Bacterial growth and antibiotic resistance patterns of ascitic fluid in a single center in the United States

Abstract

Background: Approximately 30% of patients with ascites will eventually develop spontaneous bacterial peritonitis (SBP); others will develop bacterial growth in fluid with no inflammation (bacterascites). Most patients are treated with quinolone and cephalosporin antibiotics. Very few studies have addressed the bacterial epidemiology and antibiotic resistance of SBP or bacterascites in the United States. Aims: Primary aim was to describe bacterial types and resistance patterns of ascitic fluid in a Veterans Administration Medical Center (VAMC) in the United States, secondary aim was to measure mortality rates following a positive ascitic fluid bacterial culture. Methods: We retrospectively evaluated 752 samples of ascitic fluid over a period of 4 years, from January 2007 to December 2010. Fifty-two out of the 752 samples had bacterial growth. Twentyeight of these 52 were from patients with cirrhosis and were included in the study. Samples from postoperative patients or patients undergoing peritoneal dialysis were excluded. Results: Of the 28 samples, 24 had a SAAG >1.1, and only 7 samples had PMNs >250 (qualifying for SBP). The median age of patients was 59 years, all were male. Documented etiology of cirrhosis was alcohol abuse in 40% of patients, HCV infection in 14%, and hemochromatosis in 7%. Approximately 40% of the patients had no documented cause of cirrhosis. On initial evaluation, 25% had abdominal pain, none had fever, and 40% had leucocytosis. The most frequently isolated organisms were coagulase-negative staphylococcus (42%) and streptococcus viridians (21%). Other isolated organisms included, but were not limited to, Escherichia coli (14%), Propinobacterium Acne (7%) and Klebsiella sp (3.5%). Twenty-five percent of samples grew more than one organism. Seventeen samples out of 28 had sensitivity studies performed. Of these, 12% showed resistance to quinolones, 40% to cephalosporins and 14% to amoxicillin/clavulanic acid. Mortality at 1 and 6 months was 25% and 46% respectively and was independent of presence of resistant organisms in ascitic fluid or percentage of PMNs. Interestingly, 1-month mortality in alcoholic cirrhosis was lower than in other etiologies (p=0.04). There was a tendency of higher creatinine values to be directly associated with mortality (p=0.07). Conclusions: We found a high frequency of gram positive bacterial growth and resistance to cephalosporin antibiotics in ascitic fluid. These results differ from overall reviewed literature where a higher frequency of gram negative bacteria and higher resistance to quinolones are reported. This finding, however, did not seem to be associated with mortality.