Facilitating the Exploration of Open Health-Care Data Through BOAT: A Big Data Open Source Analytics Tool

Abstract

Big Data Analytics is the use of advanced analytic techniques on very large data sets to discover hidden patterns and useful information. Many governments publicly release significant amounts of health data, including hospital ratings and patient outcomes. We propose applying Big Data Analytics to understand open health data. Ideally, citizens would use this data to choose hospitals or evaluate care options. There are major challenges, including merging data from disparate sources and applying interpretive analytics. We are building an open-source tool to facilitate analytical exploration. Such a tool could enable researchers, hospitals, insurers, and citizens to obtain integrated global-to-local perspectives on health-care expenditures, procedure costs, and emerging trends. Our tool is based on the Python ecosystem, and contains a variety of modules from database analytics to machine learning and visualization. We analyzed data from the New York Statewide Planning and Research Cooperative System and determined the distribution of costs for hip replacement surgery across the state. The mean cost over 168,676 patients is $22,700, the standard deviation is $20,900, and 88% of these patients had hip replacement costs of less than $30,000. This provides the background to understand why in a state with similar demographics, The California Public Employees’ Retirement System capped hip replacement reimbursements at $30,000, which resulted in significant medical savings. Obtaining transparency of medical costs is important to control expenditure. Even though such information is available, consumers have trouble utilizing it effectively. Our tool could be truly transformative, allowing consumers to fully use available data, and to perhaps demand access to data that ought to be made public.