Evidence-based recommendations for the diagnosis of ankylosing spondylitis: results from the Australian 3E initiative in rheumatology

Tracey Kain, Jane Zochling, Andrew Taylor, Nicholas Manolios, Malcolm D Smith, Mark D Reed, Matthew A Brown and Lionel Schachna
Med J Aust 2008; 188 (4): 235-237. || doi: 10.5694/j.1326-5377.2008.tb01592.x
Published online: 18 February 2008


Ankylosing spondylitis is a chronic disabling rheumatic condition occurring in 0.2%–1.0% of the population and characterised by pain, disability and loss of spinal mobility.1 Diagnosis is typically delayed by 5–7 years,2 often because of delayed referral from the general practitioner.3 In the modern era of effective treatment with tumour necrosis factor inhibitors, the consequences of delayed diagnosis are significant. These include inappropriate costly investigations and therapies, loss of spinal mobility and progressive decline in productivity and employability.4 As a result, education of GPs for early recognition of ankylosing spondylitis is increasingly vital.

Classification criteria of ankylosing spondylitis are fulfilled if definite radiographic changes of the sacroiliac joints occur in the setting of defined clinical symptoms and reduced measures of spinal mobility.5 However, the appearance of radiographic changes is commonly delayed by a decade or more after onset of symptoms.6 “Axial spondyloarthritis” is a recently introduced term to denote patients with symptoms of ankylosing spondylitis without radiographic sacroiliitis.7 Almost 60% of these patients will progress to radiographic sacroiliitis (and therefore ankylosing spondylitis) over 10 years.6

Ankylosing spondylitis and axial spondyloarthritis comprise up to 5% of patients with chronic back pain in primary care.8 The major challenge in general practice is to identify these patients among the large patient population presenting with chronic back pain. This article summarises a set of Australian consensus recommendations addressing early diagnosis of ankylosing spondylitis, with the goal of facilitating timely referral to a rheumatologist.

The recommendations

The objective of this study was to provide a consensus of recommendations to facilitate early recognition of ankylosing spondylitis in general practice. The importance of such recommendations is ever-increasing with the introduction into clinical practice of effective biological treatments for ankylosing spondylitis.9

Box 1 summarises the methods used. Because of the limitations of the medical literature in providing firm guidelines for evaluating patients with possible ankylosing spondylitis, a study design that incorporated the opinion of practising local rheumatologists was desirable.

The recommendations themselves are presented in Box 2. For almost all recommendations, there was strong consensus among participating rheumatologists, and therefore we propose a standardised national approach for early identification of ankylosing spondylitis.


By astute recognition of inflammatory back pain, GPs play a critical role in the early identification of ankylosing spondylitis. Inflammatory back pain represents the sine qua non of ankylosing spondylitis and has traditionally been identified by insidious onset of symptoms before the age of 40 years, symptoms persisting longer than 3 months, morning stiffness, and improvement with exercise.10 Recently, a new definition of inflammatory back pain among patients younger than 50 years with chronic back pain was proposed. This definition requires two or more of morning stiffness, improvement with exercise but not by rest, alternating buttock pain, and awakening with pain only in the second half of the night.11 These criteria appear more sensitive and specific than traditional definitions of inflammatory back pain, but remain to be validated in large populations with non-specific back pain. More importantly than explicit definitions, GPs have the ability to accurately identify patients with inflammatory back pain on the basis of typical clinical presentations.12

Non-selective and cyclooxygenase-2-selective non-steroidal anti-inflammatory drugs (NSAIDs) rapidly relieve symptoms of axial disease.13 About 75% of patients with ankylosing spondylitis show good or very good response to NSAIDs within 48 hours of commencement of treatment, compared with only 15% of patients with mechanical back pain,14 and this can be a useful discriminating feature. Failure to respond to a trial of at least two different NSAIDs at full dose over a 3-month period was thought to make the diagnosis of ankylosing spondylitis less likely. However, this finding does not rule out the diagnosis of ankylosing spondylitis, and may in effect identify a subgroup with a poorer prognosis,15 who may benefit from early referral and treatment. Similarly, raised inflammatory markers are supportive of early ankylosing spondylitis, but erythrocyte sedimentation rate or C-reactive protein are elevated in only 50%–70% of patients with active ankylosing spondylitis.16

Magnetic resonance imaging (MRI) provides an opportunity for early identification of sacroiliitis not demonstrable on plain radiography. Fat suppression sequences eliminate high signal from fat in the marrow and allow high signal from excess water at sites of inflammation to be seen. The presence of bone oedema in such images of the sacroiliac joints indicates subcortical osteitis and predicts the development of radiographic sacroiliitis with sensitivity of 85% and specificity of 47%.17 At present, there is no evidence to show that MRI abnormalities of the sacroiliac joints or spine predict the development of bony fusion of the spine, the outcome measure most associated with morbidity in ankylosing spondylitis. As a result of limited availability, high cost and non-standardisation of interpretation of images, the panel did not recommend MRI as a screening tool in general practice.

Early referral recommendations for patients with suspected ankylosing spondylitis have been recently proposed,18 but their effect on altering outcomes for ankylosing spondylitis has not been evaluated. The authors of those recommendations advocate referral of patients with chronic back pain with symptom onset before the age of 45 years with either inflammatory features or human leukocyte antigen (HLA)-B27 positivity.18 By contrast, we recommend early referral of all patients with inflammatory back pain, irrespective of age and HLA-B27 status. The positive predictive value in an unselected population of chronic back pain is low. HLA-B27 has its most useful role in patients with clinical or imaging features to suggest ankylosing spondylitis, and the panel was concerned that inaccurate interpretation of the result may delay referral to a rheumatologist. A single highly sensitive clinical parameter, inflammatory back pain, was thought to be preferable as a screening tool for referral to a rheumatologist.

Effective dissemination and implementation of these recommendations are important to standardise the approach to early diagnosis of ankylosing spondylitis. Longitudinal studies will be of interest to document the effects of these recommendations in limiting variability in clinical practice and reducing health care costs.

1 Methods

The 3E (evidence, expertise and exchange) program is a multinational effort to develop recommendations for daily management of rheumatic disorders. It combines systematic literature research (evidence) with the opinion of local rheumatologists (expertise) to ensure that recommendations are relevant to clinical practice. Local recommendations are then combined with other countries (exchange) into a set of international recommendations.

Rheumatologists from 10 countries participated in three rounds of discussion and Delphi voting to determine a set of nine questions, three for each domain of diagnosis, monitoring and treatment of ankylosing spondylitis. Questions were selected by experts with in-depth knowledge of the literature to address areas of uncertainty in clinical practice. To avoid conflict of interest with the pharmaceutical sponsor (Abbott Immunology), questions related to biological disease-modifying antirheumatic drugs were not considered.

Proposed questions for the domain of diagnosis were:

  • In individuals with back pain, what are the early clinical features that suggest ankylosing spondylitis?

  • How useful is imaging in identifying early ankylosing spondylitis?

  • Based on which clinical features should a general practitioner refer a patient to a rheumatologist for further evaluation?

A systematic literature review was conducted by international rheumatology Fellows using MEDLINE and PubMed for studies published up to August 2006. The search strategy included all relevant terms for ankylosing spondylitis combined with different sets of keywords specific for each question.

A summary of the results was presented to the 23 participants of the 3E Australian meeting, who were clinical rheumatologists experienced in the day-to-day management of ankylosing spondylitis. Following three rounds of break-out sessions, a set of recommendations was finalised after discussion and voting. The category of evidence and strength of recommendation was determined for each recommendation.

Category of evidence

Strength of recommendation

Ia: Meta-analysis of randomised controlled trials

A: Category I evidence

Ib: Randomised controlled trial

IIa: Controlled study without randomisation

B: Category II evidence or extrapolated from category I evidence

IIb: Quasi-experimental study

III: Non-experimental descriptive studies such a comparative, correlation and case–control studies

C: Category III evidence or extrapolated from category I or II evidence

IV: Expert committee reports or opinion or clinical experience of respected authorities, or both

D: Category IV evidence or extrapolated from category II or III evidence

2 Recommendations

1a. Early clinical features to suggest ankylosing spondylitis include inflammatory back pain and age at symptom onset < 45 years. (III, C, 100%)

1b. The absence of symptomatic response to an appropriate course of non-steroidal anti-inflammatory drugs makes the diagnosis of ankylosing spondylitis less likely. (III, C, 91%)

1c. Raised inflammatory markers are supportive, but their absence does not rule out the diagnosis of ankylosing spondylitis. (III, C, 90%)

2a. Despite low sensitivity to detect changes of early ankylosing spondylitis, plain radiographs of the pelvis and spine are appropriate initial imaging techniques. (IV, D, 91%)

2b. Magnetic resonance imaging is a useful imaging modality for detecting early changes of ankylosing spondylitis. (IV, D, 100%)

3. Individuals with inflammatory back pain should be referred to a rheumatologist for further evaluation. (III, C, 77%)

The parentheses contain the category of evidence, strength of recommendation, and level of agreement among the panel.

  • Tracey Kain1
  • Jane Zochling2
  • Andrew Taylor3
  • Nicholas Manolios4
  • Malcolm D Smith5
  • Mark D Reed3
  • Matthew A Brown6
  • Lionel Schachna1

  • 1 Department of Rheumatology, Austin Health, Melbourne, VIC.
  • 2 Menzies Research Institute, Hobart, TAS.
  • 3 Department of Rheumatology, Royal Perth Hospital, Perth, WA.
  • 4 Department of Rheumatology, Westmead Hospital, Sydney, NSW.
  • 5 Department of Rheumatology, Repatriation General Hospital, Adelaide, SA.
  • 6 Diamantina Institute for Cancer, Immunology and Metabolic Medicine, Princess Alexandra Hospital, Brisbane, QLD.



We acknowledge the following rheumatologists who participated in the Australian 3E meeting: Helen Cooley (Tas), Julien de Jager (Qld), Roland Ebringer (Vic), John Edmonds (NSW), Sam Engel (Vic), Marie Feletar (Vic), Chris Fong (Vic), Patrick Hanrahan (WA), Charles Inderjeeth (WA), Roman Jaworski (NSW), Geoffrey Littlejohn (Vic), Christopher Mack (Vic), Gabor Major (NSW), Prudence Manners (WA), Michael Shanahan (SA), Kathleen Tymms (ACT), and Raymond White (NSW). The 3E initiative is supported by an unrestricted grant provided by Abbott Immunology.

Competing interests:

None identified.

  • 1. Khan MA. Update on spondyloarthropathies. Ann Intern Med 2002; 136: 896-907.
  • 2. Feldtkeller E, Khan MA, van der Heijde D, et al. Age at disease onset and diagnosis delay in HLA-B27 negative vs. positive patients with ankylosing spondylitis. Rheumatol Int 2003; 23: 61-66.
  • 3. Boyer GS, Templin DW, Bowler A, et al. A comparison of patients with spondyloarthropathy seen in specialty clinics with those identified in a communitywide epidemiologic study. Has the classic case misled us? Arch Intern Med 1997; 157: 2111-2117.
  • 4. Rudwaleit M, Listing J, Brandt J, et al. Prediction of a major clinical response (BASDAI 50) to tumour necrosis factor α blockers in ankylosing spondylitis. Ann Rheum Dis 2004; 63: 665-670.
  • 5. van der Linden S, Valkenburg HA, Cats A. Evaluation of diagnostic criteria for ankylosing spondylitis. A proposal for modification of the New York criteria. Arthritis Rheum 1984; 27: 361-368.
  • 6. Mau W, Zeidler H, Mau R, et al. Clinical features and prognosis of patients with possible ankylosing spondylitis. Results of a 10-year followup. J Rheumatol 1988; 15: 1109-1114.
  • 7. Rudwaleit M, van der Heijde D, Khan MA, et al. How to diagnose axial spondyloarthritis early. Ann Rheum Dis 2004; 63: 535-543.
  • 8. Underwood MR, Dawes P. Inflammatory back pain in primary care. Br J Rheumatol 1995; 34: 1074-1077.
  • 9. Zochling J, Braun J. Developments and current pharmacotherapeutic recommendations for ankylosing spondylitis. Expert Opin Pharmacother 2006; 7: 869-883.
  • 10. Calin A. Early diagnosis of ankylosing spondylitis. Lancet 1977; 2: 1293.
  • 11. Rudwaleit M, Metter A, Listing J, et al. Inflammatory back pain in ankylosing spondylitis: a reassessment of the clinical history for application as classification and diagnostic criteria. Arthritis Rheum 2006; 54: 569-578.
  • 12. Brandt HC, Spiller I, Song IH, et al. Performance of referral recommendations in patients with chronic back pain and suspected axial spondyloarthritis. Ann Rheum Dis 2007; 66: 1479-1484.
  • 13. Zochling J, van der Heijde D, Burgos-Vargas R, et al. ASAS/EULAR recommendations for the management of ankylosing spondylitis. Ann Rheum Dis 2006; 65: 442-452.
  • 14. Amor B, Dougados M, Listrat V, et al. Are classification criteria for spondylarthropathy useful as diagnostic criteria? Rev Rhum Engl Ed 1995; 62: 10-15.
  • 15. Amor B, Santos RS, Nahal R, et al. Predictive factors for the longterm outcome of spondyloarthropathies. J Rheumatol 1994; 21: 1883-1887.
  • 16. Spoorenberg A, van der Heijde D, de Klerk E, et al. Relative value of erythrocyte sedimentation rate and C-reactive protein in assessment of disease activity in ankylosing spondylitis. J Rheumatol 1999; 26: 980-984.
  • 17. Oostveen J, Prevo R, den Boer J, van de Laar M. Early detection of sacroiliitis on magnetic resonance imaging and subsequent development of sacroiliitis on plain radiography. A prospective, longitudinal study. J Rheumatol 1999; 26: 1953-1958.
  • 18. Sieper J, Rudwaleit M. Early referral recommendations for ankylosing spondylitis (including pre-radiographic and radiographic forms) in primary care. Ann Rheum Dis 2005; 64: 659-663.


remove_circle_outline Delete Author
add_circle_outline Add Author

Do you have any competing interests to declare? *

I/we agree to assign copyright to the Medical Journal of Australia and agree to the Conditions of publication *
I/we agree to the Terms of use of the Medical Journal of Australia *
Email me when people comment on this article

Online responses are no longer available. Please refer to our instructions for authors page for more information.