Gastroesophageal reflux disease in neurologically impaired children: the role of the gastrostomy tube.

Abstract

We review our experience with gastrostomy techniques in neurologically impaired (NI) children, with or without a Nissen fundoplication. The records of 130 NI children who had a gastrostomy tube (GT) placed between January 1999 and October 2001 were reviewed. Data collected included: demographics, neurological status, operative time, time to first feed, postoperative stay, analgesic requirements, follow-up, mortality and complication rates. Open GTs were placed using the standard Stamm gastrostomy technique through a midline incision and were combined with a standard open Nissen fundoplication when indicated. Laparoscopic GTs were placed after institution of carbon dioxide pneumoperitoneum using a 2-port technique, a Mic-key G device of appropriate size and anchored to the anterior abdominal wall with 2 "U" stitches. The laparoscopic Nissen fundoplication (LNF) procedures were performed using a 5-port technique. Patients were divided into 4 groups: group I (n = 12) laparoscopic GT alone, group II (n = 44) open GT alone, Group III (n = 44) laparoscopic GT with LNF and Group IV (n = 30) open GT with Nissen fundoplication. Based on their similar characteristics, Groups I and II and Groups III and IV were compared together. Data were analysed using Student's t test, and internal review board approval was obtained. Patients ranged in age between 10 days and 17.7 years (mean 3.64 years). Their weight was between 1.2 and 63.4 kg (mean 12.8 kg). The compared groups showed similar characteristics with regard to age, weight, cause of mental impairment, and the reason for placement of the GT. The operative time for group III was significantly longer than that of group IV (P < 0.05). Time to first feed was significantly shorter for group I when compared to group II. The postoperative analgesic requirements were not statistically different. The overall short- and long-term complication rates were not statistically different when the related groups were compared, however, site-related complications and feeding problems were significantly less in group I compared to group II. Only 1 operative mortality occurred in group III. Follow-up showed less long-term morbidity and fewer complications with the laparoscopic GT compared to the open one as regard to admissions, surgery, and emergency department visits related to GT problems as well as frequency of GT change. Based on our experience, laparoscopic placement of a low-profile GT in NI children appears to be associated with less morbidity, permits earlier enteral nutrition, and has a cosmetic advantage. We believe that the laparoscopic technique should be the procedure of choice for GT placement in these children even when a Nissen fundoplication is deemed necessary.