Reducing Readmission for Sepsis by Improving Risk Prediction Algorithms.

Area of Science:

• Health Services Research
• Clinical Informatics
• Public Health

Background:

• Unplanned readmissions after sepsis occur frequently (17.5%-32%), incurring significant healthcare costs.
• Current sepsis readmission risk models, primarily using clinical indicators, lack predictive accuracy.
• Social determinants of health (SDOH) significantly impact post-discharge outcomes but are underutilized in risk algorithms.

Purpose of the Study:

• To explore the integration of social determinants of health into readmission models for sepsis survivors.
• To enhance the precision and applicability of models for predicting 30-day sepsis readmissions.
• To address health disparities by incorporating SDOH into risk prediction.

Main Methods:

• Review of existing literature on sepsis readmissions and risk modeling.
• Analysis of the impact of social determinants of health on patient outcomes.
• Exploration of methods to integrate SDOH data into predictive algorithms.

Main Results:

• Incorporating social determinants of health significantly improves the performance of predictive models.
• Underutilization of SDOH in current algorithms limits predictive accuracy for sepsis patients.
• Failure to account for SDOH exacerbates health inequities in high-risk populations.

Conclusions:

• Integrating social determinants of health into sepsis readmission risk models is a promising strategy.
• This integration can improve prediction accuracy, reduce readmissions, and optimize care for vulnerable sepsis survivors.
• Future research should focus on refining these models and exploring post-discharge monitoring strategies.