SAT0434 MINIMAL CLINICALLY IMPORTANT DIFFERENCE IN OUTCOME MEASURES FOR USE IN CLINICAL CARE AND PRAGMATIC TRIALS IN PsA

Abstract

Background:While several outcome measures have been studied for use in clinical studies of psoriatic arthritis, little is known about thresholds of meaning such as minimal clinically important improvement (MCII).Objectives:To investigate the distribution of scores for candidate outcome measures for pragmatic trials in PsA and to calculate the MCII for each outcome measure.Methods:We performed a longitudinal cohort study within the Psoriatic Arthritis Research Consortium (PARC), a multi-center study based in the US. Patients completed validated PROs (patient reported outcomes) and rheumatologists completed skin, joint, enthesis and dactylitis scores at therapy initiation and follow up 12-16 weeks later. In addition, patients completed a global assessment of response at the follow up visit, categorizing their status as improved, stayed the same, or worsened and then ratied the importance of the change on a scale from 0-7.1We then calculated and plotted the change in each of the following measures: Routine Assessment of Patient Index Data (RAPID3), clinical Disease Activity of Psoriatic Arthritis (cDAPSA), Patient Reported Outcome Measure Information System (PROMIS) Global Health short form (10a) physical health (PH) subscore, patient pain assessment, patient global assessment (0-10 NRS), and physician global assessments (0-10 NRS) of the joints and overall. We calculated the MCII as the mean change in score (with 95% confidence interval) among patients who reported improvement and rated the level of improvement as “almost none/hardly at all” or “a little important.” Additionally, we calculated Spearman’s correlation coefficients between the measures and the global assessment of response.Results:Among 148 unique patients, 233 therapy change visits were eligible for analysis. The average age was 52.5 years, 52% were female and mean BMI was 29.6. Baseline RAPID3 was 11.1 (SD 6), cDAPSA 17.9 (SD 13.9), PROMIS PH 42 (SD 8), patient global 4.2 (SD 2.5), TJC 5.9 (SD 7.5), and SJC 2.9 (SD 4.5). TNFi comprised 61% of drug initiations, 21% were IL17i and the remainder were other biologics and oral systemic therapies. At follow up, 63 (27%) patients rated themselves as improved whereas 103 (44%) stayed the same and 67 (29%) reported worsening. The mean change in each measure by patient-reported response (improved, stayed the same, or worsened) are shown in Figures 1A & B. In general, the mean score increased from ‘improved’ to ‘worsened’ as expected (with the exception of PROMIS PH which declines given a different direction of scoring). The MCII for each measure was as follows: RAPID3 -1.8 (-4.1 to 0.5), Patient Global -0.6 (-1.6 to 0.4), Physician Global -1 (-1.9 to -0.1), cDAPSA -5.7 (-9.8 to -1.7), and PROMIS PH 1.9 (-2.1 to 5.8). Correlation for each measure with the global assessment of response were: RAPID3 0.48, Patient Global 0.37, Physician Global 0.39, cDAPSA 0.51, and PROMIS PH 0.39.Figure 1A. Distribution of change (median, IQR) in RAPID3, Physician Global, Patient Global, PROMIS10a physical therapy by patient reported response.Conclusion:This is the first study to test thresholds of meaning for these particular measures in PsA. The MCII values are relatively low for all outcome measures. This may be related to the relatively low disease activity at baseline but is consistent with patients seen in clinical practice initiating therapy.2