Clinical Nutrition (1998) 17:145-152 © Harcourt Brace & Co. Ltd 1998

Review

Progress in perioperative enteral tube feeding

S. BENGMARK

Suite 361, Beta House, Ideon Research Center, Lund University, S-22370 Lund, Sweden

Key words: enteral feeding; parenteral nutrition; peri- operative nutrition; surgery

Introduction

Currently, no data support an assumption that either enteral nutrition (EN) or parenteral nutrition (PN) should always be used exclusively; instead these two feeding practices often complement one another. There are situations when EN cannot meet a patient's total nutritional needs and com- plementary PN is necessary, especially during the first 1 or 2 days after major operations. But it is also obvious that PN has not fulfilled our high expectations: neither in providing nutritional support nor in preventing septic manifestations in postoperative and post-trauma patients. Through controlled studies 50 years ago, both Mulholland and Rhoads showed that oral or gastric feeding or feeding through needle jejunostomy might have some definite advantages (1, 2). Several experimental studies as well as clinical studies performed during the 1980s support this assumption. Moore et al concluded in 1991, from a study of eight prospective randomized trials published in the literature, that early postoperative enteral feeding might partly prevent postoperative sepsis (3). In this study, the average sepsis rate in patients fed enterally was 18% com- pared with 35% in patients fed parenterally. Kudsk et al (4) demonstrated the following year, in a randomized study in abdominal trauma patients, a 76% reduction in post-tranma infection rate in patients fed enterally, compared to those fed parenterally. Also, other studies performed in more recent years have reached similar conclusions, suggesting that EN to a larger extent than PN might be the method of choice for nutrition and sepsis control in postoperative and post-tranma (POT) patients after trauma and surgery. The failure of antibiotics (including selective decontami- nation) to solve infection problems in these patients, the increasing awareness of the dangers of using antibiotics, the failure of total parenteral nutrition (TPN) alone to solve the nutritional and sepsis-preventive demands of POT patients, and the accumulating evidence of the capacity of EN to meet nutritional demands and prevent sepsis, make EN a promising alternative (5-10).

Obstacles to POT enteral nutrition

Reduced gastrointestinal motility

For almost a century, the routine has been for patients

to avoid eating sometimes up to 12-18 h before surgery, and also during one or more days after surgery or trauma, depending on the size of operation or trauma. Fear of the patient vomiting at the introduction of inhalation anaes- thesia, plus the assumption that the gastrointestinal (GI) tract is paralysed during the first days after surgery, were the main reasons for this policy. A transitory suppression of GI motility is a common phenomenon in the early post- operative period, especially after laparotomy (11). Although reduced motility mainly affects the stomach and to some extent the colon, alterations in electrical and mechanical activity also occur in the small intestine, in both expe- rimental animals (12, 13) and humans (12, 14). The reason for the pronounced problems of postoperative ileus in connection with laparotomy is that nociceptive stimulation of the peritoneum induces hyperactivity of the sympathetic nervous system, which seems to play a prominent role in the genesis of postoperative ileus (11, 13, 15, 16). Despite this, enteral feeding is almost always possible, especially if attempts are made to bypass the stomach through use of enteral feeding tubes. Not only the small intestine, but also the large intestine, almost always maintains its absorp- tive capabilities postoperatively (17, 18). Furthermore, in the future postoperative ileus may be able to be controlled by adrenergic receptor antagonists.

145

Prophylactic nasogastric decompression - a relic ?

For more than half a century the old tradition of routine gastric decompression has delayed the acceptance of GI feeding tubes among surgeons. Many surgeons are; still inclined to give priority to decompression over feeding. Gastric distension should not under normal circumstances be regarded as a contraindication to jejunal feeding, nor should high residual volumes automatically lead to cessation of tube feeding (19). Instead it is recommended (20, 21), if the gastric residual volume is high, only to withhold enteral feeding for approximately 2 h (20, 21). It is a general experience that if tube feeding is continued, the residual volume will soon decrease.

A recent controlled study of 177 so-called 'usual' GI operations compared immediate, early (within 48 h), and late removal of the gastric tube after surgery (22). No signi- ficant differences were noted in duration of hospital stay, time for return of adequate bowel function, or time before the start of an oral diet. The intubated patients, howeve, r, did

t46 PROGRESS IN PERIOPERATIVE ENTERAL TUBE FEEDING

suffer a significantly greater risk of respiratory complica- tions. In another recently-published randomized trial, 40 patients with nasogastric decompression kept in place until resumption of bowel function (group 1) were compared with 37 patients in whom the tube was removed in the recovery room (group 2) (23). Postoperative fever was noted in 23/40 (58%) in group 1 and 14/37 (38%) in group 2. Most important, the frequency of atelectasis was signi- ficantly lower in group 2 (14%), compared with group 1 (38%) (P = 0.03). Despite decompression, 11/40 (28%) in group 1 and 13/37 (35%) in group 2 developed nausea, and 5/40 (12%) and 3/37 (8%) patients, respectively, required reintubation. One-third of the patients in each group suffered abdominal distension. Gastric distention will usually resolve spontaneously, and without serious sequelae. In a randomized trial performed in patients who had stomach operations, no patient in the group with early removal of the gastric drainage required reinsertion of a tube for treatment (24). A recently-published meta-analysis based on 26 trials (a total of 3964 patients) concludes that 'routine nasogastric decompression is not supported in the literature by scientific evidence' (25). For every one patient needing a nasogastric tube left in the postoperative period, at least 20 patients will not need it. Although both patients with or without a decompression tube developed distension, this was not associated with an increase in complications or in length of stay. In fact fever, atelectasis and pneumonia were significantly less common, and days to first oral intake significantly fewer, in patients managed without nasogastric tubes (25).

Intensive perioperative EN

Several recent studies have demonstrated that early enteral feeding after larger abdominal surgery is safe and can be tolerated by the majority of patients (26, 27). Recent studies in 108 liver transplant patients showed that enteral tube feeding, starting 12-24 h after surgery, is almost always successful (28). A similar UK study in 24 patients under- going orthotopic liver transplantation reached the same conclusions (29), and a recent Canadian study in liver trans- plant patients demonstrated, in the early postoperative period, definite advantages with tube feeding over mainte- nance i.v. feeding (30); the hospital stay was 22 vs 31 days, mortality 7% vs 11%, and most interestingly, the early rejection rate 7% vs 44%.

It is, however, possible to provide EN not only imme- diately after surgery, but actually uninterruptedly, i.e. imme- diately before, during, and immediately after the surgical procedure. Such an intensive perioperative EN was first introduced in the treatment of burn patients, who frequently have multiple operations, and in whom an interruption in EN can cause significant caloric deficit and increase the risk of infections (31, 32). Thus, in these patients, it is highly desirable to abandon the traditional praxis to withhold enteral nutritional support until bowel sounds return, and to try to develop a routine of uninterrupted EN. Such a

feeding policy seems to haste been introduced by the group of Pruitt as late as 1990 (31). Although in their short abstract no details are given regarding the time and extent of perioperative nutrition, they conclude from a study in 47 patients that EN supplied uninterruptedly during the perioperative period is safe as long as the tube is maintained in the small bowel. A more recent controlled study confirms that EN is not only safe but also efficient when given during the operative procedure (32). In this study, 40 patients received EN during 161 surgical procedures and 40 had enteral support, as traditionally practised, withheld during 129 procedures. The conventionally-fed group showed a significant caloric deficit (P < 0.006), increased incidence of wound infections (P < 0.02), and a significant necessity of higher albumin supplementation (P < 0.04) compared with the group fed uninterruptedly. Other studies have shown that early or immediate EN, in addition to lower inci- dences of hospital infections and shorter hospital stays, also makes the use of stress ulcer prophylaxis unnecessary (33).

Efficient feeding tube technology

Spontaneous duodenal intubation

The main obstacle to routine use of EN during the peri- operative period has so far been the lack of appropriate feeding tubes, easy to introduce into the jejunum, which will not obstruct nor be regurgitated prematurely. Under ideal circumstances, such tubes should within minutes or a few hours be placed with the tip in the upper jejunum, without support of expensive techniques such as radiology or endoscopy. It is clear that the spontaneous transpyloric passage (STP) is unacceptably low for most tubes currently on the market. This is most likely the main obstacle to the technique receiving acceptance as a standard feeding method during routine elective surgery. Whatley et al were the first to prospectively study the rate and time course of spontaneous duodenal intubation after nasal insertion of three different weighted feeding tubes (34): Dobbhoff (Biosearch), Entriflex (Biosearch), and Duo-Tube (Argyle/ Sherwood). After 4 h, 8/20 (40%), 13/20 (65%), and 2/20 (10%), respectively, had passed the pylorus. In total, the tip of 32/60 tubes (53%) had passed the pylorus after 48 or 72 h and 35/60 (58%) after 96 h. Levenson et al, in a similar study 5 years later, compared a weighted tube, Vitafeed (Health Tek/Kendall-McGaw), with a tube simi- larly constructed but without weights (35). After 4 h the tip of 23/84 (27%) and 20/90 tubes (22%), respectively, had passed the pylorus. After 24 h, 30/84 (36%) and 31/90 (34%) respectively, had passed the pylorus; after 48 hours, 33/84 (39%), and 36/90 (40%); and after 72 hours, 35/84 (42%) and 37/90 (41%). Another similar study also pub- lished in 1988 compared three new nasoduodenal feeding tubes (Corpak): a bullet unweighted, a bullet weighted, and a bolus weighted (36). After 24 h, the tip of 7/22 (32%), 7/26 (27%) and 9/21 tubes (43%), respectively, had passed into the duodenum. Similar to the results of the previous studies, only one-third of the tubes spontaneously passed

the pylorus. Furthermore, this study did not support any advantages of using weighted tubes. Once through the pylorus, the unweighted tubes stayed in position signi- ficantly longer than did the weighted ones (P < 0.005).

Nasoenteric tube feeding is especially warranted in critically-ill patients. However, in most of these patients GI motility is no longer intact and the spontaneous transpyloric passage of feeding tubes, which is dependent on motility, can be expected to be low. In a recent study involving intensive care unit (ICU) patients requiring assisted venti- lation, transpyloric placement was successful in only 10/68 patients (15%) (37). The results of this study clearly show that at present one cannot rely on STP in ICU patients. The situation is much better in the elective surgical patient, who has intact GI motility, that can be used to transport the tip of the tube into the upper jejunum. It has been shown that weight in the tip, used in the past, will coun- teract its purpose, because the outlet of the stomach is not at its lowest part. Instead, light-weighted tips with ability to maximally use the motility of the stomach will have the best potential to be transported through the pylorus and down into the upper jejunum. This was the reason why I constructed the 'self-propelling' feeding tube, which was recently introduced on the European market (Flo-care tube, Nutricia). To maximally utilize the motility, the tip of the tube is constructed with a coil in its end. The intention is for the tube to be introduced the evening before surgery. It is expected to be in place after 4 h and ready for use during the whole perioperative period. Early studies have shown, even without pharmaceutical stimulation of motility, the STP rate for this tube is almost 100% (38).

After introduction of the feeding tube into the stomach, the motility is usually stimulated by a light meal such as a sandwich, a pizza, or an orange. Should this not be possible, as is often the case in critical care patients, motility-promoting drugs such as metoclopramide (39, 40) or erythromycin can be tried (41, 42). In a recent controlled study performed in surgical, burn and paediatric intensive care patients, metoclopramide was twice as effective when standard unweighted tubes were used compared to weighted, the 4-h STP rate was 84% and 36%, respectively (P < 00.2) (43). Passage beyond the pylorus was in most cases successful when intravenous injections of 10-20 mg meto- clopramide (in children 0.1-0.2mg/kg) 10rain before insertion of the tube were used. No difference was seen between the lower and higher doses. The macrolide anti- biotic agent erythromycin has been shown to produce similar effects on GI motility as the polypeptide mofilin. It is recommended that 3 mg/kg body-weight be infused intravenously during a 1-h period before tube placement (41). In a small controlled trial 400 mg was given in three repeated doses, 8 h after intubation of the stomach with a standard weighted tube (42). Transpylofic passage before 24 h was achieved in 6/8 (75%) in the erythromycin group compared with 0/9 in the control group. With a tube made to maximally utilize motility, STP can be expected to be even better: it should be possible to have almost 100% STP within 1 or 2 h, and the tube ready for feeding purposes

CLINICAL NUTRITION 147

almost immediately. Metoclopramide and erythromycin do have side-effects in approximately one-third of patients (42), but most occur with chronic administration. Devel- opment of other and more effective prokinetic drugs seems highly desirable.

Another important factor in tube placement is that the tube, once placed, shall remain in place until it is taken out and not be regurgitated into the stomach, with the risk of subsequently being aspirated into the respiratory tract. In general, unweighted tubes stay secured longer than do weighted ones (44). The coil of the Flo-care tube is safely anchored in the upper jejunum and so far has rarely regurgi- tated into the stomach. Should it be regurgitated, it cannot be aspirated because of the large diameter of the coil. In patients where the self-propelling capacity of such a tube cannot be used, e.g. patients with inhibited GI motility, the ability to anchor the tube in the small intestine and the low regurgitation rate will still offer advantages.

Alternatives in intestinal paralysis

There are experts who almost always manipulate manually the tip of the tube, if not into the jejunum, at least into the duodenum, especially in critically-ill patients without gastric motility. Ugo et al could, using specific patient positioning, insertion techniques, and gastric distension with air, manipulate the tip behind the pylorus in 85/103 patients (83%), within an average time of 30 min (45). Schultz et al, using fight lateral decubitus positioning plus gastric insufflation of air, reported an immediate success rate of placement in 37/42 patients (88%) (46). Zaloga et al made successful placement in 213/231 patients (92%) within an average time of 40 _+ 14 min, placement being confirmed by bile aspiration, determination of pH c]aange from acidic to basic, and blue dye injection (47). A 30 ° bend was placed 3 cm from the distal tip (6.5 cm from the end of the tube) on either Entriflex (Biosearch) or Corsafe (Corpak), and a syringe used to slowly inject air and rotate the tube down. With this method, the tip of the tube reached the proximal jejunum in 17% of cases: in the remaining 83%, the tip was placed somewhere in the duodenum, most often in the lower part. Although more expensive, fluoro- scopic and endoscopic placements are generally faster; Ott et al report placement times of 8.6 and 13.4 min respec- tively for the two techniques (48). In this study, fluoro- scopic placement was successful in 94/104 cases (90%) and endoscopic placement was successful in all cases tried (100%), even in those for which fluoroscopic placement had failed. Mathus-Vliegen et al showed that it is possible to place tubes endoscopically within an average of 13 min (49). Thus, all the above semi-invasive techniques (expert manipulation, fluoroscopy, and endoscopy) seem to give satisfactory results. It seems logical then that each medical centre use the technique that its staff perform best. Endo- scopy may be the most favourable method in patients with inhibited GI motility, but expert manipulation can occa- sionally take a long time, which might put the patient in unnecessary discomfort. However, endoscopy, although

148 PROGRESS IN PERIOPERATIVE ENTERAL TUBE FEEDING

expensive, is almost always quick and reliable. Today it can also be performed by through-the-nose gastroscopy (50, 51). Using a paediatric gastroscope and local anaesthesia to the nose, the tube was successfully placed in 82/92 patients (89%), in 10/92 patients (11%) the turbinates were bilat- erally too narrow. It should be noted that hypo-osmolar solutions of contrast media such as Gastrografin can damage the intestinal mucosa and lead to microbial trans- location, as observed in animal studies (52). For the intro- duction of PEGs and PEJs, alternative techniques such as ultrasound guidance (53) and laparoscopy (54) have been suggested as effective tools.

Oral, gastric, or enteral feeding?

Whenever possible, oral feeding is best. There are published reports of oral feeding, if not perioperatively or immediately after surgery, at least early after larger operations such as colonic resection (55). The only indication for short-term perioperative enteral tube feeding is lack of gastric motility and risk of vomiting and aspiration when eating. An effec- tive intensive perioperative EN often necessitates enteral tube feeding. Indications for long-term enteral tube feeding are different; unconsciousness, paralysis of the swallowing mechanisms, or oesophageal obstructions. In most of these conditions in which the stomach can be expected to have normal motility, both gastric and enteral feeding are alter- natives. It is unfortunate that most of the reports on tube feeding in the literature are based on mixed use of short- and long-term, acute and chronic tube feeding, which often limits the interpretation of the results.

The advantages and disadvantages of gastric and jejunal tube feeding are well summarized by Bochus (56). Montecalvo et al have compared gastric and jejunal tube feeding in a small, but not blinded, controlled study of medical and surgical ICU patients (57). The mean duration of feeding was about 10 days for the two groups. In these patients, mainly suffering trauma or cardiorespiratory failures, jejunal feeding led to a significantly higher caloric intake (P = 0.05), a greater increase in serum prealbumin concentrations (P < 0.005), and a lower rate of pneumonia, compared with gastric tube feeding. The incidence of diar- rhoea was low in both groups. Although the jejunal group had more days of diarrhoea (mean, 3.3 ___ 6.6 compared with 1.8 _+ 2.9 days), this difference was not significant. The ICU nurses also reported that the patients fed by gastric tubes were more inclined to develop gastroesophageal reflux (67.8%) and pulmonary aspiration (77%). Conse- quently, nurses in general favoured jejunal tube feeding compared with gastric (71.3% compared with 18.4%; 10.3% had no opinion on the subject). In a review (58) of the liter- ature from the period from 1978 to 1989 (58), only 2/45 papers were found to report the incidence of aspiration in gastric and jejunal feedings (59, 60): both clearly favoured jejunal feeding over gastric feeding in prevention of aspiration pneumonia. There is, however, a need for more studies.

Salivation - important for outcome?

Eating stimulates salivation and GI secretion. The 1-2 L of saliva normally secreted each day are extremely rich in infection-preventing and healing-inducing or growth factors (61). Mandel (62) has pointed out, that 'the saliva possesses a multiplicity of defence systems for antibacterial warfare that Pentagon can envy' (62). However, few studies have dealt with the eventual negative effect of lack of salivation during shorter episodes of sickness. In patients being radiated in the head and neck region, lack of sali- vation was associated with significantly lower intake of energy, vitamins and minerals (63). The normal acidity of the stomach is also important to prevent colonization of the stomach with gram-negative bacilli, because pneumonia is likely to develop from a gastric reservoir of colonizing bacilli (64-66). Even during short-term perioperative enteral tube feeding, patients would most likely benefit from inter- mittent chewing of something to stimulate salivation. It is also recommended that the acidity of the stomach should be allowed to be low intermittently, thereby providing de- contamination of the stomach (67). On the other hand, continuous nasogastric feeding has been reported to effec- tively prevent stress ulceration (68). It has also been observed that intermittent gastric tube feeding can lead to a significant deficit in caloric intake (69). Another advantage of continuous nasoenteral feeding is said to be, that it counteracts the 'hyperbilirubinemia' (70), known to occur in enteral fasting (71). However, our present knowledge is in many respects incomplete and further studies are needed.

Complications associated with tube feeding

Although the list of reported and potential complications to nasoenteric tube feeding is long, it does become shorter as we gain experience. In one studied group of ICU patients, complications were observed in 10/68 patients (15%), of which 4 were tube dislodgements and only 1 was a more serious aspiration pneumonia (37). The complication rate, with perioperative enteral tube feeding, should become significantly lower if tube dislodgement problems can be minimized. It should also be remembered that the alter- native, PN, is also fraught with its own potential complica- tions. There are currently no clinical reports of tube-related complications with solely short-term perioperative EN. In patients receiving this form of nutrition, unintentional removal of the tube and tube clogging seem to be the most frequent complications which interrupt feeding. Most of the other complications, such as severe diarrhoea, protracted vomiting, and gastric distension 'may provide special challenges to tube feeding, but are not necessarily contra- indications' (72). The diarrhoea frequently observed is less often caused by infection and more often due to the content of sorbitol in the formulas given (73), too high a content of fat (74) or the use of histamine H2-receptor antagonists (75-77), antacids (especially magnesium-containing pre- parations) (78-80), and antibiotics (81-85). For further

review, see Gottschlich et al (74). To prevent antibiotic- induced diarrhoea, supplementation with lactobacilli are increasingly recommended (5-10, 56).

The British Dietetic Association advises that, for hygienic reasons, tubes be exchanged once every 2 weeks (49). The scientific evidence for this recommendation is not known. However, it is reasonable to request that enteral feeding tubes be made to stay in place and continue to function, if needed, for at least 2 weeks.

Unintentional tube removal

In the study conducted by Mathus-Vliegen, premature re- moval of the tube occurred in 29/46 patients (63%), patients being the cause in 16 (35%) and staff in 5 (11%); there was no explanation for tube removal in 7 cases (49). Similar experiences were reported by Benson et al, when patients were the cause in 16/50 cases (32%), and staff in 9/50 (18%) (86). In a study to be published by Silk et al, non- elective tube removal occurred in most patients in whom the following brands of tubes were used: Radius (Radius International, Grayslake, IL, USA), 80%; Flexiflo (Ross Products, Columbus, OH, USA); 73.3%, and Freka (Fresenius Ltd, Runcorn, Cheshire, UK), 73.3% (87). Mathus-Vliegen et al observed that the unintentional removal rate was higher if shorter tubes (105 cm) were used compared with longer ones (145 cm), which could be positioned significantly deeper into the intestine (49). Because there are no negative consequences with use of longer feeding tubes, these tubes should probably be recommended. A slight anchoring of the tip of the tube in the duodenum, for example as provided by the Flo-care tube, might be desirable.

The high rate of unintentional removal also calls for improved methods for taping the tube to the patient's nose. Burns et al recently compared three securing methods available on the market: pink tape (Plastic Adhesive Tape, Surgical Product Corp), clear tape (Bioclusive, Johnson and Johnson) and a special tube-attachment device (Hollister Inc) (44). The mean time until failure was 100, 56, and 30 h respectively. As a consequence, these authors recom- mended that the tape be exchanged every day or every second day. It is cheap to replace tape, but costly, time- consuming, and unpleasant for the patient to replace a feeding tube.

Clogging - a remaining problem

Clogging remains an obstacle to the routine use of enteral feeding tubes, especially when tubes are used for periods longer than 1 week (86), a length of time fortunately rarely needed after most surgical operations. The time that the three brands of enteral tubes studied by Silk et al (87) remained in situ were as follows (87): Radius, 6.0 -+ 1.0 days; Flexiflo, 5.3 -+ 0.8 days; and Freka, 4.4 _+ 0.7 days. Thus, the tubes presently on the market are suitable mainly for short-term use. Mechanical tube problems or clogging occurred in 11% of the patients, no difference was observed among the three types of tubes tried. In a similar study by

CLINICAL NUTRITION 149

Mathus-Vliegen et al in which tubes made of polyurethane with open-ended tips and side-holes were used, the tubes remained in situ about 50% longer (about 10 days), but the clogging frequency was almost twice as high: 11/46 (24%) (49). Clogging could be corrected without replace- ment of the tube in 7/11 cases by stretching the tube (two episodes) or by flushing or dissolving of the clot (five episodes), but in 4/46 cases (9%), it necessitated exclhange of the tube. Many factors contribute to early luminal occlu- sion: pump malfunction (88), high formula viscosity (88), medications (89), or failure to properly irrigate the tube whenever feeding is interrupted (88). The type of tube, its construction, and the material used may influence the clogging rate, but solid research to support such an assump- tion is scarce. Gastric tubes have a greater tendency to become clogged compared with enteral tubes, because certain protein formulas clog instantly when mixed with gastric juice (pH 1.5) but not with jejunal aspirate (pH 7.6) (90). The clogging occurs at a pH of 5, when intact protein formulas are acidified, but not with formulas containing only flee amino acids and small peptides.

Although prophylaxis by frequent irrigation of the feeding tube is the single most important measure to prevent clogging, attempts should also be made to develop effective methods to unclog occluded tubes. Marcuard et al tried to unclog tubes by using different proteolytic enzymes (88). In a controlled in vitro study performed with preclotted 8 F feeding tubes, the effects of papain (Adolf's Meat Tenderizer), a mixture of trypsin, chymotrypsin, amylase and lipase (Viokase), Sprite, Pepsi, Coca Cola, and Mountain Dew were compared. The solution (5 ml) was injected into the tube, which was clamped for 5 min. Viokase adjusted by 1 tool NaOH 1/L to pH 7.9 was signi- ficantly more effective than the other solutions, however, plain, nonadjusted Viokase (pH 5.9) was ineffective. In a small subsequent clinical study all clogged tubes could be cleared except those in which the clogging was due to knotted tubes, impacted tablets, or long duration of the clogging (> 24 h). A practical way to produce a Viokase pH 7.9 solution is to crush one tablet of Viokase and one tablet of sodium bicarbonate and then add 5 ml (one teaspoon) of water.

The tube: a potential reservoir for pathogens

Numerous studies have addressed the bacterial contami- nation of enteral feeding preparations but few have evaluated in vivo the microbial colonization of the feeding tube. For review, see the article by Bussy (91). Colonization of the rinsing solutions, kept overnight in the tube, was studied in 31 consecutive patients. All proved to be con- taminated and 102 strains were cultured with a median density of 106 colony-forming units/ml (range: 10 to 101°). Of the strains detected, 48/102 (47%) were Enterobac- teriaceae, 20/102 (20%) were group D streptococci, 9/102 (9%) were Candida albicans, 9/102 (9%) were Pseudo- monas aeruginosa, and 16 were other species. Multiple

150 PROGRESS 1N PERIOPERATIVE ENTERAL TUBE FEEDING

antibiotic resistance was present in about 10% (12/102) of the strains and low sensitivity in another 25%. Throat and faeces were frequently colonized by the same bacteria as the predominant bacteria cultured from the tube-rinsing solution. Use of sterile or filtered water instead of tap water for rinsing should be considered and it should also be remembered that mineral waters are not free from patho- gens. Nosocomial legionellosis associated with nasogastric feeding and tap water was recently reported (92). Pseudo- monas organisms have been identified in 6/8 American (93) and 39/87 European (94) brands of bottled water. Although outbreaks of infections from mineral water are rare, the risk should be considered, at least for critically-ill individuals.

Hospital malnutrition needs further attention

sized by Baskin, the old adage ' if the gut works, use it' should in light of present knowledge be updated to, 'use it, or lose it' (100).

The necessary instrument for effective intensive peri- operative EN is available today, although not always used. The reason is most likely lack of information and low interest among clinicians. Enteral nutrition is becoming a specialty of its own and successful treatment is much dependent on the interest and skill of those responsible. Although it is universally recognized that nutrition support teams are extremely effective, only a minority of hospitals even in developed countries have access to such specialists. As long as EN is left to the generalist and his or her own interest in and knowledge of the potential benefits of effective EN support, it will be available to patients only at random.

It is well known that nutritional support can effectively address the problem of malnutrition, improve clinical out- come, reduce suffering, and reduce the costs of health care. Malnutrition occurs today, even at the best hospitals in the world, 40-55% of patients are either malnourished or at risk of malnutrition (95, 96): for review see article by Gallagher-Allred et al (96). Surgical patients with a like- lihood of malnutrition are three times more likely to have complications and excess mortality, they also have sub- stantially-longer hospital stays, and their hospital charges may be increased by 35-75% (96).

Most of the enteral nutrition formulas available on the market contain 1-2 kcal/ml and have an osmolarity of around 350 mosm/kg water. The amounts needed to satisfy a patient's daily energy needs is thus considerable, 50-100 ml/h depending on the concentration of the solution and mode of supply (continuous or interrupted). Although it is probably not necessary even after larger operations, it is interesting that as much as 300 kcal/h can be success- fully administered by postpyloric infusion, starting imme- diately after transfer of the patient to the recovery room (97). Studies in healthy volunteers have shown a tolerance of starter flow rates up to 150 ml/h and starter osmolarities up to 690 mosm/kg (98). It is usually recommended that immediately after the operation 50 ml/hour be given, and the volume carefully increased during the subsequent 24 h. This policy is likely far too conservative.

Conclusions

Effective programmes to counteract malnutrition and reduce health expenditure are more important than ever. Govern- ments, health maintenance organizations, insurance com- panies, doctors and nurses are increasingly aware of the importance of nutrition (99). An important part in such programs is intensive perioperative EN for which jejunal tube feeding, from the evidence available, is an important instrument. The old principles of preoperative abstinence from feeding and the 'wait for flatus' principle for post- operative nutrition should give way to an effective and uninterrupted enteral perioperative nutrition. As empha-

References

1. Mulholland J H, Tui C, Wright A M, et al. Nitrogen metabolism, caloric intake and weight loss in postoperative convalescence. Ann Surg 1943; 117:512-534

2. Riegel C, Koop C E, Drew Je t al. The nutritional requirements for nitrogen balance in surgical patients during early postoperative periods. J Clin Invest 1947; 26:18-23

3. Moore F A, Feliciano D F, Andrassy J R et al. Early enteral feeding compared with parenteral, reduces postoperative septic complications: the result of a meta-analysis. Ann Surg 1991; 216: 172-183.

4. Kudsk K A, Groce M A, Fabian T C et al. Enteral versus parenteral feeding: effects on septic morbidity after blunt and penetrating abdominal trauma. Ann Surg 1992; 215: 503-513.

5. Bengmark S, Jeppsson B. Gastrointestinal surface protection and mucosa reconditioning. Journal of Enteral and Parenteral Nutrition 1995; 19:410-415

6. Bengmark S, Larsson K, Molin G. Gut mucosa reconditioning with species-specific lactobacilli, surfactants, pseudomucus, and fiber - an invited review. Biotechnology Therapeutics 1995; 5:171-194

7. Bengmark S. Immunonutrition - A new aspect in the treatment of critically ill patients. In: A Gullo (Ed). Anaesthesia, Pain, Intensive Care and Emergency Medicine - APICE. Proc. 10th Postgraduate Course in Critical Care Medicine. Trieste, Italy, 1995:73-90

8. Bengmark S. Econutfition and health maintenance - A new concept to prevent GI inflammation, ulceration and sepsis. Clinical Nutrition 1996; 15: 1-10.

9. Bengmark S. Ecological control of gastrointestinal tract. The role of probiotic flora. Gut 1998; 47:2-7

10. Bengmark S. Intensive Perioperative Enteral Nutrition. Submitted 11. Sagrada A, Fargeas M J, Bueno L. Involvement of alpha-1 and

alpha-2 receptors in postlaparotomy intestinal motor disturbances in the rat. Gut 1987; 28:955-959

12. Appert H E, Howard J M. Differences between dogs and humans in intestinal motility following abdominal surgery. Federation Proceedings 1972; 31:828

13. Dubois A, Weise V K, Kopin U. Postoperative ileus in the rat: physiopathology, etiology and treatment. Ann Surg 1973; 178:781-786

14. Danchel J, Schang J C, Kacheloffer Je t al. Gastrointestinal myoelectical activity during postoperative period in man. Digestion 1976; 14:293-303

15. Sj/Squist A, HallerNick B, Glise H. Reflex adrenergic inhibition of colonic motility in anaesthetised rat caused by nociceptive stimuli of peritoneum - an alpha2-adrenoceptor mediated response. Dig Dis Sci 1985; 30:749-754

16. Glise H, Lindahl B-O, Abrahamsson H. Reflex inhibition of gastric motility by nociceptive intestinal stimulation and peritoneal irritation in the cat. Scand J Gastroenterol 1980; 15:673-681

17. Gluckman D L, Halser M H, Warren W D. Small intestinal absorption in the immediate post-operative period. Surgery 1966; 60:1020-1025

18. Moore E E, Jones T N. Nutritional assessment and preliminary report on early support of the trauma patient. J Am Coil Nutr 1983; 2:45-49

19. McClave S A, Snider H L, Lowen C C et al. Use of the residual volume as a marker for enteral feeding intolerance: prospective blinded comparison with physical examination and radiographic findings. Journal of Enteral and Parenteral Nutrition 1992; 16:99-105

20. Rombeau J L, Jacobs D O. Nasoenteric tube feeding. In: Rombeau J L, Caldwell M D (Eds). Enteral and tube feeding. Philadelphia: Saunders, 1984; 261-274

21. Norton J A, Otta L G, McClain C et al. Intolerance to enteral feeding in brain-injured patients. J Neurosurg 1988; 68:62-66

22. Schwartz C I, Heyman A S, Rat A C. Prophylactic nasogastric tube decompression: Is its use justified? South Med J 1995; 88:825-830

23. Petrelli N J, Stulc J P, Rodrignes-Bigas M e t al. Nasogastric decompression following elective colorectal surgery. Am Surg 1993; 59:632-635

24. Sitges-Serra A, Cabrol J, MaGubern Je t al. A randomized trial of gastric decompression after tnmcal vagotomy and anterior pylorectomy. Surgery, Gynecology, Obstetrics 1983; 158:557-560

25. Cheatham M L, Chapman W C, Key S P, et al. A meta-analysis of selective versus routine nasogastric decompression after elective laparotomy. Ann Surg 1995; 221:469-478

26. Reissman P, Teoh T-A, Cohen S M e t al Is early oral feeding safe after colorectal surgery? Ann Snrg 1995; 222:73-77

27. Carr C S, Ling K D E, Boulos Pe t al Randomized trial on safety and efficacy of immediate postoperative enteral feeding in patients undergoing gastrointestinal resection. BMJ 996; 312:869-871

28. Pescovitz M D, Metha P L, Leapman S Be t al. Tube jejunostomy in liver transplant recipients. Surgery 1995; 117:642-647

29. Wicks C, Somasundaram S, Bjarnason Ie t al. Comparison of enteral feeding and total parenteral nutrition after liver transplantation. Lancet 1994; 344:837-840

30. Sharpe M D, Pukil J, Lowndes R, et al. Early enteral feeding (EEF) reduces incidence of early rejection following orthotopic liver transplantation. London: Joint Congress of Liver Transplantation (Abstract), 1995

31. Buescher T M, Cioffi W G, Becker W K et al. Perioperative enteral feedings. Proceedings of the American Burn Association 1990; 22: 192 (abstract)

32. Jenkins M E, Gottschlich M M, Warden G D. Enteral feeding during operative procedures in thermal injuries. J Bum Care Rehabil 1994; 15:199-205

33. McDonald W S, Sharp C W, Deitch E A. Immediate enteral feeding in burn patients is safe effective. Ann Surg 1991; 214:177-183

34. Whatley K, Turner WW, Dey M e t al. Transpyloric passage of feed{hg tubes. Nutrition Support Services 1983; 3:18-21

35. Leve~nson R, Turner W W, Dyson Ae t al. Do weighted nasoenteric feeding tubes facilitate duodenal intubations? Journal of Enteral and Parenteral Nutrition 1988; 12:135-137

36. Rees R G P, Payne-James J J, King C et al. Spontaneous transpyloric passage and performance of 'Fine Bore' polyurethane feeding tubes: a controlled clinical trial. Journal of Enteral and Parenteral Nutrition 1988; 12:469-472

37. Marian M, Rappaport W, Cunningham D, et al. The failure of conventional methods to promote spontaneous transpyloric feeding tube passage and the safety of intragastric feeding in the critically ill ventilated patient. Surg Gynecol Obstet 1993; 176:475-479

38. Jeppsson B, Tranberg K-G, Bengmark S. Technical developments. A new self-propelling nasoenteric feeding tube. Clinical Nutrition 1992; 11:373-375

39. Pirola R C. Rapid duodenal intubafion with Metoclopramide. American Journal of Digestive Disease 1967; 12:913-915

40. Whatley K, Turner W W, Dey M e t al. When does Metoclopramide facilitate transpyloric intubafion? Journal of Enteral and Parenteral Nutrition 1984; 8:679-681

41. Di Lorenzo C, Lachman R, Hyman P E. Intravenous erythromycin for postpyloric intubation. J Pediatr Gastroenterol Nun 1990; 11: 45-47

42. Stern M A, WolfD C. Erythromycin as a prokinetic agent: a prospective, randomised, controlled study of efficacy in nasoenteric tube placement. Am J Gastroenterol 1994; 89:2011-2013

43. Lord L M, Weisder-Maimone A, Pulhamus M et al. Comparison of weighted vs unweighted enteral feeding tubes for efficacy of transpyloric intubafion. Journal of Enteral and Parenteral Nutrition 1993; 17:271-273

CLINICAL NUTRITION 151

44. Burns S, Martin M, Robbins V e t al. Comparison of nasogasmc tube securing methods and tube types in medical intensive care patients. Am J Crit Care 1995; 4:198-203

45. Ugo P J, Mohler P A, Wilson G L. Bedside postpyloric placement of weighted feeding tubes. Nutrition in Clinical Practise 1992; 7:284-287

46. Schultz M A, Santanello S A, Monk Je t al. An improved method for transpyloric placement of nasoenteric feeding tubes. Int Surg I993; 78:79-82

47. Zaloga G P. Bedside method for placing small bowel feeding tubes in critically ill patients - a prospective study. Chest 1991; 100: 1643-1646

48. Ott D J, Mattox H E, Gelfand D W et al. Enteral feeding tubes: placement by using fluoroscopy and endoscopy. Am J Roentgenol 1991; 157:769-771

49. Mathus-Vliegen E M H, Tytgat G N, Merkus M P. Feeding tubes in endoscopic and clinical practice: the longer the better? Gastrointest Endosc 1993; 39:537-542

50. Fregonesi D, Di Falco G. Through-the-nose gastroscopy. Am J Gastroenterol 1991; 86: 381. Letter to the Editor

51. Fregonesi D, Di Falco G. Through the nose gastroscopy for the placement of feeding tubes. Endoscopy 1993; 25:539-541

52. Feigenberg Z, Levavi H, Ben-Baruch D et al. Translocation of bacteria due to direct mucosal damage caused by Gastrografiu. Dig Dis Sci 1994, 39:157-160

53. Pugash R A, Brady A P, Isaacson S. Ultrasound guidance in percutaneous gastrostomy and gastrojejunostomy. Can Assoc Radiol J 1995; 46:196-198

54. Albrink M H, Foster J, Rosemurgy A Se t al. Laparoscopic feeding jejunostomy: also a simple technique. Surg Endosc 1992; 6:259-260

55. Moiniche S, Billow S, Hesselfeldt P et al. Convalescence and hospital stay after colonic surgery with balanced analgesia, early oral feeding and enforced mobilisation. Eur J Surg 1995; 161:283-288

56. Bochus S. When your patient needs tube feeding. Making the right decisions. Nursing 1993: July: 34-42

57. Montecalvo M A, Steger K A, Farber H H W. Nutritional outcome and pneumonia in critical care patients randomised to gastric versus jejunal tube feedings. Crit Care Med 1992; 20:1377-1387

58. Lazarus B A, Murphy J B, Culpepper L. Aspiration associated with long-term gastric versus jejunal feeding: a critical analysis of the literature. Arch Phys Med Rehabil 1990; 71:46-53

59. Burth G D, Shatney C H. Feeding jejunostomy (versus gastrostomy) passes the test of time. Am Surg 1987; 53:54-57

60. Ho C S, Yee A C N, McPherson R. Complications of surgical and percutaneous nonendoscopic gastrostomy. Review of 233 patients. Gastroenterology 1988; 95:206-210

61. Sreebny L M, Banoczy J, Edgar W M et al. Saliva: its role in health and disease. Int Dent J 1992; 42:291-304

62. Mandel I D. The function of saliva. J Dent Res 1987; 66:623~527 63. B~ickstr6m I, Funeg~rd U, Andersson I, Franzdn L, Johansson I.

Dietary intake in head and neck irradiated patients with perm~Jaent dry mouth symptoms. Oral Oncology, European Journal of Cancer 1995; 4:253-257

64. Atherton S T, White D J. Stomach as a source of bacteria colonising respiratory tract during artificial respiration. Lancet 1978; 1: 2968-2969

65. Heyland D, Mandell L A. Gastric colonisation of Gram-negative bacilli and nosocomial pneumonia in the intensive care unit patients. Chest 1992; 101:187-193

66. Toung T J K, Rosenfeld B A, Yoshiki Ae t al. Sncralfate does not reduce the risk of acid aspiration pneumonitis. Crit Care Med 1993; 21:1359-1364

67. Bonten M J M, Galliard C A, van Thiel F H et al. Continuous enteral feeding counteracts preventive measures for gastric colonisation in intensive care unit patients. Crit Care Med 1994; 22:939-944

68. Solem L D, Strate R G, Fischer R P. Antacid therapy and nutritional supplementation in the prevention of Curling's ulcers. Surgery, Gynecology, Obstetrics 1979; 148:367-370

69. Ciocon J O, Galindo-Ciocon D J, Tiessen Chet al. Continuous compared with intermittent tube feeding in elderly. Journal of Enteral and Parenteral Nutrition 1992; 16:525-528

70. Roongpisuthipong C, Heymsfield S B, Casper K et al. Continuous nasoeuteral feeding: inverse relation between infusion rate and serum levels of bilirubin. Journal of Enteral and Parenteral Nutrition 1987; 11:544-546

71. Whitmer D I, Gollan J L. Mechanisms and significance of fasting and dietary hyperbilirubinemia. Semin Liver Dis 1981; 3:42-51

152 PROGRESS IN PERIOPERATIVE ENTERAL TUBE FEEDING

72. Kirby D F, Delegge M H, Fleming C R. American Gastroenterological Association technical review on tube feeding for enteral nutrition. Gastroenterology 1995; 108:1282-1301

73. Edes T E, Walk B E, Austin J L. Diarrhoea in tube-fed patients: feeding formula not necessarily the cause. Am J Med 1990; 88:91-93

74. Gottschlich M M, Warden G D, Michel M A e t al. Diarrhoea in tube-fed bum patients: incidence, etiology, nutritional impact and prevention. Journal of Enteral and Parenteral Nutrition 1988; 12:338-345

75. Kirksey T D, Moncrief J A, Pruitt B A. Gastrointestinal complications of bums. Am J Surg 1968; 116:627-633

76. Ruddell W S J, Losowsky M S. Severe diarrhoea due to small intestinal colonisation during cimetidine treatment. BMJ 1980; 281:273

77. Kelly T W J, Patrick M R, Hillman K M. Study of diarrhoea in critically ill patients. Crit Care Med 1983; 11:7-9

78. Morrissey J F, Barreras R F. Antacid therapy. N Engl J Med 1974; 290:550-554

79. McAlhany J C, Czaja A J, Pmitt P A. Antacid control of complications from acute gastroduodenal disease after bums. J Trauma 1976; 16:645-649

80. Chernoff R, Dean J A. Medical and nutritional aspects of intractable diarrhoea. J Am Diet Assoc 1980; 76:161-169

81. Jacobson E D, Faloon W W. Malabsorptive effects of neomycin in commonly used doses. JAMA 1961; 175:187-190

82. Tetracycline diarrhoea. Editorial. BMJ 1968; 4:402 83. Wolfson A M, Saour J N, Ricketts C R et al. Prolonged nasogastric

tube feeding in critically ill and surgical patients. Postgrad Med J 1976; 52:678~82

84. Broom J, Jones K. Causes and prevention of diarrhoea in patients receiving enteral nutritional support. Journal of Human Nutrition 1981; 35:123-127

85. Levine H G, Lamont J T. Diarrhoea we can treat: antibiotic- associated colitis. Compr Ther 1982; 8:36-43

86. Benson D W, Griggs B A, Hamjilton F. et al. Clogging of feeding tubes: a randomised trial of a newly designed tube. Nutr Clin Pract 1990; 5:107-110

87. Silk D B A, Bray M J, Keele A M, et al. Clinical evaluation of a newly designed nasogastric enteral feeding tube. Gut: to be published

88. Marcuard S P, Stegall K L, Trogdon S. Clearing obstructed feeding tubes. Journal of Enteral and Parenteral Nutrition 1989; 13:81-83

89. Gora M L, Tschampel M M, Visconti J A. Considerations of drug therapy in patients receiving enteral nutrition. Nutr Clin Pract 1989; 4:105-110

90. Marcuard S P, Perkins A M. Clogging of feeding tubes. Journal of Enteral and Parenteral Nutrition 1988; 12:403-405

91. Bussy V, Marechal F, Nasca S. Microbial contamination of enteral feeding tubes occurring during nutritional treatment. Journal of Enteral and Parenteral Nutrition 1992; 16:552-557

92. Venezia R A, Agresto M D, Hanley E Met al. Nosocomial legionellosis associated with aspiration of nasogastric feedings diluted in tap water. Infect Control Hosp Epidemiol 1994; 15: 529-533

93. Hemandez-Duquino H, Rosenberg F A. Antibiotic resistance Pseudomonas in bottled drinking water. Can J Microbiol 1987; 33:286-289

94. Rosenberg F A, Hemandez-Duquino H. Antibiotic resistance Pseudomonas from German mineral waters. Toxic Assess 1989; 4:281-294

95. McWhirter J P, Pennington C R. Incidence and recognition of malnutrition in hospital. BMJ 1994; 308:945-948

96. Gallagher-Allred C R, Voss A C, Finn S C et al. Malnutrition and clinical outcomes: the case for medical nutrition therapy. J Am Diet Assoc 1996; 96: 361-366, 369

97. Moss G. Postoperative enteral hyperalimentation results in earlier elevation of serum branched-chain amino acid levels. Am J Surg 1994; 168:33-35

98. Zarling E J, Parmer J R, Mobarhan S et al. Effect of enteral formula infusion rate, osmolarity and chemical composition upon clinical tolerance and carbohydrate absorption in normal subjects. Journal of Enteral and Parenteral Nutrition 1986; 10:588-590

99. Coombs J, Wellman N. Nutrition as prevention: beyond an apple a day. HMO Magazine 1994; 25:23-28

100. Baskin W N. Advances in enteral nutrition techniques. Am J Gastroentero187: 1992; 1547-1553

Submission date: 18 November 1997 Accepted: 12 February 1998