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

PURPOSE

Readmission within a period time of discharge is common and costly. Diabetic patients are at risk of readmission because of comorbidities and complications. It is crucial to monitor patients with diabetes with risk factors for readmission and provide them with target suggestions. We aim to develop a nomogram to predict the risk of readmission within 90 days of discharge in diabetic patients.

PATIENTS AND METHODS

This is a prospective observational survey. A total of 784 adult patients with diabetes recruited in two tertiary hospitals in central China were randomly assigned to a training set or a validation set at a ratio of 7:3. Depression, anxiety, self-care, physical activity, and sedentary behavior were assessed during hospitalization. A 90-day follow-up was conducted after discharge. Multivariate logistic regression was employed to develop a nomogram, which was validated with the use of a validation set. The AUC, calibration plot, and clinical decision curve were used to assess the discrimination, calibration, and clinical usefulness of the nomogram, respectively.

RESULTS

In this study, the 90-day readmission rate in our study population was 18.6%. Predictors in the final nomogram were previous admissions within 1 year of the index admission, self-care scores, anxiety scores, physical activity, and complicating with lower extremity vasculopathy. The AUC values of the predictive model and the validation set were 0.905 (95% CI=0.874-0.936) and 0.882 (95% CI=0.816-0.947). Hosmer-Lemeshow test values were p = 0.604 and p = 0.308 (both > 0.05). Calibration curves showed significant agreement between the nomogram model and actual observations. Decision curve analysis indicated that the nomogram improved the clinical net benefit within a probability threshold of 0.02-0.96.

CONCLUSION

The nomogram constructed in this study was a convenient tool to evaluate the risk of 90-day readmission in patients with diabetes and contributed to clinicians screening the high-risk populations.