Diagnostic validity of somatosensory evoked potentials in subgroups of patients with sciatica.

The diagnostic utility of scalp-recorded somatosensory evoked potentials (SEP) in patients with sciatica has generally been regarded as low. The purpose of the present study was to determine the validity of sensory nerve SEP in different subgroups of sciatic patients. A total of 65 consecutive patients with sciatica showing disc pathology and/or facet joint hypertrophy on lumbar computed tomography (CT) and/or myelography were studied. Symptomatic myelographically compressed nerve roots were defined as truly compromised roots. Asymptomatic myelographically normal nerve roots were defined as truly normal roots. Bilateral sensory nerve SEP representing nerve roots L4, L5, and S1 were performed in all patients. Evaluation of SEP included the use of P1 latency inter-root comparison. The false-positive rate of SEP was low. Pathological L4, L5, and S1 SEP therefore strongly indicate true compromise of the corresponding nerve roots. The true-positive rate was higher in patients with facet joint hypertrophy with or without additional disc disease than in patients with disc pathology only, and highest if the sciatic sensory symptoms were present during the SEP registration. Diagnostic validity was not influenced by previous episodes of sciatica, the duration of the present episode, or the number of spinal levels with ipsilateral myelographically compressed nerve roots. Pathological SEP strongly indicate sensory radiculopathy in patients with sciatica. Diagnostic efficacy is higher in patients with facet joint hypertrophy than in patients with disc pathology only and highest when the sciatic symptoms are present during registration.