Bleeding after percutaneous endoscopic gastrostomy is linked to serotonin reuptake inhibitors, not aspirin or clopidogrel

Background: Percutaneous endoscopic gastrostomy (PEG) is an invasive procedure that can result in bleeding. Guidelines recommend discontinuing clopidogrel for 7 to 10 days, but not withholding aspirin, before PEG. Serotonin reuptake inhibitors (SRIs) have been associated with an increased risk of GI bleeding. Objective: To determine whether there is an association between periprocedural aspirin, clopidogrel, or SRI use and bleeding in patients who underwent PEG tube placement. Design: Retrospective cohort study. Setting: Large quaternary-care academic medical center. Patients: A total of 990 patients (525 men) with a median age of 69.8 years who underwent PEG from January 1999 to April 2009. Interventions: PEG tube placement. Main Outcome Measurements: GI bleeding. Results: Sixteen patients (1.6%) had evidence of bleeding during the first 48 hours after PEG, and 12 patients (1.2%) had evidence of bleeding between 48 hours and 14 days after PEG. Thirty-six patients (3.6%) received high-dose aspirin (>325 mg), 27 patients (2.7%) received clopidogrel (75 mg), and 99 patients (10%) received an SRI before PEG. Twenty-four patients (2.4%) received high-dose aspirin, 25 patients (2.5%) received clopidogrel, and 130 patients (13.1%) received an SRI after PEG. Multivariate analysis demonstrated no association between periprocedural use of aspirin (at any close) or clopiclogrel and post-PEG bleeding. However. SRIs administered 24 hours or less before PEG were associated with a significantly higher odds of post-PEG bleeding (adjusted odds ratio 4.1; 95% CI, 1.1-13.4; P = .04). Limitations: Retrospective, single-center study with limited statistical power despite a relatively large cohort of patients. Conclusions: Use of aspirin or clopidogrel before or after PEG was not associated with procedure-related bleeding. SRI use in the 24 hours before PEG was associated with an increased risk of bleeding. (Gastrointest Enclose 2011;74:22-34.)