Background and aims: Percutaneous endoscopic gastrostomy (PEG) provides enteral nutrition to patients who cannot swallow. Few studies have prospectively evaluated its long-term outcomes or eventual resumption of oral intake.
Methods: Consecutive PEG patients were prospectively recruited from a tertiary hospital over 12 months and followed until all had met the primary endpoints of death or resumption of oral diet with PEG extubation. Data was collected by standardised periodic phone interview.
Results: Forty patients (24 males, median age 74 years) were followed for up to 8.4 years (median 5.3 months, interquartile range [IQR] 13.6 months). The end-of-study mortality rate was 70% (median 6.8 months, IQR 19.9 months) and the only predictor of mortality was head injury as the indication for PEG (Cox regression HR 5.90, 95% CI: 1.2-28.4). At two years following PEG, 30% of patients had resumed oral intake (median 2.9 months, IQR 7.2 months) and 19% remained on PEG-feeding. Predictors of resumption of oral intake were the ability to tolerate some oral intake at 3 months (HR: 248.5, 95% CI: 8.7-7065.3) and 6 months (HR: 6.3, 95% CI: 1.03-38.9) but not at 12 months. Cumulative survival was highest for ear nose and throat (ENT) tumour and worst for acute head injury (log rank P = 0.048).
Conclusions: Half of all PEG patients remained alive at 2 years using PEG or have resumed full oral intake. A supervised trial of oral intake at 3 or 6 months may help predict eventual resumption of per oral diet.