HSI helps payers identify high-need-high-cost population prospectively – with 80% accuracy
Healthcare consumption is uneven across the covered population. In fact, it is extremely skewed towards the top. The bottom 50% of consumers account for only 3% of healthcare spend while the top 1% accounts for 22% of the spending.
For a payer, ACO, self-insured employer, or any risk bearer, identifying the high-utilizers prospectively (i.e. before the fact) is critically important. Not only can the risk-bearer manage financial risk better but also improve outcomes for this population. Even a small reduction in the utilization by this population, through proactive medical management, can have a meaningful impact on profitability.
Traditional approaches to identifying the HNHC population are not very effective. Risk scores (typically used for identifying HNHC) are a poor predictor of future spending – only 20-25% of the utilization is explained by member risk scores.
Health Sigma’s Artificial Intelligence to the rescue:
Health Sigma (HSI) has applied its Claims Analytics platform capabilities to this problem. HSI’s AI-driven prediction model uses historical claims, co-morbidities, risk scores, diagnosis codes, social determinants of health (SDOH) among others to drive the analysis
HSI Claims Analytics Results:
HSI’s AI-driven analytics identifies the target populations with better accuracy than other competing models.
For a Medicaid plan, HSI identified 2.8% of the total population that will have high utilization, 27% of total spend to be precise.
Prediction accuracy: Back-tested against historical data, 80% of predicted high-utilizers were in the top 10% of utilizers