Nomogram model for predicting the risk of multidrug-resistant bacteria infection in diabetic foot patients

Abstract

Objective: This study established an individualized nomogram for predicting the risk of multidrug-resistant bacterial (MDRB) infection in patients with the diabetic foot (DF), and providing a reference for clinical prevention and treatment. Methods: A total of 199 DF patients admitted to the hospital from July 2015 to December 2018 were included in this study. The pathogenic bacteria at the site of infection were detected and the factors affecting the occurrence of MDRB infection in DF patients summarized. The R software was used to draw the nomogram, and the Bootstrap Method used to internally verify the model. The calibration curve and the Harrell’s Concordance Index (C-index) were used to evaluate the predictive effect of the nomogram model. Results: Logistic regression analysis showed that age, course of diabetes, previous use of antibacterial drugs, types of antibacterial drugs, and osteoporosis were risk factors for multidrug-resistant infections in DF (P<0.05). The area under the receiver operating char-acteristic curve (AUC, Area Under Curve) of the nomogram model after internal verification was 0.773 (95% CI: 0.704–0.830). The mean absolute error between the predicted prob-ability of infection in the nomogram and the actual occurrence of MDRB was 0.032, indicating that the nomogram model had good forecasting efficiency and stability. Conclusion: The risk factors for multidrug-resistant infections in DF are age, course of diabetes, previous use of antibacterial drugs, types of antibacterial drugs used, and osteo-porosis. The nomogram model drawn on these risk factors has good predictive accuracy and can assist medical staff in formulating targeted infection prevention strategies for patients.