開發建置急診症候群即時監控暨偵測系統－－以台北市某醫院為例

Developing a Real-time Emergency Department-based Syndromic Surveillance and Outbreak Detection System-- Using Taipei City Hospitals as Example

症候群偵測系統係指透過資訊科技，蒐集大量的臨床前期症狀(prodromal phase symptoms)，利用時間──空間聚集(spatial-temporal clustering analysis)分析處理後，結合視覺化資訊呈現和異常偵測理論。提供疫情調查人員即時掌握各項症候群／疾病的流行現況，並依此主動進行疫情調查工作，採取適當的防疫因應措施。由於症候群偵測系統是使用症狀資料而不是確診的病例數，因此可用來提早偵測可能的疫情發生，及早提出因應對策。近年來症候群偵測系統已成為各國防範新興傳染病與生物恐怖攻擊之不可或缺的工具。雖然台灣疾病管制局也相繼導入相關技術，然而，醫院結合症候群偵測系統參與防疫作業的經驗仍處於起步階段。本研究嘗試在台北市建置開發急診症候群偵測系統，整合醫院急診資料，利用主訴或國際疾病代碼(ICD-9-CM)，自動監視和偵測相關疾病群組，並提供適當的權限管理和彈性的操作介面，讓院內感染控制與公共衛生決策者監視動態定義疾病症候群的分析結果，以期望及早發現疾病的流行和提升防疫能力。本研究以台北市都會區5間綜合型區域醫院為對象，整合院內急診系統的病患看診相關資料，將其上傳重整後，供症候群偵測系統運算與分析。目前系統已收集自2005年起共三年的來院病患急診相關資料共約五十萬筆；常規化監測的症候群共八種，另隨疾病流行的趨勢，加入五種動態症候群。異常偵測(aberration detection)方法採用歷史資料限制法(historical limit method)，以電子郵件通報發佈發現疾病流行異常的警訊，並可和歷年同期疾病發展之趨勢比較。疾病監視分析結果以網頁呈現為主，也可結合地理資訊系統，查看病例在時間與空間演變、聚集與擴散走向。由於本系統係以地方性疾病監視為主，涵蓋面僅限於部分台北市，未來將考慮分級訂定各種症候群的防疫階段與流程，互補地方中央防疫策略，落實公共衛生防疫政策，以發揮系統之最大功效。

Syndromic surveillance is the utilization of prodromal phase symptoms for the rapid detection of disease outbreaks. In recent years many countries have adopted syndromic surveillance systems as the frontline defense against emerging infectious diseases or bioterrorism attacks. Although Taiwan Center Disease Control (CDC) has established the system, disease control staff and investigators in health agencies and hospitals lack experiences in using syndromic surveillance systems for outbreak prevention of infectious diseases. Therefore, this study focuses on: (1) establishment of an integrated syndromic surveillance system with timely, flexibility, and automatically collecting data from hospital emergency departments (ED) in Taipei City, (2) development of effective algorithms for early detection of disease outbreaks using routinely collected chief complaints or ICD-9 CM codes, and (3) provision of friendly interfaces for presentation of surveillance data to feedback the needs of healthcare workers and decision-makers at different levels in hospitals in order to enhance their the capability in prevention of disease outbreaks. The study selected five hospitals located in different geographical areas of Taipei City. The system has been established, and collected patient data of ED visits (total about 500,000 visits) from January 1 of 2005 to June 30 of 2008. There are 8 syndrome groups for routine surveillance. In addition, the system can allow to define new groups dynamically based on the trends of disease development. Until now there are 5 dynamically defined syndrome groups for extra-targeted surveillance. The historical limit method with the short-term and long-term baselines is used for aberration detection, and the alerts are delivered via email when detected. The analyzed surveillance data can be compared with the trends of disease development in different years. The analyzed surveillance data is presented mainly using web pages. However, they can also be linked to a geographic information system to view the dynamic changes of temporal and spatial patterns, the degree of clustering, and the trends in diffusion of any interested syndromes. Since our system is mainly used for regional disease surveillance (Taipei area), in the future work, we will put our effort in setting and grading various stages of processes in disease defense strategies. So that it can work as a whole with the central government, and get better system performance.