Clinic for Intensive Care Medicine, 30262University Hospital Basel, Switzerland.
Faculty of Medicine, 30262University of Basel, Switzerland.
J Intensive Care Med. 2021 May;36(5):566-575. doi: 10.1177/0885066620913004. Epub 2020 Mar 20.
BACKGROUND/OBJECTIVE: Data regarding delirium in patients presenting with infections of the central nervous system, such as meningitis and/or encephalitis (ME), are scarce. We aimed to determine the frequency and early predictors of delirium in the acute phase of ME.
We assessed clinical, radiologic, and laboratory data of patients with ME at a Swiss academic medical center from 2011 to 2017. The highest Intensive Care Delirium Screening Checklist (ICDSC) score was assessed within 24 hours around lumbar puncture. Multivariable logistic regression was performed to identify predictors of delirium (ICDSC ≥4).
Among 330 patients with ME, infectious pathogens were identified in 41%. An ICDSC >1 was found in 28% with and 19% without identified infectious pathogens. Delirium was diagnosed in 18% with and 14% without infectious pathogens and significantly associated with prolonged in-hospital treatment and mechanical ventilation, more frequent administration of neuroleptics and anesthetics (in 96% with delirium vs 35% without), complications, and less recovery to premorbid functional baseline. Low serum albumin at presentation was the only independent predictor of delirium (area under the receiver-operating curve [AUROC] = 0.792) in patients with pathogens. In patients with infections, the AUROC was smallest for encephalitis (AUROC = 0.641) and larger for patients with meningeal infections (meningitis AUROC = 0.807; meningoencephalitis AUROC = 0.896).
Delirium in the context of ME is seen in almost every fifth patient and linked to prolonged treatment, complications, and incomplete recovery. Among clinical, radiologic, and laboratory parameters, the good calibration and discrimination of low albumin serum concentrations for the prediction of delirium in patients with ME seem promising, especially if meninges are affected.
背景/目的:关于中枢神经系统感染(如脑膜炎和/或脑炎)患者出现谵妄的数据很少。我们旨在确定 ME 急性期谵妄的频率和早期预测因素。
我们评估了 2011 年至 2017 年期间瑞士学术医疗中心 ME 患者的临床、影像学和实验室数据。在腰椎穿刺前后 24 小时内评估了最高的重症监护谵妄筛查检查表(ICDSC)评分。采用多变量逻辑回归来确定谵妄(ICDSC≥4)的预测因素。
在 330 例 ME 患者中,41%确定了感染病原体。在有和没有确定感染病原体的患者中,ICDSC>1 分别为 28%和 19%。在有和没有感染病原体的患者中,分别诊断为谵妄 18%和 14%,并与住院时间延长和机械通气、更频繁地使用神经安定药和麻醉剂(谵妄患者中 96% vs 无感染病原体患者中 35%)、并发症和恢复到发病前功能基线的程度较低相关。在有感染病原体的患者中,低血清白蛋白是谵妄的唯一独立预测因素(AUROC = 0.792)。在感染患者中,脑炎的 AUROC 最小(AUROC = 0.641),脑膜感染患者的 AUROC 较大(脑膜炎 AUROC = 0.807;脑膜脑炎 AUROC = 0.896)。
ME 患者中出现谵妄的比例接近五分之一,与治疗时间延长、并发症和不完全恢复有关。在临床、影像学和实验室参数中,低血清白蛋白浓度对 ME 患者谵妄预测的良好校准和区分似乎很有希望,尤其是脑膜受累时。