Development and Validation of a Risk Prediction Model for Prolonged Hospitalization in Patients With Diabetic Foot Ulcers

Objective  To investigate the risk factors associated with prolonged hospitalization in patients diagnosed with diabetic foot ulcers (DFU), to develop a predictive model, and to conduct internal validation of the model.

Methods  The clinical data of DFU patients admitted to West China Hospital, Sichuan University between January 2012 and December 2022 were retrospectively collected. The subjects were randomly assigned to a training cohort and a validation cohort at a ratio of 7 to 3. Hospital stays longer than 75th percentile were defined as prolonged length-of-stay. A thorough analysis of the risk factors was conducted using the training cohort, which enabled the development of an accurate risk prediction model. To ensure robustness, the model was internally validated using the validation cohort.

Results  A total of 967 inpatients with DFU were included, among whom 245 patients were identified as having an extended length-of-stay. The training cohort consisted of 622 patients, while the validation cohort comprised 291 patients. Multivariate logistic regression analysis revealed that smoking history (odds ratio OR=1.67, 95% confidence interval CI, 1.13 to 2.48, P=0.010), Wagner grade 3 or higher (OR=7.13, 95% CI, 3.68 to 13.83, P<0.001), midfoot ulcers (OR=1.99, 95% CI, 1.07 to 3.72, P=0.030), posterior foot ulcers (OR=3.68, 95% CI, 1.83 to 7.41, P<0.001), multisite ulcers (OR=2.91, 95% CI, 1.80 to 4.69, P<0.001), wound size≥3 cm² (OR=2.00, 95% CI, 1.28-3.11, P=0.002), and white blood cell count (OR=1.11, 95% CI, 1.05 to 1.18, P<0.001) were associated with an increased risk of prolonged length of stay. Additionally, a nomogram was constructed based on the identified risk factors. The areas under the receiver operating characteristic (ROC) curves for both the training cohort and the validation cohort were 0.782 (95% CI, 0.745 to 0.820) and 0.756 (95% CI, 0.694 to 0.818), respectively, indicating robust predictive performance. Furthermore, the calibration plot demonstrated optimal concordance between the predicted probabilities and the observed outcomes in both the training and the validation cohorts.

Conclusion  Smoking history, Wagner grade≥3, midfoot ulcers, posterior foot ulcers, multisite ulcers, ulcer area≥3 cm², and elevated white blood cell count are identified as independent predictors of prolonged hospitalization. Therefore, it is imperative that clinicians conduct a comprehensive patient evaluation and implement appropriate diagnostic and therapeutic strategies to effectively shorten the length of stay for DFU patients.