Challenge

Homecare patients often face hospitalization, leading to higher costs, increased intervention needs, and elevated insurance claims. Recognizing the critical need to predict and mitigate hospitalization risks, Pennant Group sought a solution to anticipate patient outcomes within 7, 14, and 21 days. This would allow families to prepare for sensitive end-of-life conversations while improving care and resource planning.

Solution

‍
Pennant partnered with v4c.ai and Dataiku to build and operationalize a hospitalization prediction model. The solution provided:

• Risk Scoring: Highlighted patients at high risk of hospitalization within set timeframes.
• Key Factors Analysis: Delivered a narrative of the driving factors behind the risk, offering actionable insights for care teams.

Impact

‍
By implementing this AI-driven model, the organization achieved:

• >75% Prediction Accuracy: The model's reliability enabled proactive intervention, reducing preventable hospitalizations and improving overall patient outcomes.
• Improved Resource Planning: With accurate predictions, the provider optimized staffing and healthcare resources, reducing inefficiencies.
• Enhanced Patient Care: Early identification of hospitalization risks allowed families to have critical conversations in a timely manner, ensuring emotional and logistical preparation for end-of-life events.

Sample Architecture

This initiative not only improved patient care but also set a new standard for leveraging predictive AI in home healthcare settings.

‍