Background/Purpose: Pain, stiffness, and fatigue have been defined as core assessment domains in polymyalgia rheumatica (PMR) by the OMERACT PMR Working Group (PMRWG). Patient partners have expressed a preference for numeric rating scales (NRS) over visual analogue scales to measure these symptoms. Numerous NRS forms have been used in different PMR clinical trials, yet evidence to support their use is lacking. We propose a standardized NRS form and provide initial results of some measurement properties.

Methods: PMRWG clinicians, methodologists, and patient partners rated two different NRS stems and anchors and provided free-text feedback. The preferred set was modified according to patient partner input. Clinical trial participants with PMR, no or low disease activity, and who received prednisolone 10 mg for ≥ 1 week prior to entry were treated with prednisolone 10 mg for 4 weeks and received each of SPI-62 and matching placebo as adjunctive therapy for 2 weeks. They completed individual NRS for pain, stiffness, and fatigue intensity, and pain chronicity, each at fortnightly trial visits with a 7-day look-back and daily at home with a 1-day look-back. Compliance at home, test-retest reliability, and correlation between the on-site and at-home versions were assessed using placebo period data from the inception trial cohort (N=12).

Results: The standardized items are shown (Figure). At-home compliance was 97%; 7/12 had 100% compliance. There were no missing items, only occasional days that participants did not respond at all. Weighted k comparing responses 2 weeks apart for on-site pain, stiffness, and fatigue intensity were 0.52, 0.45, and 0.67, and for pain chronicity was 0.54. Intraclass correlation coefficients, with intra-individual as the variance of interest, for at-home pain, stiffness, and fatigue intensity were 0.62, 0.70, and 0.84, and for pain chronicity was 0.72. Weighted k between the on-site response for 7-day look-back and the maximum at-home response for 1-day look back during any of the 7 prior days were 0.53, 0.82, and 0.80 for the intensity items and 0.55 for pain chronicity. When the on-site response was compared instead to the at-home response on the same day, weighted k were 0.51, 0.68, 0.76, and 0.64. Confidence intervals for all statistics were wide due to the limited number of participants.

Conclusion: The results provide necessary initial evidence to encourage further examination of these NRS’ psychometrics. They suggest high patient compliance for daily symptom report and moderate test-retest reliability. Patient ability to recall accurately their most severe symptoms during 7 days prior to a trial visit was strong for stiffness and fatigue, whilst moderate for pain. Symptom recall did not appear to be particularly biased by symptoms of the most recent 24 hours. The reported measurement properties point favorably toward validated symptom assessment for PMR clinical trials, lack of which currently limits conduct of patient-centric research. More data are needed to confirm these results on reliability, and to support other measurement properties such as thresholds of meaning.

« Back to ACR Convergence 2023

ACR Meeting Abstracts - https://acrabstracts.org/abstract/characterization-of-numeric-rating-scales-for-symptom-assessment-in-polymyalgia-rheumatica/