Article Text

Download PDFPDF
PWE-127 Microbiological assessment of ascitic fluid in liver disease: culture techniques, sensitivities and interpretation
  1. D Sutherland1,
  2. F Buckle1,
  3. N Lachlan1,
  4. U Altmeyer2,
  5. M Priest3,
  6. S Barclay1,
  7. A Stanley1,
  8. EH Forrest1
  1. 1Gastroenterology
  2. 2Microbiology, Glasgow Royal Infirmary
  3. 3Gastroenterology, Gartnavel General Hospital, Glasgow, UK


Introduction Diagnostic paracentesis is a routine procedure for the assessment ascitic patients and the diagnosis of spontaneous bacterial peritonitis (SBP). It is recommended that samples are sent in blood culture bottles (BC) as well as universal containers (UC) for the best culture yield. However there is uncertainty regarding the significance of bacterascites (BA). We aimed to determine the culture yield of paired BC and UC samples and to assess outcome relative to the presence of SBP, BA or culture negative neutrocytic ascites (CNNA).

Method This was a retrospective review of all the ascitic fluid samples sent for microbiology analysis in North Glasgow between November 2010 and December 2013. Measured white cell count (WCC) and bacterial cultures were noted. Positive samples were defined as follows: BApositive culture with WCC <250; CNNAWCC >250 but negative culture; SBPwas positive culture with WCC >250. Samples were classified on the basis of the initial sample or if >3 months since the previous sample.

Results A total of 4131 samples were received from patients with liver disease. Of these there were 1520 BC and UC paired cultures in 503 individual patients; the majority having alcoholic liver disease (74%). Pathogens were cultured in 118 paired samples: 47% only in BC; 34% in both BC and UC; 19% only in UC. Commensals were more common in UC: 72% compared with BC, 34%. 90 positive samples had had previous negative samples: 24 had negative samples less than 1 week prior to positive sample. 69% of pathogens cultured were sensitive to co-amoxiclav. BA and SBP were associated with a worse outcome compared with negative samples: HR 3.46 (2.0, 5.98); p < 0.0001 and HR 0.082 (0.039, 0.175); p < 0.0001 respectively. Culture of any pathogen was associated with a reduced 90 day survival compared with negative samples: HR 0.244 (0.159, 0.374); p < 0.0001.

Conclusion Both BC and UC samples yield relevant pathogen cultures and both should be requested for the best chance of a positive culture. BA is not a benign finding. Direct culture of any pathogen irrespective of WCC is associated with a worse 90 day outcome.

Disclosure of interest None Declared.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.