Validation of screening criteria for interstitial lung disease in rheumatoid arthritis in routine clinical practice:...

To evaluate the validity of the rheumatoid arthritis-interstitial lung disease screening project of the Spanish Society of Rheumatology (AR-EPIDSER) screening criteria for detecting rheumatoid arthritis (RA)-associated interstitial lung disease (ILD) and to estimate the prevalence of undiagnosed and subclinical ILD in patients with RA in Spain.

Multicentre cross-sectional observational study conducted in 15 Spanish hospitals. Adult patients with RA (American College of Rheumatology (ACR) 1987 or ACR/European Alliance of Associations for Rheumatology 2010 criteria) without a previous ILD diagnosis were included. During a rheumatology visit, the AR-EPIDSER screening criteria were applied, and all participants underwent high-resolution CT (HRCT) of the chest as the reference test, with centralised reading. Sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) were estimated with 95% CI as well as the prevalence of ILD and subclinical ILD.

A total of 477 patients were recruited; 424 had evaluable HRCT (73.6% women; median age 59 years). The prevalence of ILD on HRCT examinations was 13.0% (55/424; 95% CI 9.9% to 16.6%). Subclinical ILD (no cough, dyspnoea or 'Velcro-like' crackles) was 8.3% (35/424; 95% CI 5.7% to 11.1%), accounting for 63.6% of ILD cases. Screening was positive in 68.2% (289/424). The AR-EPIDSER criteria showed sensitivity=0.93 (95% CI 0.82 to 0.98), specificity=0.36 (0.31 to 0.41), PPV=0.18 (0.13 to 0.23) and NPV=0.97 (0.93 to 0.99).

In Spanish patients with RA, undiagnosed ILD is frequent and mostly subclinical. The AR-EPIDSER criteria are useful as an initial screening tool due to high sensitivity and excellent NPV; however, low specificity suggests that positive cases should be confirmed by HRCT.