Predicting high-need cases among new Medicaid enrollees

Area of Science:

• Health Services Research
• Health Economics
• Health Policy

Background:

• Medicaid enrollment presents an opportunity to collect health data.
• Predicting healthcare utilization and costs is crucial for resource allocation.

Purpose of the Study:

• To evaluate a short, self-reported health needs assessment (HNA) for its ability to predict future healthcare utilization and costs among Medicaid enrollees.
• To determine if HNA data improves predictions beyond sociodemographic factors.

Main Methods:

• Retrospective cohort study of 34,087 childless adult Medicaid enrollees in Wisconsin (2009-2010).
• Analyzed self-reported HNAs, medical encounters, and eligibility data.
• Used C-statistic, integrated discrimination improvement, and logistic regression to assess predictive ability for high emergency department (ED) use, inpatient admissions, and costs.

Main Results:

• HNA combined with sociodemographics achieved adequate prediction for high ED use (C-statistic=0.74) and high costs (C-statistic=0.72).
• HNA significantly improved prediction across all outcomes (ED, cost, inpatient) compared to sociodemographics alone.
• HNA enhanced sensitivity and positive predictive value without compromising specificity or negative predictive value.

Conclusions:

• Self-reported health measures collected during Medicaid enrollment are valuable for identifying high-need individuals.
• HNAs can inform targeted interventions and improve healthcare resource management within the Medicaid population.