Hospital and long-term outcome after percutaneous endoscopic gastrostomy

Background: Percutaneous endoscopic gastrostomy (PEG) has become the preferred method of providing long-term enteral nutrition. While hospitalized patients frequently have PEG inserted to facilitate enteral nutrition, little is known about these patients. The objective of the study was to determine hospital and long-term survival in patients who receive PEG while hospitalized for medical or surgical reasons. Methods: Records of all patients aged 18 years and older who underwent PEG between January 1, 1999 and December 31, 2004 at a university- affiliated community-based tertiary care center were examined. Results: 80 (11%) of 714 patients died during the index hospitalization. Older age, being married, mechanical ventilation, and dialysis were statistically significant predictors of hospital death (P < 0.05). There were nine complications and no deaths directly attributable to PEG. Overall survival was poor with 5.6% of patients dying within seven days of the procedure. Mortalities at 30, 60, and 365 days were 22%, 31% and 48%, respectively. Of the 80 patients who died prior to discharge, 40 (50%) died within one week of PEG placement. Fourteen (35%) of these 40 patients had treatment withdrawn. Kaplan-Meier median survival was 412 +/- 73 (mean +/- standard error) days. By Cox proportional hazard modeling, older age, cancer, heart disease, nonwhite race, and dialysis were significant predictors of post-PEG death (P < 0.05). Conclusions: Outcome after PEG is dependent on demographic factors and patient comorbidities. Given the very low initial complication rates, it may be advisable to delay PEG placement until just prior to discharge in order to prevent unnecessary procedures on those patients who are not likely to survive.