Background The initial diagnosis of polymyalgia rheumatica (PMR) is often made in primary care. Given the non-specific nature of PMR symptoms, diagnosis can be challenging especially as many disorders can mimic the symptoms of PMR. Although guidelines are available1, studies suggest guidelines are rarely followed2 and that diagnostic accuracy is poor3. However, previous studies have relied upon data from retrospective case review or data from referral letters.
Objectives The aim of this study was to describe how general practitioners (GPs) in the UK, report that they diagnose PMR based on symptoms, investigations and exclusion of other disorders.
Methods 5000 randomly selected GPs in the UK were mailed a questionnaire on PMR. As part of the survey GPs were asked, in their routine practice, which clinical features were most important in diagnosing PMR and, what investigations they performed. They were also asked, what they would do if the inflammatory markers were normal and which disorders they would routinely try to exclude before making a formal diagnosis of PMR.
Results 1249 (25.1%) GPs responded. Responders were generally from practices with larger list sizes and more GP principals than non-responders. GPs rated bilateral shoulder pain, raised inflammatory markers and response to corticosteroids as the most important clinical features for diagnosing PMR. Most respondents performed full blood count and ESR or CRP. Other recommended blood tests were performed less frequently (figure 1 ). 25% (n=315) of GPs reported that they would exclude PMR in patients with normal inflammatory markers, whilst 31% (n=392) would refer their patients to secondary care for specialist review. 80% (n=1001) of GPs sought to exclude giant cell arteritis (GCA). 61% (n=759) routinely excluded other rheumatological disorders, but just 55% (n=682) considered excluding relevant malignancies.
Conclusions This large GP survey suggests that despite using recommended clinical features to diagnose PMR and seeking to exclude GCA, many GPs do not routinely exclude other diagnoses as a cause of symptoms. This has significant implications for diagnostic accuracy which could result in prolonged, inappropriate treatment with corticosteroids or a missed opportunity for timely more appropriate intervention in disorders such as malignancy or rheumatoid arthritis. Further work is needed to investigate the accuracy of the diagnosis of PMR in primary care.
Rheumatology, 49(1), pp. 186-90.
Family Practice, 25(5), pp. 328-333.
Clinical rheumatology, 19(4), pp. 278-280.
Disclosure of Interest None Declared