Electronic Health Record Based Disease Mapping Using Data Analytics

Abstract

While wonderful new medical discoveries and innovations are in the news every day, uncertainties and unanswered healthcare questions are a daily reality for the decision makers who provide care. Measuring & visualizing the magnitude of disease burden on population have been considered as one of the necessary inputs for tailored healthcare program design and taking action. The overall purpose of this study is to detect disease burden and make cluster analysis by applying data analytics tools and techniques on EHR. To meet the stated objective, it has been attempted to include sources of the data, methods of data collection, techniques and tools of analysis. Particularly cluster analysis using hierarchical and kmeans algorithms are applied. The results of the study revealed, EHRs very important mainly in saving costs, infrastructure and skilled manpower are available that can support the proper handling of EHR systems. The reporting system is established from bottom up with the support of guidelines and training. However, it has been seen on different national health policies and our observation, there is less use of the system especially the ERM by the physicians and other health experts. Moreover, data analytics tools and techniques that can support EHR data analytics, especially for clustering, was also identified Most of the software on data clustering is open-source software, which is freely available. On the other hand, most of the commercial software comprises implementations of classical algorithms such as k-means or agglomerative clustering. Finally, with the application of the selected clustering techniques, metrics, tools and dataset, it has been attempted to successfully detect optimal number of disease clusters and met the objectives of the study.