Percutaneous endoscopic gastrostomy; evidence of different prognosis in various patient subgroups.

Background: As there are no prospective randomised trials about percutaneous endoscopic gastrostomy (PEG) insertion, the medical staff and caregivers encounter great difficulty in deciding when and if to perform this procedure.

Objective: To explore which variables are associated with increased mortality after PEG insertion.

Design: Prospective observational study.

Setting: Gastroenterological unit of a 500-bed community hospital.

Subjects: All patients over the age of 50 years referred for PEG insertion between January 1992 and December 2002.

Methods: Patients were studied for their indication for PEG insertion as well as their main medical problems, and demographic details and medical records were reviewed yearly until mortality.

Results: 674 patients were enrolled (mean age 80.1 years, 42% men). The median survival was worst in diabetic patients (128 days, P <0.05), patients referred from hospital (161 days, P <0.01) and patients over the age of 80 years with dementia (171 days, P <0.001). The best median survival was found among demented patients under the age of 80 (467 days, P <0.05) and women under the age of 80 referred from nursing homes (780 days, P <0.01).

Conclusions: The outcome after PEG insertion is variable, with survival of over a year in many of the patients. These data are important for the medical staff, the patients and their caregivers when deciding about PEG placement.