Poster Presentation ANZOS-Breakthrough Discoveries Joint Annual Scientific Meeting 2018

Estimating fibrosis from non-alcoholic fatty liver disease and its associations using the non-alcoholic fatty liver disease (NAFLD) fibrosis and FIB-4 scores in hospitalised inpatients: A retrospective, matched cohort study (#324)

Ramy H Bishay 1 2 3 , Dhanya Sanjeev 1 , Gideon Meyerowitz-Katz 4 , Golo Ahlenstiel 1 2 5
  1. Metabolic & Weight Loss Program, Department of Endocrinology and Diabetes, Blacktown Hospital, Blacktown, NSW, Australia
  2. School of Medicine, University of Western Sydney, Sydney, NSW, Australia
  3. School of Medicine, University of Sydney, Sydney, NSW, Australia
  4. Western Sydney Diabetes, Integrated and Community Health Directorate, Blacktown Hospital Department of Endocrinology and Diabetes, Western Sydney Local Health District, Sydney, NSW, Australia
  5. Storr Liver Centre, Westmead Millennium Institute, Westmead Hospital, University of Sydney, Sydney, NSW, Australia

INTRODUCTION Diagnosing non-alcoholic fatty liver disease (NAFLD) is impeded by poorly corroborative serological markers and invasiveness of ‘gold standard’ liver biopsy. Scoring systems (NAFLD fibrosis score [NAFLDFS], FIB-4) have been utilised with some success in the community, yet their use in hospitalised obese individuals has not been evaluated. METHODS Retrospective data extraction for obesity-related conditions were obtained from all admissions to Blacktown-Mt Druitt hospitals (April 2016-February 2017). NAFLD fibrosis and FIB-4 scores were applied to estimate the risk of liver fibrosis for patients with both obese-related (Ob) admissions vs an age and gender matched non-obese (NOb) related admission cohort. RESULTS Of 43,212 admissions, 244 had an Ob-related diagnosis. Compared with NOb patients, the Ob cohort (mean age 55+17 vs 56+17 yrs; 54 vs 53% female; mean weight 82+25 vs 126+37 kg; BMI 29+8 vs 46+12 kg/m2) featured significantly greater comorbidities (median 14 vs 4, P<0.001), 2-5x greater prevalence of Type 2 diabetes (T2D), cardiopulmonary disease, pharmacologic burden, length of stay and cost (all P<0.001). There were no differences in NAFLDFS or FIB-4 scores between the NOb and Ob patients. Compared with a low NAFLDFS or FIB-4 score (<-1.455; <1.45), a high score (>0.675; >3.25) was respectively associated with ischaemic/coronary artery disease (P=0.006, OR 2.48; P<0.001, OR 3.41) and hyperlipidaemia (P<0.001, OR 3.41; P=0.001, OR 2.53), whereas NAFLDFS was additionally associated with pulmonary hypertension (P=0.047, OR 3.36), hypertension (P<0.001, OR 3.20), obstructive sleep apnoea (P<0.001, OR 2.74) and T2D (P<0.001, OR 3.03). Of those with high NAFLDFS (n=142) or FIB-4 (n=22), only 4% (n=6) and 0% had a diagnosis of NAFLD, respectively. CONCLUSION The use of NAFLDFS and FIB-4 correlate with well-known cardiovascular risk phenotypes. High scores may prompt an opportunity to improve clinical inertia in the diagnosis of this disease in at-risk hospitalised patients.