Background: To compare magnetic resonance imaging (MRI) and electrophysiologic investigation as prognostic methods in acute head injury.
Methods: Fifty-seven patients suffering moderate to mild (Glasgow Coma Scale score > 8) or severe (Glasgow Coma Scale score < 8) head injury were included. Both groups were analyzed as a total and separately for outcome as assessed by Glasgow Outcome Score. Two outcome groups were separated (Glasgow Outcome Score 1-3: unfavorable vs. Glasgow Outcome Score 4-5: favorable). MRI scans (T1-, T2-, T2*-sequences; transverse, coronal, and sagittal slices) were obtained 1 to 39 days after trauma (mean, 14 days). Electrophysiologic investigations consisted of median nerve-evoked somatosensory responses to assess corticosubcortical function and brain stem auditory-evoked potentials and brain stem reflexes for brain stem function. Recordings were performed 24 to 72 hours after trauma and repeated every 3 to 5 days.
Results: Evaluation of all patients revealed a prognostic significance of MRI lesions within the corpus callosum, the basal ganglia, the hippocampus, the midbrain, and the pons. In the severe head injury subgroup, significance was limited to lesions within the corpus callosum, the basal ganglia, and the midbrain. Among the electrophysiologic findings, dysfunction of the corticosubcortical region as well as of the midbrain were linked to an unfavorable outcome. In severe head injury, prognostic significance was restricted to bilateral corticosubcortical dysfunction. A statistical test for diagnostic convergence of both methods indicated a distinct convergence only for lesions of the midbrain and the pons.
Conclusion: MRI scans performed early after head injury provide several indicators for unfavorable outcome. Electrophysiologic investigations add to this prognostic evidence. Both methods have comparably high specificity. However, because of the higher density of prognostic information obtained, MRI seems superior to electrophysiologic testing.