Duration of clinically meaningful improvement in pain as a patient-centered endpoint in acute pain trials.

Current US Food and Drug Administration (FDA) guidance recommends the sum of pain intensity differences (SPID) as the primary efficacy endpoint for acute pain trials, which lacks validated thresholds for clinically meaningful change and requires imputation of pain scores after rescue medication use. We introduce the responder outcome over time (ROOT), defined as the proportion of study time a participant experiences clinically important improvement in pain intensity without recent rescue medication use or early discontinuation. The ROOT combines the clinical interpretability of a responder-based outcome with the statistical efficiency of a continuous measure, without requiring imputation of pain scores after rescue. We reanalyzed data from phase 2 and 3 analgesic trials submitted to the FDA to assess concordance between SPID and ROOT. For SPID, pain scores after rescue were imputed using the established practice in FDA submissions, windowed last-observation-carried-forward (LOCF) for 6 hours. For ROOT, meaningful improvement was defined as ≥50% reduction in pain intensity; participants were classified as nonresponders for 6 hours after rescue. In 34 trials (N = 11,028), SPID and ROOT reached statistically concordant conclusions for 92.2% of the 204 treatment comparisons. Calculation of SPID required LOCF imputation of pain scores after rescue medication for a median of 27% of study time in active drug groups and 46% in placebo groups. Reanalyses showed that LOCF distorted the temporal pattern of treatment effects during periods of high rescue use, producing apparent treatment differences not corroborated with ROOT. These results show that ROOT provides a clinically interpretable, patient-centered alternative to SPID with similar ability to detect statistically significant treatment effects.