Clinical, Electroencephalographic, and Neuroradiological Outcome Predictors in Acute Nonhypoxic Encephalopathy

Marked impairment of consciousness, brain lesion on neuroimaging, and nonreactive electroencephalographic (EEG) background activity are established outcome predictors in patients with hypoxic encephalopathy. In this observational cohort study, we aimed to assess the predictive value of clinical, neuroimaging and EEG characteristics for outcome in different types of acute nonhypoxic encephalopathic patients. All adult intensive care unit patients from a tertiary academic medical care center with clinical and EEG evidence of acute nonhypoxic encephalopathy were included from 2004 to 2012. Clinical data, neuroimaging studies, EEG characteristics, and outcome were assessed. In-hospital death was the main outcome. Median age of 262 patients was 65 years (range 18-98 years). Mortality was 12.6%. In Poisson regression analyses, older age (P =.02), intracranial hemorrhage (P =.008), coma (P =.012), and nonreactive EEG background activity (P <.0001) were independently associated with death with nonreactive EEG being the strongest predictor (relative risk 3.74; 95% confidence interval 2.02-6.91). Subgroup analysis revealed no significant effect modification for the predictive value of nonreactive EEG by the presence or absence of coma and different types of acute brain lesions. In conclusion, this study identifies and quantifies the independent predictive value of older age, intracranial hemorrhage, coma, and nonreactive EEG for death, in patients with different types of acute nonhypoxic encephalopathy. These results add further credence to the limited body of evidence that EEG provides important prognostic information in different types of acute encephalopathy not related to hypoxic brain injury. Further studies are warranted to analyze the robustness of this predictor in larger subpopulations with specific etiologies of acute nonhypoxic encephalopathies.