摘要：

目的: 分析淋巴细胞/高密度脂蛋白胆固醇比值（LHR）与中至重型颅脑创伤患者入院后28 d内全因死亡率的关系，评估该项指标对早期死亡风险的预测价值。方法: 收集2020年1月1日至2024年8月31日首诊于郑州大学第一附属医院的163例中至重型颅脑创伤患者的临床资料及入院后首次LHR。采用单因素和多因素逐步法Cox比例风险回归分析中至重型颅脑创伤患者入院后28 d内全因死亡率影响因素，绘制受试者工作特征（ROC）曲线并计算曲线下面积，验证LHR的预测效能。结果: 共163例患者入院第28天时生存128例（78.53%），死亡35例（21.47%）。多因素Cox比例风险回归分析：模型1（未调整混杂因素）、模型2[调整年龄、创伤至入院时间、既往糖尿病和入院时Glasgow昏迷量表（GCS）评分]、模型3（进一步调整性别、致伤原因、既往高血压等全部潜在混杂因素）均显示，LHR降低是中至重型颅脑创伤患者入院后28 d内全因死亡率增加的危险因素（HR=0.028，95% CI：0.005~0.156，P=0.000；HR=0.048，95% CI：0.008~0.289，P=0.000；HR=0.032，95% CI：0.004~0.234，P=0.000）。ROC曲线进一步验证LHR的预测效能，其曲线下面积为0.767（95% CI：0.681~0.854，P=0.000），灵敏度为80.50%、特异度为62.90%，最佳截断值为0.460。结论: 入院后首次LHR是简便、易获得的中至重型颅脑创伤患者入院后28 d内全因死亡率的早期预测指标。

关键词: 脑损伤，创伤性, 淋巴细胞, 胆固醇，HDL, 死亡率, 比例危险度模型, ROC曲线

Abstract:

Objective: To analyze the relationship between lymphocyte to high-density lipoprotein cholesterol ratio (LHR) and 28 d all-cause mortality after admission in moderate-severe traumatic brain injury (TBI) patients, and evaluate the predictive value of indicator for mortality risk in moderate-severe TBI patients. Methods: Clinical data and initial LHR after admission of total 163 patients with moderate-severe TBI, who were first diagnosed at The First Affiliated Hospital of Zhengzhou University between January 1, 2020 and August 31, 2024, were collected. Univariate and multivariate stepwise Cox proportional hazards regression analyses were used to identify the influencing factors for 28 d all-cause mortality after admission in moderate-severe TBI patients. The predictive value of LHR was further validated by plotting receiver operating characteristic (ROC) curve and calculating the area under the curve (AUC). Results: By the 28 d after admission, 128 patients (78.53%) survived, and 35 patients (21.47%) died. Multivariate stepwise Cox proportional hazards regression analysis: Model 1 (unadjusted for any confounding factors), Model 2 [adjusted for age, time from injury to admission, diabetes, and Glasgow Coma Scale (GCS) score at admission], and Model 3 (further adjusted for all potential confounding factors such as sex, cause of injury, and hypertension) all showed that a lower LHR was a risk factor for increased 28 d all-cause mortality after admission in moderate-severe TBI patients (HR = 0.028, 95%CI: 0.005-0.156, P = 0.000; HR = 0.048, 95%CI: 0.008-0.289, P = 0.000; HR = 0.032, 95%CI: 0.004-0.234, P = 0.000). The predictive value of LHR was further validated using the ROC curve, with AUC of 0.767 (95%CI: 0.681-0.854, P = 0.000). The optimal cutoff value was 0.460, with a sensitivity of 80.50% and a specificity of 62.90%. Conclusions: The initial LHR after admission is a simple and easily accessible early predictor of 28 d all-cause mortality after admission in moderate-severe TBI patients.

Key words: Brain injuries, traumatic, Lymphocytes, Cholesterol, HDL, Mortality, Proportional hazards models, ROC curve