Session: Advancing Knowledge in Diabetes

Predicting hospital readmission risk of people with diabetes
Daniel Rubin, Temple University

Hospital readmissions are a costly and undesirable outcome. Given the global prevalence of diabetes and the increased risk of readmission associated with diabetes, people with diabetes suffer a huge burden of readmissions. An understanding of readmission risk is critical to optimize the efficiency of interventions for reducing this risk. Qualitative and quantitative studies have described numerous sources of readmission risk, including health system failure, poor health literacy, social determinants of health, diabetes features, comorbidities, hospital factors, and abnormal laboratory results. Tools designed to predict the readmission risk of individual patients have been developed. One such tool specifically designed for people with diabetes is the Diabetes Early Readmission Risk Indicator (DERRI TM), which has been internally and externally validated. Adding variables to the DERRI TM improves predictive accuracy but decreases the feasibility of use at the point of care. We are working on embedding this tool in an electronic health record system to allow for more complex and accurate models while increasing the feasibility of obtaining readmission risk predictions in routine clinical practice. Using tools like the DERRI TM to identify patients at high risk of readmission enables the targeting of resources for reducing the risk of patients most in need.