Risk factors and complications following percutaneous endoscopic gastrostomy: a case series of 1041 patients.

BACKGROUND

Most studies exclude patients with severe coagulation disorders or those taking anticoagulants when evaluating the outcomes of percutaneous endoscopic gastrostomy (PEG).

OBJECTIVE

To investigate complications and risk factors of PEG in a large clinical series including patients undergoing antiplatelet and anticoagulant therapy.

METHODS

During a six-year period, 1057 patients referred for PEG placement were prospectively audited for clinical outcome. Exclusion criteria and follow-up care were defined. Complications were defined as minor or severe. Uni- and multivariate analyses were used to evaluate 14 risk factors. No standardized antibiotic prophylaxis was given.

RESULTS

A total of 1041 patients (66% male, 34% female) with the following conditions underwent PEG: neurogenic dysphagia (n=450), cancer (n=385) and others (n=206). No anticoagulants were administered to 351 patients, thrombosis prophylaxis was given to 348 while full therapeutic anticoagulation was received by 313. No increased bleeding risk was associated with patients who had above-normal international normalized ratio values (OR 0.79 [95% CI 0.08 to 7.64]; P=1.00). The total infection rate was 20.5% in patients with malignant disease, and 5.5% in those with nonmalignant disease. Severe complications occurred in 19 patients (bleeding 0.5%, peritonitis 1.3%). Cirrhosis (OR 2.91 [95% CI 1.31 to 6.54]; P=0.008), cancer (OR 2.34 [95% CI 1.33 to 4.12]; P=0.003) and radiation therapy (OR 2.34 [95% CI 1.35 to 4.05]; P=0.002) were significant predictors of post-PEG infection. The 30-day mortality rate was 5.8%. There were no procedure-related deaths.

CONCLUSIONS

Cancer, cirrhosis and radiation therapy were predictors of infection. Post-PEG bleeding and other complications were rare events. Collectively, the data suggested that patients taking concurrent anticoagulants had no elevated risk of post-PEG bleeding.