Mortality predicted using this model was compared with observed mortality. Results: We included 366 cirrhotic patients admitted to a hospital with infection (56y; 58% males, 29% and 25% with alcoholic and HCV cirrhosis, respectively). Median (IQR) MELD at admission was 16.5 (10-23); 66% of patients developed at least 1 organ failure (18%, 7% and 8% with HDAC inhibitor review 2, 3 and 4 organ failures, respectively). Observed 30- and 90-day mortality was 18.6% and 29.2%, respectively. Patients with higher predictive model scores had higher mortality. However, the surgery model overestimated

mortality in patients at risk for infection-ACLF (table). Conclusion: The observed 30- and 90-day mortality in cirrhotic patients with infection-related ACLF is lower than that predicted for surgery-related ACLFThis suggests that mortality in ACLF depends not only on severity of liver disease and organ failure but also on the precipitating event. Observed ACLF mortality compared to predicted values Disclosures: Patrick S. Kamath – Advisory Committees or Review Panels: Sequana Medical Jacqueline G. O’Leary – Consulting: Gilead, Jansen K. Rajender Reddy – Advisory Committees or Review Panels: Genentech-Roche, Merck, Janssen, Vertex, Gilead,

BMS, Novartis, Abbvie; Grant/Research Support: Merck, BMS, Ikaria, Gilead, Janssen, AbbVie Florence Wong – Consulting: Gore Inc; Grant/Research Support: Grifols Michael B. Fallon – Grant/Research Support: Bayer/Onyx, triclocarban Eaisi, Gilead, Grifolis Jasmohan S. Bajaj – Advisory Committees or Review Panels: Salix, Merz, otsuka, ocera, grifols, american college signaling pathway of gastroenterology; Grant/Research Support: salix, otsuka, grifols The following people have nothing to disclose: Siddharth Singh, Guadalupe Gar-cia-Tsao, Scott W. Biggins, Benedict Maliakkal, Ram M. Subramanian, Heather M. Patton, Leroy Thacker Aims: Infection is

associated with high mortality in cirrhosis. Malnutrition is a known risk factor for infection, but the risk associated with obesity is unknown. The study aim was to evaluate the association between infection and nutritional status in cirrhotics. Methods: We reviewed the Nationwide Inpatient Sample, years 2009-2011. Patients under age 18, with HIV, or post-liver transplant were excluded. Patients and infections were identified using International Classification of Diseases 9th revision (ICD-9) codes. Subjects were grouped as malnourished, obese, and morbidly obese. Infections evaluated for included bacteremia, sepsis, cellulitis, urinary tract infection (UTI), lower respiratory infection (LRI), Clostridium diffiicile infection (CDI), and spontaneous bacterial peritonitis (SBP). We adjusted for patient co-morbidities using the Deyo modification of the Charlson index and for liver decompensation using the Baveno IV criteria. We evaluated for risk factors for infection and mortality using multivariable logistic regression.