6533b822fe1ef96bd127cb25

RESEARCH PRODUCT

Early prediction of coma recovery after cardiac arrest with blinded pupillometry

subject

description

Objectives Prognostication studies on comatose cardiac arrest (CA) patients are limited by lack of blinding, potentially causing overestimation of outcome predictors and self-fulfilling prophecy. Using a blinded approach, we analysed the value of quantitative automated pupillometry to predict neurological recovery after CA. Methods We examined a prospective cohort of 103 comatose adult patients who were unconscious 48 hours after CA and underwent repeated measurements of quantitative pupillary light reflex (PLR) using the Neurolight-Algiscan® device. Clinical examination, electroencephalography (EEG), somatosensory evoked potentials (SSEP) and serum neuron specific enolase (NSE) were performed in parallel, as part of standard multimodal assessment. Automated pupillometry results were blinded to clinicians involved in patient care. Cerebral Performance Categories (CPC) at 1 year was the outcome endpoint. Results Survivors (n=50 patients; 32 CPC 1, 16 CPC 2, 2 CPC 3) had higher quantitative PLR (median 20 [range 13-41] vs. 11 [0-55] %, p<0.0001) and constriction velocity (1.46 [0.85-4.63] vs. 0.94 [0.16-4.97] mm/sec, p<0.0001) than non-survivors. At 48 hours, a quantitative PLR <13% had 100% specificity and positive predictive value to predict poor recovery (0% false-positive rate), and provided equal performance than that of EEG and SSEP. Reduced quantitative PLR correlated with higher serum neuron specific enolase (Spearman's r -0.52, p<0.0001). Interpretation Reduced quantitative PLR correlates with post-anoxic brain injury and, when compared to standard multimodal assessment, is highly accurate in predicting long-term prognosis after CA. This is the first prognostication study to show the value of automated pupillometry using a blinded approach to minimize self-fulfilling prophecy. This article is protected by copyright. All rights reserved.