BACKGROUND Early identification of sepsis in the emergency department (ED) triage is both valuable and challenging. Numerous studies have endeavored to pinpoint clinical and biochemical criteria to assist clinicians in the prompt diagnosis of sepsis, but few studies have assessed the efficacy of these criteria in the ED triage setting. The aim of the study was to explore the accuracy of clinical and laboratory markers evaluated at the triage level in identifying patients with sepsis. METHODS A prospective study was conducted in a large academic urban hospital, implementing a triage protocol aimed at early identification of septic patients based on clinical and laboratory markers. A multidisciplinary panel of experts reviewed cases to ensure accurate identification of septic patients. Variables analyzed included: Charlson comorbidity index, mean arterial pressure (MAP), partial pressure of carbon dioxide (PetCO2), white cell count, eosinophil count, C-reactive protein to albumin ratio, procalcitonin, and lactate. RESULTS A total of 235 patients were included. Multivariable analysis identified procalcitonin ≥1 ng/mL (OR 5.2; p < 0.001); CRP-to-albumin ratio ≥32 (OR 6.6; p < 0.001); PetCO2 ≤ 28 mmHg (OR 2.7; p = 0.031), and MAP <85 mmHg (OR 7.5; p < 0.001) as independent predictors for sepsis. MAP ≥85 mmHg, CRP/albumin ratio <32, and procalcitonin <1 ng/mL demonstrated negative predictive values for sepsis of 90%, 89%, and 88%, respectively. CONCLUSIONS Our study underscores the significance of procalcitonin and mean arterial pressure, while introducing CRP/albumin ratio and PetCO2 as important variables to consider in the very initial assessment of patients with suspected sepsis in the ED. CLINICAL RELEVANCE Early identification of sepsis since the emergency department (ED) triage is challenging Implementing the ED triage protocol with simple clinical and laboratory markers allows to recognize patients with sepsis with a very good discriminatory power (AUC 0.88).

中文翻译：

---

在急诊科分诊时预测脓毒症：在第一次护理评估中实施临床和实验室标记，以提高诊断准确性。


背景 在急诊科 (ED) 分诊中早期识别脓毒症既有价值又具有挑战性。许多研究都致力于确定临床和生化标准，以帮助临床医生及时诊断脓毒症，但很少有研究评估这些标准在急诊分诊环境中的有效性。该研究的目的是探讨在分类水平上评估的临床和实验室标志物在识别脓毒症患者方面的准确性。方法 在一家大型学术城市医院进行了一项前瞻性研究，实施分诊方案，旨在根据临床和实验室标志物早期识别脓毒症患者。多学科专家小组审查了病例，以确保准确识别脓毒症患者。分析的变量包括：查尔森合并症指数、平均动脉压 (MAP)、二氧化碳分压 (PetCO2)、白细胞计数、嗜酸性粒细胞计数、C 反应蛋白与白蛋白比率、降钙素原和乳酸。结果 总共纳入 235 名患者。多变量分析确定降钙素原≥1 ng/mL（OR 5.2；p < 0.001）； CRP 与白蛋白比率 ≥32（OR 6.6；p < 0.001）； PetCO2 ≤ 28 mmHg（OR 2.7；p = 0.031）和 MAP <85 mmHg（OR 7.5；p < 0.001）作为败血症的独立预测因子。 MAP ≥85 mmHg、CRP/白蛋白比值 <32 和降钙素原 <1 ng/mL 表明脓毒症的阴性预测值分别为 90%、89% 和 88%。结论 我们的研究强调了降钙素原和平均动脉压的重要性，同时将 CRP/白蛋白比值和 PetCO2 作为对急诊科疑似脓毒症患者进行初步评估时需要考虑的重要变量。 临床相关性 由于急诊科 (ED) 分诊具有挑战性，因此脓毒症的早期识别具有挑战性 使用简单的临床和实验室标志物实施 ED 分诊方案可以以非常好的辨别力 (AUC 0.88) 识别脓毒症患者。