Better data, better outcomes and better value with analytics.

Abstract

BACKGROUND/ INTRODUCTION: Healthcare executives and administrators have a wealth of data accessible through several data sources. Yet they typically remain unsure as to how much of that data is usable or even how they can use that data to improve outcomes. Data science is now a widely adopted approach to leveraging such data sets in the healthcare domain to enhance quality and reduce cost of care. This project develops a model to predict readmission occurrences across various clinical specialties. It starts with understanding the data and their dictionaries as applicable, then will work on cleaning the data and preparing it for the third stage which involves the use of data science techniques and tools. METHODLOGY: This research is exploratory in nature using a multiple case study approach. The chosen cases focus on spinal surgery, colorectal, urology, cardiothoracic, and orthopaedic. Further, the study consists of 6 key phases as follows: Phase 1: Understanding the structures of the various data sets, Phase 2: Data wrangling, Phase 3: Modelling, Phase 4: Evaluations, Phase 5: deployment and Phase 6: Final evaluation and feedback exchange. RESULTS AND CONCLUSIONS: This is a work in progress. To date, we have developed the appropriate conceptual model which identifies the important key data elements that must be incorporated, from where and how they will be collected as well as key aspects of their respective data structures. In addition, we are fine turning the necessary data preparation strategies. given the changes to the healthcare environment and the likelihood of a bundles payment structure to be adopted, similar to the US healthcare environment, this study represents a strategic necessity so that healthcare providers can be prepared and ready to operate effectively and efficiently in a new value-based healthcare environment; ie so they have better data, better outcomes and better value.