Key points are not available for this paper at this time.
Background: Perianal infection is a frequent and detrimental complication in patients undergoing hematopoietic stem cell transplantation (HSCT), leading to significant morbidity and extended hospital stays. Despite its clinical significance, a substantial gap remains in our ability to proactively identify high-risk individuals. However a validated predictive model is actually lacking and it is still unclear how to identify these patients before the onset of infection. Objective: This study aimed to develop and to validate a risk prediction model for the accurate assessment of perianal infection risk following HSCT. Methods: A total of 353 patients who underwent HSCT between January 2020 and June 2023 were retrospectively enrolled in this study. The dataset was randomly split into a training cohort (n = 247) and a validation cohort (n = 106) at a ratio of 7:3. Univariate and multivariate logistic regression analyses were employed to identify independent risk factors. Based on these factors, a nomogram prediction model was constructed. Restricted cubic splines were used to assess nonlinear relationships. Model validation included discrimination (AUC), calibration (Hosmer-Lemeshow test), and clinical utility (decision curve analysis). Results: The independent risk factors of perianal infection following HSCT are the type of HSCT, history of perianal disease, diarrhea, neutropenia length, and preoperative prognostic nutritional index. In the training group, the model achieved an AUC of 0.846 (95% CI: 0.775-0.918), an accuracy of 81.0%, a sensitivity of 77.1%, a specificity of 81.9%, and a Youden's index of 0.590. Subsequently, the validation group was assessed, showing an AUC of 0.829 (95% CI: 0.696-0.963), an accuracy of 73.6%, a sensitivity of 87.5%, a specificity of 71.1%, and a Youden's index of 1.586. Model validation showed good calibration agreement. Decision curve analysis demonstrated clinically meaningful net benefit across a wide range of probability thresholds. Restricted cubic spline analysis revealed nonlinear associations of neutropenia length and preoperative PNI with perianal infection risk, with threshold analysis identifying clinically relevant inflection points. Conclusion: This study has constructed a simple, feasible, and visual nomogram for evaluating and early warning of perianal infection following HSCT.
Wu et al. (Thu,) studied this question.