Quantitative sensory testing cannot differentiate simulated sensory loss from sensory neuropathy

FREEMAN R, CHASE KP, RISK MR

NEUROLOGY

LIPPINCOTT WILLIAMS & WILKINS

Lugar: Minneapolis; Año: 2003 vol. 60 p. 465 - 470

0028-3878

OBJECTIVE: To differentiate the quantitative sensory testing (QST) results of subjects simulating small and large fiber sensory loss from those of normal subjects and subjects with sensory peripheral neuropathy. BACKGROUND: QST is used to measure sensory thresholds in clinical, epidemiologic, and research studies. It is not known whether there are objective test results that characterize the subject seeking to deceive the examiner. METHODS: The Computer Aided Sensory Examination IV 4, 2, and 1 stepping algorithm was used to determine vibration and cold perception in nine naïve subjects. Subjects were asked to simulate sensory loss (on two occasions) and to respond normally on one occasion. Test results were compared to those of subjects with diabetic sensory neuropathy. Each QST trial was performed three times. RESULTS: Reproducibility, measured by the intraclass correlation coefficient, was similar in all groups for the vibration perception test (simulation 1: 0.68 [95% CI 0.31, 0.91], simulation 2: 0.82 [95% CI 0.54, 0.95], normal response: 0.77 [95% CI 0.47, 0.94], and subjects with peripheral neuropathy: 0.76 [95% CI 0.18, 0.95]) and the cold perception test (simulation 1: 0.53 [95% CI 0.12, 0.85], simulation 2: 0.82 [95% CI 0.55, 0.95], normal subjects: 0.67 [95% CI 0.30, 0.90] and subjects with peripheral neuropathy: 0.88 [95% CI 0.57, 0.97]), all just noticeable difference units. There were no differences between performance characteristics in the two simulation trials. Responses to null stimuli did not differentiate between groups. CONCLUSION: Test performance characteristics do not permit discrimination among subjects simulating sensory loss, subjects with normal responses, and subjects with peripheral neuropathy.