Prediction of Poor Outcome in Coma Patients Resuscitated from Cardiac Arrest due to Submersion

ABSTRACT

PURPOSE: This study was conducted to predict poor outcomes in coma patients resuscitated from cardiac arrest due to submersion. METHODS: We retrospectively investigated 27 submersion victims who were comatose for more than 6 hours after cardiac arrest. Vegetative state and death (GOS 1-2) were classified as poor outcomes whereas all other outcomes (GOS 3-5) were classified as good outcomes. Demographic, clinical, laboratory, radiologic, eletrophysiologic, and treatment variables were evaluated for their usefulness in predicting outcome. RESULTS: Of the 27 patients, 9 (33%) presented with good outcomes. Among the many variables, a GCS score of 3-5 in the first 24 hours, absence of pupillary reactions in the first 24 hours or on day 3, motor response to pain flexion or less (M1 - 3) or no motor response to pain (M1) in the first 24 hours or on day 3, bilateral absence of N20 on SEP showed significant differences between the two outcome groups (p<0.05). Prognostic factors with 100% specificity for prediction of poor outcome included absence of pupillary reactions in the first 24 hours or on day 3, no motor response to pain (M1) on day 3, and bilateral absence of N20 on SEP. CONCLUSION: The factors including chest radiography, body temperature, and laboratory findings which were associated with submersion showed no significant differences between the good and the poor outcome groups. Absence of pupillary reactions in the first 24 hours or on day 3, no motor response to pain (M1) on day 3, and bilateral absence of N20 on SEP helped predict poor outcome in submersion patients resuscitated from cardiac arrest.