Characterization of pathogenic microorganisms in diabetic foot infections and development of a risk prediction model

Abstract

This study aimed to investigate the distribution of pathogenic microorganisms in diabetic foot infections (DFIs) and develop a nomogram to predict DFIs. It included 136 diabetic foot (DF) patients hospitalized at Henan University Huaihe Hospital from November 2020 to November 2024, with 86 (63.23%) having confirmed infections. Infections were predominantly caused by Gram-positive cocci (54.65%) and Gram-negative bacilli (43.02%). The nomogram incorporated age, C-reactive protein (CRP), Wagner grade, lower extremity arterial disease (LEAD), and peripheral neuropathy (PN). The predictive model exhibited robust discriminatory capacity, achieving an area under the curve (AUC) of 0.803 (95% confidence interval (CI) 0.735-0.878) with internal cross-validation stability (AUC = 0.804). Goodness-of-fit was confirmed by the Hosmer-Lemeshow test (χ² = 5.014, p = 0.756), with excellent net benefit shown by decision curve analysis. Our findings indicate a high infection rate in DF patients, mainly caused by Gram-positive cocci. The nomogram incorporating age, CRP, Wagner grade, LEAD, and PN parameters enables rapid DFIs screening, facilitating timely antibiotic initiation through early infection detection to enhance clinical management.

Keywords: Diabetic foot infection; Pathogenic microorganism; Risk prediction model.