Background/Objective Patients with polymyalgia rheumatica experience flares and require a lengthy course of glucocorticoid treatment. Ultrasound application is often used for diagnosing polymyalgia rheumatica. This study aimed to determine whether polymyalgia rheumatica diagnosed with ultrasound complementation has a more favorable clinical course compared with that of only clinically diagnosed patients.

Methods In this cohort study, we retrospectively identified 152 patients with polymyalgia rheumatica from January 2008 to December 2018. We extracted patients' clinical and ultrasound information, and hazard ratio and propensity-score matched analyses were performed.

Results Among 152 patients with polymyalgia rheumatica, the flare, methotrexate add-on, and C-reactive protein normalization rates were 15.9 (95% confidence interval, 8.8-23.1)/100 person-years, 9.3 (3.6-15.0) /100 person-years, and 70.3 (61.3-79.2) /100 person-months, respectively. Age (p=0.01), C-reactive protein levels (p=0.03), and absence of peripheral joint pain (p=0.03) were significantly different between 81 and 71 patients with and without ultrasound complementation, respectively. The hazard ratio showed that ultrasound complementation did not contribute to the clinical course; flare, methotrexate add-on, and C-reactive protein level normalization yielded values of 0.88 (p=0.64), 1.93 (p=0.056), and 0.94 (p=0.72), respectively. Propensity-score-matched analysis showed a similar clinical course between 51 pairs: flare (p=0.45), methotrexate add-on (p=0.15), and C-reactive protein normalization (p=0.94).

Conclusions Age, C-reactive protein, and involved joint distribution were factors leading to ultrasound complementation at the time of polymyalgia rheumatica diagnosis. Ultrasound complementation at PMR diagnosis is useful for differential diagnosis but may not affect the clinical course after GC introduction.