Background One of the most frequent desires of patients with polymyalgia rheumatica (PMR) is to individualize their treatment based on comorbidities, co-medication and prognostic factors. The upcoming ACR/EULAR guidelines for the management of PMR will counsel individualized treatment decisions in clinical practice.

Objectives To review the evidence on the value of prognostic factors predicting the outcome of PMR informing the ACR/EULAR PMR guidelines development group

Methods We used GRADE methodology as framework for the guideline development process. For the systematic literature review we searched MEDLINE, Embase, PubMed, CINAHL, Web of Science and the Cochrane Library (1970 through June 2013) for relevant articles on the prognostic value of age, sex, baseline inflammatory markers, disease activity/severity, presence of peripheral arthritis, symptom duration, concomitant conditions relevant to PMR and/or disease management with glucocorticoid (GC) therapy, response to GC, shared patients' management by primary and secondary care providers and optimal control management. The quality of identified articles was appraised by the Quality In Prognosis Studies (QUIPS) tool

Results Out of 10078 records, we identified 37 articles meeting our inclusion and exclusion criteria. We found no studies investigating the prognostic value of symptom duration, shared patients' management and optimal control management. One study [with low risk of bias (LoB) in 5/8 categories of the appraisal] reported lower risk of health care resource (HCRes) use in patients at older age. Females, patients with a high erythrocyte sedimentation rate (ESR) and patients with peripheral arthritis may have a higher risk of relapse (1 study with LoB in 5/8, 4 studies with LoB in 4-5/8 and 1 study with LoB in 5/8 categories, respectively). Females and patients with a high ESR also may have a lower probability to discontinue GCs (1 study, LoB in 4/8 and 1 study, LoB in 5/8, respectively) or may have longer duration of therapy (1 study, LoB 3/8 and 2 studies, LoB 2-5/8, respectively). Several studies of varying quality, however, failed to prove the associations reported above between clinical predictors and outcomes. Cumulative GC dose (1study, LoB 4/8) and GC side effects (3 studies, LoB 2-4/8) were increased in females, whereas HCRes use was lower than in males (1 study, LoB 6/8). Patients with high ESR had lower quality of life during follow-up (1 study, 8/8) and a lower HCR use than patients with low ESR (1 study, 6/8). More active disease at baseline predicted worse function and lower quality of life at 12 months (1 study, 8/8). A rapid response to GC was associated with fewer GC side effects (1 study, LoB 2/8) but failed to show any relation to other outcomes.

Conclusions Limited evidence exists for a worse prognosis for PMR patients with female sex, high ESR and peripheral arthritis at disease outset.

Disclosure of Interest C. Dejaco Grant/research support: Pfizer, MSD, AbbVie, Speakers bureau: Pfizer, MSD, AbbVie, UCB, Roche, BMS, Y. Singh: None declared, P. Perel: None declared, A. Hutchings: None declared, E. Matteson Grant/research support: Novartis, Centocor, genentech, Pfizer, UCB Group, Hoffmann-LaRoche, Schering, B. Dasgupta Consultant for: Merck, Roche, Napp, Mundipharma, Servier

DOI 10.1136/annrheumdis-2014-eular.2316