Test-retest and interrater reliability of experimental within-subject variability of pain reports as assessed by the focused analgesia selection test.

Abstract

Within-subject variability (WSV) of pain intensity reports has been shown to predict the placebo response. The focused analgesia selection test (FAST), which allows to experimentally assess WSV of pain reports, has been used as a screening tool to identify participants who are likely to have a strong placebo response in drug-development clinical trials. Yet, the reliability of FAST has not been reported. To assess test-retest and interrater reliability of the FAST outcomes. To mimic pharma-sponsored clinical trials, we enlisted inexperienced assessors who underwent limited training. Healthy volunteers performed the FAST twice within a week and were randomly assigned to either the test-retest group or the interrater group. T-tests, partial Pearson correlations, intraclass correlations (ICC), and Bland-Altman plots were generated to assess FAST outcomes' reliability. Sixty-three participants completed the study and were assigned to the test-retest (N = 33) or interrater (N = 30) arms. No statistically significant differences in the FAST outcomes were detected between the 2 sessions, except for the FAST covariance (FAST CoV) in the interrater assessment (P = 0.009). Test-retest reliabilities of the FAST-main outcomes were r = 0.461, ICC = 0.385 for the FAST R 2 and r = 0.605, ICC = 0.539 for the FAST ICC and in the interrater cohort, they were FAST R 2: r = 0.321, ICC = 0.337 and FAST ICC: r = 0.355, ICC = 0.330. Using inexperienced assessors, the FAST outcomes test-retest ranged from moderate to strong, whereas the interrater reliability ranged from weak to poor. These results highlight the importance of adequately training study staff members before using this tool in multicentre clinical trials.