OBJECTIVE: To determine the value of multimodality-evoked potential recordings in predicting outcome in comatose children. DESIGN: Prospective series and literature review. SETTING: Pediatric ICU in a university hospital. PATIENTS: Forty-one children with a Glasgow Coma Scale score of less than 8, who were admitted to the pediatric ICU between 1984 and 1989. INTERVENTIONS: Forty-one patients underwent brainstem auditory-evoked potential testing within 72 hrs of admission. Of these patients, 37 also had somatosensory-evoked potential testing at the same time. Four patients did not receive somatosensory-evoked potential testing for various nonmedical reasons. MEASUREMENTS AND MAIN RESULTS: Multimodality-evoked potential recordings were used to predict outcome in these comatose children. Outcomes were categorized as bad (death or chronic vegetative state) or good (all other outcomes). Survivor outcomes were determined at discharge and on subsequent follow-up visits from 1 to 3 yrs later. There were no false pessimistic predictions, and two false optimistic predictions in this series. A comprehensive literature review of coma outcome prediction, using multimodality-evoked potential recordings, revealed 20 series with 982 additional patients in whom the predictive errors of false optimism and false pessimism could be determined. Five cases of false pessimism and 99 cases of false optimism were identified in the 982 additional patients. If neonates are excluded, the false pessimism number is reduced to three. CONCLUSIONS: A bad outcome can be reliably predicted using multimodality-evoked potential recordings with little chance of a false pessimistic prediction. The acceptable error of false optimism occurs frequently, since patients often die of progressive neurologic and nonneurologic problems that may or may not be present at the time of the evoked potential recordings. Thus, in comatose children, multimodality-evoked potential recordings are a useful adjunct to clinical examination and other diagnostic aids in predicting outcome and in making decisions regarding the degree of intervention to offer.
OBJECTIVE: To determine the value of multimodality-evoked potential recordings in predicting outcome in comatosechildren. DESIGN: Prospective series and literature review. SETTING: Pediatric ICU in a university hospital. PATIENTS: Forty-one children with a Glasgow Coma Scale score of less than 8, who were admitted to the pediatric ICU between 1984 and 1989. INTERVENTIONS: Forty-one patients underwent brainstem auditory-evoked potential testing within 72 hrs of admission. Of these patients, 37 also had somatosensory-evoked potential testing at the same time. Four patients did not receive somatosensory-evoked potential testing for various nonmedical reasons. MEASUREMENTS AND MAIN RESULTS: Multimodality-evoked potential recordings were used to predict outcome in these comatosechildren. Outcomes were categorized as bad (death or chronic vegetative state) or good (all other outcomes). Survivor outcomes were determined at discharge and on subsequent follow-up visits from 1 to 3 yrs later. There were no false pessimistic predictions, and two false optimistic predictions in this series. A comprehensive literature review of coma outcome prediction, using multimodality-evoked potential recordings, revealed 20 series with 982 additional patients in whom the predictive errors of false optimism and false pessimism could be determined. Five cases of false pessimism and 99 cases of false optimism were identified in the 982 additional patients. If neonates are excluded, the false pessimism number is reduced to three. CONCLUSIONS: A bad outcome can be reliably predicted using multimodality-evoked potential recordings with little chance of a false pessimistic prediction. The acceptable error of false optimism occurs frequently, since patients often die of progressive neurologic and nonneurologic problems that may or may not be present at the time of the evoked potential recordings. Thus, in comatosechildren, multimodality-evoked potential recordings are a useful adjunct to clinical examination and other diagnostic aids in predicting outcome and in making decisions regarding the degree of intervention to offer.