Diagnostic accuracy of lung ultrasound for the detection of interstitial lung disease in rheumatoid arthritis: a systematic review and meta-analysis

To evaluate the diagnostic accuracy of lung ultrasound (LUS) for detecting rheumatoid arthritis-associated interstitial lung disease (RA-ILD), using high-resolution computed tomography (HRCT) as the reference standard. A systematic review and diagnostic test accuracy meta-analysis was conducted according to PRISMA-DTA. PubMed, Embase, Scopus, and Google Scholar were searched from January 2012 to December 2025. Observational studies evaluating LUS for RA-ILD in adults were eligible. Sensitivity and specificity were jointly pooled using a bivariate random-effects model, deriving positive and negative likelihood ratios. The primary analysis was restricted to studies using the common threshold of ≥ 5 B-lines. An exploratory all-threshold analysis included one estimate per study, selecting the lowest eligible threshold. Fourteen studies were included; 13 assessed B-lines and were eligible for meta-analysis. In the primary analysis (six studies), pooled sensitivity was 84.0% (95% CI 72.4–91.3), specificity 75.5% (95% CI 66.2–82.9), and AUC 0.858; positive and negative likelihood ratios were 3.45 (95% CI 2.62–4.65) and 0.21 (95% CI 0.13–0.33). Heterogeneity was moderate-to-high, mainly affecting specificity, with no major influence of individual studies on leave-one-out analysis. In the all-threshold analysis, pooled sensitivity was 86.8% (95% CI 80.1–91.4), specificity 75.4% (95% CI 62.3–85.1), and AUC 0.890. LUS shows good diagnostic accuracy for RA-ILD at the common threshold of ≥ 5 B-lines. Threshold selection, scanning protocol, and clinical setting influence performance. The likelihood ratios indicate LUS is more informative for ruling out than ruling in RA-ILD, supporting its use primarily as a screening tool, with HRCT remaining necessary to confirm a positive scan. What is already known on this topic • Lung ultrasound (LUS) has been increasingly explored as a radiation-free, bedside tool for detecting interstitial lung disease in rheumatic diseases, with the most consistent evidence accumulated in systemic sclerosis. • In RA, individual studies have reported associations between B-lines and HRCT-defined interstitial changes, but diagnostic accuracy estimates have been heterogeneous, and no diagnostic test accuracy meta-analysis had pooled sensitivity and specificity. What this study adds • This is the first diagnostic test accuracy meta-analysis of LUS for RA-ILD: at the common threshold of ≥5 B-lines, LUS showed high pooled sensitivity (84.0%) and moderate specificity (75.5%), with likelihood ratios indicating it is more informative for ruling out than for ruling in RA-ILD. • Diagnostic performance was substantially influenced by B-line threshold, scanning protocol, and clinical setting, with shorter standardized protocols offering the most favorable balance between accuracy and feasibility, although this finding is hypothesis-generating given the small number of studies involved. How this study might affect research, practice or policy • The findings support the use of LUS primarily as a screening or triage tool to identify patients who warrant confirmatory HRCT, rather than as a replacement for it. • They highlight the need for standardized LUS definitions and protocols, and for validation of pleural line assessment in RA-ILD, to clarify how LUS can be integrated into the diagnostic work-up.