ULTRASOUND is an effective screening tool for steatotic liver disease (SLD), provided that inter- and intra-observer comparisons are incorporated to enhance diagnostic consistency. SLD, an abnormal triglyceride accumulation within the hepatocyte, is a chronic condition that leads to an increased risk of cancer-related mortality. Timely and cost-effective detection is therefore essential, and ultrasound could provide a practical and affordable alternative to liver biopsy.
A new analysis from the Chile BiLS study evaluated the reliability of ultrasound imaging for diagnosing SLD in a high-risk cohort of Chilean women with gallstones. These patients commonly exhibit obesity and related metabolic disorders, which are key risk factors for the development of SLD.
A board-certified radiologist (Observer 1) evaluated a subset of 425 baseline images and compared their assessments with those originally recorded by radiology technicians involved in the Chile BiLS cohort. Observer 1 then reanalysed 34 blinded duplicates and two Chile BiLS radiology technicians (Observers 2 and 3) independently reviewed these images to assess intra-observer reproducibility.
Findings revealed considerable variability in diagnostic agreement. Observer 1 demonstrated only slight inter-observer agreement (κ=0.12; 95% CI: 0.08–0.15; 41.0% agreement). In contrast, the two Chile BiLS technicians showed fair agreement (Observer 2: κ=0.29, 64.7% agreement; Observer 3: κ=0.32, 63.6% agreement). In terms of intra-observer reliability, Observer 1 achieved moderate agreement (κ=0.45; 95% CI: 0.08–0.82; 81.3% agreement), while Observer 2 exhibited substantial consistency across repeated image reviews (κ=0.64; 95% CI: 0.37–0.90; 81.8% agreement).
These findings underscore the variability inherent in ultrasound-based diagnosis of SLD, highlighting the importance of inter- and intra-observer comparisons for optimal diagnostic consistency.
Reference
Spencer-Sandino M et al. Inter- and intra-observer agreement in ultrasound diagnosis of steatotic liver disease: implications for screening in resource-limited settings. Sci Rep. 2025;15:29819.
