NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

National Collaborating Centre for Acute Care (UK). Nutrition Support for Adults: Oral Nutrition Support, Enteral Tube Feeding and Parenteral Nutrition. London: National Collaborating Centre for Acute Care (UK); 2006 Feb. (NICE Clinical Guidelines, No. 32.)

Cover of Nutrition Support for Adults

Nutrition Support for Adults: Oral Nutrition Support, Enteral Tube Feeding and Parenteral Nutrition.

Show details

9Enteral tube feeding in hospital and the community

9.1. Introduction

For the purposes of these Guidelines, enteral tube feeding (ETF) refers to the delivery of a nutritionally complete feed (containing protein or amino acids, carbohydrate +/− fibre, fat, water, minerals and vitamins) directly into the gut via a tube. The tube is usually placed into the stomach, duodenum or jejunum via either the nose, mouth or the direct percutaneous route37. ETF is not exclusive and can be used in combination with oral and/or parenteral nutrition. Patients receiving ETF should be reviewed regularly to enable re-instigation of oral nutrition when appropriate. Most enteral feeding tubes are introduced at the bedside but some are placed surgically, at endoscopy or using radiological techniques, and some are inserted in the community. Whenever possible the patient should be aware of why this form of nutrition support is necessary, how it will be given, for how long, and the potential risks involved. There may be considerable ethical difficulties in deciding if it is in a patient’s best interests to start a tube feed.

Innumerable questions regarding best ETF practice could be asked but for these guidelines, reviews were restricted to studies providing potential guidance on the indications for ETF, studies on the benefits of ETF compared to oral or parenteral nutrition, and studies on some technical aspects of delivering enteral feeds. No studies on different types of enteral feed were reviewed.

9.2. General Indications for Enteral Tube Feeding in hospital and the community

9.2.1. Introduction

Enteral tube feeding (ETF) is used to feed patients who cannot attain an adequate oral intake from food and/or oral nutritional supplements, or who cannot eat/drink safely. The aim is to improve nutritional intake and so improve or maintain nutritional status. It is used most commonly in patients with dysphagia either because they cannot meet their nutritional needs despite supplements and/or modifications to food texture/consistency, or because they risk aspiration if they try to do so. The GI tract must be accessible and functioning sufficiently to absorb the feed administered. Common indications for ETF are listed in Table 18, although this is not necessarily an exhaustive list. If ETF is unsafe or unlikely to be successful (e.g. inaccessible GI tract, severe malabsorption, excessive gastrointestinal losses), parenteral nutrition is likely to become the therapy of choice.

Table 18. Indications for enteral tube feeding.

Table 18

Indications for enteral tube feeding.

9.2.2. Relevant Studies

Most studies on indications for ETF (rather than timing, type of tube, type/amounts of nutrients etc) exclude all patients with the most common clinical indication for ETF (i.e. those with a functional GI tract but unsafe swallow, who would starve or require PN if ETF were not used). The findings from these studies do not therefore provide help with decision making for routine clinical practice. The recommendations were therefore derived using expert opinion.

9.3. Recommendations for clinical practice

9.3.1. Indications for enteral tube feeding

Healthcare professionals should consider enteral tube feeding in people who are malnourished38 or at risk of malnutrition39 and have:

  • inadequate or unsafe oral intake, and
  • a functional, accessible gastrointestinal tract. [D(GPP)]

Enteral tube feeding should be stopped when the patient is established on adequate oral intake. [D(GPP)]

9.4. Enteral tube feeding versus standard care

9.4.1. Introduction

Some patients are put at potential risk of malnutrition (or worsening of pre-existing malnutrition) through a limitation of oral intake or absorptive capabilities from effects of their disease or direct and indirect consequences of surgery (e.g. nausea or ileus and/or clinical practice of restricting post surgical oral intake). If this limitation is severe and long-lasting, nutrition support using ETF or PN will be needed but ETF could also be beneficial for patients who are likely to have limited intake for only a few days (as in most post-operative patients), especially if they already malnourished. However, the benefits from using ETF in this elective, supplementary role is uncertain and it is possible that the risks might outweigh any clinical benefits. Two reviews were therefore conducted to identify:

  • RCTs comparing patients who received ETF (with or without oral intake) vs. patients receiving standard care (e.g. normal hospital diet and/or oral nutrition supplements) and
  • RCTs that included patients receiving elective early post-operative ETF vs. no early postoperative nutrition (i.e. nil by mouth post-surgical dietary care with simple IV fluids until clinical signs of returning GI function).

9.4.2. Studies of ETF vs. standard care

The review conducted identified 10 RCTs20,52,143,180,228,315,318,339,340,361 (Table 41). Four of these compared the effect of patients receiving 12 to 24 hours of nasogastric tube feeding plus continued normal hospital diet with patients receiving a standard hospital diet only 143,228,339,340. Two studies compared nasogastric/nasoduodenal feeding with standard hospital diet 52,180. One study compared nasogastric feeding with standard hospital diet plus ad lib snacks 20, while another had two intervention arms in which patients received a nasogastric feed with amino acids alone or a nasogastric feed containing amino acids plus carbohydrates 218. The control group continued on a normal hospital diet. A further study compared oesophagostomy tube feeding with a clear liquid diet, advancing to a normal diet as tolerated 315 and one investigated the benefits of pre-operative ETF (nasogastric tube feeding) compared with routine hospital diet 318. The final study examined the effect of perioperative nutrition in malnourished head and neck cancer patients 361 using three intervention arms: one group received no preoperative and standard postoperative ETF; another group received standard preoperative and postoperative ETF; and the third group received arginine supplemented preoperative and postoperative tube feeding.

The patients included in the studies were orthopaedic hip fracture patients (four studies covering 337 patients)20,143,339,340, people who were generally malnourished (one study covering 86 patients)228, malnourished surgical patients (one study covering 110 patients) 318, total laryngectomy patients (one study covering 67 patients)315, malnourished patients undergoing surgery because of a head and neck malignancy (one study covering 49 patients) 361 and patients with alcoholic liver disease (two studies covering 66 patients)52,180.

9.4.3. Clinical evidence ETF vs. standard care

The main outcomes reported were nutritional intake achieved, changes in nutritional status, mortality, length of stay and complications associated with tube feeding (e.g. tolerance of the feeding tube).

The difference in nutritional intake (usually reported as energy and/or protein intake) between the enterally tube fed patients and those receiving standard care was reported in six studies52,143,180,228,339,340. In all six studies, the enterally fed group achieved a significantly greater nutritional intake (range p<0.0001 to 0.012).

Five studies reported changes in measures of nutritional status 20,180,228,318,361 with three showing improvement20,228,318 (range p=0.001 to p=0.05) while two showed no differences 180,361.

Mortality was reported in 8 studies20,52,143,180,318, 339,340,361. Four showed no differences between groups 20,339,340,361 but one 52 did show significantly lower mortality in the ETF group (p=0.02) and two further studies reported lower mortalities but with no significance values given 180,318. One study, 143 noted a higher mortality rate for the patients who were tube fed but again no p-value was reported.

There were no significant differences in postoperative complications reported in four studies315,339,340,361; nor in the incidence of pressure sores in one study143; diarrhoea in one study 180, or infection rates in one study 52. In one study 318 the incidence of wound infection, nausea and vomiting were lower in the ETF group although no p-value was reported.

Five studies reported that ETF had no influence on length of hospital stay 180,315,339,340,361; although in one study 20, median time to independent mobility was lower in the ETF group (p 0.02 –0.04).

Three studies20,228,318 provided information on patient’s tolerance of ETF but no p-values were reported. In two studies 22%20 and 30%228 of study participants experienced problems tolerating the nasogastric tube. In the third study 318 7 out of 67 patients receiving ETF (10.5%) needed it to be discontinued due to uncontrollable diarrhoea, vomiting or severe aversion to the smell and taste of the feed.

9.4.4. Cost-effectiveness evidence ETF vs. standard care

Four studies were found that reported a cost comparison104,210,234,256: two RCTs, one retrospective cohort study and a study that constructed a simple model on the basis of two small trials (Table 56).

One RCT210 evaluated insertion of double-lumen gastrojejunostomy tube compared with routine care by the surgeon after pancreatico-duodenectomy. Half the patients in the routine care arm received PN; and the other group probably received NG feeding (but the route of feeding was unclear). The study found significant reductions in gastro-paresis and in costs. The second RCT104 compared early nasogastric enteral feeding with early oral feeding after colorectal resection in cancer patients. They found that early oral intervention was safe but there were no cost savings or improvements in clinical outcomes.

The aim of the retrospective study 234 was to test whether there were cost savings in using tube-feeding rather than a carer manually feeding the patient (which requires expensive staff time and risks causing aspiration) for patients with advanced dementia. The results showed that the total costs were higher for the patients with feeding tubes compared with those without tubes (£5,600 vs. £3,100, p=0.04). The difference was due to tube feeding placement cost and hospital costs arising from complications directly related to tube feeding. However, the sample size of this study was small (11 patients in each group) and potentially biased since it was a convenience sample. Costing was also made using Medicaid and Medicare reimbursement rates, which may not be applicable to the UK NHS setting.

The fourth cost-effectiveness study evaluated the cost of preoperative enteral nutrition 256. ETF (10–21 days) was compared with no ETF. The study was a sensitivity analysis based on the two small trials with the largest reduction in complication rate. Incremental cost per complication averted was between £9,000 and £94,500 with hospital preoperative ETF, depending on the assumptions made. However, they found that home preoperative ETF is more likely to be cost saving.

There were no economic studies evaluating pre and post-operative ETF.

9.4.5. Studies of early post-operative ETF vs. no early post-operative nutrition

We identified one systematic review206 that looked at early post-operative feeding (oral or enteral) versus post-operative ‘nil by mouth’. There were 11 studies included in this review: 6 on early post-operative enteral feeding versus no early post-operative nutrition 27,58,148,301,310,367 and 5 on early post-operative oral feeding versus postoperative ‘nil by mouth’ 32,144,258,291,334 (included in the oral chapter 8). In this section we have included the six studies from the systematic review that looked at the effect of early postoperative ETF. In addition to the studies from this systematic review, we identified 17 further studies that looked at the effect of early post-operative ETF versus no early post-operative nutrition. The RCTs were analysed according to the type of surgical patients included in the studies.

Five studies included patients undergoing upper GI surgery 45,148,263,341,367 (Table 51). Three studies included patients undergoing lower GI surgery 215,301,310 (Table 51). Six studies included both upper and lower surgery 27,58,160,298,321,328 (Table 53). Three studies included patients undergoing hepatobiliary surgery 117,145,164 (Table 54). Six studies included acute trauma patients 65,98,175,216,238,281 (Table 45).

We extracted data on seven outcomes: vomiting, anastomotic dehiscence, pneumonia, death, intra-abdominal abscess, wound infection and hospital length of stay (LOS) where available. Where appropriate we pooled the data for these outcomes. We were unable to pool the data for LOS as the studies reported the data in different units and information needed to convert these units was not available.

9.4.6. Clinical evidence: early post-operative ETF vs. no early post-operative nutrition

Analyses for each of the surgical subgroups showed no statistically significant differences in any of the outcomes extracted. The P value from tests for heterogeneity was greater than 0.1 for all outcomes in all the groups.

We also conducted a combined analysis which included all the surgical studies (Appendix Eight: Meta-Analyses Enteral versus Nil Post Operative Nutrition Support). This also identified no statistically significant differences in any of the outcomes extracted which included vomiting, anastomotic dehiscence, pneumonia, intra-abdominal abscess, wound infection and mortality. The data on lengths of hospital stay reported in fourteen studies 58,98,117,145,148,216,238,263,281,301,321,328,341,367 were not adequate to permit a combined analysis but statistically significant differences were only detected in two studies with one showing that early feeding led to fewer days in hospital (p< 0.05)301 whilst the other showed it extended length of stay (p< 0.01)328.

Table 19Outcomes reported in studies of early enteral tube feeding

No. patients (early feeding/late feeding)RR (fixed) 95% CI
Vomiting (reported in four studies27,148,175,216298/2801.27 [0.92, 1.75]
P value from test for heterogeneityP= 0.21
Anastomotic dehiscence (reported in 10 studies 27,148,215,263,301,310,321,328,341,367257/2640.60 [0.33, 1.10]
P value from test for heterogeneityP= 0.79
Pneumonia (reported in 9 studies 27,98,148,216,238,263,310,321,328355/3610.76 [0.53, 1.08]
P value from test for heterogeneityP= 0.36
Death (reported in 10 studies 27,98,148,216,263,281,301,310,321,328368/3750.72 [0.45, 1.15]
P value from test for heterogeneityP=0.37
Intra-abdominal abscess (reported in eight studies 27,98,148,238,301,310,321,328250/2560.60 [0.32, 1.14]
P value from test for heterogeneityP=0.69
Wound infection (reported in 12 studies 27,98,117,148,163,216,263,281,301,321,328,341402/4080.92 [0.68, 1.23]
P value from test for heterogeneityP= 0.26

9.4.7. Cost effectiveness evidence: early post-operative ETF vs. no early post-operative nutrition

We identified three cost-effectiveness analyses for ETF compared to nil nutrition postoperatively27,145,147, although all three were small and potentially biased due to methodological weaknesses. Results were inconsistent although all reported a lower number of infections in the ETF groups compared to the nil groups. Estimated effects on cost were as follows:

A non-randomised prospective US study of patients undergoing bowel resection147 showed a cost saving (the magnitude and statistical significance is unclear due to poor reporting) with jejunal feeding tube placed during surgery and feeding initiated within 12 hours of surgery compared with usual care (which was not detailed). The cost savings were due to a reduction in infections.

A small Danish RCT27 reported a non-significant difference in (median) cost of about £1,500 for a 4 day nasoduodenal intervention compared with placebo after major abdominal surgery. Mean costs, which are more relevant than median costs, were not reported.

A small US RCT comparing nasojejunal tube feeding from 12 hours after surgery with maintenance iv fluid after liver transplantation 145 found a non-significant incremental cost of £1,200, despite a 50% reduction in infections. Control patients that were moved to tube feeding were excluded.

9.4.8. Conclusions

ETF in patients where there is some doubt about the adequacy of oral intake is effective in increasing nutritional intake over and above the intake observed with standard care and/or oral supplements and this usually leads to an improvement in nutritional status. However, this does not seem to produce consistent benefit in terms of length of stay or mortality rates and tube tolerance is sometimes a problem in these patients. The evidence of benefit related to complications, quality of life, costs and cost-effectiveness is very limited and ETF use in older people with dementia could be more expensive than oral feeding. The cost-effectiveness of preoperative enteral nutrition is unclear but might be improved if administered in the patients’ home. However, oral nutrition support is likely to be more cost-effective, when this can be tolerated by the patient.

The studies on early post-operative ETF compared to standard practice of nil by mouth until return of GI function, do not support the use of early ETF although most did not focus on very malnourished patients who might benefit from this approach. There may be cost benefits associated with the use of post-operative jejunostomy feeding in some circumstances but more research is needed.

The studies that examined elective ETF in malnourished patients prior to surgery suggest that they benefit in terms of nutritional status. However, much larger trials are needed to determine whether there are any benefits in lengths of hospital stay or mortality.

9.4.9. Rationale for recommendation(s)

Although ETF does increase nutritional intakes in patients the evidence that this benefits outcomes such as length of hospital stay or mortality is not clear.

9.5. Recommendations for clinical practice

9.5.1. Indications for enteral tube feeding

Enteral tube feeding should not be given to people unless they are malnourished40 or at risk of malnutrition41 and have; inadequate or unsafe oral intake and a functional, accessible gastrointestinal tract, or they are taking part in a clinical trial. [A]

Enteral nutrition support for surgical patients: Surgical patients who are: malnourished22 and have; inadequate or unsafe oral intake and a functional, accessible gastrointestinal tract and are due to undergo major abdominal procedures, should be considered for pre-operative enteral tube feeding. [B]

General surgical patients should not have enteral tube feeding within 48 hours post-surgery unless they are malnourished22 or at risk of malnutrition23 and have; inadequate or unsafe oral intake and a functional, accessible gastrointestinal tract. [A]

9.6. Enteral tube feeding routes of access

9.6.1. Introduction

Many types of enteral feeding tubes can be used to deliver nutrition into the stomach or upper small intestine. Choices depend on the proposed/expected period of feeding, clinical condition, and anatomy. Nasogastric (NG) tubes are used most frequently but others include nasoduodenal or nasojejunal tubes and gastrostomies or jejunostomies placed by endoscopic, radiological or surgical means. Nasogastric tubes

NG tubes are used mainly for short-term support in patients who do not have problems such as vomiting, gastro-oesophageal reflux, poor gastric emptying, ileus or intestinal obstruction, although they can also be used for longer term support where other enteral access is not possible or carries a risk. NG tubes are potentially dangerous in patients with an unsafe swallow and those who need to be nursed prone or flat and a risk/benefit assessment should be carried out before placement. Fine bore (5 – 8 FrG) NG tubes should be used for ETF unless there is a need for repeated large volume gastric aspiration i.e. gastric decompression. NG tubes should be placed by appropriately trained staff.

There is a small risk that NG tubes can be misplaced on insertion or move out of position at a later stage. Position of NG tubes should be verified on initial placement and before each use. Guidance from the National Patient Safety Agency 246 advocates aspiration of gastric contents and the use of pH graded indicator paper. It is recommended that a pH <5.5 is consistent with gastric placement. If aspirate cannot be obtained or the pH is >5.5 feeding should not commence. The NG tube should be left in place, the patient’s position changed and the aspirate re-tested in one hour. The feed itself can increase the pH in the stomach, so aspiration should take place at least 1 hour after the feed has been stopped. Radiography (a chest x-ray) is not recommended routinely, but it is suggested that local policies be developed for high risk groups (e.g. intensive care or neonatal units) or for where an aspirate is not obtained. Radiography in these circumstances would depend on the clinical situation and failure of aspiration checks. N.B. Gastric antisecretory drugs can cause the gastric acid pH to be altered. Clinical judgement needs to be exercised in this situation together with local guidance. Nasoduodenal and nasojejunal tubes

Nasoduodenal (ND) and nasojejunal (NJ) tubes are those placed into the gastrointestinal tract with the distal tip lying beyond the stomach in the duodenum or jejunum respectively. These tubes can be placed at the bedside or with endoscopic/radiological assistance but the position needs to be confirmed by abdominal X-ray after placement (unless placed under fluoroscopic guidance). Gastrostomy and jejunostomy

Gastrostomy tubes pass through the abdominal wall directly into the stomach. They are usually used for patients who require medium to long-term feeding or where NG access is difficult. Gastrostomy tubes are usually placed endoscopically (Percutaneous Endoscopic Gastrostomy - PEG) but they can also be placed radiologically or surgically

Jejunostomy tubes pass through the abdominal wall into the jejunum and are usually placed surgically. However, many percutaneous jejunostomy tubes are placed endoscopically or radiologically via gastric puncture with an extension through the pylorus into the duodenum or jejunum (Percutaneous Endoscopic GastroJejunostomy PEGJ)

Gastrostomy feeding does not negate the risks associated with reflux and aspiration, although risks may be lower than with NG feeding. In patients at high risk of aspiration, jejunostomy tubes or PEGJ tubes should be considered since they probably do reduce aspiration risks.

9.6.2. Nasogastric (NG) versus nasoduodenal (ND) or nasojejunal (NJ) tubes Introduction

Patients receiving ETF via the naso/orogastric route can have problems tolerating their enteral feeding regimen due to gastro-oesophageal reflux or delayed gastric emptying. As a result, patients may experience reflux or vomiting which may cause aspiration pneumonia and also result in a reduced nutrient intake. When these problems occur despite drug intervention, nasoduodenal or nasojejunal feeding should be considered. Studies on Nasogastric (NG) versus nasoduodenal (ND) or nasojejunal (NJ) tubes

We identified 14 RCTs (707 patients) that compared nasogastric feeding with nasoduodenal or nasojejunal feeding (Table 42) 34,76,77,96,134, 149,153,179,191,200,235,236,247,338. Twelve studies included intensive care patients34,76,77,96,134,149,153,179,191,235,236,247, one study malnourished neurological patients338 and one study was in healthy people 200. In five of these studies the intervention and comparison arms used the naso/orogastric route but did not specify the number of patients for each.

The main outcomes reported included aspiration96,153,179,247, pneumonia76,77,179,191,235, 236,338, vomiting77,235,236,247, diarrhoea76,77,179, 235,236 and percentage of target energy received34,77,96,134,235. Other outcomes reported included: length of stay in ICU and in hospital, mortality and change in nutritional status. Clinical evidence

No significant differences were found for mortality, length of stay in intensive care or hospital, incidence of pneumonia, vomiting or diarrhoea. Two studies reported the mean weight change, one showed no significant difference 179 while the other reported a significant weight gain for the nasogastric group 251. However, the weight change for the latter study was only recorded for 21 of the 38 patients entered into the study. Four out of the five studies reported no significant difference in the percent of prescribed calorie intake34,77,96,134 but one showed the nasojejunal patients achieving a significantly higher percent of their daily goal caloric intake than the nasogastric patients235. Cost-effectiveness

No study reporting cost or cost-effectiveness was found. Conclusions

Feeding patients with a nasogastric tube is usually as effective as a post-pyloric tube (nasoduodenal/nasojejunal) for delivering nutrients to patients (especially to patients on intensive care). The expected problems of gastric feeding in patients with gastro oesophageal reflux and delayed gastric emptying are not apparent in these studies.

It must be noted, however, that for ethical reasons randomised studies have not been performed in the patient groups usually considered for post pyloric feeding, although some information about the effectiveness and safety of post pyloric feeding in these patients may be gained from trials that compare post-pyloric feeding to parenteral nutrition. Rationale for recommendation(s)

The gastric route is usually technically simpler and in most circumstances achieves similar nutrient delivery with similar risks. Clinical studies have failed to show any clear advantage in feeding post-pylorically.

9.7. Recommendations for clinical practice Route of access

People in general medical, surgical and intensive care wards who are malnourished42 or at risk of malnutrition43 and have; inadequate or unsafe oral intake and a functional, accessible gastrointestinal tract should be fed via a tube into the stomach unless there is upper gastrointestinal dysfunction. [A]

People who are malnourished42 or at risk of malnutrition43 and have; inadequate or unsafe oral intake and a functional, accessible gastrointestinal tract with upper gastrointestinal dysfunction (or an inaccessible upper gastrointestinal tract) should be considered for post-pyloric (duodenal or jejunal) feeding. [D(GPP)]

9.8. Percutaneous Endoscopic Gastrostomy (PEG) versus Nasogastric (NG) Feeding

9.8.1. Introduction

For some patients with acute or chronic conditions requiring enteral feeding there is the option of feeding through a nasogastric tube or a gastrostomy (usually a PEG). Nasogastric tube feeding is usually successful but problems include dislodgement of the tube with the need for replacement which can be invasive and uncomfortable. For some patients the location and securing by tape of the nasogastric tube can also be irritating and may raise ethical issues surrounding patient restraint. For some patients the tube itself may also cause discomfort in the back of the throat and occasionally swallowing problems

In contrast, a gastrostomy tube cannot be dislodged as easily and is more comfortable. However, there are potential difficulties and risks in placement; feed aspiration can still occur and there can be greater difficulties surrounding any decision to withdraw gastrostomy feeding compared to NG/NJ feeding (although from the ethical stand-point there is no distinction to be made between short and long-term tubes, nor between withdrawing feeding compared to not instigating it in the first place (section 5.3). Since gastrostomy feeding is increasingly considered for patients likely to require long-term ETF we undertook a review of studies comparing the two access techniques. Studies considered for this review

Our review compared percutaneous endoscopic gastrostomy with nasogastric feeding (Table 43). Three small published RCTs 15,251,267 and a large multi-centre randomised controlled trial 343 met the inclusion criteria. One study looked at neurological, surgical and ear, nose and throat (ENT) patients 15, while the multi-centre study and the other two studies focused on stroke patients with accompanying dysphagia 251,267,343.

The main outcomes reported in the studies were absolute risk of death and risk of death or poor outcome (using the Modified Rankin Scale - MRS), treatment failure, amount of feed received, weight change, mortality, GI - haemorrhage and pressure sores. Other outcomes reported were: the time needed for tube insertion, length of hospital stay, convenience of care, quality of life, fixation of tube to patient and the incidence of aspiration or pneumonia. Clinical evidence

There were some methodological problems with two of the smaller studies. One15 had more sick patients in the PEG group than did the NG group suggesting a possible allocation bias between groups, while in another 267 most of the patients in the NG arm crossed over to the PEG arm less than halfway through so that by day 28 of the study period, 18 out of the 19 patients had switched to PEG feeding.

Two studies251,267 reported significantly greater intake of prescribed feed and consequently significantly greater weight gain in PEG patients. In three studies15,251,267 there was a non-significant increase in treatment failure in the nasogastric group.

Mortality was reported for all of the trials. One of them 267 showed no difference between study groups, one showed significantly higher mortality in the nasogastric arm than the PEG arm 251 and two 15,343 reported higher mortality in the PEG group especially if inserted within the first two weeks following a stroke. In addition to the small increase in risk of death demonstrated by the large multi-centre randomised trial 343, this study also showed an increased risk of poor outcomes, although for secondary outcomes such as GI haemorrhaging, PEG patients fared better. Cost-effectiveness evidence

We did not find any study reporting cost or cost-effectiveness. Conclusions

The results of the largest multi-centre trial showed that significant benefit of a PEG over an NG tube is very unlikely and there is a significant mortality/morbidity from PEG insertion. However, patients generally prefer a PEG to a NG tube for long term treatment as it less likely to displace, can remain unseen and is more comfortable. A PEG should therefore be considered after a patient has been shown to tolerate gastric feeding via a nasogastric tube for 2–4 weeks or in patients unable to tolerate a nasogastric tube despite the tube being well secured. After an acute neurological event such as a stroke, insertion of a PEG should be delayed until; the prognosis/QOL of the patient can be better predicted.

‘If the patient cannot decide for themselves, the doctor must provide such treatment and care as are in the patient’s best interests including the duration for which treatment is to be provided. In determining what constitutes best interests the doctor should have regard to the views expressed by carers and the appropriate multidisciplinary health team (see Section 5.3).’ A similar group should decide whether feeding should be stopped. In clinical practice it is more difficult to stop feeding through a PEG than though an NG tube although the same ethical/moral considerations apply.

9.9. Recommendations for clinical practice

9.9.1. People with dysphagia

In the acute setting, for example following stroke, people unable to swallow safely or take sufficient energy and nutrients orally should have an initial 2–4 week trial of nasogastric enteral tube feeding. Healthcare professionals with relevant skills and training in the diagnosis, assessment and management of swallowing disorders should assess the prognosis and options for future nutrition support. [A]

9.9.2. Route of access

Gastrostomy feeding should be considered in people likely to need long-term (4 weeks or more) enteral tube feeding. [D(GPP)]

9.10. Commencing enteral tube feeding after insertion of a percutaneous endoscopic gastrostomy

9.10.1. Introduction

Percutaneous endoscopic gastrostomy (PEG) is a relatively common procedure but it has a significant mortality/morbidity (NCEPOD report). The length of time one should wait before commencing feeding after insertion of the tube has been subject to controversy. Many clinicians believe that feeding should be delayed for at least 24 hours post-insertion but others use PEGs much earlier. Delays in starting PEG feeding may result in unnecessary prolongation of hospital stay and costs. A review was therefore performed to assess the safety of early PEG feeding (within four hours of installation) compared with delayed feeding (more than 24 hours after installation).

9.10.2. Studies considered for this review

Four published RCTs (including 290 patients) met the inclusion criteria46,63,223,333 (Table 46). The more recent studies were of higher methodological quality. The mean age of patients in all studies was more than 60 years.

9.10.3. Clinical evidence

No significant differences were reported for mortality (three studies) or complication rates (4 studies), although two studies reported more gastric distension which had resolved by day three after insertion.

9.10.4. Conclusion

Since none of the studies detected a significant difference or trend between the early or late groups it can be assumed that in an uncomplicated patient there is no reason to delay the start of feeding for more than 4 hours after insertion of a new PEG tube.

9.11. Recommendation for clinical practice

9.11.1. Route of access

Percutaneous endoscopic gastrostomy (PEG) tubes which have been placed without apparent complications can be used for enteral tube feeding 4 hours after insertion. [A]

9.12. Types of enteral feeds

Most enteral feeds come as ready to use liquid microbial free preparations that contain energy, protein, vitamins, minerals, trace elements and fluid +/− fibre. They are usually nutritionally complete within a specific volume. A ready to use standard feed will usually contain 1 kcal and 0.04g protein per ml but many other types of enteral feed preparations are available with differing energy: protein ratios and types of fat or protein.

The GDG did not undertake a formal review of the literature related to different types of enteral feed, however a summary is provided in Table 20.

Table 20. Types of enteral feed.

Table 20

Types of enteral feed.

9.13. Mode of delivering Enteral Tube Feeding

9.13.1. Bolus v continuous

Administering an enteral feed into the stomach rather than small intestine permits the use of hypertonic feeds, higher feeding rates and bolus feeding. Enteral feeding pumps are available to alter rates and in patients with doubtful GI motility, the stomach may be aspirated every 4 hours. If aspirates are high (e.g. exceed 200 – 300 mls depending upon local policy), the pump rate may be reduced and/or prokinetic drugs considered. This is usually recommended in the critical care setting though an aspirate of under 400 ml correlates poorly with the risk of aspiration or pneumonia225. Enteral feeding delivery is usually increased gradually over the first 24 hours (or slower in the very malnourished, see section 6.6).

When using NG feeding, enteral feeds can be delivered continuously over a variable number of hours or intermittently as boluses (or as a combination of both methods). There are potential advantages and disadvantages to both methods. We therefore identified studies that compared different modes of delivering enteral feeds. The RCTs found were categorised into continuous v bolus and continuous (24hr) v continuous (16–18hr). The rationale for non-continuous feeding is that it is more physiological and allows the stomach to completely empty and hence may reduce bacterial colonisation of the stomach which may be safer should an episode of aspiration occur.

9.13.2. Studies on bolus vs. continuous

Nine studies compared continuous v bolus regimens in neurological dysphagic patients, patients with injuries to the head, post-operative cancer patients, critically ill patients314, older patients and healthy adults24,35,66,154,188,201,276,332 (Table 44). Most regimens described in the studies compared 24 hourly continuous feeding with 3–6 hour bolus feeds (250 –500ml). The main outcomes reported were: abdominal discomfort, aspiration pneumonia, change in nutritional status, clogged tubes, nurse preference and biochemical changes.

9.13.3. Clinical evidence on bolus vs. continuous

For abdominal discomfort, aspiration pneumonia and nurse preference there was no evidence of benefit between the continuous and bolus fed group54,66,201,332. However, in one study276 the continuous group were found to have a significant improvement in nutritional status (body weight and arm circumference) compared to the bolus fed group (p<0.01), while in another 66 there was less clogging of nasogastric tubes with bolus feeding (p=0.01).

9.13.4. Continuous vs. cycled continuous

Five studies compared continuous ETF (24hours) v cycled continuous ETF (16–18hours) with daily breaks (2 – 4 hours) or even intermittent ETF (e.g. 4–6 hours feeding then 2 hours rest). Studies were undertaken in critically ill, ventilated patients and post surgical patients35,55,134,323,360. The main outcomes reported were; length of hospital stay, duration of enteral feeding, mortality, ventilator associated pneumonia, gastric pH and rate of gastric colonisation.

9.13.5. Clinical evidence: continuous vs. cycled continuous

There were no significant differences between the 24 hour continuous feeding groups and the 16–18 hour feeding groups in either mortality or ventilator associated pneumonia; 35,134,360, and rates of gastric colonisation and levels of gastric pH were also similar35,323. In one study however 360 there was a significant reduction in hospital stay for a 16 hour fed group compared to a 24 hour continuous group (p=0.04).

9.13.6. Cost-effectiveness

No study reporting cost or cost-effectiveness was found.

9.13.7. Conclusions

Bolus feeding is as effective as continuous (16–24 hours) feeding. Overall, however, the mode of feed delivery can be dictated by practical issues. For example, in patients who pull or dislodge nasogastric tubes regularly, bolus feeding can be used as a practical safe alternative to continuous feeding, while in intensive care the severity of illness and issues of gastric emptying, metabolic stability and control of glucose levels favour continuous feed administration.

9.14. Recommendations for clinical practice

9.14.1. Mode of delivery

For people being fed into the stomach, bolus or continuous methods should be considered, taking into account patient preference, convenience and drug administration. [B]

For people in intensive care, nasogastric tube feeding should usually be delivered continuously over 16–24 hours daily. If insulin administration is needed it is safe and more practical to administer feeding continuously over 24 hours. [D(GPP)]

9.15. Motility Agents

9.15.1. The use of enteral motility agents

If patients with impaired gastrointestinal motility are fed enterally they may develop symptoms of abdominal distension vomiting, gastro oesophageal reflux, pulmonary aspiration, pneumonia or sepsis. They may also have large gastric aspirates and impaired fluid and nutritional intakes. The administration of prokinetic agents is used widely to help with these problems by promoting gastric emptying and improving intestinal motility. We conducted a review to identify studies comparing patients receiving enteral feeds with and without motility agents to see whether this approach is of benefit.

9.15.2. Studies on enteral motility agents in ETF

Ten studies were identified and were categorised into 5 groups according to the type of prokinetic agent administered; erythromycin, metaclopromide and or cisapride (Table 47 and Table 48). However, since cisapride has now been withdrawn, the studies using that drug are not reported here. Most of the studies included patients on intensive care in whom gastrointestinal feed intolerance is associated with a worse outcome and the development of aspiration pneumonia. However, this association is not considered to be causal and the inclusion of these high risk patients in the studies makes interpretation difficult. Erythromycin v placebo

5 studies were included in which erythromycin was administered intravenously either as a single dose 61,212 or every six hours for a minimum of five days30,289,381 (Table 46). Four studies included intensive care patients and one pancreaticoduodenectomy patients. In 2 studies patients were only recruited if they demonstrated intolerance to enteral feeding61,212. The outcomes assessed included mortality, pneumonia, length of stay, complications, gastric emptying, residual gastric volume and feed tolerance.

One study30 detected no significant differences in mortality, pneumonia or length of stay between the intervention and control group and two studies 30,381 reported similar complication rates. Gastric residual volumes were lower with erythromycin in one study289 but there were no differences reported in another 381. Improved tolerance to enteral feeds in the intervention group, was observed in one study30, p=0.001 during the first 48 hours of feeding but there were no significant differences by the end of the study period. In another study61 enteral feeding was more successful in the intervention group after 1hour, p=0.05 and 12 hours, p=0.01 of a single initiating dose of erythromycin but there were no significant differences 24 hours after the dose. Metoclopramide v placebo

Three studies were included176,212,380 one of which also had an additional arm for erythromycin212 (Table 47). All the studies included intensive care patients who were tube fed, with one study 212 only recruiting patients who were not tolerating enteral feeds. The metoclopramide was administered intravenously in one study176 and via a naso/orogastric tube in the other two212,380. No differences were found in intensive care mortality or nosocomial pneumonia, however, this could be due to the inadequate power of the studies. Gastric emptying rates were higher with metaclopramide (p=0.04) in one study 176 but similar in another 212.

9.15.3. Cost-effectiveness

Motility agents could be cost-effective, if they get the gut working without having to resort to parenteral nutrition in a substantial proportion of patients. No study reporting cost or cost-effectiveness was found.

9.15.4. Additional considerations

Prior to administration of motility agents healthcare professionals should review the patient’s need for drugs with known effects in delayed gastric emptying, such as opiates. A reduction in the dose of these drugs may itself improve intolerance to enteral feeds. Within intensive care elevating the head of the patient above 30 degrees is recommended at all times for ETF also turning on the right side may improve gastric emptying.

Patients with moderate to mild gastric motility problems should be offered oral/enteral/IV erythromycin unless there is a high probability of intolerance. Patients with severe gastric problems and those who do not respond to oral agents after 48 hours, should be offered IV motility agents and alternative methods of nutrition support such as post-pyloric ETF or PN may be needed.

9.15.5. Conclusions

Metaclopromide and erythromycin appear to be effective in improving gastric motility and may improve tolerance to enteral feeds for a limited period. However, the studies do not provide evidence of benefit for important long term clinical end points. In the intensive care population care should be taken to consider the risk of drug interactions and side-effects (e.g. dystonic reactions in older people with metoclopramide).

9.16. Recommendations for clinical practice

9.16.1. Motility agents

For people in intensive care with delayed gastric emptying who are not tolerating enteral tube feeding, a motility agent should be considered, unless there is a pharmacological cause that can be rectified or suspicion of gastrointestinal obstruction. [A]

People in other acute care settings who have delayed gastric emptying and are not tolerating enteral tube feeding should also be offered a motility agent unless there is a pharmacological cause that can be rectified or suspicion of gastrointestinal obstruction. [D(GPP)]

If delayed gastric emptying is severely limiting feeding into the stomach, despite the use of motility agents, post-pyloric enteral tube feeding and/or parenteral nutrition should be considered. [D(GPP)]

9.17. Complications of enteral tube feeding

9.17.1. Introduction

Although the GDG did not conduct a formal review of the literature, it is important to recognize that Enteral Tube feeding is associated with a number of complications. These are summarised in Table 21.

Table 21. Complications of enteral tube feeding.

Table 21

Complications of enteral tube feeding.

In view of the above, placement of all enteral tubes should only be undertaken by suitably trained individuals. The position of all NG tubes should be confirmed after placement and before each time of using aspiration and pH paper (with X-ray if necessary) as per the advice from the National Patient Safety Agency 246. This advice should be incorporated in local protocols which should also address the clinical criteria (e.g. unchanged length of tube, absence of any apparent ETF related problems) which will allow ETF to proceed when the ability to repeat checks of position are limited (aspiration and pH checking may be impossible or unhelpful due to gastric acid suppression and repeated X-rays before every feed are not practical). The initial placement of post-pyloric tubes requires X-ray with clinical checks before repeated use. All patients receiving ETF should be closely monitored, particularly early after instigation. Monitoring allows quantification of losses to enable daily estimation of replacement requirements, maintenance of metabolic balance, detection of toxicity/deficiency states, and early detection of complications (see Chapter 7). NG tubes should be replaced in the time frame recommended by the manufacturers.

9.18. Recommendations for clinical practice

9.18.1. Management of tubes

People requiring enteral tube feeding should have their tube inserted by healthcare professionals with the relevant skills and training. [D(GPP)]

The position of all nasogastric tubes should be confirmed after placement and before each use by aspiration and pH graded paper (with X-ray if necessary) as per the advice from the National Patient Safety Agency (NPSA 2005). Local protocols should address the clinical criteria that permit enteral tube feeding. These criteria include how to proceed when the ability to make repeat checks of the tube position is limited by the inability to aspirate the tube, or the checking of pH is invalid because of gastric acid suppression. [D(GPP)]

The initial placement of post-pyloric tubes should be confirmed with an abdominal X-ray (unless placed radiologically). Agreed protocols setting out the necessary clinical checks need to be in place before this procedure is carried out. [D(GPP)]

9.19. Research recommendations

What are the benefits to Intensive care patients likely to stay for >5 days, who are offered ETF only compared to ETF and PN if they fail to tolerate >60% of their target nutritional needs 2 days after starting ETF in terms of survival, complications and hospital costs?

This is an area of common practice but where the benefits of these interventions are unclear and poorly reported.

What are the benefits to malnourished surgical patients who have indications for ETF being offered ETF only compared to ETF and PN if they fail to tolerate >60% of their target nutritional needs two days after starting ETF in terms of survival, complications and hospital costs?

Currently patients who present with the indications for enteral feeding are being given PN early when it seems that they are not tolerating enough enteral feed to meet requirements, however the benefits of fairly early intervention with PN are unclear.

What are the benefits of enteral tube feeding compared to no enteral tube feeding in people with dysphagia and early to mid stage dementia in terms of reduced complications associated with swallowing, improved nutritional status, delay onset of advanced stage dementia, hospital admissions, cost effectiveness and survival?

Much of the research tends to tends to focus or concentrate on tube feeding people with advanced dementia or those who may be in terminal stages of the disease. Depending upon the types of dementia a person has swallowing disorders may occur at an earlier stage in the disease, for example vascular dementia. The benefits and complications of tube feeding may be quite different in people in the earlier stages than those who are in the advanced stage of dementia.



Enteral feeding tubes may also be used for the administration of drugs, frequently on an unlicensed basis. Information and choice on suitable drug preparations can be obtained from local pharmacy or Medicines Information Departments. Further information can also be obtained from ‘Guidance in administering drugs via enteral feeding tubes’ from www​


Malnourished: BMI <18.5 kg/m2, unintentional weight loss >10% within the last 3–6 months, a BMI<20 kg/m2 and unintentional weight loss >5% within the last 3–6 months.


At risk of malnutrition: eaten little or nothing for more than 5 days and/or likely to eat little or nothing for the next 5 days or longer or poor absorptive capacity, and or high nutrient losses and or increased nutritional needs from causes such as catabolism.


Malnourished: BMI <18.5 kg/m2, unintentional weight loss >10% within the last 3–6 months, a BMI<20 kg/m2 and unintentional weight loss >5% within the last 3–6 months.


At risk of malnutrition: eaten little or nothing for more than 5 days and/or likely to eat little or nothing for the next 5 days or longer or poor absorptive capacity, and or high nutrient losses and or increased nutritional needs from causes such as catabolism.


Malnourished: BMI <18.5 kg/m2, unintentional weight loss >10% within the last 3–6 months, a BMI<20 kg/m2 and unintentional weight loss >5% within the last 3–6 months.


At risk of malnutrition: eaten little or nothing for more than 5 days and/or likely to eat little or nothing for the next 5 days or longer or poor absorptive capacity, and or high nutrient losses and or increased nutritional needs from causes such as catabolism.

Copyright © 2006, National Collaborating Centre for Acute Care.

Apart from any fair dealing for the purposes of research or private study, criticism or review, as permitted under the Copyright, Designs and Patents Act, 1988, no part of this publication may be reproduced, stored or transmitted in any form or by any means, without the prior written permission of the publisher or, in the case of reprographic reproduction, in accordance with the terms of licences issued by the Copyright Licensing Agency in the UK. Enquiries concerning reproduction outside the terms stated here should be sent to the publisher at the UK address printed on this page.

The use of registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant laws and regulations and therefore for general use.

The rights of National Collaborating Centre for Acute Care to be identified as Author of this work have been asserted by them in accordance with the Copyright, Designs and Patents Act, 1988.

Bookshelf ID: NBK49253


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...