Is It Safe to Place Percutaneous Gastric Feeding Tubes in Cirrhotic Patients?

Abstract

Background: A dilemma arises in deciding whether to perform percutaneous endoscopic gastrostomy (PEG) and percutaneous radiologic gastrostomy (PRG) in cirrhotic patients, because they may have varices or ascites and are more prone to complications. The literature on this topic is limited to case reports. Design: Case control study was conducted of all cirrhotics (age > 18) who underwent PEG or PRG between 1996 and 2007 at 2 tertiary care hospitals. The control group, selected from the same patient population, was matched for age, sex and type of procedure (PEG or PRG) and had multiple co-morbidities, as expected in those requiring feeding tubes. Baseline, post-procedure and 3 month follow-up labs were collected. Model for end-stage liver disease (MELD) scores were calculated. Charts were reviewed for complications. Mixed effect model of regression analysis assessed differences, both within the cirrhotic group and between the cirrhotic and control groups. Results: 40 cirrhotic patients had PEG (n=10) or PRG (n=30) [University hospital (n=29) or VA hospital (n=11)]. Mean age was 59.4, with 6 females and 34 males. Pre-procedure evaluation showed esophageal varices in 9, gastric varices in 2, and ascites in 4. The mean follow-up was 11.8 months. No complications were noted during the follow-up period in cirrhotics who underwent PEG. Two cirrhotics experienced significant pain in the immediate post-PRG period (6.7% complication rate), for a 5% overall complication rate in the cirrhotic group. No acute hepatic decompensations were recorded during follow-up. The complication rate in the control group was 18% (7/40), most minor. In cirrhotics with PEG, MELD decreased from 8.7 to 7.8 (p = 0.28) in the immediate follow-up period and was 6 (p = 0.77) at 3 month follow-up. Cirrhotics with PRG had a MELD drop from 11.9 to 11.5 (p = 0.53) during immediate follow-up, and it was 11.9 (p = 0.77) at 3 months. At baseline, MELD was 11.3 for the combined PEG/PRG cirrhotic group and 11.5 in the control group (p = 0.47). In the immediate post-procedure period, cirrhotic group MELD was 10.8 vs 13.6 for the control group (p = 0.04). At 3 months, cirrhotic group MELD was 11.4 vs 11.8 for the control group (p = 0.14). Conclusions: This is the largest evaluation to date of cirrhotic patients who have undergone PEG and PRG placement. Complication rate in the cirrhotic group was not greater than the control group or that quoted in the literature. This study is limited as selection bias invariably occurred in the cirrhotic group; however, this study found no evidence to support a poorer outcome in the selected cirrhotic patients after PEG or PRG tube placement.