Coma prognosis in children. Part I: definitional and methodological challenges.

: Prediction of outcome from coma is a frequent and important task of neurologists. It is difficult enough in adult patients and even more difficult in children. Part I of this review considers some of the methodological problems and caveats besetting clinical research in this field: the very definition of coma, definition of the study population and outcome variables, study design, the fallacy of self-fulfilling prophecy, early death rate from nonneurologic causes resulting in low statistical power, and invalid attempts to compensate for that by combining outcome categories, lumping together age groups, short and inhomogeneous follow-up, and failure to provide confidence intervals. Part II reviews the clinical pediatric coma-prognosis literature, first according to etiology and then according to electrodiagnostic tests.