種主動偵測,及五項強化的例行偵測(入境旅客發燒篩檢、症候群偵測、國際疫情偵蒐、輿情與訛傳偵蒐、飲食衛生監測),同時加以不同季節、不同活動群集性質且長期間舉行的台北市國際花卉博覽會的症候群偵測進行比較。各項偵測結果,再以算數移動平均法(moving average, MA)、指數加權移動平均法(Exponentially weighted moving average, EWMA)與累計和(cumulative sum method, CUSUM)三統計法,比較其異常值偵測情形。 結果發現新建置完成的傳染病疫情調查資訊系統,以流行病學疫情調查題庫為核心(648問卷問題、43個危險因子模組與13個情境模組),每一問卷提供六大方面題庫(人口學、臨床特徵、流行病學暴露特性、環境條件、宿主因子與微生物檢測),且內建35種初階分析功能,包括可繪製流行曲線圖、人口學分析表,以及疾病危險因子分析。因此,此項新型傳染病疫調整合性資訊系統具有設計問卷、自動整併與排序問題、初階分析、接觸者管理與群聚分析等多項功能。分析2006至2008年之疫情調查資料,發現整體的疫情調查時間均縮短,並以人員應變時間與設計新問卷所需的人機介面互動時間縮短幅度最明顯。人員應變時間於系統正式啟用後,由前四個月在9.8至28.8天的區間內,進步到小於十天,並且持續維持(於2008年為0.88 ± 1.52天)。最重要的是其對新興傳染病的時效提升,如設置網路版新興傳染病問卷所需的人機介面互動時間於2007年禽流感H5N1問卷的2.6人時與2008年屈公病(Chikungunya)的3.4人時,均顯較2003年嚴重急性呼吸症候群(severe acute respiratory syndrome, SARS)時的1142.5人時短。此外,公共衛生效益上,以德國麻疹為例,隨著愈晚發生的群聚,其群聚延續的時間縮短(p = 0.019),同時與每個群聚的人員應變時間之縮短(64.8 ± 47.3時至25.2 ± 38.2時)呈顯著正相關(p < 0.0001)。 在建立整備臺灣因應大型群眾集會傳染病偵測與防治能力部分:2009年世運籌備期進行的第一階段風險評估,結果發現高度風險疾病為登革熱;中高度風險疾病有麻疹、腸病毒;而中度風險疾病有新型流感、H5N1流感、屈公病、諾羅病毒(norovirus)、流行性腦脊髓膜炎、西尼羅腦炎(West Nile encephalitis)、德國麻疹與桿菌性痢疾。於2009年7月16至26日世運期間,入境旅客發燒篩檢與定點主動偵測雖均未偵測到法定傳染病或群聚事件,但由回溯性症候群偵測資料分析發現急性呼吸道感染合併發燒有異常值,且與當時疫調發現高市社區中某宗教團體新型流感群聚事件發生期間吻合。此外,回溯採用三統計模式應用於入境旅客發燒、法定傳染病、症候群偵測及蟲媒指數四種偵測資料,發現2009世運期間MA法、EWMA與CUSUM之C1-Mild三法均偵測到同於7月19日有急性結膜炎出現異常值,同時前二法又測得入境旅客發燒僅有一日異常。其他症候群偵測中,只有急性腹瀉、手足口症與疱疹性咽峽炎由EWMA法測得異常值。蟲媒偵測上,成蚊指數與容器指數僅以EWMA法測得異常值,而布氏與家戶兩指數可由MA與EWMA二法測得異常值。另方面,對於不同季節、不同群聚性質且長期間舉行的台北市國際花卉博覽會症候群偵測資料分析,發現只有腸胃道症狀與急性腹瀉由EWMA法與CUSUM(C2-Medium與C3-Ultra)法同時測得於2011年1月份第2周出現異常值,其餘症候群(類流感與手足口症)的異常值僅被EWMA法測得。 結論是新研發的傳染病疫情調查資訊系統可良好整合於現有的公共衛生體系,並能協助各級政府傳染病防治與因應。此新資訊系統所提供的彈性工具與友善介面,有助於防疫人員進行即時性疫情調查資料蒐集、整合與分析,縮短人員應變時間、填補防疫人員經驗斷層、提供更有效率的作業流程,並協助防疫人員提早找出無症狀與輕症的個案,以減少疫情繼續傳播的機會。此外,整備因應大型群眾集會傳染病偵測與防治上,整體而言,比較三統計法上,EWMA法具有最佳的偵測率,而CUSUM法對於平緩的疫情較具鑑別力。因此,針對未來大型群眾集會建議的偵測做法是:(一)宜多種不同的偵測系統與異常偵測模式多管齊下、(二)針對不同症候群的數據特質,應採用不同的異常偵測模式偵測(如:持續平穩發生的腸胃道感染症候群等,可以用CUSUM法,若疫情起伏較大的症候群,可以EWMA法偵測)。未來,將以流行病學疫情調查及大型群眾集會傳染病偵測資訊數據,運用數學模式,以推估傳染病疾病負擔、時空風險度、擬訂衛生政策與前瞻疫情走向。 Background Epidemiologic investigation and contact tracing have been helpful in containing epidemics of infectious diseases. When organizing mass gathering events, host countries must assess the potential risks of importing communicable diseases and confronting bioterrorism attack; these events might lead to the subsequent explosive outbreaks if they were not detected and managed early. This dissertation describes two examples of public health experiences in Taiwan: (1) the establishment of an adjustable epidemiologic information system (AEIS) to assist Taiwan’s public health network in responding to infectious disease outbreaks and to evaluate the performance of AEIS, and (2) establishing an enhance infectious disease surveillance for the 2009 World Games in Kaohsiung, and to evaluate the feasibility of different statistical models that can be applied to other mass gathering events. Methods and Results AEIS - AEIS contains a bank of template questions consists of 648 questions, featuring 43 modules specific to types of risk factors and 13 modules representing for epidemic scenarios. The modules of template questions contain six categories of risk factors: (1) demographics, (2) clinical manifestations (3) epidemiologic characteristics of exposures, (4) environmental conditions, (5) host attributes, and (6) microbiological findings. AEIS has 35 built-in first-tier analysis functions to generate an epidemic curve, demographic tables and a correlation analysis between risk factors and disease. Thus, the innovative AEIS was equipped with functions of questionnaire design, integration/sorting questions, first-tier data analysis, contact management and cluster analysis. After the implementation of the system, the overall response times (RTs) were shortened, strikingly shortening of personnel response time (PRT) and the time needed to draft a new questionnaire that incurred as personnel-system interface (PSI); PRT dropped from a fluctuating range of 9.8 ~ 28.8 days in the first four months to < 10 days in the following months and remained low till 2008 (0.88 ± 1.52 days). The PSIs for newly emerged infectious diseases were 2.6 and 3.4 person-hours for H5N1 in 2007 and chikungunya in 2008, respectively, a much improvement from 1142.5 person-hours for SARS in 2003. The duration of each rubella epidemic cluster showed a shortening trend (p = 0.019) that positively concurred with the shortening of PRT from 64.8 ± 47.3 to 25.2 ± 38.2 hours per cluster (p < 0.0001). Enhanced surveillance for mass gathering – We initiated two-stage strategic risk assessment (SRA) process prior to and during the World Games to identify epidemic-prone diseases. The first stage SRA identified the diseases with three important levels of risk for the 2009 World Games, including dengue as high risk, measles and enterovirus as medium to high risk, and 2009 pandemic H1N1 influenza, avian H5N1 influenza, Chikungunya, norovirus, Neisseria meningitis, West Nile encephalitis, rubella, and shigellosis as medium risk. Although no notifiable diseases and events were detected by the active surveillance and inbound airport fever screening during the 2009 World Games (July 16th-26th, 2009), the acute respiratory infection with fever was identified from hospital ED-based syndromic surveillance that was consistent with a pandemic H1N1 influenza outbreak in a Church group confirmed by epidemiologic investigation through syndrome surveillance in Kaohsiung City. For the aberrations detection, three statistical methods - moving average (MA), Exponentially weighted moving average (EWMA) and cumulative sum method (CUSUM) - were employed. All of the three statistical methods - MA, EWMA and CUSUM (C1-Mild) detected an aberration of acute hemorrhagic conjunctivitis (AHC) on July 19th, 2009, and both of MA and EWMA methods detected one flag of aberration with slightly increased value through inbound airport fever screening on July 21st, 2009. For the syndromic surveillance, aberration flags of acute diarrhea, hand-foot-mouth disease and herpangina were detected by EWMA method. Entomologic surveillance found that all of aberrations of the four mosquito indices (adult, Breteau, house and water-filled container indices) were detected by EWMA method whereas only abnormal Breteau and house indices were also found by MA method. In addition, as for the comparison with another mass gathering event in different season, city, characteristics, and longer duration, we employed the second data set from the Taipei International Flora Exposition (TaipeiExpo). Both gastric illness and acute diarrhea showed aberrations during the second week of January were detected by EWMA and CUSUM (C2-Medium and C3-Ultra) methods, and the influenza-like illness and hand foot mouth disease were detected by EWMA method. Conclusions This first evaluation of the novel instrument, AEIS, demonstrate that it was well integrated into the existing public health infrastructure and can assist Taiwan’s multi-level government for infectious diseases control,. It provided flexible tools and computer algorithms with friendly interface for timely data collection, integration, and analysis; as a result, it shortened RTs, filled in gaps of those personnel lacking sufficient experiences, created a more efficient flow of response, and identified asymptomatic/mild cases early to minimize further spreading. Our experiences on preparedness of enhanced surveillance for mass gathering found that, in general, EWMA method provided the best detection rate, and CUSUM method, on the other hand, worked better for gradually elevated epidemics. Therefore, we recommend for future mass gathering surveillance need: (1) to establish multi-surveillance systems and various aberration detection methods to monitor daily data; (2) to apply the most appropriate aberration detection method according to the data characteristics of the syndrome. In combination with further development, AEIS is anticipated to be useful in the application of other acute public health events needing immediate orchestrated data collection and public health actions. Other possibilities for future directions will include application of mathematical model to better estimate of the true disease burden, proactively respond to the epidemics, evaluate temporal-spatial risk, and formulate public health policy. |
URI: | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/48237 |
Fulltext Rights: | 有償授權 |
Appears in Collections: | 流行病學與預防醫學研究所 |
Files in This Item:
File | Size | Format | |
---|---|---|---|
ntu-100-1.pdf Restricted Access | 3.82 MB | Adobe PDF |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.