Data blending in health care : Evaluation of data blending

Sammanfattning: This report is aimed at those who are interested in data analysis and data blending. Decision making is crucial for an organization to succeed in today’s market. Data analysis is an important support activity in decision making and is applied in many industries, for example healthcare. For many years data analysts have worked on structured data in small volumes, with traditional methods such as spreadsheet. As new data sources emerged, such as social media, data is generated in higher volume, velocity and variety [1]. The traditional methods data analysts apply are no longer capable of handling this situation. Hence scientists and engineers have developed a new technology called data blending. Data blending is the process of merging, sorting, joining and combining all the useful data into a functional dataset [2]. Some of the well-known data blending platforms include Datawatch, Microsoft Power Query for Excel, IBM DataWorks and Alteryx [3]. Synergus AB is a consulting company engaged in health economics, market access and Health Technology Assessment (HTA) [4]. The company does analysis for their clients. Unfortunately the way they work is not efficient. New tools and methods need to be applied in the company. The company has decided to apply data blending in their daily work. My task in this project was to build datasets for analysis and create workflows for future use with a data blending platform. For my interest, I did a research on data blending to understand how this new technology works. During the project I have worked with four data sources. These were Microsoft Excel worksheet, CSV file, MS Access database and JSON file. I built datasets the company needs. I also preceded a case study on data blending process. I focused on the three steps of data handling, namely input, process and output. After the project, I reached a conclusion that data blending offers better performance and functionality. It is easy to learn and use, too.