Background/Purpose:

In early arthritis, after a two years follow up, rheumatologist diagnosis of rheumatoid arthritis agrees well with 2010 ACR/EULAR criteria (Saraux A et al. Arthritis care Arthritis Res (Hoboken). 2013;65:1227-34). Today, we do not have straightforward rules to diagnose spondyloarthritis (SpA) after a two year follow up, and we do not know the gap between rheumatologist diagnosis and SpA defined using all potential methods to classify patients having inflammatory low back pain (IBP).

Methods:

We used the nationwide, longitudinal, prospective cohort (DESIR) of patients with IBP suggestive of SpA at baseline. After 2 years, patients were classified based on: imaging (MRI, X-rays), the certainty with which the rheumatologist diagnosed SpA (evaluated on 0-100 visual analogue scales), treatment used (non-steroidal anti-inflammatory drugs [NSAID] and/or TNF-alpha blockers) and classification criteria (any among axial Assessment of SpA international Society [ASAS] criteria, European SpA Study Group [ESSG] and Amor). First, we described clinical, biological and imaging characteristics, the items of all classification criteria and the diagnosis established by the rheumatologists after 2 years. Second, we evaluated the certainty in the diagnosis of SpA at inclusion and after a 2 year follow-up. Third, using ROC curves, we determined the best end point of the diagnosis certainty using the validated classification criteria (AMOR, ESSG, and ASAS) as gold standards. Then, we evaluated agreement between all potential methods to classify patients having IBP based on Cohen’s kappa coefficient in the whole group and in the group having a MRI.

Results:

On the 708 patients initially included, 548 had information on rheumatologist’s certainty after 2 years. Using ROC curves, we found that a certainty of diagnosis ≥ 75% gave the best balance of sensitivity and specificity. This certainty of diagnosis increased with the follow up [357 of 548 (65.1%) patients had a certainty≥75% after a two year follow up versus 265 (48.3%) at inclusion]. Certainty of diagnosis≥75% after a two year follow up was statistically associated with all classification criteria (AMOR p: 0.005; ESSG p: 0.003; ASAS p<0.0001) and the ASAS criteria had the best agreement, although it was low (kappa 0.09, 0.11 and 0.25 for AMOR, ESSG and ASAS, respectively). None of the various other potential classifications items had a better agreement.

Conclusion:

Rheumatologist diagnosis of SpA certainty after 2 years does not agree well with the various previously published criteria for SpA.

---

« Back to 2014 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/how-to-classify-spondyloarthritis-after-a-two-year-follow-up-results-from-the-french-recent-onset-spondyloarthritis-cohort/