Value of inflammatory biomarkers in predicting amputation risk in diabetic foot: a multicenter prospective cohort study with external validation

BACKGROUND: Diabetic foot amputation severely impacts patient outcomes, while existing risk assessment tools have limited predictive capacity. This study explores the value of inflammatory biomarkers in predicting amputation risk among diabetic foot patients and constructs a high-precision prediction model.METHODS: A multicenter prospective cohort study included 856 diabetic foot patients from three tertiary hospitals (training cohort, January 2022-June 2023) and 324 patients from two additional hospitals (validation cohort, July 2023-June 2024). Inflammatory biomarkers including high-sensitivity C-reactive protein (hs-CRP), interleukin-6 (IL-6), and neutrophil-to-lymphocyte ratio (NLR) were measured. Primary outcome was amputation within 18 months. Competing risk models analyzed risk factors and constructed prediction models with external validation.RESULTS: Training cohort amputation rate was 31.2% (267/856); validation cohort 28.4% (92/324). Fine-Gray competing risk regression identified five independent risk factors: age (sHR=1.032, 95%CI: 1.021-1.044), Wagner grade ≥3 (sHR=3.247, 95%CI: 2.318-4.543), hs-CRP ≥18.5 mg/L (sHR=2.683, 95%CI: 1.947-3.699), IL-6 ≥22.3 pg/mL (sHR=2.154, 95%CI: 1.576-2.943), and NLR ≥7.8 (sHR=1.782, 95%CI: 1.314-2.415). The prediction model achieved C-index 0.847 in training and 0.823 in validation cohorts, with sensitivity 90.5%, specificity 72.8%, and negative predictive value 94.7%.CONCLUSIONS: The inflammatory biomarker-based prediction model demonstrates excellent discrimination and calibration, significantly superior to traditional Wagner classification, providing an important tool for individualized risk stratification and treatment decision-making.