The widespread use of therapeutic hypothermia for infants with encephalopathy has raised several questions.
Prior to this new(ish) technique, the neurological examination of the infant was thought to have a reasonably good predictive accuracy for long term outcomes. Infants who were never worse than a stage 1 encephalopathy did well, infants who were comatose (stage 3) almost all did very badly. For the ones in between, the addition of other findings, such as seizures that were difficult to control, widespread damage on imaging, effects on other organ systems, etc. increased their risks.
Now that we have an effective treatment (induced hypothermia) that reduces mortality, and reduces long term morbidity, does the clinical exam have the same predictive value? The NICHD Neonatal Research Network has published a few articles from the cohort entered into their hypothermia trial to address this question. The most recent is: Shankaran S, Laptook AR, Tyson JE, Ehrenkranz RA, Bann CM, Das A, Higgins RD, Bara R, Pappas A, McDonald SA et al: Evolution of encephalopathy during whole body hypothermia for neonatal hypoxic-ischemic encephalopathy. J Pediatr 2012, 160(4):567-572 e563. http://www.sciencedirect.com/science/article/pii/S0022347611009322
They recorded serial examinations during the study, half of the babies received treatment with hypothermia. 23% of the 100 surviving babies in the hypothermia group had severe encephalopathy at 24 hours, by 48 hours 94 were still alive, and 21% had severe encephalopathy, by 72 hours there were 89 surviving babies, 21% of them had severe encephalopathy. The authors showed that if a baby still had severe encephalopathy by 72 hours the chance of a poor outcome in both groups, hypothermia and controls, was high.
The cooling trials were launched with many commentators worrying that children would survive with very severe handicap, and that this could be worse than not surviving at all for some children. The trials have been comforting in all showing a decrease in mortality and a decrease in handicap, however some babies already have such severe damage that hypothermia does not help them. This new paper suggests that after 24 hours of cooling, and in particular by 72 hours, if a baby still has signs of severe encaphalopathy the chances of a poor outcome, at least as assessed by a Bayley less than 70 or disabling cerebral palsy as assessed at 18 to 22 months of age, are high.
The Network haf already shown that the neurological exam prior to 6 hours of age is less predictive in infants treated with hypothermia. Ambalavanan N, Carlo WA, Shankaran S, Bann CM, Emrich SL, Higgins RD, Tyson JE, O’Shea TM, Laptook AR, Ehrenkranz RA et al: Predicting Outcomes of Neonates Diagnosed With Hypoxemic-Ischemic Encephalopathy. Pediatrics 2006, 118(5):2084-2093. http://pediatrics.aappublications.org/content/118/5/2084.abstract. In that study the positive predictive value of severe encephalopathy, at 6 hours of age, for death or severe disability was was 0.85 among control infants and only 0.72 among infants receiving hypothermia. That is a big difference, but the thing that caught my eye is the 0.85. That is, 15% of the babies with severe encephalopathy are not dead or severely disabled (keep in mind that severe disability includes a Bayley MDI of less than 70 at 18 to 22 months, which in preterm babies is not very predictive of long term problems (it may be different in term asphyxiated babies).)
That predictive value is less than in other studies with follow up of babies with HIE, probably because of the early timing of the determination. Other studies ( Reviewed in: Perlman M, Shah PS: Hypoxic-Ischemic Encephalopathy: Challenges in Outcome and Prediction. The Journal of Pediatrics 2011, 158(2, Supplement 1):e51-e54. http://www.sciencedirect.com/science/article/pii/S0022347610009662 ) have shown that it is difficult to predict who will do will based on an examination at 6 hours, Yet others have shown that examination later than 6 hours is more predictive. In fact the ‘seminal’ paper of Sarnat and Sarnat which gave us our staging system, started their evaluations between 12 and 24 hours of age.
So basically we know that examining babies in the first 6 hours of life has some predictive ability, but it is by no means perfect, and that examining the babies 6 hours later is better but they need to be cooled by then, so generally giving the benefit of the doubt and initiating cooling is usually the best approach; a decision that can be revisited later if there is no improvement.