Published data regarding the prognostic utility of conventional MRI in neonates with hypoxic ischemic encephalopathy are reanalyzed by researchers at University of Oxford, UK. Severe abnormalities on conventional MRI in the first week have a sensitivity of 71% and specificity of 84% for very adverse outcome in infants with moderate or Sarnat stage 2 encephalopathy. MR biomarkers alone are not sufficiently accurate to direct treatment-limitation decisions. Limitations in existing prognostic studies include small sample size, selection bias, and overly inclusive outcome assessment. MRI or MR spectroscopy may have a role in combination with other prognostic markers to identify infants with very adverse outcome. Meta-analysis studies do not provide a clear definition of those HIE infants who will have a very severe outcome if they survive. 
COMMENT. Meta-analysis (Thayyil S et al. 2010, 2008) has provided important evidence on the use of MR biomarkers for identification of neonates with moderately severe HIE who might benefit from neuroprotection or early developmental intervention. MR in isolation is not sufficiently accurate to direct treatment-limitation decisions, but used in combination with Sarnat clinical staging (1976) and electroencephalography, MRI may be more reliable. In the current report, Wilkinson provides a list of 77 references related to this topic.