Improving patient matching: Single patient view for Clinical Decision Support using Big Data analytics

In this era of open information and data explosion, Healthcare industry is on a tipping point. Big Data plays a major role in this new change. One of the biggest challenges that the healthcare industry faces while it steps up digitization is the sheer size of the data, speed of generation of this data and complexity arising out of multiple & non-standard formats. Patient data residing in disparate systems is a roadblock to having the right information at the right time. Clinical Decision Support systems need a single view of the patient for making better diagnosis and treatments. Patient identification and matching is a critical challenge in interfacing to the Electronic Health Record (EHR). Different documents and results from various disparate systems like laboratory, pharmacy, claims systems etc. need to be linked to the correct patient record. At this point when healthcare organizations share patient information internally as well as externally, patient records from numerous disparate databases should be connected effectively to guarantee that the decisions made by the clinicians are based on correct patient records and minimizing duplicate information and overheads. This arises the need of improving patient matching for better decision support using single patient view. This paper attempts to study the problem of matching patient records from disparate systems and proposes a solution by using Big Data Analytic techniques like Fuzzy Matching algorithms & MapReduce for better clinical decision support. The main benefits of the proposed system are scalability, cost-effectiveness, flexibility of using any fuzzy algorithm and handling of any data source.