Proposals and evaluation of different thresholds for defining a clinically meaningful improvement in an absolute change in the rheumatoid arthritis impact of disease (RAID) score. a. Proposals of threshold according to different techniques and different times of evaluation. b. Evaluation of external validity (versus DAS28-ESR) of different proposed thresholds. c. Evaluation of external validity (versus patient's global assessment) of different proposed thresholds. *Thresholds, proposal based on the following techniques and time of evaluation: a, 75th percentile technique at week 4; b, empirical technique and correct classification at week 4; c, smallest detectable change and 75th percentile at week 12; d, ROC technique at week 4; e, empirical technique and correct classification at week 4; f, correct classification at week 12; g, ROC technique at week 12. +Positive likelihood ratio (higher values are indicative of better performing thresholds. See Methods for further explanation). DAS28-ESR, Disease Activity Score 28-erythrocyte sedimentation rate; ROC, receiver operating characteristic.