Gastrostomy tube feeding: when to start, what to feed and how to stop


Many severe and/or chronic diseases of childhood can be accompanied by undernutrition, and when dietary measures are insufficient to fulfil the nutritional requirements of these children, enteral nutrition (EN) may be needed. In clinical practice, when the digestive tract is functioning, EN support is preferred over parenteral nutrition. EN is a safe and effective method of nutrition support for a wide range of paediatric diseases, and can alleviate undernutrition in several chronic conditions.

Indications for enteral tube feeding in children include incapacity or limited ability to eat, inability to meet nutritional requirements by oral intake alone, high risk of aspiration (as frequently observed in neuro-muscular disorders or psychomotor retardation), increased nutritional losses, altered metabolism, and in some cases as primary disease management (see Table 1).

Table 1 Indications for enteral tube feeding in paediatric patients

The goal of EN support in children should be to maintain or improve the child's nutritional status, while minimizing gastrointestinal symptoms and improving/maintaining the quality of life of both the child and the caregiver. Secondary advantages are easier administration of fluids and medications and so better compliance with treatments. EN can also preserve more time for education and rehabilitation; it is, however, important to encourage oral feeding for pleasure.

Routes of delivery

Enteral tube feeds can be administered into multiple sites in the gastrointestinal tract, depending on the functional status of the gastrointestinal tract and the risk of aspiration. Possible routes of nutrient administration include (1) gastric (nasogastric or gastrostomy) and (2) transpyloric (nasoduodenal, naso-jejunal, gastrojejunostomy or jejunostomy).

Transpyloric feeding is indicated in children with congenital gastrointestinal abnormalities, gastric dysmotility, severe vomiting resulting in growth faltering, gastric surgery, and those at increased risk of aspiration.

Gastric feeding is the most preferred route of tube feeding, for it most closely simulates physiological digestive and hormonal responses, retains the microbial properties of the stomach in the feeding process, and allows for easier tube insertion and tolerance of large osmotic loads (ESPGHAN Committee on Nutrition, in press). Naso-gastric tubes of silicone or polyurethane are efficient and well tolerated for short-term EN.

When long-term tube feeding is anticipated and/or enteral tube feeding exceeds 4–6 weeks, a gastrostomy tube should be considered (ESPGHAN Committee on Nutrition, in press). Gastrostomy feeding is generally well accepted with children as it is more comfortable and prevents the need for frequent tube changes. Percutaneous endoscopic gastrostomy is the technique of choice for children requiring longer-term EN. Percutaneous endoscopic gastrostomy is nowadays a routine procedure in children of all ages, including neonates weighing as little as 2.5 kg. After a period of a couple of months or more, when the gastrostomy site has healed, the gastrostomy tube can be replaced by a more convenient device known as a gastrostomy button. Alternatively, per-endoscopic one-step button procedures have become recently available.

Gastrostomy feeding has been found to improve weight and mid-arm circumference (Cook et al., 2005; Sullivan et al., 2005; Craig et al., 2006; Ramelli et al., 2007), reduce feeding time (Cook et al., 2005; Sullivan et al., 2005), improve the quality of life of caregivers and children (Sullivan et al., 2004; Cook et al., 2005; Kindermann et al., 2008; Mahant et al., 2009), and result in other health gains (Sullivan et al., 2005). The prospect of gastrostomy feeding, however, is seen as a major intervention by parents and therefore detailed preparation and discussion are needed. Factors that need to be taken into account include practicalities for the family as well as potential psychological issues for the child. Patient associations or parent-to-parent contact groups with children in the same situation can be helpful.

Home enteral tube feeding

In clinical practice, two different situations can be observed:

  1. 1

    Acute conditions wherein undernutrition can occur rapidly or can be predicted. These cases should be recognized and treated (with EN) early before the negative consequences of undernutrition take hold. Such situations can occur at any age and from many different conditions: the neonatal period with severely sick infants (premature with digestive disease, respiratory failure, and so on), traumatic or surgical stress, acute infection in myopathic children, intensive chemotherapy, bacterial super-infection in cystic fibrosis, and critical illness in children.

  2. 2

    Chronic conditions wherein nutritional support is required for months or even years. In these circumstances home enteral nutrition (HEN) is the best solution for the child and their family. Interest in HEN support has expanded rapidly worldwide. It is a safe and efficient method of feeding that is more compatible with family integration as well as social circumstances including schooling. However, it is dependent on children and parents being adequately trained and attending follow-up assessments with relevant health-care professionals.

In our experience in 416 children on HEN, the indications included digestive disorders (including cystic fibrosis) (35%), neurological and muscular disorders (35%), malignancy (11%), failure to thrive (8%) mainly for neonatal reason (bronchopulmonary dysplasia), and miscellaneous causes such as cardiopulmonary and renal diseases (9%; Daveluy et al., 2005). Furthermore, in our experience, the mean age of children commencing HEN was 5 years. At commencement of HEN, 27% of children were under 1 year of age and 25% were older than 11 years. The age varied from less than 2 years in patients with chronic cholestasis, neonatal digestive disease and bronchopulmonary dysplasia, to 9–13 years in patients with cystic fibrosis and inflammatory bowel disease (Daveluy et al., 2005).

Many health-care professionals should be involved in the screening, assessment and care of children who require HEN. These children usually require the involvement of paediatricians from almost every specialty, as well as nurses, the family and nutritionists/dietitians (Agostoni et al., 2005). The ESPGHAN Committee on Nutrition recently recommended implementation of nutrition support teams in paediatric units to improve screening for nutritional risk and to identify patients who require nutritional support (Agostoni et al., 2005).

To prevent acute hospital-acquired undernutrition and its complications, the risk of nutritional depletion needs to be identified at the time of admission, so that appropriate nutritional interventions can be initiated at an early stage. Ideally, the nutritional screening tool should be a simple scoring system, easy to use for non-specialized staff and adapted to the population in question. The first paediatric nutritional risk-screening tool was developed and validated a few years ago by colleagues from Paris (Sermet-Gaudelus et al., 2000). It is based on severity gradation of disease and presence of pain and/or reduction of food intake assessed within the first 2 days after admission to hospital. The nutritional risk score is classified as low, moderate or high and adapted nutritional intervention is proposed. However, the main obstacle for using such a tool is the difficulty to implement it as routine practice in every department of the hospital.

Providing education and information to the children and their families is a key consideration before starting HEN. Different support mechanisms can be used and have to be adapted to the age of the child as well as the socio-cultural status of the family.

Significant evolution of HEN in childhood has been observed during the last few years. The annual number of patients treated with HEN in our centre has increased dramatically from 16 in 1990 to 200 in 2000, with more than 65 new patients every year since 1999. The British Artificial Nutrition Survey 2007 report similar figures ( In the same period, the mean age at the commencement of HEN (based on BAPEN data) decreased from 6.2±1.4 to 4.8±0.7 years ( The use of naso-gastric tubes also decreased from 63% in 1990 to 35% in 1998, and the use of gastrostomy increased from 20% in 1990 to 50–60% since 1994 (Daveluy et al., 2006).

There are probably several reasons for such evolution. Paediatric indications for EN increased, along with the increasing recognition of the clinical efficacy of nutritional support in the treatment of severe and chronic diseases in childhood. Improved management techniques for most of these conditions have also led to the emergence of a large group of surviving children presenting with chronic malnutrition. The second factor is the increase in duration of EN (595±719 days in our experience), which is explained by the increased life expectancy of children with chronic diseases requiring HEN (for example, cystic fibrosis and neurological disabilities) and is partly related to improvements in their nutritional status. New modes of health-care delivery have also favoured the development of HEN, such as reimbursement of expenses by health-care providers, shorter hospitalization, the development of child home-care services and the development of nutrition support teams. Finally, changing practices reflect the development of improved pumps, improved enteral feeds and improved delivery techniques such as percutaneous endoscopic gastrostomy.

Organization of transfer to adult care, careful evaluation of education, assessing the consequences of EN based on the psycho-social condition, and the quality of life of children and their parents (carers) remain challenges for the future.

What to feed

Thanks to the recent development of industrialized (EN) formulations adapted for children, HEN has become easier and safer. It is strongly recommended to use all-in-one closed systems that avoid the risk of manipulation errors and bacterial contamination (Bott et al., 2001). Industrialized paediatric EN formulations actually meet the requirements in most paediatric situations. Enteral feeds for children over 1 year of age may consist of paediatric polymeric, semi-elemental or elemental feeds, and may vary in energy density, fibre content, and macro- and micro-nutrient distribution. In 2000, only 29% of HEN patients required home-made formulations (Daveluy et al., 2006).

The choice of feed depends on many factors, including the nutritional status and energy/protein requirements of the child. Factors to consider include inability to tolerate large volumes of fluid (energy-dense feeds 1.5 kcal/ml), excess weight gain (low-energy feeds 0.75 to 1 kcal/ml), increased energy requirements (energy-dense feeds 1.5 kcal/ml), constipation (high-fibre feeds) and gastro-oesophageal reflux (whey dominant feeds).

How to stop

Developmental aspects of feeding

Weaning a child from tube feeding can be difficult. A clear understanding of the developmental progression of feeding ability is a prerequisite both for avoiding potential difficulties and for successful management of tube withdrawal.

As early as the 7th week of gestation, the developing brain is receiving sensory signals from the oro-pharyngeal cavity and by 12 weeks oral movements are evident. At the end of the 3rd month of gestation the fetus is sucking its fingers and swallowing amniotic fluid.

At birth the neuromotor apparatus required for reflex and automatic feeding activity is in place and oral feeding for the newborn infant is entirely reflexive. Rooting, nipple latching, sucking and swallowing do not appear to require suprabulbar activity. The sucking reflex, triggered by any sensory stimulation of the lips and the tongue, allows feeding during the first half of infancy.

There are also protective reflexes such as the gag reflex, which protects the upper respiratory tract from foreign bodies. This powerful reflex is triggered when a foreign body penetrates the mouth; there is a strong contraction of the soft palate and the pharyngeal constrictors, the mouth opens and the tongue protrudes to eject the foreign body before it gets into the pharynx. As sucking and neurological maturation progresses, this reflex is gradually inhibited and, by the age of 6 or 7 months, will only be triggered to induce vomiting.

Children with motor disabilities or medical problems that interfere with normal oral exploration may develop hypersensitivity to oro-pharyngeal sensory stimulation. They may gag when a spoon is placed on the tongue or pull away as the spoon approaches because of previous unpleasant experiences, and refuse to try any new taste, texture or method of feeding.

Although feeding begins as a reflex, it gradually becomes a voluntary act, with only the pharyngeal and oesophageal parts of the swallow remaining under reflex control. Growth and neurological maturation have a role in feeding development, but experiential learning is crucial. This learning process begins in the neonatal period and is dependent not only on experiential opportunities but also on sensory inputs and suprabulbar neurological maturation. When the supranuclear structures have completely taken over the control of the brain stem reflex activities, the infant's reflexive sucking–swallowing decreases as mature voluntary oral feeding is established around the age of 7–8 months. Chewing is not established until the age of 18–24 months and becomes mature between 3 and 6 years of age.

Oral feeding that requires suckling, swallowing and breathing coordination is, therefore, the most complex sensorimotor process the newborn infant undertakes. Furthermore, critical and sensitive periods are believed to exist in the development of normal feeding behaviour (Illingworth and Lister, 1964). Thus, when infants with major physical and physiological problems are prevented from initiating oral feeding in the way their more typically developing peers do, many demonstrate prolonged delays and significant difficulty with oral feeding.

This may be a particular problem in preterm infants. Infants born at less than 30 weeks gestation stay in hospital on average for about 3 months. During this period, oral-motor stimulation may consist primarily of necessary medical procedures, for example, insertion of an endotracheal tube, nasal continuous positive airway pressure, naso- or oro-gastric tubes, and suctioning of the airway. Often little priority is given to preparing the infants for oral feeding and they are potentially at greater risk of having feeding difficulties (Fucile et al., 2002).

After a period of tube feeding, the majority of preterm, neurologically impaired children or those in the postoperative period take up oral feeding without any problem as long as tube feeding does not last longer than 2–3 weeks. This is especially true when they are stimulated with a non-nutritive sucking treatment during tube feeding. For instance, Fucile et al. (2002) have shown that an early oral stimulation programme accelerates the transition to full oral feedings in preterm infants and that this was associated with greater overall intake and feeding efficiency in the intervention group when compared with the control. In neonates and infants who have never been orally fed from birth and in those for whom tube feeding lasts for a longer period (from 6 weeks to over 1 year), however, without any progressive oral stimulation programme, the reintroduction of oral feeding can be an ordeal for the child and the caregiver alike. The absence or involution of the afferent sensory input (tactile, olfactory and taste) from the oro-pharynx, which normally occurs during oral feeding, means that, when attempts are made to start oral feeding, the presence of food in the mouth is likely to be misinterpreted as a noxious stimulus and to stimulate the gag reflex and lead to oro-aversion. Thus, early oral stimulation is a crucial component of feeding management in the exclusively tube-fed neonate and intraoral massage, for instance, has been shown to be useful in inhibiting the gag reflex and helping it to mature into an adult pattern (Senez et al., 1996).

Weaning a child from gastrostomy tube feeding

Promotion of a positive caregiver–child relationship is fundamental if weaning from tube feeding is to be achieved. Problems with parent–child interaction occasionally manifest themselves as feeding problems even in children who are not tube fed. For those children who are tube fed, it is essential that when weaning is attempted the caregiver approach should be responsive, sensitive and skilful (Satter, 1990). A critical component of the process of determining a child's readiness for oral feeding, therefore, is to establish the extent to which their caregiver is ready. This in turn depends on the caregiver not only having a clear understanding of behavioural feeding techniques but also of being able to adopt considerable consistency, patience and perseverance in their application (Schauster and Dwyer, 1996). Furthermore, it is important to appreciate that the time required to make this transition may take weeks or months.

In addition, it is important to establish that the child is also ready for the transition. It is possible that there could be difficulties during the transitional period that could compromise nutritional intake. Thus, it would be inappropriate to begin the process of transition until the child's nutritional status is secure. This is less likely to happen if there is a significant degree of oral-motor dysfunction (as, for example, in severe cerebral palsy); hence, an assessment of oral-motor function is necessary. Similarly, an assessment of the safety of the swallow as judged by contrast video fluoroscopy is required. When airway protection is compromised by neurological impairment and aspiration occurs often (especially when this has been associated with repeated chest infections), it would be unwise to attempt to reinstate oral feeding as the exclusive source of nutrient intake. Contrast video fluoroscopy may show that a thicker consistency of food can be safely swallowed and then a clinical judgement will have to made as to whether or not adequate nutritional intake can be provided safely by mouth.

When a decision is made to begin the process of normalizing feeding, this will be more successful if approached as a multi-disciplinary process. The speech and language therapist will provide oral stimulation and deal with eating-related maladaptive behaviour, such as grimacing, mouth closure and gagging. The dietitian will advise on food texture and consistency and assist in the development of a behavioural feeding plan. This plan may involve feeding regulation by allowing hunger–satiety cues to develop so that tube feeds can gradually be decreased. In certain cases, especially when there has been prolonged dependency on naso-gastric tube feeding, the process of hunger provocation has been shown to be an effective tool for use in weaning children from tube feeds (Kindermann et al., 2008). Education of the main caregiver (usually the mother) is essential and they must be shown how to promote an optimal feeding environment. Finally, before removal of a gastrostomy feeding tube, it may be necessary to test the ability of the child to maintain their nutritional status with oral feeding alone for a few weeks.

Managing feeding problems and especially weaning a child off tube feeding are major challenges for the parents and for all health workers involved, including dietitians, speech therapists, psychologists, nurses and paediatricians.


  1. Agostoni C, Axelson I, Colomb V, Goulet O, Koletzko B, Michaelsen KF, et al., ESPGHAN Committee on Nutrition; European Society for Paediatric Gastroenterology (2005). The need for nutrition support teams in pediatric units: a commentary by the ESPGHAN committee on nutrition. J Pediatr Gastroenterol Nutr 41, 8–11.

  2. Bott L, Husson MO, Guimber D, Michaud L, Arnaud-Battandier F, Turck D et al. (2001). Contamination of gastrostomy feeding systems in children in a home-based enteral nutrition program. J Pediatr Gastroenterol Nutr 33, 266–270.

  3. Cook S, Hooper V, Nasser R, Larsen D (2005). Effect of gastrostomy on growth in children with neurodevelopmental disabilities. Can J Diet Pract Res 66, 19–24.

  4. Craig GM, Carr LJ, Cass H, Hastings RP, Lawson M, Reilly S et al. (2006). Medical, surgical, and health outcomes of gastrostomy feeding. Dev Med Child Neurol 48, 353–360.

  5. Daveluy W, Guimber D, Mention K, Lescut D, Michaud L, Turck D et al. (2005). Home enteral nutrition in children: an 11-year experience with 416 patients. Clin Nutr 24, 48–54.

  6. Daveluy W, Guimber D, Uhlen S, Lescut D, Michaud L, Turck D et al. (2006). Dramatic changes in home-based enteral nutrition practices in children during an 11-year period. J Pediatr Gastroenterol Nutr 60, 240–244.

  7. ESPGHAN Committee on Nutrition. Braegger C, Decsi T, Dias JA et al. Practical approach to paediatric enteral nutrition. A commentary by the ESPGHAN Committee on Nutrition. J Pediatr Gastroenterol Nutr (in press).

  8. Fucile S, Gisel E, Lau C (2002). Oral stimulation accelerates the transition from tube to oral feeding in preterm infants. J Pediatr 141, 230–236.

  9. Illingworth RS, Lister J (1964). The critical or sensitive period, with special reference to certain feeding problems in infants and children. J Pediatr 65, 839–848.

  10. Kindermann A, Kneepkens CM, Stok A, van Dijk EM, Engels M, Douwes AC (2008). Discontinuation of tube feeding in young children by hunger provocation. J Pediatr Gastroenterol Nutr 47, 87–91.

  11. Mahant S, Friedman JN, Connolly B, Goia C, Macarthur C (2009). Tube feeding and quality of life in children with severe neurological impairment. Arch Dis Child 94, 668–673.

  12. Ramelli GP, Aloysius A, King C, Davis T, Muntoni F (2007). Gastrostomy placement in paediatric patients with neuromuscular disorders: indications and outcome. Dev Med Child Neurol 49, 367–371.

  13. Satter E (1990). The feeding relationship: problems and interventions. J Pediatr 117 (Part 2), S181–S189.

  14. Schauster H, Dwyer J (1996). Transition from tube feedings to feedings by mouth in children: preventing eating dysfunction. J Am Diet Assoc 96, 277–281.

  15. Senez C, Guys JM, Mancini J, Paz PA, Lena G, Choux M (1996). Weaning children from tube to oral feeding. Childs Nerv Syst 12, 590–594.

  16. Sermet-Gaudelus I, Poisson-Solomon A, Colomb V, Brusset MC, Mosser F, Berrier F et al. (2000). Simple pediatric nutritional risk score to identify children at risk of malnutrition. Am J Clin Nutr 72, 64–70.

  17. Sullivan PB, Juszczak E, Bachlet AM, Lambert B, Vernon-Roberts A, Grant HW et al. (2005). Gastrostomy tube feeding in children with cerebral palsy: a prospective, longitudinal study. Dev Med Child Neurol 47, 77–85.

  18. Sullivan PB, Juszczak E, Bachlet AM, Thomas AG, Lambert B, Vernon-Roberts A et al. (2004). Impact of gastrostomy tube feeding on the quality of life of carers of children with cerebral palsy. Dev Med Child Neurol 46, 796–800.

Download references

Author information

Correspondence to F Gottrand.

Ethics declarations

Competing interests

PB Sullivan has received consulting fees from Danone Limited, lecture fees from Nutricia Ltd and grant support from Numico Research Foundation. F Gottrand has received consulting fees from Nutricia, Danone, Nestle, Movetis, Naturalpha and Astra-Zeneca, and lecture fees from Nutricia, Danone, Lactalis and Astra-Zeneca. F Gottrand has also received grant support from VLM, AFM, Danone and Nutricia.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Gottrand, F., Sullivan, P. Gastrostomy tube feeding: when to start, what to feed and how to stop. Eur J Clin Nutr 64, S17–S21 (2010).

Download citation

Further reading