蝮蛇咬伤创面感染的独立危险因素及联合模型的诊断效能分析

Analysis of Independent Risk Factors and Diagnostic Efficacy of a Combined Model for Wound Infection in Patients with Viper Bites

摘要: 目的：探讨蝮蛇咬伤患者继发创面感染的危险因素，并评估相关指标的预测价值。方法：回顾性分析2022年1月~2024年12月某院收治的244例蝮蛇咬伤患者资料，按是否发生创面感染分为感染组(n = 53)与非感染组(n = 191)。采用单因素分析筛选变量，并进行多因素Logistic回归分析；对独立相关指标绘制ROC曲线并构建联合预测模型。结果：创面感染发生率为21.63%。多因素Logistic回归显示，NLR、TT、SSS评分及皮下出血为创面感染的独立危险因素。ROC分析中，NLR预测效能较高(AUC = 0.945)，SSS评分次之(AUC = 0.828)；由NLR、TT、SSS评分及皮下出血构建的联合模型AUC = 0.963。结论：蝮蛇咬伤患者创面感染发生率较高，NLR、TT、SSS评分及皮下出血与感染发生独立相关；联合模型可进一步提高感染风险识别能力，为临床早期分层管理与干预提供依据。

Abstract: Objective: To investigate the risk factors for secondary wound infection in patients with viper bites and to evaluate the predictive value of relevant indicators. Methods: A retrospective study was conducted on 244 patients with viper bites admitted to a hospital between January 2022 and December 2024. According to the occurrence of wound infection, the patients were divided into an infection group (n = 53) and a non-infection group (n = 191). Univariate analysis was first performed to identify potential variables associated with wound infection, followed by multivariate Logistic regression analysis to determine independent risk factors. Receiver operating characteristic (ROC) curves were generated for independent predictors, and a combined predictive model was established. Results: The incidence of wound infection was 21.63%. Multivariate Logistic regression analysis identified neutrophil-to-lymphocyte ratio (NLR), thrombin time (TT), Snakebite Severity Score (SSS), and subcutaneous hemorrhage as independent risk factors for wound infection. ROC analysis showed that NLR had the highest predictive value (AUC = 0.945), followed by SSS (AUC = 0.828). The combined model incorporating NLR, TT, SSS, and subcutaneous hemorrhage demonstrated an AUC of 0.963. Conclusion: Wound infection is relatively common in patients with viper bites. NLR, TT, SSS, and subcutaneous hemorrhage are independently associated with the occurrence of wound infection. The combined model may further improve early identification of patients at high risk of infection and provide a basis for early risk stratification and clinical intervention.