A prediction model of risk factors of poor wound healing after craniocerebral surgery

OBJECTIVE: To explore the independent risk factors of poor wound healing after craniocerebral surgery, and to generate a risk prediction model. METHODS: A single-center retrospective observational analysis of 160 patients who underwent craniocerebral surgery in The 904th Hospital of the Joint Logistics Support Force of the PLA from February 2018 to February 2021 was carried out. Patients were divided into Group-A (n=70) and Group-B (n=90) according to postoperative wound healing outcome. Logistic regression was used to analyze the independent risk factors, and a nomogram prediction model was constructed using R software. The receiver operating characteristic (ROC) curve was used to test the predictive ability of the model, and the fitting effect was verified by Hosmer Lemeshow. RESULTS: The duration of operation, surgical site infection, diabetes mellitus, and the time of intubation in Group-B were significantly lower than Group-A (P<0.05). Serum albumin (ALB) and hemoglobin (HGB) in Group-B were significantly higher than those in Group-A (P<0.05). Logistic regression analysis showed that long operation duration, surgical site infection, duration of drainage tube, ALB <35g/L, and abnormal HGB were independent risk factors for poor wound healing (P<0.05). The area under the ROC curve (AUC) predicted by the model was 0.932, 95%CI (0.862~1.000). The Hosmer-Lemeshow goodness of fit test showed that the expected probability calculated by the model matched the actual probability (P>0.05). CONCLUSIONS: Long operation duration, surgical site infection, duration of drainage tube, ALB <35g/L, and abnormal HGB were risk factors for poor wound healing. The nomograph model based on these factors showed good discrimination, calibration, and clinical effectiveness in predicting poor wound healing.