Context:
Hospitals face financial penalties and increased costs when patients are readmitted within 30 days of discharge. Diabetes is a major driver of readmissions. One of our healthcare clients wanted us to build a predictive model using R programming to identify high risk patients and guide timely interventions.

Resolution:
Our team developed predictive models using logistic regression and random forest to identify patients at high risk of readmission. Random forest provided more accurate predictions than logistic regression.

We also built a Shiny dashboard where clinicians can input patient details and view readmission risk scores, enabling practical application of the model in clinical settings.

Result:
The hospital reduced 30-day readmissions by 12 percent after targeting high risk patients with follow up calls, home visits, and medication reconciliation. This improved patient outcomes and lowered healthcare costs.