Supporting nutrition and promoting growth of vulnerable children is central to paediatrics and child health. Children whose nutritional needs cannot be met naturally must be artificially fed either enterally or parenterally. The use of the percutaneous endoscopic gastrostomy (PEG) technique has become a standard practice in adult and paediatric settings in overcoming swallowing and other feeding-related problems. The advantages over nasogastric feeding are, for example, better feed tolerance, reduction in spontaneous extubation, distress with repeated reinsertion, and misplacement of nasogastric tubes. For these obvious reasons, PEG has come to be the preferred option for providing medium and long-term nutritional support in patients with impairment of feeding ability leading to undernutrition.
In addition to improving the nutrition and growth of subjects (1,2), the use of the PEG can significantly reduce feeding time and ease drug administration. Improvement in the social functioning, mental, general health perception, and quality of life of caregivers has been demonstrated in prospective cohort studies (3). The nutritional status of unwell children is a common cause of anxiety for parents and feeding times can be stressful (4). The impact of PEG feeding is positive with many parents reporting a high level of satisfaction (5) and wishing the intervention to have taken place earlier (6).
Nutritional support with use of the PEG has been demonstrated in children with neurodisability (6,7), cystic fibrosis (8,9), neonatal pulmonary disease (10), congenital heart disease (CHD) (11,12), Crohn disease (13), oncological conditions (14), metabolic disease, genetic-chromosomal, and degenerative disease (1,15).
We reviewed the indications for PEG in a UK tertiary paediatric gastroenterology service and ascertained procedure-related complications and outcomes for a 5-year period.
Prospectively collected data relating to PEG insertion, removal, or change to low profile button devices (LPBD) were collated from endoscopy records held at the gastroenterology department of the Alder Hey Children's National Health Service Foundation Trust, Liverpool, UK, between 2002 and 2006. Additionally, information was verified using the hospital computerised patient information system, which provides secure online intranet access to histopathology, radiology, nursing, and operating theatre reports. Case notes were reviewed to confirm clinical course and complications. The study was approved by the clinical audit department of Alder Hey Children's National Health Service Foundation Trust.
Decisions to insert a PEG in these patients were taken following multidisciplinary assessment at either a neurodisability or cardiac feeding clinic where children were assessed by a gastroenterologist, neurologist or cardiologist, speech-language therapist, dietician, and a specialty liaison nurse. Subjects with neurological disease had a detailed clinical assessment of swallowing ability and if necessary a video fluoroscopy (VFS) was performed, to define the safety of swallow and exclude anatomical abnormalities. Due to different indications for PEG insertion, the preoperative workup was tailored per individual patient needs. For example, children with cerebral palsy (CP) underwent (in the majority of instances) VFS to assess safety of swallow. An unsafe swallow indicated PEG insertion. VFS was not needed in children with congenital cardiac disease or in the majority of children with other indications. Twenty-four-hour pH studies were available on a significant proportion of the children with CP. As a departmental policy, we have not been keen on 24-hour pH studies because they document acid reflux only and, therefore, based decisions regarding PEG placement on a multidisciplinary assessment of clinical symptoms, growth, swallowing abilities (speech-language therapy assessment), and nutrition (dietetics). Investigations were chosen to complement assessment to plan management. Undernutrition in infants with CHD due to fatigue on feeding, fluid restriction, and increased metabolic expenditure due to cardiac failure were the usual indications for referral to the cardiac feeding clinic. Interdepartmental and regional referrals for feeding support were met through the service for a variety of other conditions and clinical indicators.
The Corflo, size 12 F PEG tube (Merck Ltd, UK) with a collapsible internal retention dome is the standard device used at our centre. It is inserted endoscopically using the “pull” technique as described by Gauderer et al (16) under general anaesthesia. We have preferred this device over the FREKA PEG device (Fresenius Kabi) because it has a noncollapsible internal bolster and, therefore, needs to be endoscopically removed. All patients received 1 dose of intravenous cefotaxime at PEG insertion or appropriate endocarditis prophylaxis if there was underlying CHD. Permanent PEG removals were performed when children no longer required the PEG for feeding support as decided following a multidisciplinary assessment. The PEG was replaced with a LPBD when long-term feeding support was desired. The “Mickey button” (Kimberley Clark) size 14 F has been our standard LPBD (of appropriate length as measured at PEG removal), but we have used the Bard button in individual instances for repeated inadvertent removal of the Mickey button. The Bard device (Bard Access Systems) has a malecot-style internal bolster, which is inserted by stretching it over a metal introducer. It has an internal antileak valve.
PEG removal was performed by traction under general anaesthesia, predominantly by a specialist nurse. However, a proportion of children underwent endoscopic removal during the first 2 years of the study period before traction removal was adopted as the departmental policy. Microsoft Excel 2003 (Microsoft Corp, Redmond, WA) and SPSS version 13 (SPSS Inc, Chicago, IL) were used to analyse data.
Six hundred one PEG-related procedures were performed during the study period (384 insertions, 165 conversions to LPBD, 49 PEG removals, and 3 PEG to PEG conversions). The main indications for PEG insertion (Table 1) were neurodisability (160/384; 41.6%, median age at insertion 3.56 years) and CHD (115/384; 30%, median age at insertion 0.39 year). Cerebral palsy was the single most common indication for PEG insertion (Table 2). Among those subjects with CHD (Table 3), atrioventricular septal defect in children with Down syndrome was the predominant indication (21/115) followed by isolated ventricular septal defect (19/115). The age of insertion of PEG (Table 4) was significantly different between the 2 main groups (neurodisability 3.56 years vs CHD 0.39 year; P < 0.001 [t test]). Among the children with CHD the age of insertion was not significantly different between groups of cyanotic or acyanotic CHD disease (cyanotic CHD group [n = 47; 0.39 year; 0.1–6.85] vs acyanotic CHD group [n = 68; 0.38 year; 0.14–4.77]). Eight of the 384 patients (2%) proceeded to an open gastrostomy. The indications were poor transillumination due to abnormal gastric position in 4 patients with CP, previous scarred abdomen from peritoneal dialysis in 1 patient, previous abdominal surgery for omphocele in 1 patient, and probable colonic interposition in the remaining 2. The duration of hospital stay following PEG insertion was 1 day (mode) with a range of 1 to 14 days. No patient remained in hospital beyond this time having had a PEG as the sole procedure. Fifty-nine of the 384 patients were discharged on the same day of PEG insertion without complication (day case gastrostomy). Most of these patients were fed by a pump through a nasogastric tube and, therefore, families did not require further training in tube feed management.
Erythema at the PEG insertion site was the most common complication after PEG insertion (59/384; 15.4%). In most instances this was treated as presumed infection with oral antibiotics in 36 and intravenous antibiotics in 4 instances. Other complications included abscess at the insertion site in 2 patients. This was treated with intravenous antibiotics. Surgical drainage was not necessary. Bumper migration into the subcutaneous tissue occurred in a 10-year-old boy with a severe seizure disorder. He presented with pain after feeding 1 month after PEG insertion. On endoscopy, the internal bolster was not identified within the stomach. Entanglement of PEG tube or attempts at inadvertent PEG removal are likely explanations. He needed PEG replacement. Inadvertent PEG removal occurred in 3 patients consequent to these children pulling on the PEG tube. Foley catheters were placed as an emergency measure before replacing them with LPBDs.
Tube migration into the pylorus was noticed in 1 instance, when, on a subsequent endoscopy the internal bolster was noted at the pylorus rather than being in the body of the stomach. The redundant portion of the PEG tube was in situ. PEG was replaced with an LPBD.
An exploratory laparotomy was needed in a 6-year-old girl with medulloblastoma who underwent PEG insertion. Severe abdominal pain and possible intestinal obstruction in the hours following the procedure indicated laparotomy. There were no findings at laparotomy. Improved feed tolerance was noted with resolution of postoperative ileus.
No procedure-related mortality ensued. Among the patients who had CHD, recorded complications were episodes of presumed infection at the PEG site treated with oral antibiotics in 14 and intravenous antibiotics in 1 instance. One child had recurrent PEG site infections, which remitted with PEG removal. Complications from PEG insertion and removal are summarised in Table 5.
Two hundred seventeen children underwent PEG removal during the study period. In 165 patients, the PEG was substituted with an LPBD, whereas in 49 children the PEG was removed because it was no longer required for feeding support. Three patients underwent PEG-to-PEG conversions. PEG removal was performed endoscopically in 51 and by traction in 166 children, respectively. This was a reflection of change in departmental policy with more PEG removals done by traction in the latter 3 years of the study period. The median duration between the PEG insertion to LPBD conversion was 0.83 year (0.12–3.86).
Over the study period 49 patients had the PEG permanently removed because it was no longer required for feeding support (median duration after PEG insertion 1.31 years). The indication for the PEG insertion in 24 of 49 children was CHD (49%: median duration of insertion 1.29 years; range 0.31–5.03 years). Twelve had an underlying neurodisability (24%; median duration of insertion 1.3 years; range 0.32–3.46). The following diagnoses were present in the neurodisability group in which the PEG could be permanently removed: CP, 4; seizure disorder, 2; head injury following road traffic accident, 2; hydrocephalus, 4 (tubercular 1, Dandy Walker syndrome 1, postmeningitic 1, other 1). Four children with cleft lip and palate, 2 with cystic fibrosis, 4 subjects with faltering growth, and 1 subject each with medulloblastoma, underlying metabolic disease, and chronic renal failure also recovered feeding skills and underwent permanent PEG removal.
Most patients went home on the same day (mode) after the PEG removal or conversion to LPBD (range 0–9 days). Complications from traction removal of PEG included bumper separation in 2 cases (allowed to pass per rectum, uneventfully), failure to insert button requiring PEG to PEG conversion in 1 patient, an enterocutaneous fistula requiring surgical closure of the stoma in 1 patient, and misplacement of a button device in an individual patient requiring surgical intervention. Disruption of the stoma track/creation of a false track occurred in 2 patients who underwent endoscopic PEG removal. Complications from PEG removal are summarised in Table 6.
Neurodisabilty and CHD were the main indications for PEG insertion in our experience. Cerebral palsy was the single most important indication for PEG insertion. Children with CHD were recognised as requiring feeding support at an earlier age than children with neurodisability irrespective of cyanotic or acyanotic CHD. Although CHD was the indication for PEG insertion in up to 30% of instances, a greater percentage of PEG removal (50%) was seen in this group during the study period. A low incidence of complications was noted in this group despite being younger at the time of PEG insertion. Feeding support in infancy was also required in patients with cleft lip/palate and chronic lung disease in contrast to those with cystic fibrosis and oncological conditions who were older at the time of PEG insertion. The differences in the age of PEG insertion for the major indications reflect on the critical times of nutritional support needed by these varied groups of children. Our experience establishes congenital heart disease (CHD) as 1 of the major indications for PEG insertion in children. A low incidence of complications in children with CHD despite needing PEG insertion in infancy compared to children with neurodisability is also noteworthy. Over a period of time, greater proportion of children with CHD feed independently as evidenced by higher permanent PEG removal rates.
The most common indication for an open gastrostomy was poor transillumination due to presumed abnormal gastric position in CP, but also occurred with previous abdominal surgery. The length of hospital stay following PEG insertion was 1 day, whereas most patients undergoing PEG removal were discharged on the same day. No mortality or serious morbidity was reported with PEG-related procedures. Presumed infection was the most common complication following gastrostomy insertion.
Craig et al (1) have reported PEG experience in a north London cohort where the predominant indication for insertion of PEG was CP followed by genetic syndromes, metabolic syndromes, and progressive degenerative disorders. An inability to swallow was the principal indication for PEG insertion in a South African series (17), whereas neuromuscular and metabolic causes (18) and faltering growth (19) were the most important indication in other studies.
This study has not set out to analyse the long-term outcomes of children requiring nutritional support delivered by PEG, like improvements in growth or weight gain or effects on gastroesophageal reflux. Repeat investigations have been performed on individual patients to reassess and manage their condition. The risk of gastroesophageal reflux was not studied systemically.
Although we acknowledge these shortcomings, it does define our population of patients with feeding problems, the multidisciplinary approach to evaluating patients and the procedures used to offer nutritional support to vulnerable children.
PEG is a safe and effective means of achieving nutritional support and we believe in the CHD group of patients, its use encourages the early adoption of voluntary feeding skills, resulting in PEG removal. In the majority of instances, we believe this has happened following corrective cardiac surgery. Early experiences with PEG feeding have been published from our centre in 37 children with CHD where significant improvements in standard deviation scores for weight were recorded (11). Similar outcomes have been observed elsewhere (12).
In conclusion, the role of PEG is well established, enabling nutritional support in vulnerable patients. Our experience with insertion and removal of the PEG has been generally safe, with low procedure-related morbidity in children with a wide range of complex medical conditions.
1. Craig GM, Carr LJ, Cass H, Hastings, et al. Medical, surgical, and health outcomes of gastrostomy feeding. Dev Med Child Neurol
2. Corwin DS, Isaacs JS, Georgeson KE, et al
. Weight and length increases in children after gastrostomy placement. J Am Diet Assoc 1996; 96:874–879.
3. Sullivan PB, Juszczak E, Bachlet AM, et al
. Impact of gastrostomy tube feeding on the quality of life of carers of children with cerebral palsy. Dev Med Child Neurol 2004; 46:796–800.
4. Sloper P, Turner S. Risk and resistance factors in the adaptation of parents of children with severe physical disability. J Child Psychol Psychiatry 1993; 34:167–188.
5. Avitsland TL, Kristensen C, Emblem R, et al
. Percutaneous endoscopic gastrostomy in children: a safe technique with major symptom relief and high parental satisfaction. J Pediatr Gastroenterol Nutr 2006; 43:624–628.
6. McGrath SJ, Splaingard ML, Alba HM, et al
. Survival and functional outcome of children with severe cerebral palsy following gastrostomy. Arch Phys Med Rehabil 1992; 73:133–137.
7. Sullivan PB, Juszczak E, Bachlet AM, et al
. Gastrostomy tube feeding in children with cerebral palsy: a prospective, longitudinal study. Dev Med Child Neurol 2005; 47:77–85.
8. Richter T, Meier C, Steppberger K, et al
. Experiences with enteral nutrition of patients with cystic fibrosis (CF) via a percutaneous endoscopic gastrostomy (PEG). Klin Padiatr 2001; 213:325–328.
9. Williams SG, Ashworth F, McAlweenie A, et al
. Percutaneous endoscopic gastrostomy feeding in patients with cystic fibrosis. Gut 1999; 44:87–90.
10. Guimber D, Michaud L, Storme L, et al
. Gastrostomy in infants with neonatal pulmonary disease. J Pediatr Gastroenterol Nutr 2003; 36:459–463.
11. Ciotti G, Holzer R, Pozzi M, Dalzell M. Nutritional support via percutaneous endoscopic gastrostomy in children with cardiac disease experiencing difficulties with feeding. Cardiol Young 2002; 12:537–541.
12. Hofner G, Behrens R, Koch A, et al
. Enteral nutritional support by percutaneous endoscopic gastrostomy in children with congenital heart disease. Pediatr Cardiol 2000; 21:341–346.
13. Cosgrove M, Jenkins HR. Experience of percutaneous endoscopic gastrostomy in children with Crohn's disease. Arch Dis Child 1997; 76:141–143.
14. Skolin I, Hernell O, Larsson MV, et al
. Percutaneous endoscopic gastrostomy in children with malignant disease. J Pediatr Oncol Nurs 2002; 19:154–163.
15. Casswall TH, Backstrom B, Henstrom L, et al
. Percutaneous endoscopic gastrostomy (PEG) in children: long term outcome and nutritional aspects. J Pediatr Gastroenterol Nutr 2004; 39(Supplements 1):S492–S493.
16. Gauderer MW, Ponsky JL, Izant RJ Jr. Gastrostomy without laparotomy: a percutaneous endoscopic technique. J Pediatr Surg 1980; 15:872–875.
17. Van der Merwe WG, Brown RA, Ireland JD, et al
. Percutaneous endoscopic gastrostomy in children—a 5-year experience. S Afr Med J 2003; 93:781–785.
18. Yeh TC, Yeung CY, Sheu JC, et al
. Percutaneous endoscopic gastrostomy in children: 15 cases experience. Acta Paediatr Taiwan 2003; 44:135–139.
19. Eger R, Reif S, Yaron A, et al
. Percutaneous endoscopic gastrostomy (PEG) in children: indications, the procedure, outcomes, short and long-term complications. Harefuah 2008; 147:21–24.