TY  - JOUR
TI  - Development and validation of a novel risk assessment model for necrotizing fasciitis in patients with diabetic foot
AU  - Li, Xun
AU  - Zhang, Xiaoru
AU  - Cheng, Chao
AU  - Ni, Xia
AU  - Huang, Qiuhong
AU  - Tang, Ziwei
AU  - Zhao, Wenrui
AU  - Du, Zhipeng
AU  - Zeng, Qinglian
AU  - Cheng, Qingfeng
PY  - 2026
JO  - Exploration of Endocrine and Metabolic Diseases
VL  - 3
SP  - 101463
DO  - 10.37349/eemd.2026.101463
UR  - https://www.explorationpub.com/Journals/eemd/Article/101463
AB  - Aim: Necrotizing fasciitis (NF) is a severe and early challenging-to-identify complication of diabetic foot (DF). This study aimed to develop and validate a novel risk assessment model for NF with DF patients utilizing conventional clinical indicators. Methods: A retrospective analysis was conducted on 815 DF patients admitted to the First Affiliated Hospital of Chongqing Medical University between October 2018 and April 2022. Based on the presence of NF, patients were stratified into a DF group (n = 703) and a DF complicated with NF (DNF) group (n = 112). Clinically and statistically significant variables were converted into categorical form. A new risk assessment for DNF (NRADNF) nomogram was developed via multivariable stepwise logistic regression. Model performance was evaluated using the area under the receiver operating characteristic curve (AUC) for discriminative ability, the Hosmer-Lemeshow goodness-of-fit test for calibration, decision curve analysis (DCA) for clinical utility, and bootstrap resampling for stability. Results: The final NRADNF model incorporated six indicators: age < 60 years, body temperature ≥ 38°C, foot skin necrosis, neutrophil-to-lymphocyte ratio (NLR) ≥ 8.5, hypersensitive C-reactive protein > 20 mg/L, and hemoglobin ≤ 100 g/L. The model demonstrated favorable predictive performance with an AUC of 0.815 (95% CI: 0.773, 0.857), and it was significantly superior to the RADNF model by our team (P = 0.027). Calibration curves and the Hosmer-Lemeshow test indicated good accuracy. DCA confirmed the model’s clinical net benefit, and internal validation via bootstrap resampling supported its stability. Conclusions: Based on its favorable predictive performance and accessible indicators, the NRADNF model is suitable for preliminary screening of DNF in clinical practice.
ER  -