Nephrology Division, Yitzhak Shamir Medical Center, Zerifin, Affiliated with the Sackler Faculty of Medicine, Tel Aviv University, 70300, Zerifin, Israel.
Emergency Medicine Department, Yitzhak Shamir Medical Center, Zerifin, Affiliated with the Sackler Faculty of Medicine, Tel Aviv University, Zerifin, Israel.
Sci Rep. 2023 Apr 4;13(1):5463. doi: 10.1038/s41598-023-32270-8.
We tested whether CRP combined with the neutrophil-to-lymphocyte ratio (NLR) optimizes the prediction of infectious inflammation in hemodialysis patients. We conducted a retrospective study of 774 (mean age 71.1 ± 12.8 years, 35% women) hemodialysis patients from our institution, hospitalized between 2007 and 2021 for various reasons, with CRP levels available at admission. Infection was defined according to the International Sepsis Definition Conference criteria. An algorithm for the optimal CRP and NLR cutoff points for predicting infection was developed based on a decision tree analysis in the training cohort (n = 620) and then tested in the validation cohort (n = 154). A CRP level above 40 mg/L (obtained as the cutoff point in predicting infections in the training group, using ROC curve analysis) predicted an infection diagnosis with a sensitivity of 75% and a specificity of 76% with an odds ratio (OR) of 9.37 (95% CI: 5.36-16.39), according to a multivariate logistic regression analysis. Whereas, CRP levels above 23 mg/L together with an NLR above 9.7 predicted an infection diagnosis with a sensitivity of 69% and a specificity of 84% with an OR of 25.59 (95% CI: 9.73-67.31). All these results were reproduced in the validation set. Combined use of CRP with NLR may lower the CRP cutoff point in distinguishing between infectious and noninfectious inflammation in hemodialysis patients.
我们测试了 C 反应蛋白（CRP）联合中性粒细胞与淋巴细胞比值（NLR）是否可以优化对血液透析患者感染性炎症的预测。我们对我院 774 名（平均年龄 71.1±12.8 岁，35%为女性）血液透析患者进行了回顾性研究，这些患者因各种原因于 2007 年至 2021 年期间住院，入院时可检测 CRP 水平。感染根据国际脓毒症定义会议标准进行定义。基于训练队列（n=620）中的决策树分析，开发了用于预测感染的最佳 CRP 和 NLR 截断点的算法，然后在验证队列（n=154）中进行了测试。CRP 水平高于 40mg/L（在训练组中使用 ROC 曲线分析预测感染的截断点）预测感染诊断的敏感性为 75%，特异性为 76%，优势比（OR）为 9.37（95%CI：5.36-16.39），这是根据多变量逻辑回归分析得出的。相比之下，CRP 水平高于 23mg/L 加上 NLR 高于 9.7 预测感染诊断的敏感性为 69%，特异性为 84%，OR 为 25.59（95%CI：9.73-67.31）。所有这些结果在验证组中都得到了重现。CRP 与 NLR 联合使用可能会降低 CRP 截断点，以区分血液透析患者的感染性和非感染性炎症。
