Related Experiment Videos

A prediction model for targeting low-cost, high-risk members of managed care organizations

Henry G Dove1, Ian Duncan, Arthur Robb

  • 1Division of Health Policy and Administration, Department of Epidemiology and Public Health, Yale University, New Haven, Conn 06520-8034, USA. dove@worldnet.att.net

Insights

This study developed a predictive model to identify health maintenance organization (HMO) members at high risk for future high healthcare costs. The model accurately targets low-cost individuals likely to incur significant expenses within 12 months.

Area of Science:

  • Health Services Research
  • Predictive Analytics
  • Healthcare Management

Background:

  • Identifying high-cost healthcare consumers is crucial for effective resource allocation and cost containment within health plans.
  • Predictive modeling offers a data-driven approach to proactively identify at-risk individuals before significant healthcare expenditures occur.

Purpose of the Study:

  • To develop and validate a predictive model for identifying health maintenance organization (HMO) members likely to incur high healthcare costs.
  • To enable targeted interventions for cost management and improved patient outcomes.

Main Methods:

  • A split-sample multivariate regression analysis was employed using claims data from a 350,000-member HMO.
  • The model incorporated clinical and behavioral variables from 1998 and 1999 claims data to predict costs in subsequent years.
  • Prospective testing involved applying the model to identify low-cost patients (costing <$2000) at high risk for exceeding $2000 in costs the following year.

Main Results:

  • The predictive model identified 47.8% of high-cost patients among top-ranked low-cost individuals in 1998, compared to 14.2% in a randomly selected control group.
  • In a prospective validation, 39.7% of top-ranked low-cost patients in 1999 incurred high costs in 2000, versus 12.2% of the control group.

Conclusions:

  • The developed predictive model effectively identifies low-cost, high-risk HMO members with a high likelihood of incurring substantial medical expenses within the next 12 months.
  • This capability supports targeted disease management and cost-saving strategies for health plans.
Abstract

Related Concept Videos