white) race (OR=1.34, 95% CI: 1.09–1.66), pre-LT diabetes (OR 1.23, 95% CI: 1.08–1.48) or HF (OR 2.21, 95% CI: 1.58–3.09) and discharge to a skilled nursing facility (vs. home) (OR 2.99, 95% CI: 2.63–3.40) after index LT. Findings were consistent for 90d readmissions. Mean length of stay for a CVD-related rehospitalization was 7.0 ± 10.0 days. Thirty-day in-hospital mortality was 0.57% and comorbid CVD conditions were present in 91.7% of these deaths. CONCLUSIONS: Cardiovascular disease is a leading contributor to both 30- and 90-day readmission after LT and is primarily due to non-ischemic

etiologies. This study identifies patients at high risk for readmission after LT with CVD comorbidity that may benefit from medical optimization through a tailored multidisciplinary care LEE011 purchase pathway prior to discharge. Disclosures: Josh Levitsky – Grant/Research Support: Salix, Novartis; Speaking and Teaching: Gilead, Salix, Y-27632 manufacturer Novartis The following people have nothing to disclose: Lisa B. VanWagner, Marina Serper, Raymond Kang, Brittany Lapin, Donald M. Lloyd-Jones, Anton I. Skaro, Samuel Hohmann Background: Whilst non-invasive biomarkers have been shown to accurately predict significant liver fibrosis their ability to provide information on clinical prognosis is less well developed. Aim: To investigate whether non-invasive biomarkers are prognostic factors for clinical outcomes (e.g. all-cause & liver mortality and the development of liver cancer) in patients with chronic

liver disease.

Methods: A systematic review of evidence published between 1st January 2002 and 1st October 2012 identified from Embase, MEDLINE and Pubmed Central was performed using the following terms: ‘encephalopathy,’ ‘death,’ ‘liver transplant,’ ‘mortality,’ check details ‘ascites,’ ‘cancer,’ ‘variceal’ OR ‘varices’ between 1st January 2002 and 1st October 2012. Studies were included if >1 non-invasive biomarkers (APRI (AST:platelet ratio index), Fib-4, AST:ALT ratio, BARD, NFS, ELF, Hepascore, Fibrotest, Fibrometer, Forns, Fibroscan or transient elastography) were examined in relation to >1 clinical outcomes in the search criteria. Where possible Hazard Ratios (HRs) for each biomarker were extracted and if appropriate, pooled across studies using a random effects model. Results: The search identified 1456 results. After removal of 298 duplicates, 31 unique studies met selection criteria. There was significant study heterogeneity regarding choice of biomarker (and cut-off), disease aetiology, choice of clinical outcome, analysis method and reporting standards. The commonest markers assessed were APRI (13 studies, 7842 patients), Fib-4 (6 studies, 4385 patients) and AST:ALT ratio (6 studies, 1716 patients). Three studies from which HR information for overall survival could be extracted (either directly or from log-rank information) were analyzed. For an APRI cut-off of >1.5–2.0 in patients with viral hepatitis C (HCV) a summary HR for overall mortality of 2.51 (1.37–4.60) and 4.43 (1.64–11.