The effect of prednisolone initiation and short-term discontinuation on the clinical assessment of polymyalgia rheumatica.

Abstract

Approximately half of individuals with suspected PMR initiate prednisolone before rheumatological evaluation, but it is unknown if prednisolone reduces the ability to diagnose PMR. This study evaluated the clinical assessment of a PMR diagnosis before and after prednisolone initiation and after short-term prednisolone discontinuation. Data from a prospective cohort of treatment-naïve patients with PMR (n = 66) and non-PMR (n = 27) were used to create case report forms (CRFs). All patients were assessed at baseline and patients with PMR were reassessed after 8 weeks of prednisolone treatment (n = 57) and after short-term prednisolone discontinuation at week 10 (n = 48). In total, 198 CRFs were created containing patient demographics, current symptoms, physical examination and blood test results. Six rheumatologists, blinded to diagnosis, prednisolone treatment and visit timing, assessed each CRF, rating the probability of a PMR diagnosis from 0 to 10 and classifying them as PMR or non-PMR. At baseline, CRF assessors effectively distinguished PMR from non-PMR, demonstrating median probabilities of PMR of 7.8 and 2.8, respectively, and a sensitivity/specificity of 85.9%/82.7% for the PMR diagnosis. After prednisolone initiation, the median probability of PMR was 1.0 and sensitivity was 3.5%, followed by a median probability of PMR of 2.3 and a sensitivity of 25% after short-term prednisolone discontinuation. This study shows that clinically assessing PMR becomes challenging after prednisolone initiation, and that short-term prednisolone discontinuation does not improve clinical assessment to pre-treatment levels. Consequently, the study supports recommendations of deferring prednisolone initiation until after specialist evaluation in individuals suspected of having PMR.