Estimating Change in Tremor Amplitude Using Clinical Ratings: Recommendations for Clinical Trials
Tremor rating scales are the standard method for assessing tremor severity and clinical change due to treatment or disease progression. However, ratings and their changes are difficult to interpret without knowing the relationship between ratings and tremor amplitude (displacement or angular rotation), and the computation of percentage change in ratings relative to baseline is misleading because of the ordinal nature of these scales. For example, a reduction in tremor from rating 2 to rating 1 (0–4 scale) should not be interpreted as a 50% reduction in tremor amplitude, nor should a reduction in rating 4 to rating 3 be interpreted as a 25% reduction in tremor. Studies from several laboratories have found a logarithmic relationship between tremor ratings R and tremor amplitude T, measured with a motion transducer: logT = α·R + β, where α ≈ 0.5, β ≈ –2, and log is base 10. This relationship is consistent with the Weber–Fechner law of psychophysics, and from this equation, the fractional change in tremor amplitude for a given change in clinical ratings is derived: (Tf−Ti)/Ti=10α(Rf−Ri)−1, where the subscripts i and f refer to the initial and final values. For a 0–4 scale and α = 0.5, a 1-point reduction in tremor ratings is roughly a 68% reduction in tremor amplitude, regardless of the baseline tremor rating (e.g., 2 or 4). Similarly, a 2-point reduction is roughly a 90% reduction in tremor amplitude. These Weber–Fechner equations should be used in clinical trials for computing and interpreting change in tremor, assessed with clinical ratings.