Development a risk prediction nomogram for multidrug-resistant bacterial and fungal infection in gastrointestinal fistula patients during the perioperative period.

BACKGROUND

This study aims to develop a risk prediction model for multidrug-resistant bacterial and fungal infections in patients with gastrointestinal fistulas during the perioperative period.

METHODS

A retrospective cohort study was conducted at Anhui No. 2 Provincial People's Hospital from January 2022 to July 2024. We analyzed the distribution, resistance patterns, and mechanisms of multidrug resistance. Univariate and multivariate logistic regression analyses were performed to identify independent risk factors. A nomogram was constructed based on these risk factors, and its performance was evaluated using calibration curves, receiver operating characteristic (ROC) curves, and decision curve analysis (DCA).

RESULTS

A total of 266 patients were included, with 157 (59.02%) testing positive for multidrug-resistant infections. We isolated 329 pathogenic strains: 84 Gram-positive (25.53%), 215 Gram-negative (65.35%), and 30 fungal strains (9.11%). The most common isolate was (57 strains, 17.33%). Patients were divided into a training cohort (n = 177) and a validation cohort (n = 89). Multivariate analysis identified six key indicators: secondary surgery, length of hospital stay, preoperative white blood cell (WBC) count, preoperative neutrophil count, postoperative WBC count, and postoperative C-reactive protein (CRP) levels. The nomogram demonstrated excellent predictive ability, with an area under the curve (AUC) of 0.905 in the training cohort and 0.793 in the validation cohort. Calibration curves indicated high consistency between predicted probabilities and observed values. DCA confirmed the clinical utility of the nomogram.

CONCLUSION

Our study shows that multidrug-resistant infections in patients with gastrointestinal fistulas are predominantly caused by Gram-negative bacilli, especially carbapenem-resistant Enterobacteriaceae. Key risk factors include secondary surgery and various blood count parameters. The developed nomogram provides robust predictive accuracy, aiding healthcare providers in implementing targeted infection prevention strategies.