Diagnostic performance of lung ultrasound for the detection of interstitial lung disease in patients with rheumatoid arthritis: comparison with risk scoring systems

Introduction: Interstitial lung disease (ILD) is a frequent and severe complication of rheumatoid arthritis (RA), often remaining subclinical and associated with substantial morbidity and mortality. Given the limitations of high-resolution computed tomography (HRCT) for widespread screening, alternative tools for early detection are needed. Aim: The objective of this study was to assess the diagnostic performance of lung ultrasound (LUS) for the detection of ILD in patients with RA, compared with HRCT and established clinical risk scores, stratified by symptom status. Methods: A diagnostic test study was conducted in RA patients fulfilling the ACR 2010 criteria who had undergone HRCT within the previous six months. LUS was performed using a 14-point protocol, with abnormal findings defined as >5 B-lines or pleural irregularities in ≥2 windows. Two validated RA-ILD risk scores were applied. LUS findings were compared with HRCT, and logistic regression was used to identify clinical and imaging predictors of ILD. Results: A total of 147 patients were included (23.8% male; mean age 62.3 years; mean disease duration 12.3 years), of whom 68 had HRCT-confirmed ILD. Abnormal LUS findings were present in 58.5% of patients. Factors associated with RA-ILD included male sex (OR 2.8), dyspnea (OR 2.5), and velcro crackles (OR 47.5). Abnormal LUS (OR 59.8), >5 B-lines (OR 37.6), and pleural irregularities (OR 29.5) were strongly associated with ILD. In asymptomatic patients, LUS showed high sensitivity (95.4%) and specificity (77.7%), with a +LR of 4.3 and a −LR of 0.06. Risk scores showed limited diagnostic value (Spanish consensus ≥5: +LR 0.97; Pierre ≥25: +LR 1.45). Combining LUS with risk scores increased sensitivity to 100%. Conclusion: LUS outperforms clinical risk scores for the detection of RA-associated ILD, including in asymptomatic patients. Its integration with risk scores may optimize screening strategies for the early identification of RA-ILD.