{
  "abstract": "Introduction Liver ultrasounds are reported by clinicians with varying experience and in different settings and machines. Anecdotally reporting terminology is diverse especially for steatosis and fibrosis. We aimed to evaluate our local accuracy and terminology within our real-life NHS setting.Method Retrospective data collection of ultrasound reports within 6 months of liver biopsy. Liver biopsies were collected between 2013-2024 in NHS Tayside. As grades of steatosis or fibrosis is not reported on ultrasound, preselected terminology was considered positive or negative. For example, for steatosis, if reported as increased echogenicity or equivalent, this was counted as positive for steatosis. Reported normal liver or equivalent was considered negative. If a word did not fall within the pre-set terminology then this was decided between two authors. P-values calculated by chi square test.Results In 101 ultrasound reports average age was 54.8, 57.4% were female, 49.5% were reported by a radiologist and 50.5% by a sonographer. 66.3% (67/101) were performed in a tertiary hospital and 33.7% (34/101) in a district hospital. The sensitivity of ultrasound at each steatosis grade was mild S1 64%; moderate S2 100%; severe S3 88.9%. ≥S2 steatosis sensitivity was 95.8% and for ≥S1 was 79.6%. The tertiary hospital performed better for ≥S2 steatosis, sensitivity 100% (p value 0.23) and ≥S1 75.9% (p value 0.65) compared to the district hospitals, 90% and 70% respectively but was not statistically significant.Analysis for cirrhosis (F4) on biopsy found poor sensitivity of individual ultrasound parameters except for liver echotexture, 88.8% sensitivity. Irregular liver outline, shrunken liver size and enlarged spleen had sensitivity of 50%, 0% and 72.7% respectively. 3/11 of F4 reports (27.3%) reported a normal spleen size. Omissions in reports were high. Liver outline, liver size, portal vein flow, ascites presence/absence and mention of cirrhosis were reported in 30.8%, 46.2%, 75.0%, 69.2%, and 30.8% respectively. Liver echotexture, focal lesions and spleen size were the most reported (all 84.6%).For liver parenchyma description we found multiple phrases reporting similar findings such as ‘echobright’, ‘bright’, ‘increase in reflectively’. For those who had ≥S1 on biopsy, the most common phrase used in 22.4% (11/49) was ‘coarse echotexture,’ followed by ‘increase in echoity’ 26.3% (8/49). For F4 patients, again ‘coarse echotexture’ was the most used phrase 54.5% (6/11), conversely 27.3% (3/11) of F4 patients were labelled as echobright.Conclusion Ultrasound diagnosis of steatosis is good but parameters for cirrhosis are not always present and cannot be relied on for diagnosis. There are omissions in reporting and wide variance in terminology. Improvements in these may help to improve diagnosis.",
  "authors": [
    {
      "affiliations": [
        "University of Dundee, Dundee, United Kingdom"
      ],
      "name": "Jennifer Cathcart"
    },
    {
      "affiliations": [
        "University of Dundee, Dundee, United Kingdom"
      ],
      "name": "Anu Pillai"
    },
    {
      "affiliations": [
        "University of Dundee, Dundee, United Kingdom"
      ],
      "name": "Aisha Ahmed"
    },
    {
      "affiliations": [
        "NHS Tayside, Dundee,"
      ],
      "name": "Ruairi Lynch"
    },
    {
      "affiliations": [
        "University of Dundee, Dundee, United Kingdom"
      ],
      "name": "John F Dillon"
    }
  ],
  "title": "P117 High sensitivity of liver ultrasound for steatosis but low sensitivity for fibrosis: biopsy-comparison study of ultrasound parameters and terminology variability",
  "uid": "997a26ca-00f4-506a-a2e5-b93b6f20c6e2"
}