ABSTRACT
BACKGROUND: The prevalence of spontaneous bacterial peritonitis (SBP) in hospitalised cirrhotics with ascites is 10-30%. Treatment for refractory ascites includes paracenteses, transjugular intrahepatic portosystemic shunt or drain placement; the latter is discouraged due to a perceived infection risk. AIM: This study aimed to evaluate the risk of bacterial peritonitis (BP) with peritoneal drains in patients with Child-Pugh class B or C cirrhosis and determine their impact on survival. METHODS: We conducted a retrospective review of end-stage liver disease (ESLD) patients with non-malignant, refractory ascites who had peritoneal drains placed for ≥3 days at Loyola University between 1999 and 2009. Cell counts were performed at drain placement and within 72 h. BP was defined as ascitic polymorphonuclear neutrophils >250/mm(3) . Univariate analysis assessed the association between demographics, laboratory markers and development of BP. Kaplan-Meier curve estimates by infection were constructed and survival distributions were compared using log-rank statistic. RESULTS: There were 227 drain placements during the study period. Twenty-two per cent were diagnosed with BP (12% had SBP at drain placement; 10% developed BP within 72 h). There was no association between BP and baseline characteristics. Patients who developed BP within 72 h of drain placement had 50% mortality at 5 months compared with 50 months in those without infection (log-rank P ≤ 0.003). CONCLUSION: In ESLD patients who received an indwelling peritoneal catheter, there was 10% risk of developing BP and significant mortality increase. Though placing drains is not the mainstay of treatment for refractory ascites, we confirm the theoretical adverse risk of peritoneal drains on infection and survival in cirrhotics.
