NIIS中央資料庫建立及更新機制之研究

The Study of Building NIIS Central Database and Refreshment Mechanism

多年來，行政院衛生署疾病管制局持續努力於國內疾病預防及維護國人保健防疫的工作，對國人自嬰幼兒時期起，由基層衛生局所及合約醫療院所實施多種預防接種措施，擬將原本分別記錄於其所屬縣市衛生局的全國預防接種資訊管理系統之個別資料庫，著手建置一個在中央集中控管的整合型預防接種中央資料庫系統。本研究提出一個有效建置全國預防接種整合型中央資料庫系統及其更新問題解決的機制，從不同層面而言，提供以跨資料庫間的資訊整合架構為重要目的，將分散於各縣市衛生局、衛生所之預防接種相關資料庫進行整合並建置一個有效支援預種管理及決策分析之中央資料庫。系統離型建置測試顯示，資料庫選擇性複製策略在運作上兼具系統整體效能及儲存成本；透過中繼資料處理對應交易紀錄檔之更新機制的自動化處理更提高了資料使用效益與資料取得效率，其效率對微軟之SQL資料庫管理系統進行標竿測試比較結果，在少量資料更新時，雖稍有遜色，但當資料量超過某一數量時，其效率則有超越之勢，但兩者並無明顯的差異。建置完成後，其效益評估顯示，它不僅可提供全國性預防接種快速進行統計分析的資訊化管理功能，更可對疫苗採購庫存管理及配送之最佳化規劃提供最完整的資料來源。

For the past decades, the Center for Disease Control (CDC), Department of Health, Executive Yuan, has continually operated the implementation of the various immunization policies in Taiwan for the prevention of national disease, and maintenance of national health. The immunization records of individual citizen are currently stored at the local database of National Immunization Information System (NIIS) located at each county's and city's bureau of health. Therefore, the CDC launches the project to centralize and integrate the database of NIIS for managing and controlling the database systems more effectively and efficiently. The purpose of this research is to provide a well-solution for implementing experience of building the integrated Central Database with its refresh mechanism for NIIS. In consideration from different perspectives, the objective of this study is to provide an integrated central database implementing and refreshing framework to integrate inoculation relative databases. The resulting benefits show that the system implementation of adopting a selective replication strategy for database design could have the advantages on both effectively enhancing system performance and reducing the computer storage cost. In additions, by using transaction logs with metadata mapping to build an automatic refresh mechanism could obtain better result of enhancing the benefit of data utilization and the efficiency of data retrieval. We also conduct the benchmark testing with Microsoft SQL Server. The result shows when the transaction data are small, the refresh efficiency of our mechanism is a little bit worse, however, when transaction data are larger than a certain amount, our mechanism is apparently better than Microsoft SQL Server, but two systems have no significant difference. After completing the implementation, the central database not only may provide the functions for promptly proceed to generate the statistical analysis for national immunization management, enhance the completion rate of immunization but also may provide the data resource to conduct the optimal planning of vaccine procurement, inventory management and distribution.