术前NLR联合对数转换降钙素原(log(PCT)) 预测急性重度胆囊炎的价值

Value of Preoperative NLR Combined with log-Transformed Procalcitonin (log(PCT)) in Predicting Severe Acute Cholecystitis

摘要: 目的：探讨对数转换降钙素原(log(PCT))和中性粒细胞–淋巴细胞比值(NLR)单独及联合应用对急性重度胆囊炎的预测价值。方法：回顾性纳入2018年1月至2023年1月本院确诊的急性胆囊炎患者200例，根据东京指南(TG18)严重程度标准分为重度组(58例)和非重度组(142例)。收集患者入院时的PCT、NLR等临床资料。采用单因素及多因素Logistic回归分析筛选急性重度胆囊炎的独立预测因素，并通过受试者工作特征(ROC)曲线评估各指标及联合模型的预测效能。通过决策曲线(Decision Curve Analysis, DCA)分析评估该联合模型在不同阈值概率下的临床净获益。结果：重度组患者的log(PCT)、NLR水平显著高于非重度组(P < 0.001)。多因素Logistic回归分析显示，校正混杂因素后，log(PCT)和NLR是急性重度胆囊炎的独立预测因子(OR = 1.343, 95% CI: 1.020~1.769, P = 0.036; OR = 1.117, 95% CI: 1.025~1.217, P = 0.011)。ROC曲线分析显示，log(PCT)、NLR及联合模型(log(PCT) + NLR)预测急性重度胆囊炎的曲线下面积(AUC)分别为0.804、0.830和0.851。联合模型的AUC显著高于任一单一指标(DeLong检验，P均 < 0.05)，其敏感度和特异度分别为77.6%和81.0%。决策曲线分析显示，在0.05~0.70的阈值概率范围内，联合模型的平均净获益为0.136，高于单用log(PCT) (0.118)和NLR (0.114)。结论：log(PCT)和NLR是急性重度胆囊炎的独立预测因子，由log(PCT)和NLR构成的联合预测模型能更有效地预测急性重度胆囊炎的发生，有助于临床早期识别高风险患者并指导治疗决策。决策曲线分析显示，在广泛的阈值概率范围内(0.06~0.98)，该联合模型均优于“全部治疗”和“不治疗”策略，证实了其良好的临床实用性。

Abstract: Objective: To investigate the predictive value of log-transformed procalcitonin (log(PCT)) and neutrophil-to-lymphocyte ratio (NLR), both individually and in combination, for severe acute cholecystitis. Methods: A total of 200 patients diagnosed with acute cholecystitis at our hospital from January 2018 to January 2023 were retrospectively enrolled. According to the severity criteria of Tokyo Guidelines 2018 (TG18), patients were divided into severe group (n = 58) and non-severe group (n = 142). Clinical data including PCT and NLR at admission were collected. Univariate and multivariate Logistic regression analyses were performed to identify independent predictors of severe acute cholecystitis, and receiver operating characteristic (ROC) curves were used to evaluate the predictive efficacy of each indicator and the combined model. The clinical net benefit of the combined model at different threshold probabilities was evaluated using Decision Curve Analysis (DCA). Results: The severe group exhibited significantly elevated log(PCT) and NLR levels compared to the non-severe group (both P < 0.001). Following adjustment for confounders, multivariate Logistic regression identified log(PCT) and NLR as independent predictors of severe acute cholecystitis (OR = 1.343, 95% CI: 1.020~1.769, P = 0.036; and OR = 1.117, 95% CI: 1.025~1.217, P = 0.011). ROC analysis demonstrated AUCs of 0.804, 0.830, and 0.851 for log(PCT), NLR, and their combination, respectively. The combined model significantly outperformed individual markers (DeLong’s test, all P < 0.05), achieving 77.6% sensitivity and 81.0% specificity. Decision curve analysis demonstrated that within the threshold probability range of 0.05~0.70, the combined model achieved a mean net benefit of 0.136, which was superior to that of log(PCT) alone (0.118) and NLR alone (0.114). Conclusion: Both log(PCT) and NLR are independent predictors of severe acute cholecystitis. The combined prediction model constructed with log(PCT) and NLR can more effectively predict the occurrence of severe acute cholecystitis, facilitating early identification of high-risk patients and guiding clinical treatment decisions. Decision curve analysis demonstrated that the combined model outperformed both the “treat-all” and “treat-none” strategies across a wide range of threshold probabilities (0.06~0.98), confirming its favorable clinical utility.