Percutaneous endoscopic gastrostomy outcomes: can patient profiles predict mortality and weaning?

INTRODUCTION

The benefits of percutaneous endoscopic gastrostomy (PEG) remain controversial. Patient selection is important to identify those who will benefit from PEG. This study aims to identify patient factors that may help in patient selection for PEG.

METHODS

Patients who underwent PEG at the Gastroenterology Unit of Tan Tock Seng Hospital, Singapore, from January 1998 to December 1999, were identified. Multiple logistic regression was used to predict patient's outcomes.

RESULTS

There were 106 (61 male) patients with a mean follow-up period of 465 days (range 3-1,410 days). The mean patient age was 64.5 years (range 17-94 years). The 30-day, six-month, one-year and two-year mortality rates were 7.5 percent, 26.4 percent, 35.8 percent and 46.2 percent, respectively. Older age (p-value is 0.023), presence of bedsores (p-value is 0.042) and abnormal nutritional status based on body mass index less than 20 kg/square metres (p-value is 0.001) were predictive of mortality. 26 percent of patients were able to wean off PEG in an average period of 185 days (range 3-870 days). Patients were generally younger (p-value is 0.003) and had better renal function (p-value is 0.047).

CONCLUSION

Older age, poor nutritional status and presence of bedsores were predictors of poor outcome. Younger age and preserved renal function were significant predictors of weaning off PEG feeding.