Late awakening, prognostic factors and long-term outcome in out-of-hospital cardiac arrest - results of the prospective Norwegian Cardio-Respiratory Arrest Study (NORCAST).

Abstract

Outcome prediction after out-of-hospital cardiac arrest (OHCA) may lead to withdrawal of life-sustaining therapy if the prognosis is perceived negative. Single use of uncertain prognostic tools may lead to self-fulfilling prophecies and death. We evaluated prognostic tests, blinded to clinicians and without calls for hasty outcome prediction, in a prospective study. Comatose, sedated TTM 33-treated OHCA patients of all causes were included. Clinical-neurological/-neurophysiological/-biochemical predictors were registered. Patients were dichotomized into good/poor outcome using cerebral performance category (CPC) six months and > four years post-arrest. Prognostic tools were evaluated using false positive rates (FPR). We included 259 patients; 49 % and 42 % had good outcome (CPC 1-2) after median six months and 5.1 years. Unwitnessed arrest, non-shockable rhythms, and no-bystander-CPR predicted poor outcome with FPR (CI) 0.05 (0.02-0.10), 0.13 (0.08-0.21), and 0.13 (0.07-0.20), respectively. Time to awakening was median 6 (0-25) days in good outcome patients. Among patients alive with sedation withdrawal >72 h, 49 % were unconscious, of whom 32 % still obtained good outcome. Only absence of pupillary light reflexes (PLR) -and N20-responses in somato-sensory evoked potentials (SSEP), as well as increased neuron-specific enolase (NSE) later than 24 h to >80 μg/L, had FPR 0. Malignant EEG (burst suppression/epileptic activity/flat) differentiated poor/good outcome with FPR 0.05 (0.01-0.15). Time to awakening was over six days in good outcome patients. Most clinical parameters had too high FPRs for prognostication, except for absent PLR and SSEP-responses >72 h after sedation withdrawal, and increased NSE later than 24 h to >80 μg/L.