, Volume 58, Issue 3, pp 768-776
Date: 25 Sep 2012

Percutaneous Endoscopic Gastrostomy in Cancer Patients: Predictors of 30-Day Complications, 30-Day Mortality, and Overall Mortality

Rent the article at a discount

Rent now

* Final gross prices may vary according to local VAT.

Get Access

Abstract

Background

Cancer patients benefit from percutaneous endoscopic gastrostomy (PEG) in many ways including nutritional support and venting in cases of malignant obstruction. Lack of high-quality studies with adequate follow-up has led to limited information regarding risk stratification and predictors of morbidity and mortality.

Aims

Elucidate predictors of complications and mortality with long-term follow-up in cancer patients undergoing PEG.

Method

Retrospective review of all patients undergoing PEG placement at MD Anderson Cancer Center from January 1, 2004 to December 31, 2006. Statistical analysis included descriptive statistics, Kaplan–Meier survival estimates, and Cox proportional hazards regression analyses.

Results

A total of 218 subjects underwent PEG. Those with American Society of Anesthesiology (ASA) scores of 4, 4E, or 5E were at significant risk of a major complication in the first 30 days. Multivariate analysis revealed ASA scores ≥4, elevated WBC count, and advanced tumor stage to be independent predictors of mortality in the first 30 days and INR >1.5 and diversion/venting as an indication for PEG placement to be independent predictors of overall mortality.

Conclusions

Patients with high baseline illness severity are more likely to have complications and are at increased risk of mortality after PEG. Our study results suggest that particular attention be directed to ASA score, INR, WBC counts, transfusion requirements, presence of advanced malignancies, and the indication for PEG placement when determining risk of complications or death. Patients undergoing venting PEG are expected to have short post-PEG survival but improvement in quality of life likely justifies the risks associated with PEG placement.