The History of Nonsurgical Enteral Tube Feeding AccessDepartment of Surgery, Medical College of Georgia, Augusta, Georgia Correspondence: Gail Cresci, MS, RD, CNSD, LD, Medical College of Georgia, Department of Surgery, Room 4072, 1120 15th Street, Augusta, GA 30912. Electronic mail may be sent to gcresci{at}mcg.edu. Specialized nutrition support, particularly enteral feeding, has been used for centuries. Technologic advancements have affected the provision of enteral feeding. Feeding solutions and devices, as well as the techniques to place the feeding devices, have evolved. This article reviews the history of bedside placement methods for short-term enteral access devices. Over the past several decades, most published guidelines regarding nutrition support in the critically ill recommend the use of enteral nutrition (EN) rather than parenteral nutrition (PN), according to an accumulating body of evidence suggesting EN is associated with better clinical outcomes than PN.1–6 However, long before these technical methods of providing nutrition to those who cannot self-consume adequate nutrients were developed and fine tuned, EN was provided via various techniques. This article will focus on the history of bedside placement methods for short-term enteral access devices.
Reports of gastric feeding date as far back as the 16th century. Over the years, many different devices were used to deliver the feedings. The most popular device was a long tube with a funnel or syringe attached to the outside end.7 In 1617, Fabricius ab Aquapendente reported the use of a silver tube passed through the nostril into the nasopharynx for feeding patients with tetanus.8 Fabricated flexible leather catheters were introduced into patients' stomachs for feedings in 1646.9 In the 1790s, Hunter10 described the use of a hollow tube introduced into the stomach for feeding. He successfully treated a patient with paralysis of the muscles of deglutition by using a tube made of a whalebone probe covered with an eel skin and attached to a bladder. The eel skin–covered probe was placed orogastrically, and the suggested feedings were jellies, whipped eggs, sugar, milk, or wine. The stomach pump, which became the first consistent method of intragastric feeding, was initially provided for gastric lavage and gastric decompression.8 In 1867, Kussmaul introduced a flexible orogastric tube for gastric decompression, and in 1874, Ewald and Oser introduced the soft rubber tube for gastric intubation.8 The stomach pump was used for feeding severely mentally ill English patients during the first half of the 19th century. Complications associated with hypopharyngeal feedings were noted by Reeve11 in 1851 and included stomach lacerations and drowning with beef broth. He perceived that "the circumstances of forcible injection of food or medicine into the stomach requires patience; it is always a troublesome task and if confided to ignorant keepers is also a perilous one."
Gastric Feeding Tube Insertion Methods To date, the NGT remains the most-used short-term enteral feeding access due to its simplicity and low cost. The tubes vary in diameter size (large bore, >14-Fr; small bore, 8–12-Fr). The larger tubes are made of a stiff polyvinyl material and have less likelihood of being malpositioned, lower incidence of clogging, and are easier and more reliable when aspirating to check for residuals.15 These larger-bore tubes are typically inserted at the patient's bedside by a nurse.
Hippocrates was the first physician documented to use rectal feedings when gastric feeding was not feasible or tolerated. Some practitioners believed that colonic absorption through reverse peristalsis could provide a person's nutrition needs.7 Rectal feeding devices evolved from a piece of pipe with a bladder tied to one end, used by Hippocrates, to long pieces of rubber tubing attached to funnels or wooden syringes.11 In 1891, adequate absorption was claimed by slowly infusing fluids with no return through a rectally inserted rubber tube that was one-eighth inch in diameter and 1-  feet
long.16 Advocates of rectal feeding claimed that providing a piece of pancreatic gland in the feeding solution was necessary. In 1878, Kaufman claimed that a normal life could be lived for up to 9 months when rectally fed an enemata of eggs, milk, beef broth, and pancreatic gland.17 In 1878, a case report of rectal feedings in a patient with esophageal stenosis claimed adequate provision of fluids and nutrients by giving food enemas pushed into the rectum twice daily with a wooden syringe consisting of fat-free raw beef without connective tissue and hog's pancreas.18 Other foods given as enematics included tobacco, meat mixed with wax and starch, brandy, and red wine.7 Defibrinated blood was sometimes used in rectal feeding, but it was reported that rapid putrefaction of the blood and rectal irritation occurred, thus limiting its use.7,17 One of the most famous reports of rectal feeding was that of President Garfield in 1882.19 President Garfield was rectally infused with peptonized beef broth, beef peptonoid, and whiskey every 4 hours for almost 79 days after receiving a gunshot wound. Up until 1940, the rectal route was still used for provision of water, saline, and glucose solutions.7
A major advance in enteral feeding was introduced in 1910 by Einhorn.20 He noted that rectal feeding, provided when oral or gastric feeding was not possible, was unsatisfactory due to rectal irritation and inadequately used infused materials. Einhorn engaged a tube he had developed as a pump for sampling duodenal contents for feeding into the duodenum or proximal jejunum.8 The tube was rubber and had a 10- to 12-g metal weight attached to the end. The tubing was introduced into the stomach, and feeding was begun as soon as it was in the duodenum. The tube was left in place for 8–12 days and was more comfortable for the patient compared with previously used tubes. Patients were fed 240- to 300-mL solutions consisting of milk, raw egg, and lactose every 2 hours, in addition to rectally infusing a quart of physiologic salt solution.7,8 After following 3 patients, Einhorn concluded that duodenal feeding had many advantages over rectal feeding, stating the rectum and colon were simply organs for the expulsion of feces and for the absorption of possibly remaining liquids; the duodenum was the organ of important secreted digestive juices.8 In the early 1900s, many other physicians adopted Einhorn's technique, with modifications, including Morgan21 and Jones.22 Jones noticed that some patients did not tolerate bolus feedings and recommended a gradual volume increase in feedings, along with modifications in the foodstuffs delivered. During this time, scientific understanding of nutrient absorption increased and feeding delivery accompanied this, with alterations in the formulations provided. In 1915, Gross and Held23 introduced their own tube. It was made of rubber, had a 7-mm diameter, and included a 10- to 11-g ball on the distal end. Patients with a gastric or duodenal ulcer were placed in a chair, and the tube was introduced while the patient was in this sitting position. He reported that with the assistance of gravity, it took 15–20 minutes for the tube to reach the duodenum compared with Einhorn's tube taking 3–12 hours, with the sole use of the stomach's propulsive action. Jejunal feeding was introduced in 1918 when Andersen reported immediate feeding, on the operating room table, after gastroenterostomy.24 He passed the Rehfuss gastroduodenal tube at the time of surgery through the nares until the distal end was well into the jejunum after completing the gastroenterostomy. Jejunal feedings consisted of 200 mL of milk, 15 g of dextrose, and 8 mL of whiskey at 2-hour intervals. Neither rectal infusions nor IV fluids were necessary to maintain fluid balance in these patients. Refinement of the nasojejunal feeding tube and solution can be credited to 2 groups. In 1939, Abbot and Rawson25 constructed a double-lumen tube with 1 opening in the stomach for gastric suction and 1 opening in the jejunum for feeding. Also in 1939, Stengel and Ravdin26 described an orojejunal method of feeding for maintenance of nutrition in surgical patients. The device consisted of 2 tubes tied together and passed nasogastrically, the longer tube being passed well into the jejunum during surgery. The goal was for gastric emptying and simultaneous jejunal feeding. Later, a double-lumen tube was substituted for the 2 tubes tied together. Stengel and Ravdin26 strongly recommended a controlled delivery of feeding flow with a feeding machine. Later, Stengel and Ravdin26 adapted the Abbott tube, a double-lumen tube but without the valves originally proposed by Abbott. The next several decades led to the experimentation and implementation of surgically placed enteral feeding tubes, which will be described in detail in an upcoming issue. With the technologies of endoscopy and radiology came the development of new enteral feeding tubes and access methods. Below is a review of the various methods and techniques described over the past several decades for nonsurgically placed nasoenteric feeding tubes (NEFTs).
Bedside Spontaneous Methods The goal for gaining enteral access is for the method to be quick, effective, and inexpensive, with low morbidity. In attempts to achieve this goal, many bedside manual techniques of inserting NEFTs have been described, with variable success rates. Some use various types of tubes, medications, and manipulation techniques. To follow the original feeding tubes described by Einhorn and others, modern feeding tubes may or may not contain a weighted tube tip. Limited studies have evaluated the success rates of the weighted tubes spontaneously migrating postpylorically. Intuitively, as Gross and Held23 reported, one would think that a weighted tube would migrate and remain postpyloric more often than a nonweighted tube. However, when small-bore (8 Fr) weighted and nonweighted feeding tubes were inserted and allowed to advance spontaneously, the passage into the duodenum was low, with similar rates for both types of tubes.27 When the tube positions were checked radiographically, only 32% of the tubes were in the small bowel, and the nonweighted tubes remained postpylorical significantly longer than the weighted tubes. When larger tubes (10-Fr) were placed, there was not any difference in postpyloric spontaneous passage of weighted and nonweighted tubes (57% vs 67%, p = .27).28 Due to the lack of success, prokinetic agents such as metoclopramide and erythromycin have been used to facilitate spontaneous passage of feeding tubes from the stomach into the intestine. Lord et al29 showed that 84% of nonweighted tubes passed postpylorically within 4 hours of placement compared with 36% of weighted tubes (p = .002) when 20 mg of IV metoclopramide was given before insertion. Other studies investigating prokinetic agents have mixed results. Giving 10 mg IV metoclopramide 10 minutes immediately after tube insertion improved success rates in diabetic patients but not in other patient populations.30 After administering metoclopramide to critically ill ventilated patients, weighted tubes spontaneously passed into the duodenum in only 14% of the patients.31 Whatley et al32 showed improved postpyloric placement when 20 mg of metoclopramide was administered 10 minutes before tube insertion, whereas Heiselman et al33 found no benefit to giving this amount and timing of metoclopramide. Erythromycin provision was shown to significantly increase duodenal tube placement after 24 hours (75% vs 0%) when provided at a dose of 400 mg every 8 hours.34 Kalliafas et al35 found a difference in placement success within 30 minutes of tube insertion when patients were provided with IV erythromycin (61% vs 35%). Because the success rates for spontaneous passage, use of weighted tube tips, and prokinetic agents alone are not ideal, other methods for gaining timely and effective enteral access have been investigated.
Air-Insufflation Techniques Salasidis et al36 conducted a similar randomized trial in critically ill patients. Subjects were placed in a supine right lateral decubitus position. After inserting a 12-Fr weighted feeding tube into the fundus, 500 mL of air was injected into the stomach in a single bolus. The tube was then advanced its entire length, except for the last 10 cm. An abdominal radiograph was obtained within 2 hours of procedure completion. If the tube was not in the duodenum, then 10 mg IV metoclopramide was provided and a repeat radiograph was obtained the next day. On initial radiograph, 66% of tubes were postpyloric, increasing to 77% on repeat radiograph. The procedure was performed by a single investigator and on average took 4 minutes to actually insert the tube; however, the time is longer if not properly positioned on the first attempt. Schulz et al38 used a similar technique and reported an 88% success rate.
Other Variations
Novel Techniques Continuous stomach electrocardiogram (ECG), along with air insufflation and erythromycin provision, was described by Slagt et al44 for bedside NEFT placement. Forty consecutive critically ill patients were provided with 200 mg IV erythromycin 30 minutes before insertion of the feeding tube. The tube was inserted into the stomach, followed by 500 mL of air insufflation. ECG was performed, and during further insertion of the tube the QRS complex was continuously monitored for a change in polarity, suggesting passage across the midline through the pylorus. At the end of the procedure, aspirate was obtained and checked for an alkaline pH. Postpyloric placement was achieved in 65% of cases on first passage and 88% of cases on second passage, with a median time of 15 minutes (7–75 min). Change in QRS polarity had 94% sensitivity in predicting postpyloric tip placement, and of the 32 alkaline pH aspirates, 31 were postpyloric. Gabriel and Ackermann45 describe a technique using a 12-Fr polyurethane feeding tube that was modified by placing a small magnet in the distal tip. After inserting the tube through the nares into the esophagus, an external handheld magnet was used to draw the tube tip beyond the pyloric sphincter and potentially further into the small bowel. Placement was verified by plain abdominal radiograph. After 329 intubations, 89% were placed beyond the pyloric sphincter, with 42% in the distal portion of the duodenum or jejunum, in a mean procedure time of 15 minutes. A single physician performed all the placements.
Endoscopic and Radiologic Methods Fluoroscopic assistance has been widely used and historically has been the preferred method of NEFT placement in many centers when bedside maneuvers have been unsuccessful. Directed tube placement is performed under fluoroscopic guidance, using otherwise similar tubes and placement methodology to what is described above. Success rates of 53% for jejunal positioning have been documented with this approach, with a mean of 22 minutes of fluoroscopy room time being reported.46 Positioning in the distal duodenum has been achieved in 86% of cases, with reported aspiration rates of only 2% with such placement.47 Placement in the ICU setting under C-arm fluoroscopy is an option for critically ill patients. Bedside ultrasound guidance has more recently been described as an alternative to fluoroscopic assistance.48 This technique appears to have the advantage of being highly portable and involves no radiation exposure. Placement is done using a stylet in place and with intermittent saline or water injection, both of which facilitate ultrasonic tube localization. This technique has been reported to improve postpyloric placement rates substantially compared with standard bedside methods, including a jejunal placement rate of 42%. Average procedure time was 18 minutes.48 Endoscopic placement options can be grouped into several categories. Early attempts involved using a suture at the end of the NEFT, which was grasped with endoscopic biopsy forceps and used to direct the tube endoscopically into the desired location (the so-called pull technique).49 Overall success rates with this technique have been poor because tube positioning is often lost due to the NEFT's being pulled back as the endoscope is withdrawn. This can sometimes be avoided by adding fluoroscopic visualization to the endoscopic effort; however, the benefit of such an approach over simple fluoroscopic-guided placement is not established. Another method involves the Seldinger technique for tube positioning. In this approach, a guidewire is placed in the small intestine via the endoscopic biopsy channel.50 The endoscope is then withdrawn while maintaining guidewire position and an oronasal transfer performed, such as is used in the placement of a nasobiliary drain at the time of endoscopic retrograde cholangiopancreatography (ERCP). The NEFT is then passed over the wire. A variation on this technique involves passage of the NEFT into the stomach, after which a guidewire is passed through the same and directed into the small bowel using endoscopic biopsy forceps. Completion of tube placement is then accomplished by advancing the NEFT over the guidewire to the desired more distal location. Again, fluoroscopy may be beneficial with these techniques in preventing looping in the stomach or tube displacement as positioning is achieved or the endoscope withdrawn. A third approach involves simple stiffening of the NEFT with guidewire (typically a 0.052-inch or 2 0.035-inch wires), with simple manual advancement of the tube under endoscopic visualization.51 The stiffening may lessen the likelihood of inadvertent tube displacement as the endoscope is withdrawn, but the advantages of this method over fluoroscopic-guided placement, particularly given the typical requirement for conscious sedation with endoscopy, is not apparent. One exception to this might be the situation where upper endoscopy is indicated for other purposes and concomitant NEFT placement is desired. With the advent of smaller-caliber endoscopes for transnasal foregut endoscopy, the potential for NEFT placement with the above approaches may be possible without requiring oronasal transfer maneuvers or perhaps even conscious sedation.52 A prospective, randomized trial of transnasal endoscopic vs fluoroscopic placement in 100 consecutive patients has been reported in a critical care setting.53 Successful positioning was in the 90% range with both methods; however, the endoscopic procedure took an average of 12 minutes, compared with 19 minutes for the fluoroscopic approach (p < .001). Another approach to endoscopic placement involves use of a therapeutic gastroscope with larger biopsy channel. This allows placement of an 8- to 10-Fr NEFT via the biopsy channel after the endoscope has been advanced into the small bowel.54,55 The endoscope is then withdrawn over the tube. With this technique, the potential for the endoscope to displace the tube during withdrawal by drag in the esophagus, such as is mentioned above with other techniques above, may be less likely. With the increasing availability of new endoscopic tools that allow tissue fixation via the endoscopic biopsy channel, including clips and suturing techniques, novel approaches to NEFT placement with endoscopic assistance are beginning to be reported.56 A recent report involved using endoscopic positioning into the small bowel of a NEFT after the same was advanced into the stomach, using the pull technique described above. Once the greatest postpyloric position achievable had been reached endoscopically, an endoscopic clip was used to fix the suture on the end of the tube to the small bowel mucosa at that location. Successful placement was achieved in 100% of cases, the clip was able to be used for radiologic visualization of tube position, and no spontaneous tube migrations were reported in 21 patients. Such techniques may help overcome some of the challenges reported with endoscopic methods of tube placement in the past. Techniques have been used that achieve jejunal placement via either preexisting or synchronously placed percutaneous endoscopic gastrostomy tubes using some of the above endoscopic positioning techniques.57,58 These approaches can allow for simultaneous gastric decompression and jejunal nutrition access without requiring ongoing nasal or oral tube access. A more detailed review of surgical and percutaneous approaches to tube placement will be available in a separate article in the December issue.
The developments and advancements in specialized nutrition support have come a long way since the first reports of alternate feeding methods for those unable to consume adequate nutrients orally. Currently, there is no one ideal method for gaining enteral access at the bedside that suits every facility. The various methods described and their reported success rates are based upon the local expertise available. As technology advances and changes in healthcare continue, innovative methods of gaining enteral feeding access will undoubtedly follow as other members of the healthcare team (eg, dietitians, nurses) performing these functions continue to become involved.
Nutrition in Clinical Practice, Vol. 21, No. 5,
522-528 (2006) This article has been cited by other articles:
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
|
Top
Gastric Feeding
Rectal Feeding
