Gastrostomy complications in childhood

Abstract

Background The use of surgical and percutaneous endoscopic gastrostomy for complete or partial enteral feeding in childhood is increasing. Despite this there is little information about complications of gastrostomy and gastrostomy use in children. Aim To investigate the time course, prevalence and causative factors of the most common gastrostomy complications: granulation tissue (GT) and leakage. Methods Telephone and enface survey of all patients who received a gastrostomy the 5years from 1 June 1996 to 31 December 2000. Patients Eighty of 88 patients were surveyed (42 M: 38 F, mean age 6.3years±6months) The most common reasons for gastrostomy were neurological disorders n=28, cystic fibrosis n=7, cardiac disease n=6, cranio-facial n=6. A Nissen Fundoplication was performed on the majority of these 59 of 80. The most common device was the Bard® n=52. Results Leakage was seen in 62 of 80 (78%) and GT in 63 of 80 (79%) patients. GT was an early complication and present in 43 of 80 (54%) within 2months of procedure. GT formation was not associated with surgeon, type of device, socio-economic status, passive exposure to tobacco smoke, formula or mode of delivery of feed (continuous vs. bolus). Leakage and GT were commonly seen but the percentage of patients with GT and leakage was similar to the percentage of patients with GT and no leakage 52/64 v 12/17 (2=0.138; P=0.71). The most widely used treatment was AgNO3 n=61, in 90% GT resolved or improved. n=4 had surgical avulsion. Most common other therapies used were Kenacomb® n= 10 and Paw-Paw ointment n=14. Two patients required the device to be removed because of florid GT. Conclusions Granulation tissue occurs very commonly following childhood gastrostomy and usually in near proximity to procedure. Factors such as type of device, type of formula, mode of feeds and leakage do not influence granulation tissue formation. Most patients achieve satisfactory results with AgNO3 and removal of gastrostomy is seldom necessary.