An audit of percutaneous endoscopic gastrostomy in a general hospital in Singapore.

OBJECTIVE

We conducted an audit on 50 percutaneous endoscopic gastrostomies (PEGs) performed by physician endoscopists from January 1996 up to November 1997.

DESIGN

A retrospective cohort study was conducted.

RESULTS

The mean age of the patients was 68.5 years (range 20-101) and the main indications were cerebrovascular accident in 40 (80%), neurological dysphagia in six (12%) and head injury in three (6%). The interval between the diagnosis of dysphagia and PEG was > 60 days in 19 patients (38%), 31-60 days in eight (16%) and < or = 30 days in 23 (46%). The commonest reason for PEG insertion was intolerance to nasogastric tube in 49 patients (98%). Twelve patients had antibiotics given concurrently for other infections and two had antibiotics given specifically to cover PEG insertion. There was evidence of post-PEG infection in two of 14 patients given antibiotics (14%) and in 14 of 36 patients not given antibiotics (39%). By Cox regression, the adjusted relative risk of infection in patients receiving antibiotics versus those not receiving was 0.6927 (95% CI 0.3396-1.4130; not significant). The 30-day mortality was 7/48 patients (15%), with two patients lost to follow-up before 30 days. There were no deaths directly attributable to PEG.

CONCLUSIONS

PEG is still not adequately requested by doctors for patients needing enteral feeding for more than 30 days. The use of antibiotics in this retrospective cohort study failed to show any benefit in reducing the rate of infective complications.