Jeremy D. P. Bland

Introduction/Aims Previous studies have reported weak correlations between neurophysiological measurements and subjective severity of symptoms in carpal tunnel syndrome, with Pearson r ≤ 0.26. We hypothesize that this resulted in part from patient‐to‐patient variability in the assessment of subjective severity using tools such as the Boston Carpal Tunnel Questionnaire. To compensate for this, we aimed to assess within‐patient differences in symptom and test result severity. Methods In our study we used retrospective data from 13 005 patients with bilateral electrophysiological results and 790 patients with bilateral ultrasound imaging drawn from the Canterbury CTS database. Measures of neurophysiological (nerve conduction studies [NCS] grade) and anatomical (cross‐sectional area on ultrasound) severity within individual patients were compared between the right and left hands, eliminating individual variation in the way in which patients interpret the questionnaire. Results There was a correlation found between right‐hand NCS grade and symptom severity score (Pearson r = −0.302, P < .001, n = 13,005), but not between right‐hand cross‐sectional area and symptom severity (Pearson r = 0.058, P = .10, n = 790). In the within‐subject analyses, there were significant correlations between symptoms and NCS grade (Pearson r = 0.6, P < .001, n = 6521) and between symptoms and cross‐sectional area (Pearson r = 0.3. P < .001, n = 433). Discussion The simple correlation between symptomatic and electrophysiological severity was comparable with previous studies, but within‐patient analysis revealed that the relationship was stronger than previously reported and strong enough to be clinically useful. The relationship between symptoms and cross‐sectional area measurement on ultrasound imaging was weaker.