Predictive Value of Electrophysiology in Children with Hypoxic Coma Leena D. Mewasingh, MD*, Catherine Christophe, MD † , Christine Fonteyne, MD ‡ , Bernard Dachy, MD § , France Ziereisen, MD  , Florence Christiaens, MD ¶ , Paul Deltenre, MD, PhD # , Viviane De Maertelaer , PhD  , and Bernard Dan, MD, PhD †† Assessment of prognosis of children in hypoxic coma is difficult. The value of clinical evaluation is often lim- ited. The usefulness of electrophysiologic tests has been documented mostly in adults and neonates and in cases of traumatic coma. We reviewed retrospectively 39 consecutive children with nontraumatic hypoxic coma to assess the prognostic value of EEG, visual, and auditory evoked potentials. Correlation between elec- trophysiology and neurologic outcome after mean fol- low-up period of 30 months was significant (r s  0.6, P < 0.001). In contrast there was no correlation between Pediatric Risk of Mortality score (PRISM) and out- come (r s 0.42, P  0.8). Combining magnetic resonance imaging with electrophysiology further en- hanced their prognostic value (r s  0.69, P < 0.001). Neuroimaging was highly sensitive but less specific, and electrophysiologic tests were highly specific but less sensitive. We conclude that early electrophysiology can contribute to predicting outcome in pediatric hy- poxic coma. © 2003 by Elsevier Inc. All rights reserved. Mewasingh LD, Christophe C, Fonteyne C, Dachy B, Ziereisen F, Christiaens F, Deltenre P, De Maertelaer V, Dan B. Predictive value of electrophysiology in children with hypoxic coma. Pediatr Neurol 2003;28:178-183. Introduction Coma in childhood has serious clinical implications [1-6]. Neurologic outcome is often of foremost concern to parents and physicians as it may range from absence of impairment to severe disability or death. Clinical exami- nation on its own has been found to be neither sensitive nor specific in determining prognosis [4]. This finding has led to the use of paraclinical procedures to assess the extent of brain injury. Thus neurophysiologic tests and neuroimaging are commonly undertaken in the acute phase. The prognostic value of these tests has been studied mostly in neonates [7,8] and adults [9-12], demonstrating the usefulness of evoked potentials. Less is known about the potential benefit of these tests in children in whom coma after head injury has been more frequently docu- mented than nontraumatic coma [1,13]. The presence of hypoxia is documented as signifying worse long-term outcome in both traumatic [13] and nontraumatic brain injury [5], although the causes of hypoxic coma can be varied in the pediatric age group (beyond immediate neonatal period until early teens). In a previous study, we analyzed the prognostic value of magnetic resonance imaging (MRI) in children with non- traumatic coma admitted to the pediatric intensive care unit (PICU) of a tertiary children’s university hospital [14]. We found a significant correlation between first MRI and neurologic outcome (sensitivity 96%, positive predic- tive value [PPV] 82%), with maximal PPV when MRI was performed within 72 hours (sensitivity 89%, PPV 100%). In the present retrospective study, we reviewed the same cohort to assess the prognostic value of electrophysiologic investigations (EPI) carried out in routine in this setting. This finding was compared with a standardized tool known to correlate with morbidity/mortality, the Pediatric Risk of Mortality Score (PRISM), used as a predictor of outcome in traumatic coma in children [15]. Methods Study population The study group consisted of 39 children admitted consecutively with hypoxic coma to the PICU of Ho ˆpital Universitaire des Enfants Reine From the * ¶†† Department of Neurology; † Neuroradiology and ‡ Intensive Care Unit, Children’s University Hospital Queen Fabiola, Brussels, Belgium; the §# Department of Neurophysiology, Brugmann University Hospital, Brussels, Belgium; and the  IRIBHM, Statistical Unit, Free University of Brussels, Brussels, Belgium. Communications should be addressed to: Dr. Dan; Professor of Developmental Pediatrics; Children’s University Hospital Queen Fabiola; Ave JJ Crocq 15; 1020 Brussels, Belgium. Received March 19, 2002; accepted September 5, 2002. 178 PEDIATRIC NEUROLOGY Vol. 28 No. 3 © 2003 by Elsevier Inc. All rights reserved. doi:10.1016/S0887-8994(02)00509-X ● 0887-8994/03/$—see front matter