Comparison of CT-detected hepatic steatosis in adults between two Middle Eastern centers: a retrospective dual-center study

Abstract Background Hepatic steatosis can be detected incidentally on non-contrast CT, but Middle Eastern data are limited, and incidental steatosis is often underreported. Objectives To compare CT-detected hepatic steatosis and its mention in radiology reports between two centers with different levels of care, and to explore the liver-to-muscle attenuation ratio (LMAR). Methods This retrospective study included 682 adults (Site A, secondary care, n = 352; Site B, tertiary care, n = 330) who underwent non-contrast abdominal CT (Jan–Dec 2025). Steatosis was defined as mean liver attenuation < 45 HU (primary) and liver–spleen attenuation difference <− 10 HU (secondary). Radiology reports were reviewed for any mention of fatty liver. Reproducibility was tested with 240 repeat measurement sets. Results Steatosis frequency (15.6% vs 19.1%; p = 0.232; adjusted OR 0.77, 95% CI 0.52–1.16) and mean liver attenuation (52.9 vs 53.7 HU; p = 0.386) did not differ significantly. Spleen, aortic, and psoas attenuation were 1.8–3.2 HU lower at Site A, and the liver–spleen criterion identified fewer cases there (8.8% vs 16.7%; p = 0.002). Reports mentioned fatty liver less often at Site A (3.7% vs 11.5%; p < 0.001); sensitivity for moderate-to-severe steatosis was 34.2% vs 64.4% ( p = 0.008), with specificity of at least 98.9%. Liver and spleen ICCs were 0.95–0.996. LMAR did not differ between centers ( p = 0.173). Conclusion Radiology reports often omitted CT-detected steatosis, more often at the secondary-care center, although CT-defined steatosis showed no detectable difference. Reference-tissue attenuation differed between centers, so LMAR remains exploratory.