The State of Patient Reported Outcome Measures in Rheumatology.

Abstract

No prior studies in rheumatology have evaluated the quality or timeliness of patient reported outcome (PRO) measure reporting. Clinical trials that informed new Food and Drug Administration (FDA) approvals for the first rheumatological indication between 1995-2021 were identified. Whether collected PROs were published, met minimal clinically important (MCID) difference or statistical significance (p < 0.05) thresholds, or were consistent with CONSORT-PRO reporting standards was recorded. Hazard ratios and the Kaplan-Meier estimate were used to assess the time from FDA approval to PRO publication. Thirty one FDA approvals corresponded with 110 pivotal trials and 262 reported patient reported outcome measures (PROs). One of the included studies (1.1%) met all five recommended items, 10/90 (11.1%) met 4 items, 17/90 (18.9%) met 3 items, 21/90 (23.3%) met 2 items, 26/90 (28.9%) met 1 item, and 15/90 (16.7%) met none of the reporting standards. Most PROs met MCID thresholds (149/262, 56.9%) and were statistically significant (223/262, 85.1%). Of our subset analysis, one third of PROs were not published upfront (70/212, 33.0%) and one of nine (22/212, 10.4%) remained unpublished four or more years after initial trial reporting. Publication rates were highest for HAQ-DI (97.4%) and lowest for SF-36 (81.8%). Less than half of these published PROs met MCID and statistical significance thresholds (94/212, 44.3%). One in nine PROs remained unpublished for more than 4 years after initial trial reporting, and compliance with CONSORT PRO reporting guideline was poor. Efforts should be made to ensure PROs are adequately reported and expeditiously published.