A real headache: A cross-sectional survey of symptoms associated with giant cell arteritis in a primary care setting

Research output: Contribution to journalAbstractpeer-review


Background: GCA is the commonest form of large vessel vasculitis. Its association with permanent visual loss in up to 23% of patients makes prompt recognition and treatment essential. However, the high prevalence of headache and the lack of specificity of a raised ESR (two key items of the current classification criteria) pose a challenge to identify GCA appropriately in primary care. In addition, using classification criteria for the purposes of diagnosis risks the misinterpretation of inference. In this study we aimed to investigate the prevalence of symptoms associated with GCA in a general practice population with the objective of identifying key discriminating features that identify the disease in this setting.

Methods: A cross-sectional survey of all individuals aged 55 years or older from a large general practice of two sites in Norfolk, UK, was carried out. The questionnaire was developed to include the full range of clinical features that have been related to GCA, identified from a literature review. All GP diagnoses of GCA identified on the GP database and clinical care records were reviewed for the purpose of case classification against 1990 ACR criteria. Key clinical features were identified through step-wise logistic discrimination adjusting for age and sex. Likelihood ratios of individual features in the final selected model were estimated.

Results: The analysis is based on completed responses from 2277 individuals from an invited sample of 4728 individuals aged 50 years or older (response rate 49%). The median age was 68 years old. There was no statistical difference between responders and non-responders for age structure and gender ratio. 15 participants self-reported a diagnosis of GCA (8 of who were confirmed on their care record). Of the total responses, 2.8% of males and 6.7% of females reported having suffered severe headaches lasting longer than 7 days. In the logistic discrimination model (adjusted for age and sex) the following clinical features provided the most predictive power: headache, unexplained weight loss and early morning shoulder pain and stiffness (receiver operator characteristic = area under curve 96.8%). Likelihood ratios for individual symptoms as predictors of GCA are displayed in Table 1.

Conclusion: This is the first study to survey an entire population for symptoms commonly associated with GCA in a community setting. The data show that GCA accounts for a small proportion of those with prolonged headache. As expected the co-occurrence of jaw, tongue pain and scalp tenderness with headache are important predictors of GCA in this setting. It is of particular interest that the highest positive likelihood ratio is related to unexplained weight loss. This is not included in current classification criteria but may be of diagnostic value in identifying GCA in the in the general practice setting.
Disclosure statement: The authors have declared no conflicts of interest.
Original languageEnglish
Pages (from-to)i38-i39
Number of pages2
Issue numbersuppl_1
Publication statusPublished - 1 Apr 2015
EventAnnual Meeting of the British Society for Rheumatology and British Health Professionals in Rheumatology - Manchester, United Kingdom
Duration: 28 Apr 201530 Apr 2015

Cite this