Minimal Clinically Important Difference in TETRAS Score for Essential Tremor
Synopsis
Using data from the TRANQUIL multicenter randomized controlled trial, this study estimated minimal clinically important differences (MCIDs) for the modified activities of daily living (mADL) and mADL11 scales of TETRAS in essential tremor via anchor-based approaches (CGI-I and PGI-I) and ROC analyses, recommending patient-report-based MCIDs of -6.2 for mADL and -4.1 for mADL11.
Interpretation
It provides the first estimates of MCIDs for the mADL and mADL11 scales, two frequently used primary outcome measures in essential tremor clinical trials. MCIDs for these TETRAS-modified measures had not been established previously, and this study fills that gap. Based on data from the TRANQUIL multicenter randomized controlled trial, using CGI-I and PGI-I anchor-based approaches as primary methods, supplemented by ROC analyses.
MCID thresholds differ across methods: mADL ranges from -3.5 to -6.2, and mADL11 ranges from -1.0 to -4.1. It reveals that ROC-derived estimates produce the lowest thresholds, whereas PGI-I-based estimates yield the highest. Anchor measures demonstrated acceptable Spearman correlations with score changes, supporting the applicability of the anchor-based approach.
Recommended patient-report-based MCID values are -6.2 for mADL and -4.1 for mADL11. It offers a patient-perception-grounded threshold reference for interpreting results of future essential tremor clinical trials. The recommended values derive from the PGI-I patient-reported anchor, representing an estimate at the level of patients' subjective experience.
Perspective
This study provides an interpretive framework for score changes on the mADL and mADL11 scales in essential tremor clinical trials, applicable to trials using these two TETRAS-modified scales as primary outcomes, and is especially useful for researchers and readers judging whether observed changes reach a patient-perceptible clinical significance.
The current text is abstract-level and does not present details such as sample size, specific correlation coefficient values, confidence intervals, or area under the ROC curve; the thresholds vary widely across anchor methods (mADL from -3.5 to -6.2), so which threshold to apply in practice still requires consideration of the specific trial context and reader judgment.
