Although long-term nasogastric tube feeding has become a standard therapy in the treatment of chronically sick children, it still encounters several problems. Dislocations of nasogastric tubes often require replacement or even cause severe aspiration. Permanent nasopharyngeal irritation may result in pronounced mucus production, which is particularly important in patients taking anticonvulsant drugs that are already associated with increased levels of mucus production. In children with transient impairment of swallowing, training to feed orally becomes more difficult because of local pharyngeal irritation, and outpatients are affected by the negative aspect of having a nasogastric tube in place while in public.
Since Gauderer et al. introduced the percutaneous endoscopic gastrostomy (PEG) procedure, long-term tube feeding has changed dramatically (1). Compared with conventional nasogastric tube feeding it has proved to be safer, more convenient for the patient, and much cheaper for health maintenance organizations (2). Surgical gastrostomy is rarely needed anymore. Although results in numerous studies have demonstrated efficacy, safety, and reliability of PEGs in adults, only a few studies have reported on PEGs in children (2-8). Here we report about our experience in the placement and maintenance of PEGs in a large group (139 patients) of children and adolescents.
PATIENTS AND METHODS
A total of 139 patients (79 males and 60 females, mean age 4.4 years (range, 0.06-36.5 years), mean body weight 14 kg (range, 3.1-60 kg) underwent PEGs between April 1989 and October 1995 in one of our three centers (Erlangen, n = 99; Leipzig, n = 27; and Altdorf, n = 13). Eleven patients more than 18 years old with severe cerebral palsy and dystrophy were included. Because of these long-term problems they resembled more the pediatric than the adult population.
Indications for the PEG are listed in Table 1. Most patients with swallowing disorders originating in the central nervous system had cerebral palsy, except one patient whose impaired swallowing was caused by a meningomyelocele associated with Arnold-Chiari syndrome and another who had a traumatic spinal cord injury. In 26 children with complex congenital cardiac malformations or chronic renal failure (n = 21) and neoplasms or cystic fibrosis (n = 5), PEG was placed to facilitate alimentation. In some of these patients general dystrophy was already present, whereas the other patients were provided with a PEG before becoming severely dystrophic. Seven patients with cystinosis, maple syrup urine disease, or chronic inflammatory bowel disease underwent a PEG because they did not tolerate oral feeding with the required (unpleasant tasting) special diet and became dependent on long-term tube-feeding. In one infant with congenital gastric volvulus, we tried to fix the stomach to the abdominal wall through the PEG. No patient had fundoplication.
PEGs were placed by using the pull-through method described by Keymling et al. (9) Except for patients with chronic vomiting in whom PEGs were placed in the duodenum, PEGs were generally located in the stomach. We used either a polyurethane (Freka 9 or 15 F, Fresenius; Bad Homburg, Germany) or a silicone set (Flocare 10 or 14 F, Pfrimmer-Nutricia; Erlangen, Germany). The endoscopies were performed with flexible endoscopes (FG 23 H, FG 24 X, FG 29 H, Pentax Handelsgesellschaft; Hamburg, Germany; and GIF P 20, GIF XP 20, Olympus Optical, Hamburg, Germany).
PEGs were performed under local anesthesia (1% mepivacaine hydrochloride) combined with either intravenous sedation (at Erlangen, 0.06-0.3 mg/kg midazolam and etomidate, either as a single or a repeat dose; at Leipzig, 1-2 mg/kg midazolam at first in combination with ketamine, later with midazolam alone); at Altdorf, 5-10 mg diazepam). No patient was given general anesthesia. No prophylactic antibiotics were given except to prevent endocarditis in patients with congenital cardiac malformations. After placement of the PEG, the patients were monitored for 24 hours. For analgesia they received paracetamol suppositories or continuous intravenous medication with tramadol (2-3 mg/kg/day). Tube feeding began 12 hours after the placement.
Sixty patients underwent repeat endoscopies for indications unrelated to the PEG. Among those, the outcome was evaluated retrospectively to determine whether esophagitis related to gastroesophageal reflux had appeared or had increased macroscopically after the procedure.
The clinical impact of the treatment was assessed 6 to 12 months after placement of the PEG. We asked the patients, the parents, the guardians, or the caregivers about changes in the patients' physical status and for their thoughts about the handling and care of these patients.
In 118 patients PEGs were placed in the stomach with a mean procedural duration of 12 minutes (range, 5-45 minutes). Twenty-one patients received a duodenal PEG with a mean procedural duration of 46 minutes (range, 24-105 minutes). Unfortunately, in 6 patients duodenal placement proved impossible. Two of these patients received a direct percutaneous endoscopic jejunostomy and the remaining 4 patients a gastric PEG. Initially, patients (n = 94) were given the small PEG (9 or 10 F), but in recent years the larger systems (14 or 15 F) have been used almost exclusively (n = 45) because of the lower risk of occlusion.
At the end of the follow-up period, PEGs had been in place in 60% of the patients for more than 1 year (mean period, 16 months). Sixty patients underwent repeat endoscopy for reasons unrelated to insertion of the PEG (gastrointestinal bleeding or signs of ulcerations). Seven had either macroscopically progressive esophagitis or new-onset esophagitis. In six patients esophagitis improved. In the majority (n = 47) there were no signs of inflammation, either initially or at the final follow-up examination.
We observed 19 notable complications, which we categorized as either clinical or technical (Table 2). In five patients, the duodenal tube was dislocated, with subsequent recurrence of vomiting. These patients underwent successful replacement of the tube. In two patients, a symptomatic pneumoperitoneum occurred. In both patients, perforations closed within 24 hours after adaptation of the inner and outer retention disk. Signs of systemic inflammation could be managed successfully within 2 days by intravenous antibiotic regimens (cephalosporins with aminoglycosides). Three patients developed a severe local inflammation, requiring removal of the PEG in two children, whereas it was left in situ after intravenous antibiotic treatment in the third child. In one patient with a gastric PEG sustained vomiting developed. After an unsuccessful attempt to exchange the gastric PEG for a duodenal PEG, he underwent a surgical jejunostomy. Technical problems were seen in eight patients: In four the inner retention disk became disconnected. Those patients underwent endoscopic extraction of the disrupted intragastral part of the tube with subsequent placement of a new system. In another four patients occlusion of the tube developed (after application of drugs through the feeding tube). In these patients the system was exchanged for a larger tube.
The patients with feeding difficulties could be trained more easily without the irritating nasogastric tube. In this group there was a marked weight gain of up to 52% in 8 to 13 months of nutrition by PEG. The excess of mucus production and rate of upper airway infections decreased markedly. In some of the children with cerebral palsy, muscular spasms decreased. Finally, children who required special diets were no longer handicapped in their daily life by the nasogastric tube.
Asked about their attitude toward the PEG, all involved (patients, parents, and caregivers) voted in favor of the PEG and considered it a major improvement in long-term tube feeding compared with former procedures. They would definitely choose this treatment again in a similar situation.
Today, tube feeding is a very common method of enteral feeding in patients who cannot eat or drink by themselves. New systems, like PEG, must be proved better that current methods before being widely implemented. The potential benefit of a PEG depends very much on the underlying disease and cannot be easily assessed on the basis of one result (e.g., weight gain). For instance, in patients who are more or less immobile, a major weight gain would cause more problems than benefits in everyday care. Other children who require PEGs for long-term tube feeding of special (unpleasant) diets usually do not have dystrophy.
In accordance with reports in the literature, the majority of our patients required a PEG for long-term tube feeding because of central dysphagia (3). Most of these severely handicapped patients need lifelong tube feeding. The nasogastric tube and anticonvulsant therapy for seizures in the majority of these patients lead to production of a large volume of mucus with recurrent upper airway infections and even aspiration. In our patients the mucus production and the rate of infections decreased markedly. Avoidance of repeated insertion of nasogastric tubes, which caused pain in the hypopharynx, could be the reason for the demonstrated decrease in muscle tonus. Asked about the general welfare of these patients, parents and caregivers agreed in all cases that the PEG had resulted in a major improvement in everyday life. With a PEG there was more time for the general care of the patients. A similar beneficial effect was described by Heine et al. in their patients (8).
Less attention has been paid to the treatment of malnutrition in infants with chronic cardiac dysfunction or renal failure. These children represent a completely different group of patients. They require up to 30% more calories than healthy subjects (10); therefore tube feeding is essential for prevention or treatment of severe dystrophy. In contrast with dystrophic patients, these children are mentally normal and are less restricted in their physical activities. In the months before the placement of the PEG, infants extracted their nasogastric tubes, sometimes several times a day. This situation was extremely stressful for both the infants and the parents. In these groups the main benefit was reduced stress for patients and parents and, only secondarily, a gain in weight.
A completely different situation exists in adolescents with dystrophy caused by neoplasms. Almost all of these patients refused the option of feeding by nasogastric tube because of its negative cosmetic aspect. Long-term tube feeding by PEG achieved a major improvement in the patients' nutrition and in their psychosocial situation. One patient showed a weight gain from 50 up to 76 kg in 13 months.
In accordance with reports in the literature (7), we also found a beneficial effect of PEG in the treatment of patients with cystic fibrosis or muscular dystrophy, if normal alimentation became impossible because of severe dyspnea. Weight gain was limited because of end-stage disease, but the situation was improved subjectively by undisturbed breathing. Probably the PEG should have been introduced earlier in these patients.
Finally, one of our patients had chronic vomiting with subsequent dystrophy caused by a gastric volvulus. We tried to achieve a caudal fixation of the stomach with the PEG to improve the passage, as described in the literature (11). Although there was a major reduction in vomiting with subsequent adequate weight gain, this patient finally required surgical gastropexy because of refractory singultus.
In the majority of our patients (n = 118) the PEG was placed in the stomach without intraoperative problems. In the case of delayed gastric emptying or severe gastroesophageal reflux, a duodenal placement was necessary-a much more difficult procedure than gastric placement, and sometimes even impossible. Therefore, one patient with chronic vomiting required a surgical catheter-jejunostomy. In two additional patients we performed a direct endoscopic jejunostomy. To the best of our knowledge, these are the first such procedures in children reported in the literature. During a follow-up period of 10 to 12 months, no problems occurred in these children. These results contradict the unfavorable results reported in adult patients (12,13).
In the beginning we used small PEG systems with an outer diameter of 9 or 10 F. Later, we changed to larger systems. They allow later button implantation or insertion of small tubes into the duodenum, if a gastric tube feeding is no longer possible. Additionally, occlusion of the tube was avoided with the larger system. On the basis of our experience we recommend the application of the larger system in all patients.
There is no procedural difference between infants and children in surgical placement of the PEG. In our experience (14) the PEG was placed under sedation or intravenous anesthesia without difficulty. No patient was given general anesthesia, in contrast with reports in the literature (2-4).
Complications after a PEG can be divided into clinical and technical problems (Table 2). Minor clinical complications include local infections and pain at the entry site. Minor inflammation of the skin at the entry site is common and usually resolves without specific treatment. Dislocations of the duodenal part of the tube into the stomach occurred in five patients with a duodenal PEG. This experience demonstrates the necessity in duodenal PEGs of placing the tip of the tube beyond the ligament of Treitz. In our work, dislocations no longer occurred with this method.
Major clinical complications (bleeding, perforation, peritonitis, fistulae, ileus) have been reported in adult populations in 2.5% of the reported cases (15-22) and in pediatric cohorts in 9% of the cases (6-54%; 3-5). In our study severe complications occurred in six patients (4%). At the beginning of our experience two patients had a perforation (symptomatic pneumoperitoneum) after implantation of a 15-F system. Both patients developed an acute abdomen immediately after creation of the PEG, accompanied by peritonitis in one of them. These complications however did not require surgical treatment. After adaptation of the inner and outer retention disk and antibiotic treatment, both children became completely asymptomatic within 48 hours; and tube feeding began without further problems. McKealey et al. reported a similar patient who was cured with peritoneocentesis only (23). The low rate of local infections in our cohort (in three patients) is remarkable, in that we did not perform prophylactic antibiotic treatment or oral decontamination of the nasopharynx. Today, decontamination is not practiced in adult patients, as well, without subsequent increase of the infection rate (Keymling, 1994, personal communication, ).
The mortality rate with PEG has been reported in the literature as 1% (4,6,7,15-19,21,22,24,25). Although we performed this procedure in very young infants, no death in our series was related to the PEG.
Technical problems include fracture of the catheter and disruption of the interior retention disk. The mean rate of these complications in adult patients is 4.3% (15-22). Data on complications in infants and children are presently not available in the literature. In our series those complications were observed at a slightly higher rate (6%) than that reported in adults. In four patients we found a disconnection of the interior retention disk. The connection of the tube and the interior retention disk has been changed in the meantime, however, and disconnections have not been observed in the last 5 years. Overall, in recent years there have been no major clinical or technical problems although the illnesses of the patients have progressed, and they have lost weight steadily.
Gastroesophageal reflux related to the PEG is controversial. It depends on the method of diagnostic (continuous pH measurement or endoscopy with or without histologic examination [8,26] and on the feeding method (bolus versus continuous feeding ). Gastroesophageal reflux did not occur in those patients continuously fed through the intragastric tube. In our cohort of 60 patients who underwent repeat endoscopy 6 to 12 months after creation of the PEG, only 7 patients (12%) showed esophagitis that had not been noted before the PEG or that had increased after the procedure. Furthermore, it must be remembered that especially patients with cerebral palsy carry an increased risk of developing this complication in the natural course (8,28,29). In our opinion the rate of gastroesophageal reflux esophagitis is probably not increased after creation of a PEG.
According to our experience PEG is superior to nasogastric tubes in children requiring long-term tube feeding. Since it is associated with a low rate of complications it should be considered more often even in infants.
1. Gauderer MWL, Ponsky JL, Izant RJ. Gastrostomy without laparotomy: A percutaneous endoscopic technique. J Pediatr Surg 1980;15:872-5.
2. Cory DA, Fitzgerald JF, Cohen MD. Percutaneous nonendoscopic gastrostomy in children. American Journal of Roentgenology 1988;151:995-7.
3. Gauderer MWL. Percutaneous endoscopic gastrostomy: A 10-year experience with 220 children. J Pediatr Surg 1991;26:288-92.
4. Küster P. Erfahrungen mit der perkutanen endoskopischen Gastrostomie in der Pädiatrie. Monatsschrift für Kinderheilkunde 1994;142:101-5.
5. Mago H, Chen CL, Wesson DE, Filler RM. Incisionless gastrostomy for nutritional support. J Pediatr Gastroenterol Nutr 1986;5:66-9.
6. Marin OE, Glassman MS, Schoen BT, Caplan DB. Safety and efficacy of percutaneous gastrostomy in children. Am J Gastroenterol 1994;89:357-61.
7. Steinkamp G, von der Hardt H. Improvement of nutritional status and lung function after long-term nocturnal gastrostomy feedings in cystic fibrosis. J Pediatr 1994;124:244-9.
8. Heine RG, Reddihough DS, Catto-Smith AG. Gastro-oesophageal reflux and feeding problems after gastrostomy in children with severe neurological impairment. Dev Med Child Neurol 1995;37:320-9.
9. Keymling M, Schlee P, Wörner W. Die perkutane endoskopisch kontrollierte Gastrostomie. Dtsch Med Wochenschr 1987;112:182-3.
10. Barton JS, Hindmarsh PC, Scrimgeour CM, Rennie MJ, Preece MA. Energy expenditure in congenital heart disease. Arch Dis Child 1994;70:5-9.
11. Eckhauser ML, Ferron JP. The use of dual percutaneous endoscopic gastrostomy in the management of chronic intermittent gastric volvulus. Gastrointest Endosc 1985;31:340-2.
12. DiSario JA, Foutch PG, Sanowski RA. Poor result with percutaneous endoscopic jejunostomy. Gastrointest Endosc 1990;36:257-60.
13. Henderson JM, Strodel WE, Gilinsky NH. Limitations of percutaneous endoscopic jejunostomy. J Parenter Enteral Nutr 1993;17:546-50.
14. Behrens R, Seiler A, Rupprecht T, Lang T. Sedierung versus All-gemeinnarkose in der pädiatrischen Endoskopie. Klin Pädiatr 1993;205:158-61.
15. Honneth J, Nehen HG. Perkutane endoskopische Gastrostomie. Dtsch Med Wochenschr 1991;116:1532-3.
16. Hull MA, Rawlings J, Murray FE, et al. Audit of outcome of long-term enteral nutrition by percutaneous endoscopic gastrostomy. Lancet 1993;341:869-72.
17. Keymling M. Perkutane endoskopisch kontrollierte Gastrostomie. Z Gastroenterol 1989;27(suppl):65-8
18. Langhans M, Henneking K. Enterale Ernährung durch perkutane endoskopische Gastrostomie (PEG). Endoskopie heute 1993;6:91.
19. Markgraf R, Geitmann K, Pientka L, Scholten T. Langzeitergebnisse enteraler Ernährungstherapie über perkutane endoskopische Gastrostomie bei polymorbiden internistischen Patienten. Z Gastroenterol 1993;31(suppl):21-3.
20. Meier R, Bauerfeind P, Gyr K. Die perkutane endoskopische Gastrostomie in der Langzeiternährung. Schweiz Med Wochenschr 1994;124:655-9.
21. Ponsky JL, Gauderer MWL, Stellato TA. Percutaneous endoscopic gastrostomy. Review of 150 cases. Arch Surg 1983;118:913-4.
22. Wenk A, Krauss EC, Markreiter M, Holzapfel P. Komplikationen der perkutanen endoskopisch kontrollierten Gastrostomie (PEG)-Erfahrungsbericht über 180 Patienten. Verdauungskrankheiten 1994;12:61-6.
23. McKealey WDC, McCallion WA, Boston VE. Tension pneumoperitoneum: A potentially life threatening complication of percutaneous endoscopic gastrojejunostomy. J Pediatr Gastroenterol Nutr 1996;22:334-5.
24. Kirby DF, Delegge MH, Flemming CR. American Gastroenterological Association technical review on tube feeding for enteral nutrition. Gastroenterology 1995;108:1282-1301.
25. Gossner L, Ludwig J, Hahn EG, Ell C. Risiken der perkutanen endoskopischen Gastrostomie. Dtsch Med Wochenschr 1995;120:1768-72.
26. Grunow JE, Al-Hafidh A-S, Tunell WP. Gastroesophageal reflux following percutaneous endoscopic gastrostomy in children. J Pediatr Surg 1989;24:42-4.
27. Coben RM, Weintraub A, DiMarino AJ, Cohen S. Gastroesophageal reflux during gastrostomy feeding. Gastroenterology 1994;106:13-8.
28. Flake AW, Shopene C, Ziegler MM. Anti-reflux gastrointestinal surgery in the neurologically handicapped child. Pediatric Surgery International 1991;6:92-4.
29. Sondheimer JM, Morris BA. Gastroesophageal reflux among severely retarded children. J Pediatr 1979;94:710-4.
In the Next Issue
Small Bowel and Fecal Microbiology in Children Suffering from Persistent Diarrhea in Bangladesh
Pradip K. Bardhan, M. John Albert, Nur H. Alam, Shah M. Faruque, Prodyut K. B. Neogi, and Dilip Mahalanabis
Vomiting and Gastroesophageal Motor Activity in Children with Disorders of the Central Nervous System
Alberto M. Ravelli and Peter J. Milla
Pundendal Nerve Function in Normal and Encopretic Children
Stephen M. Sentovich, Stuart S. Kaufman, Rebecca L. Cali, Peter M. Falk, Garnet J. Blatchford, Dean L. Antonson, Alan G. Thorson, and Mark A. Christensen
Vitamin E Status in Preterm Infants: Assessment by Plasma and Erythrocyte Vitamin E-Lipid Ratios and Hemolysis Tests
D. van Zoeren-Grobben, N. J. M. Jacobs, E. Houdkamp, J. H. N. Lindeman, G. F. Drejer, and H. M. Berger
Nitric Oxide and Inflammatory Bowel Disease: Evidence for Local Intestinal Production in Children with Active Colonic Disease
Jeremiah J. Levine, Michael J. Pettei, Elsa Valderrama, David M. Gold, Bradley H. Kessler, and Howard Trachtman