Nomogram Predicting 90-Day Readmission in Patients with Diabetes: A Prospective Study

Area of Science:

• Clinical Medicine
• Diabetes Research
• Health Services Research

Background:

• Hospital readmission is a significant concern, particularly for diabetic patients due to comorbidities.
• Effective risk stratification is crucial for managing diabetic patient readmissions and associated healthcare costs.

Purpose of the Study:

• To develop and validate a nomogram for predicting the risk of 90-day hospital readmission in adult patients with diabetes.

Main Methods:

• A prospective observational study involving 784 diabetic patients from two tertiary hospitals.
• Patients were randomly assigned to training (70%) and validation (30%) sets.
• Multivariate logistic regression was used to build the nomogram, with assessments of depression, anxiety, self-care, physical activity, and sedentary behavior.

Main Results:

• The 90-day readmission rate was 18.6%.
• Key predictors identified for the nomogram included prior admissions, self-care scores, anxiety, physical activity, and lower extremity vasculopathy.
• The nomogram demonstrated strong predictive performance with AUC values of 0.905 (training) and 0.882 (validation), and good calibration.

Conclusions:

• The developed nomogram is a practical tool for assessing 90-day readmission risk in diabetic patients.
• It assists clinicians in identifying high-risk populations for proactive management and intervention.