| Summary: | 碩士 === 國立陽明大學 === 生物醫學資訊研究所 === 101 === The usages of Big Data among medical researches and studies have been increased tremendously by years. One of those Dig Data is called the longitudinal medical claims data. In usual, medical claims data are hold by patients themselves or health insurance companies, but there is a difference in Taiwan. The National Health Insurance (NHI) Administration of Taiwan was established since March 1995; in other words, NHI manages over 99% medical claims data of citizens in past 18 years. All those data are stored in the National Health Insurance Research Database (NHIRD), which becomes an important data source of Evidence-based medicine (EBM) studies.
According to statistics, more and more studies are based on the NHIRD. Due to the information overload and lack of domain-specific analysis tools of NHIRD, it is hard for researchers to extract valuable information from the database without learning any Structured Query Language (SQL). To improve the qualities and efficiencies of NHIRD related researches, this study aims to design a friendly and reusable web-based user interface, which allows users to interact with NHIRD directly without any prerequisite.
The user interface is built on Ruby on Rails web framework and running on Ruby for cross platform compatibility. It runs with the data of Longitudinal Health Insurance Database 2005 under PostgreSQL in production mode.
We present a flexible web interface that users can easily query database and do elementary analysis without programming expertise. It also dynamically draws statistical charts and calculates estimate number of total entries for every query result. Furthermore, it provides several pre-built query conditions for variety purposes and generates the download link of result data set, which can be used to do advanced analysis. It greatly simplifies the data access of NHIRD and assist associated studies more effectively.
|
|---|
