Reducing Avoidable Admissions in Medicaid Populations

This case study details the efforts of a large Integrated Delivery Network (IDN) in the Southwest to reduce avoidable admissions among high-risk Medicaid plan members. The IDN implemented Lightbeam AI’s 30-day Avoidable Admission model to identify members at risk of avoidable admissions. The model utilizes patient-specific clinical and socio-economic data for risk stratification, allowing care managers to prioritize outreach and intervention. Over a 12-month period, the IDN identified an average of 4,200 high-risk members monthly and intervened with approximately 150 members each month. The intervention led to a 5.3% reduction in the admission rate, translating to a 43.3% relative reduction and an estimated cost savings of $637,000. The study also evaluates the AI model's performance, noting that the top 8% of high-risk members accounted for over 50% of potential admissions, demonstrating the effectiveness of targeted interventions. The findings underscore the importance of integrating AI in healthcare for improved patient outcomes.