A retrospective review of 25 patients with traumatic brachial plexus injuries was undertaken to compare pre-operative diagnoses of pre-ganglionic versus post-ganglionic avulsions with intra-operative nerve studies utilizing somatosensory evoked potentials. Preoperative diagnoses were made on the basis of clinical examination, pre-operative imaging (CT myelography or MRI), and electrodiagnostic testing. Of 55 roots considered to have pre-ganglionic avulsion by pre-operative testing, 14 were found to be intact by use of intra-operative nerve testing with somatosensory evoked potentials (25%). This enabled those roots to be utilized for reconstruction (cable grafting, neurolysis, or neurotization). For those roots tested which had been considered intact by preoperative assessment, only 2 were found to be avulsed by intraoperative nerve testing. Of the 14 roots found to be intact, 11 were used for reconstruction in 7 patients. M3-M5 function was achieved in 82% of those reconstructions with an average follow-up of 22 months.
While many questions abound about the components of the pre-operative work-up and whether or not direct visual inspection during neurolysis also may contribute to a change in intra-operative diagnosis, intra-operative nerve testing was seen to change diagnosis and provide useful spinal root for reconstruction in 25% of presumed pre-ganglionic root avulsions and 1/3 of the patients. The authors used their data to calculate a positive predictive value of preoperative diagnosis or root avulsion of 71% and a preoperative positive predictive value of a nerve root being intact 93%. The sensitivity of preoperative exam to nerve root avulsion was 95%, but the preoperative specificity in their study for preoperative nerve avulsion was 65%. They cite this low sensitivity as being the main reason that intraoperative nerve testing is necessary. While their preoperative work-up components are not defined for the cases where diagnostic discrepancy existed, they do import data from another study by Oberle et al (Oberle J, Antoniadis G, Kast E, Richter HP. Evaluation of traumatic cervical nerve root injuries by intraoperative evoked potentials. Neurosurgery 51:1182-8, 2002; Discussion 8-90.) to calculate similar PPV’s, sensitivities, and specificities. While this study is certainly not prospective and it utilizes only one component of a battery of intraoperative nerve testing performed by the authors, it does demonstrate a significant proportion of misdiagnosed nerve injuries that may exist based on preoperative assessment and that, while not universally accepted, there may be a role for intraoperative nerve testing to verify the integrity of brachial plexus nerve roots. It certainly calls for further study using prospective designs.