Somatosensory evoked potentials are not a sensitive indicator of potential positioning injury in the prone patient

This study monitored somatosensory evoked potentials (SSEP) at the median and ulnar nerves in awake volunteers placed in a simulated position for prone surgery. Neurologic symptoms were used as a surrogate endpoint for position related peripheral nerve injury; the occurrence of these symptoms was correlated with the presence or absence of SSEP changes in median and ulnar nerves.

Median and ulnar nerve SSEP data was obtained from awake volunteers in the supine and prone positions. With the head midline in the prone position, SSEPs were measured as the arms were advanced in four cephalad increments. Symptoms, defined as tingling, numbness, or aching in the hand, forearm, or upper arm, were recorded at each position. SSEP changes were considered significant when amplitude decreased 60% and/or latency increased 10% compared with baseline prone measurements. Symptoms were correlated with SSEP changes using chi-squared analysis (p < 0.05), and Fisher's exact analysis (p < 0.07).

Data were collected on 14 patients, mean age 34 +/- 3 years. Seven (50%) subjects reported symptoms with changes in position, while four (29%) subjects displayed SSEP changes. There was no statistically significant association between symptoms and SSEP changes. There were no false positives (no symptoms in the presence of significant SSEP changes), but there were 3 (21%) false negatives (positive symptoms without SSEP changes).

While all SSEP changes were associated with symptoms, the development of symptoms in 3 of 7 patients without SSEP changes suggests that SSEPs may be an imperfect monitor for the detection of positioning injury. The limited sensitivity of SSEPs in this study may be due to the duration of the monitoring, sample size, or the validity of symptoms as a surrogate for nerve injury.