Uncovering Trends in Healthcare Cost and Prevalent Health Issues in Australia through Health Insurance Claims Data Analysis

As healthcare costs continue to rise, one way that private health insurance funds attempt to keep insurance fees affordable is by rationalising their cover programs. This study analysed the claims database from a mid-sized industry fund in order to uncover (1) current and future trends in healthcare costs amongst its members, (2) leading indicators of key conditions needing costly interventions, and (3) a performance evaluation of chronic disease management programs proposed to its members. A de-identified claims database covering a 10-year period between 2008 and 2018 formed the core dataset of this study. An initial phase aimed to verify the consistency of data through time and across various tables, such as membership and claims. Following a descriptive analysis of the most prevalent conditions across various demographic groups, the healthcare cost analysis identified major individual (member level) and overall (fund level) cost contributors. Then, a time series forecasting approach helped predict future trends. The leading indicators analysis employed three methods to detect early warnings of future costly conditions, namely association rule mining, sequential rule mining and a heuristic method. The performance evaluation used a propensity score-based matching method.