Clinical and electrophysiologic correlates of quantitative sensory testing in patients with incomplete spinal cord injury.

Objective: To determine the degree of association among indices of preserved sensation derived from quantitative sensory testing (QST), somatosensory evoked potentials (SEPs), and the clinical characteristics of patients with spinal cord injury (SCI).

Design: A controlled correlational study of diverse measures of preserved sensory function.

Setting: Regional SCI rehabilitation center in Ontario, Canada.

Participants: Thirty-three patients with incomplete SCI and 14 able-bodied controls.

Interventions: Not applicable.

Main Outcome Measures: QST measures of perceptual threshold for temperature and vibration, American Spinal Injury Association sensory scores (light touch, pinprick), and tibial nerve SEPs.

Results: There was a low degree of association (kappa) between QST results and sensory scores (|kappa|=.05-.44). QST measures yielded greater numbers of patients with SCI being classified as impaired, suggesting a greater sensitivity of QST to detect more subtle sensory deficits. QST measures of vibration threshold generally corresponded to the patients' SEP recordings. QST measures of modalities conveyed within the same tract were significantly (P<.05) correlated (|r|=.46-.84) in patients with SCI, but not in controls, whereas those modalities mediated by different pathways had lower and generally nonsignificant correlations (|r|=.05-.44) in both patients and controls.

Conclusions: The low degree of association between QST measures and sensory scores is likely attributable to measurement limitations of both assessments, as well as various neuroanatomic and neuropathologic factors. QST provides more sensitive detection of preserved sensory function than does standard clinical examination in patients with incomplete SCI.