The predictive value of cranial ultrasound and of somatosensory evoked potentials after nerve stimulation for adverse neurological outcome in preterm infants.

Thirty-nine preterm infants were studied to compare the predictive value of somatosensory evoked responses (SEPs) following median-nerve and posterior tibial-nerve stimulation with the predictive value of cranial ultrasound. With regard to the SEP, a normal median-nerve response was by no means a guarantee of a normal outcome. A normal posterior tibial-nerve response, however, almost guaranteed a normal outcome, but the test was very time consuming and the number of false positive responses was high (sensitivity 95.6%, specificity 50%). The presence of parenchymal involvement, either due to a haemorrhage or cystic leukomalacia predicted cerebral palsy with a sensitivity of 95.6% and specificity of 68.5%. The combination of an abnormal posterior tibial response and the presence of parenchymal brain lesions had the best predictive value with a sensitivity of 91.3% and a specificity of 81.2%. These results show that, although posterior tibial-nerve responses have a better predictive value than median-nerve responses, these values were lower than that of cranial ultrasound. The best prediction was obtained when a combination of posterior-tibial responses and cranial ultrasound was used.