Abstract

Enteral feeding is now standard and routine practice in intensive care. The use of a nasogastric tube for enteral feeding is generally considered to be safe, but tubes with small bores can sometimes lead to aspiration or passage clogging when malpositioned in sedated patients who are on long-term mechanical ventilation. Thus, accurate confirmation of correct placement is mandatory in such patients. This is not always the case, but this faulty practice can lead to serious complications in the absence of potential bezoar-forming medicines or gastrointestinal pathology. We present here one such interesting case of a patient who developed esophageal bezoar due to a malpositioned nasogastric tube for administering a casein-containing feed. In addition, we present a review of the literature.

Keywords

bezoars; enteral nutrition; esophagus; nasogastric tube;

1. Introduction

Enteral feeding through a nasogastric tube (NGT) is standard routine practice in patients in intensive care because of its ease, cost-effectiveness, convenience, and established safety profile.1, 2, 3, 4 The use of an NGT for feeding is generally considered to be safe, but silicone NGTs of smaller bores, used as a standard tool for reducing gastroesophageal reflux, can sometimes lead to aspiration or passage clogging if malpositioned.1, 2, 3, 5 Thus, accurate confirmation of correct placement is mandatory before their use, especially in sedated patients under assisted ventilation.1, 2 However, this practice is not always followed, in spite of serious complications.2, 5 We present here one such rare interesting case of a patient who developed an esophageal bezoar due to malpositioning on an NGT used for administering a casein-containing feed, without any gastrointestinal pathology being present, or any ingestion of bezoar-forming medicines on the part of the patient.

2. Case report

A 20-year-old man with a provisional diagnosis of meningoencephalitis was transferred to our intensive care unit (ICU) for further management. He had recently had sustained a tonic clonic seizure and had been resuscitated after cardiac arrest for 20 minutes. At the end of treatment, the patient had received tracheal intubation for assisted ventilation and was being sedated with morphine and midazolam infusions.

An NGT (14 French gauge; Flexiflo; Pennine Health Care, London, UK) was inserted to start both oral medications and enteral feeding. The position of the NGT was confirmed using the traditional auscultatory method, and both the treating doctor and the assisting nurse were satisfied with the result. The patient was then started on enteral feeding (Ensure; Abbott, Columbus, Ohio, USA) in accordance with our ICU protocol. As his family refused a diagnostic lumbar puncture, he was treated conservatively with piperacillin and tazobactam, acyclovir along with sodium valproate, levetiracetam, thiamine, and omeprazole.

He responded favorably to the conservative management, but we could not wean him from mechanical ventilation as he was unable to wake up completely even after sedation had been stopped. This was possibly due to hypoxic brain damage incurred during the cardiac arrest, which was confirmed by magnetic resonance imaging. Thus, the patient underwent a tracheostomy, and the weaning process was started. He developed ventilator-associated pneumonia which responded to meropenem, and we were able to wean him from the ventilator about 2 months later. His family refused insertion of gastrostomy tube for long-term feeding. Thus NGT feeding was continued, and the patient was transferred to the general ward with a tracheostomy tube in place.

After another 2 weeks, the patient developed aspiration pneumonia (confirmed radiologically) and was readmitted to ICU. He was again put on a ventilator under sedation. In the meantime, the on-duty nurse found that the NGT was blocked. We met with some resistance while removing the blocked NGT, but it was finally pulled out with some force. The NGT was found to be completely intact but was blocked internally. Insertion of a fresh NGT was difficult due to resistance in the middle of the esophagus; even the senior anaesthesiologist failed to insert one using conventional methods. As we reviewed the previous chest X-ray, we found that the tip of the old NGT was anchored just above the gastroesophageal junction (Fig. 1). We then consulted the gastroenterologist for endoscopic evaluation of the upper gastrointestinal tract.

Download full-size image

Fig. 1. X-ray showing the tip of the nasogastric tube lying in the distal esophagus.

The endoscope could only be advanced up to 20 cm. Further advancement was not possible due to complete obstruction of the esophagus by some solid white material (Fig. 2). The consulted gastroenterologist tried to remove this to clear the passage but failed. He then decided to take it out piecemeal and to wash the esophagus with sodium bicarbonate solution. This resulted in successful clearing of the esophagus although with great difficulty. The stomach was found to be completely free of obstruction, and a fresh NGT (14 Fr) was easily passed under endoscopy. This time, correct placement of the NGT was confirmed radiologically.

Download full-size image

Fig. 2. Endoscopic image showing blockage of the esophagus by the bezoar and clearing of the same after washout.

The material that had been removed was sent for pathological analysis. This showed it to be a mixture of amorphous materials formed from bacteria, foci of calcification, and food particles, along with squamous hyperplasia. We hypothesized that the white solid material was a bezoar formed due to solidification of stagnant enteral feeds in combination with medicines. We were finally able to wean the patient from the ventilator and transfer him to the local hospital for palliative care as he no longer required tertiary care.

3. Discussion

Enteral feeding is used in critically ill patients because of its cost-effectiveness, convenience, and safety profile.1, 2, 3, 4 The use of an NGT is generally considered to be relatively safe.1, 2, 5 However, the thinner bore NGTs, which are now commonly used to reduce esophageal reflux, can lead to aspiration or passage clogging due to pulmonary insertion and esophageal malpositioning of the NGT in sedated, intubated, or tracheostomized patients who have a decreased gag reflex and are on long-term mechanical ventilation.1, 2, 3, 5 Thus, confirmation of correct positioning of the NGT is mandatory before use.1, 2

Commonly used epigastric auscultation does not always ensure correct positioning and has been proved to be misleading as sound can be transmitted even if the NGT is in the distal esophagus or the lung.2, 5 Most clinicians have thus recommended abandoning this method.² Chemical testing of the NGT aspirate (pH or bilirubin) has also not been found to be foolproof to show correct placement.2, 5 Radiographic confirmation is still considered to be the gold standard.2, 5 Even so, this is not routinely performed in many centers because of cost, exposure risk, and delay.2, 5 Correct reporting of X-rays and visualization of the NGT tip are also problematic.² However, even after radiographic confirmation, soft and smaller bore NGTs have been reported to become malpositioned in agitated and confused patients during bouts of retching, vomiting, or coughing.1, 5 Thus regular rechecking and the use of more than one method of confirmation have been recommended.2, 5

Another rare but potentially serious complication of enteral feeding is bezoar formation.³ Bezoars are concretions of partly or totally undigested food or other ingested materials that are mostly found to form in stomach or upper small intestine, patients often present with signs and symptoms consistent with gastrointestinal obstruction.4, 5, 6, 7 Depending on their source of origin, bezoars are classified as phytobezoars (resulting from undigested food materials, the commonest type), trichobezoars (formed by ingested hairs), lactobezoars (seen in neonates fed on improper formula feed), and pharmacobezoars (resulting from undigested medicines such as cholestyramine, aluminium-containing antacids, sucralfate, enteric-coated nifedipine and aspirin, bulk laxatives, and enteral feed).3, 6, 7

Bezoars are rarely reported to form in the esophagus in patients with structural or functional anomaly.⁴ Solidification of enteral feed in presence of concurrent bezoar-forming medicines in patients with a structural and/or functional anomaly of the esophagus has been reported as the main reason.⁴ However, this is very rarely reported in patients with a normal esophagus who are not taking any such bezoar-forming medicines.³

It is postulated that casein-containing enteral feeds, such as Ensure, could solidify in the presence of refluxed acidic gastric juice (pH less than 5) that is deficient in pepsin and pancreatic enzymes, which regurgitates in patients who are on prolonged mechanical ventilation along with sedative analgesic drugs known to hamper gastrointestinal motility.⁴ This phenomenon is not seen with non-casein-containing feeds.⁴ This type of complication can be prevented by feeding the high-risk patient in the semi-recumbent position, preferably with a non-casein-containing feed, and avoiding the concurrent use of bezoar-forming medications, except for prokinetic and gastric acid-reducing agents.⁴

Although upper gastrointestinal bezoars can be identified radiologically, endoscopy is considered to be the best diagnostic as well as therapeutic tool.3, 4 Bezoars can be removed endoscopically by dissolution, fragmentation, and washout using different forceps, snares, lytic enzymes (papain, N-acetylcysteine, cellulose), and effervescent solutions (Coca-Cola, sodium bicarbonate).3, 4, 6 Surgery is now very rarely used.³

In our case, formation of the esophageal bezoar was facilitated by feeding a casein-containing enteral feed through an NGT whose tip was unfortunately malpositioned in the distal esophagus. Initially, the NGT position was checked by two on-duty professionals who were satisfied with the placement and had the NGT fixed at an external naris at the 50 cm mark. We used a smaller bore NGT in line with the current protocol to reduce gastroesophageal reflux. Because these thinner bore NGTs are extremely flexible, they are more prone to malpositioning, leading to such complications as happened in our case.

In our patient, the position of the NGT was confirmed neither radiologically nor by an aspiration test, which was a grave fault on the part of our care providers. As the patient was well adjusted to such feeding during his first stay in ICU, nobody paid any attention to this trivial yet important issue. We were unable to track what happened on the ward after his return with a tracheostomy. When he was again admitted to ICU, where the nurses routinely checked the position of the NGT and found that it was blocked, with the sticking point at the 40 cm mark at the naris. As the NGT was found to be blocked, it was removed and an attempt was tried to insert a new one for feeding. It was at this juncture that the distal esophagus was also found to be blocked.

In our case, it was possible that the residual enteral feeds that stagnated in the NGT in the distal esophagus had solidified after reacting with the regurgitated acidic gastric juice. Usually, when the NGT is appropriately placed in the stomach, the enteral feeds, if immovable, would stagnate only in the stomach. If bezoars have formed, which usually occurs in the presence of gastric pathology, these will pile up in the stomach or move downward into the intestine. A correctly placed NGT rarely leads to an accumulation of feeds in the distal esophagus unless there is gastroesophageal pathology.

In our case, malposition of the NGT in the esophagus was confirmed, leading to an accumulation of feeds in the distal esophagus that afterward solidified in the presence of refluxed gastric juice, as is commonly seen in recumbently ventilated patients. This is possibly facilitated by the use of high doses of sedatives and analgesics. Our patient was initially given omeprazole, which was later stopped once he had become accustomed to NGT feeding. None of the other drugs used in our patient has been implicated in bezoar formation.

In our case, the actuality of whether the NGT was malpositioned initially or displaced later remained debatable. Both wrong placement at the start and later displacement are common occurrences in such patients. The latter is more logical as the NGT obstruction occurred before readmission to ICU. A major mistake in our procedure was our failure to use X-ray or a second method to confirm the correct position of the NGT during the initial placement or at the time of readmission. This reminds us that confirmation of correct placement of an NGT cannot rely solely on one method, and the position of the NGT should always be confirmed, preferably radiologically, before starting feeding.