Limited correlations between clinician-based and patient-based measures of essential tremor severity.

Abstract

We investigated the relation between changes in clinician-based and patient-based measures of tremor severity, within the Fahn-Tolosa-Marin Tremor Rating Scale (TRS) and Visual Analogue Scale (VAS) in essential tremor patients. Thirty-seven patients were assessed twice: on- and off-medication. Clinician-based, objective TRS assessments, consisting of part A (postures/movements) and part B (tremor-inducing tasks) were conducted by a blinded assessor using video-tapes. Patients completed TRS part C (limitations in activities of daily life) and indicated subjective tremor severity using VAS. Patients' total TRS and VAS scores improved on-medication (both p<0.001). Mean improvement was 6.3 (sd 5.4) points on the total TRS and 2.3 (sd 2.3) points on the VAS score. Within the TRS, we found moderate correlations between changes in clinician-based TRS-B and patient-based TRS-C scores (ρ=0.387, p=0.011), but not between changes in clinician-based TRS-A and TRS-C scores (ρ=0.128, p=0.232). Moreover, changes in subjective VAS scores correlated with changes in total TRS (ρ=0.422, p=0.007), changes in TRS-C scores (ρ=0.367, p=0.015) and, more weakly, with changes in TRS-B scores (ρ=0.281, p=0.049), but again: not with changes in TRS-A scores (ρ=-0.008, p=0.482). We found no correlation between changes in clinician-based TRS-A, and patient-based measures TRS-C or VAS scores, and a weak correlation between clinician-based TRS-B and VAS scores. The limited correlations between changes in clinician-based and patient-based measures of tremor severity suggest that the different scales measure different aspects of tremor severity and support the additional use of subjective patient-based assessments in clinical practice and clinical trials.