Can Quality of Care for Patients With Cirrhosis Be Measured?

systems for data collection. The views expressed in this abstract are those of the authors and do not represent the U.S. Navy or Department of Defense. Mo1026

OBJECTIVE: Bacterial translocation, causing intestinal inflammation, is the key mechanism in the pathogenesis of hepatic encephalopathy (HE) and spontaneous bacterial peritonitis (SBP) in patients with cirrhosis. The presence of calprotectin in feces quantitatively relates to neutrophil migration into the gastrointestinal tract and is therefore considered as a valid marker of intestinal inflammation. We aimed to assess the role of fecal calprotectin concentrations (FCCs) in diagnosing the onset and severity of HE and SBP. METHODS: 61 patients with cirrhosis were prospectively included in this study. 42 subjects served as noncirrhotic controls. Complications of cirrhosis including ascites, HE and SBP were diagnosed by reference methods. Stool samples were collected for measuring FCCs. Patients revealing other causes of abnormal calprotectin results e.g. GI bleeding or inflammatory bowel disease were excluded from the study. Multivariate analysis of cirrhosis-associated complications and their relation to FCCs was performed. RESULTS: FCCs were significantly higher in cirrhotics compared to controls (p<0.001). Among patients with cirrhosis, FCCs were significantly elevated dependent on the severity of liver disease as assessed by Child- and MELDscores. A significant correlation emerged between elevated FCCs and HE grading as measured by West-Haven criteria and critical flicker frequency (p<0.001; sensitivity >0.93, specificity >0.89) and SBP (p<0.02; sensitivity =0.71, specificity = 0.79). FCCs were also significantly higher in cirrhotic subjects with additional extraintestinal inflammation (p<0.01; sensitivity = 0.65, specificity =0.8). There was no significant influence of parameters of systemic inflammation (CRP, leucocytes) on FCCs. CONCLUSIONS: FCCs may serve as a screening tool to identify cirrhotic patients with HE and SBP. Furthermore, assessment of FCCs may faciliate grading of HE-severity.

Mo1024 Serum Retinol Binding Protein 4 (RBP-4) is a Unique Surrogate Marker of END Stage Liver Disease With Complications. A Pilot Clinical Trial P Patrick Basu, Nithya Krishnaswamy, Niraj James Shah, Chris Tang Background: Retinol binding protein 4 (RBP-4), a 21 kilodalton plasma protein, is secreted from the liver and adipose tissue, and is known to transport retinol in the blood. RBP-4 levels have an inverse relationship with disease progression and decompensation. This study evaluates the role of RBP-4 on assessing the severity of decompensated liver disease Methods: 85 patients (n=85) (age=49.6±11.4, mean±SD) with moderate to -severe liver disease (mean MELD of 18) were recruited into three groups: Group A- Chronic liver disease with cirrhosis (n=30):20/30 HCV (67%) 8/30 HBV (27%) 2/30 Misc (6.6%) with mean MELD of 22; Group B- non-cirrhotic (n=30) 13/20 HCV (43%), 4/20 HBV (13.3%), 5/30 Alcoholics (17%), and 5/30 misc (17%) with mean MELD of <10 Group C- controls (n=25) with MELD <10, Liver biopsy, insulin resistance (HOMA-score), MELD score, adiponectin, Leptin and serum RBP-4 (ELISA), APRI, Hyaluronic acid, and Hepascore for liver fibrosis were analyzed Results: There was a significant difference in RBP-4 across the cirrhotic, non-cirrhotic, and control groups (F(2, 82)=47.9, p<0.001; one-way ANOVA). Posthoc Bonferroni tests over controls (n=25, 4.21+/1.60; mean+/-SD). RBP4 is significantly decreased in Cirrhotics (p<0.001; n=30, 1.46+/0.51) over the non-Cirrhotics (p=0.015; n=30, 3.37+/0.97) and controls. Cirrhotics also had decreases in HOMA (1.09±0.09 vs. 1.51±0.60; p=0.001, Student's t-test),adiponectin (0.93±0.40 vs. 1.26±0.53; p=0.01), but with an increase in TNFalpha (11.23±3.81 vs. 3.72±3.37; p<0.001) compared to the non-cirrhotic patients. Leptin (8.20±3.23 vs. 10.93±9.44; p=0.14) and hepatic steatometry (BURNT score<4)-(0.73±0.83 vs. 1.10±1.21; p=0.19 showed no difference. RBP 4 has no relevant correlations with TNFalpha (r=-0.62, p<0.001), leptin (r=0.12, p=0.38) and hepatic steatometry (BURNT score<4)(r=0.12, p=0.41). RBP-4 correlated negatively with MELD score (r=-0.57, p=0.001) Conclusions: This pilot study proposed that RBP-4 has a role as a biomarker for endstage liver disease, with an independent, inverse relationship with severity of fibrosis, hepatic decompensation, MELD score, and complications of liver failure. There is a positive correlation with insulin resistance, metabolic syndrome, and Adipokines, but not with hepatic steatometry. A larger prospective clinical trial needs to be conducted to validate the unique role of RBP-4 in fibrosis and decompensated liver disease.

Mo1027 Clinical Presentation and Outcomes of Patients With Cirrhosis and Hemorrhagic Ascites Nathalie H. Urrunaga, Amit G. Singal, Jennifer A. Cuthbert, Don C. Rockey Background: Hemorrhagic ascites can pose diagnostic and therapeutic dilemmas in patients with cirrhosis. Therefore, we aimed to explore the clinical characteristics and outcomes of patients with cirrhosis and hemorrhagic ascites. Methods: We retrospectively reviewed the records of all patients admitted to an urban county hospital with cirrhosis and ascites who underwent paracentesis between January 2003 and June 2010. Hemorrhagic ascites was defined as an ascitic red blood cell (RBC) count ≥ 10,000/μl. Causes of hemorrhage in the ascites were defined as due to an underlying process (hepatocellular carcinoma (HCC), trauma, or iatrogenesis) or unknown (idiopathic). We also examined a control group of cirrhotic patients with ascites who had an ascitic RBC count < 10,000/μl. These patients were all admitted during the same time period and matched using age and gender (3 controls: 1 case). Patients with and without hemorrhagic ascites were compared using Student's ttest for continuous variables and chi square analysis for categorical variables. Results: Of the 1,113 cirrhotic patients who underwent paracentesis, 214 (19%) had hemorrhagic ascites; 115 (54%) were idiopathic in origin. The median ascitic RBC count was 35,175 (range 10,050-4,225,000) with 133 patients having a RBC count of 10,000-50,000/μl and 81 patients having a RBC count ≥ 50,000/μl. The median age at presentation was 51 years and 22% were female. The most common causes of liver disease were alcoholic cirrhosis and/or hepatitis C, which accounted for 179 (84%) cases. The median bilirubin level was 3.1 mg/dL and median MELD score was 18. The median hemoglobin level was 10.2 g/dL, although 32 (15%) patients presented with a hemoglobin level less than 8 g/dL. Hemorrhagic ascites was more common in patients of minority race or ethnicity (p=.001) and with a history of HCC (p<0.001). Patients with hemorrhagic ascites had a significantly higher rate of portal hypertensive bleeding (p<0.001), spontaneous bacterial peritonitis (SBP) (p< 0.001), higher MELD scores at presentation (p<0.001), acute kidney injury (AKI) (p<0.001), and were significantly more likely to require ICU-level care (p=0.01) compared to patients without hemorrhagic ascites. Furthermore, patients with hemorrhagic ascites had a significantly higher one-month mortality than controls, 47/214 (22%) vs. 91/642 (14%), respectively (p=0.007). Differences in the rate of SBP, AKI, and one-month mortality persisted in patients with idiopathic ascites (all p<0.001).Conclusions: Hemorrhagic ascites occurs most often in patients with advanced liver disease and more frequent complications of portal hypertension. These patients have a poor prognosis with a significantly higher rate of ICU level care, acute renal failure and overall mortality.

Mo1025 Can Quality of Care for Patients With Cirrhosis Be Measured? John T. Bassett, Michael Volk Recently, specific indicators have been proposed to measure the quality of care delivered to patients with cirrhosis (Clin Gastro Hep 2010;8:709). However, the feasibility of using these indicators in routine practice has not been determined. Methods: We applied the top four of these quality indicators to 114 hospitalizations occurring in 111 cirrhotic patients from June 2006-May 2009: 1) If hospitalized patients with ascites have an ascitic fluid polymorphonuclear count of 250 cells/mm3, then they should receive empiric antibiotics within 6 hours of the test result. 2) If patients with cirrhosis present with upper gastrointestinal (GI) bleeding, then they should receive upper endoscopy within 24 hours of presentation. 3) If patients with cirrhosis are found to have bleeding esophageal varices, then they should receive endoscopic variceal ligation (EVL) or sclerotherapy at the time of index endoscopy. 4) If patients with cirrhosis survive an episode of acute variceal hemorrhage, then they should receive one of the following therapies to prevent recurrence of variceal hemorrhage: EVL every 1-2 weeks until obliteration, beta-blockers, or a combination. Results: Of the 111 cirrhotic patients meeting criteria for study inclusion, the mean age was 55 years and 59% were male. There were 44 cases of SBP identified with empiric antibiotics given in 24 cases (54%) within 6 hours of presentation. Reasons for late administration included a delay from the time the order was written to the time the antibiotic was administered as well as a limited intravenous access. There were 70 cases of upper GI bleeding with tachycardia, hypotension, requirement for blood transfusion, or >2g drop in hemoglobin. Upper endoscopy was performed at our institution within 24 hrs in 30 (42.3%) of cases. Of the 40 cases in which early EGD was not performed at our institution, 9 (12.9%) had been performed at an outside hospital prior to transfer. Banding or sclerotherapy was performed in 38 (54.3%), and a repeat EGD was performed within 1-2 weeks and/or beta-blocker given in 52 (74%) of cases. However, it was not always clear when these interventions were medically appropriate. Only two patients had active variceal bleeding at the time of endoscopy and 23 had an esophageal red wale or nipple sign, while 5 had stigmata of gastric variceal bleeding. Conclusion: Applying performance indicators to measure the quality of care for patients with cirrhosis is fraught with logistical problems, including numerous denominator exclusions or unclear denominator, shared care between institutions, and lack of automated

Mo1028 Predictors for the Mortality of Spontaneous Bacterial Peritonitis(SBP) Salyavit Chittmittrapap, Panudda Srichomkwun, Suporn Sakarin, Sombat Treeprasertsuk, Piyawat Komolmit Aim: Spontaneous bacterial peritonitis (SBP) is serious complication of cirrhosis. Despite standard treatment, mortality remained high. We aimed to evaluate the predictors for the mortality in patients with SBP. Methods: Consecutive patients with ascites culture positive SBP admitted to King Chulalongkorn Memorial Hospital in 2006-2009 were enrolled. Univariate and multivariate analysis were used to identified the associated factors of death. Results: Of 154 episodes of suspected SBP, 94 patients had SBP with culture positive ascites (61 %). Etiology was HBV in 26 %, HCV 22 %, alcohol 22 %, cryptogenic 25 % and others 5 %. Gram positive and gram negative bacteria were found in 41 and 59 % respectively. Gram negative SBP showed quinolone resistant rate of 40 %. In-hospital mortality rate in Antibiotics resistant group was higher (39 % vs susceptible organism 27 %). Recurrent SBP (n=27) had 19 % mortality (compared to 34 %). Other parameters were showed in Table1. Univariate

S-957

AASLD Abstracts

AASLD Abstracts

Is Fecal Calprotectin a Useful Screening Parameter for Hepatic Encephalopathy and Spontaneous Bacterial Peritonitis in Cirrhosis? Felix Gundling, Holger Seidl, Christian Pehl, Wolfgang Schepp