Introduction

Providing adequate nutrition and maintenance of body condition is a key component to improving patient outcome in hospitalized patients.¹ Enteral nutrition is preferentially used in patients whose gastrointestinal function permits digestion and absorption of food but whose caloric ingestion is inadequate.¹ Advantages include maintenance of intestinal structure and function, decreased permeability to endotoxins, and reduced cost and decreased risk when compared with total parental nutrition.¹ Nasoenteric (nasogastric [NG] or nasoesophageal) tube placement is a commonly used technique in both human and veterinary patients to provide enteral nutrition.

Within the human medical literature, various complications associated with blind misplacement of narrow-bore nasoenteric tubes into the respiratory tree have been reported.² These complications include pneumonia, acute respiratory failure, pneumothorax, tracheal perforations, and death.^2,3 Fatalities as a result of nasoenteric tube misplacement are uncommon but pose a potential risk in both human and veterinary patients. A large study reviewing complications of blindly placed narrow-bore nasoenteric tubes in human patients reported that, of 187 misplacements into the tracheobronchial tree, 35 resulted in pneumothorax.² Pneumothorax was identified as the most common cause of death secondary to tracheobronchial misplacement with five out of the eight total deaths reported being a direct result of pneumothoraces.² Other causes of death reported in the human literature include sepsis secondary to alimentation delivered into the pleural space, respiratory failure during placement, and fatal hemorrhagic shock due to esophageal perforation.^2,4

Similar complications associated with nasoenteral tubes have been reported in limited numbers in the veterinary literature, including misplacement into the tracheobronchial tree, vomiting, regurgitation, aspiration pneumonia, epistaxis, dacryocystitis, rhinitis, pneumomediastinum, hemothorax, isocalothorax, and subcutaneous emphysema.^5,6 Although multiple complications have been described, to the best of our knowledge, no fatalities have been documented directly due to misplacement of nasoenteric tubes in the tracheobronchial tree in veterinary patients.^5,6 A tension pneumothorax secondary to nasojejunal feeding tube misplacement in a canine patient who ultimately died has been described; however the cause of death in this case was suggested to be due to nonrespiratory complications of the dog’s underlying carcinoma diagnosis that may have led to multi-organ failure.⁷

This case report describes fatal complications attributed to misplacement of NG tubes in two intubated veterinary patients and reviews potential options for earlier detection of nasoenteric tube misplacement in order to reduce the risk of future fatalities.

Case Report

Case 2

A 9 yr old male neutered Pomeranian presented to the internal medicine service at an academic teaching hospital for an exploratory laparotomy including a partial pancreatectomy, splenectomy, and potential cholecystectomy and cystotomy. With the intent to provide postoperative nutritional support, an NG tube^a was placed while the animal was under general anesthesia following the conclusion of the surgery and before recovery.

Standard placement of the NG tube was followed by an immediate postplacement left lateral radiograph of the thorax (Figure 1). The preliminary radiographic report concluded that the NG tube was correctly placed in the esophagus (this was later found to be incorrect). The weighted tip of the NG tube was thought to be superimposed with the region of the esophagus, extending caudodorsally before turning cranioventrally and summating with the stomach. In the cervical region, small gas bubbles were detected within the esophagus, dorsal to the endotracheal and NG tube. Both lungs were moderately hypoinflated with corresponding mild diffuse increase in unstructured pulmonary opacification. Upon recovery from surgery, the dog’s oral mucous membranes became cyanotic, which improved with oxygen supplementation. The dog was placed in an oxygen cage overnight. The following day, feeding via the NG tube was initiated with a Perative^b continuous rate infusion at one-third the daily resting energy requirement.

View Figure

• Download as Powerpoint
• Download Figure

Thirty-six hours following initiation of feeding, the dog was found to be mildly tachypneic with increased respiratory effort and abdominal distension. Aspiration of the NG tube revealed 40 mL of pink tinged fluid, and the Perative CRI rate was decreased. The dog’s condition continued to deteriorate, and over the next 4 hr, the dog became tachycardic and tachypneic with increased respiratory effort. Mild to moderate amounts of pleural fluid were noted bilaterally on point-of-care thoracic ultrasound. Cytology of the pleural fluid revealed a large number of neutrophils, macrophages, and mesothelial cells with no observed bacteria. Because of the presumptive fluid overload, IV fluids and feedings via the NG tube were discontinued at this time and furosemide^c (2 mg/kg IV) was administered, but no improvement was observed.

Three-view thoracic radiographs revealed tracheobronchial placement of the NG tube with mild to moderate pneumohydrothorax, pneumoperitoneum consistent with the previous celiotomy, central venous catheter placement, and, given the patient’s history, presumptive right caudal lung lobe perforation and isocalothorax (Figure 2). Bilateral thoracostomy tubes were placed. Removal of the NG tube at this time led to a continuous pneumothorax, which was managed with vacuum-assisted drainage of the thoracostomy tubes.

View Figure

• Download as Powerpoint
• Download Figure

Four days after initial presentation, the patient’s condition declined further with recurrence of increased respiratory rate and effort. A focused thoracic ultrasound found no evidence of pleural effusion and a positive glide sign, making pneumothorax unlikely, but numerous B-lines (indicating interstitial edema) were observed bilaterally. The dog suffered cardiac arrest, and cardiopulmonary resuscitation was attempted but discontinued at the owner’s request.

Discussion

Within this large academic teaching hospital, nasoenteric feeding tubes are commonly used because of their low cost and efficiency at providing enteral nutrition to patients. Over the past year, 231 NG tubes have been placed at our institution, of which 2 have resulted in fatalities. Although the fatality rate is low (less than 1%), it is still significant given that these deaths may have been prevented if other safety measures had been in place.

The first case in this report describes an intraoperative misplacement fatality. A severe pneumothorax developed acutely after removal of an unsuccessful blind NG tube placement. This likely occurred because when the misplaced NG tube was removed, a defect was left in the respiratory tree that was no longer plugged by the tube. Upon the tube’s removal, air entered the pleural space, leading to subsequent formation of a pneumothorax. Ultimately, despite placement of a thoracostomy tube following NG tube removal and vacuum assisted suction, the patient’s pneumothorax was unable to be adequately managed, leading to poor ventilation and cardiopulmonary arrest. Intraoperative use of laryngoscopy for nasoenteric tube placement in anesthetized patients allows direct visual confirmation of placement within the esophagus, rather than through the rima glottidis, before being fully advanced into the stomach. If the tube is found to be inadvertently entering the trachea, it can be retracted and redirected into the esophagus before being advanced to a distance that could cause pulmonary perforation. Once esophageal intubation is confirmed, direct palpation of the NG tube at the lower esophageal sphincter can confirm correct placement within the stomach. Since this mortality, our hospital has instituted mandatory direct visualization of the larynx during intraoperative placement of NG tubes.

In contrast to case 1, case 2 represents an example of a nonintraoperative misplacement fatality. When the patient’s respiratory condition began to deteriorate, enteral feeding and IV fluids were immediately discontinued, and medications were administered in attempt to alleviate the pleural effusion. Because the patient was refractory to treatment, thoracocentesis was performed and follow-up three-view thoracic radiographs were obtained to further investigate the cause of the respiratory decline (Figure 2). Upon reviewing the thoracic radiographs, it became evident that the NG tube had been misplaced in the pleural space. Bilateral thoracostomy tube placement was performed before removal of the NG tube in an attempt to prevent an acute tension pneumothorax. The order of operations for handling the misplacement of the NG tube was a key detail taken from the events in case 1. Ensuring placement of one or multiple thoracostomy tubes before removal of the NG tube was critical because of the risk of rapid and potentially fatal development of tension pneumothorax.

Despite these efforts, the dog in case 2 succumbed to complications from the feeding tube misplacement. Although the pleural effusion and pneumothorax were able to be successfully managed, the patient experienced continued respiratory compromise with worsening of B-lines. B-lines are hyperechoic comet tail artifacts that indicate widening of the interlobular septa by fluid accumulation, inflammation, or fibrosis, as seen with cardiogenic pulmonary edema, acute respiratory distress syndrome, interstitial lung diseases, or pneumonia.⁸ Reported causes of death of patients with isocalothorax in the human medical literature largely include septic pleural effusion and empyema.⁹ Given the sterile nature of the pleural effusion before thoracostomy tube placement and lack of pleural effusion at the time of worsening respiratory distress, pleural effusion was unlikely to be the major factor in our patient. Although we cannot entirely rule out a pneumothorax, lung ultrasound has superior diagnostic sensitivity to thoracic radiographs in four recently published meta-analyses in human medicine and one report in veterinary medicine.^10–14 We suspect that significant inflammation secondary to isocalothorax-induced pleuritis and pneumonitis may have led to this patient developing acute respiratory distress syndrome and ultimately undergoing cardiopulmonary arrest.

The cases in this report highlight the trends observed in human medicine that being mechanically ventilated or intubated increases the risk of nasoenteric tube misplacement. This contradicts the popularly regarded mistruth that the presence of a cuffed endotracheal (ET) tube makes it virtually impossible to misplace the NG tube into the respiratory tract.^2,15 In a large review of 9931 nasoenteric tubes blindly placed in humans, 60% of tracheobronchial misplacements occurred in patients who were mechanically ventilated with a cuffed ET or tracheostomy tube in place. It is thought that the combination of altered mentation with subsequent lack of gag reflex, the presence of the ET tube itself preventing normal glottic closure and inhibiting swallowing, and the low-pressure high-compliance ET tube cuff allow styletted feeding tubes to easily bypass the larynx and ET tube.^2,15 Therefore, clinicians should be remain vigilant when placing nasoenteric tubes in intubated patients.^9,15 Additional risk factors for tracheobronchial misplacement of nasoenteric tubes in people include altered gag or cough reflex and multiple attempts at placing the tube.^2,15

Three cases of iatrogenic pneumothorax associated with nasoenteric tube misplacement have previously been reported in the veterinary literature.^7,16 One dog was mechanically ventilated, one was sedated, and one was obtunded. In all cases, multiple unsuccessful blind attempts were made to place the tube. All three dogs developed pneumothorax requiring thoracocentesis with one undergoing cardiopulmonary arrest with successful resuscitation. In all three of these cases, techniques used to confirm appropriate placement included radiographic evaluation, aspiration of the feeding tube for negative pressure, and palpation of the feeding tube in the stomach in the dog undergoing a celiotomy. Unfortunately, these techniques failed to detect tracheobronchial misplacement before significant damage and pulmonary perforation occurred.

Previously described techniques to confirm appropriate nasoenteric tube positioning such as abdominal auscultation of an air bolus, evaluation of coughing during tube placement or after saline flushing, and aspiration of fluid to assess pH have inherent disadvantages and lack reliability.^17,18 The current gold standard in human pediatric patients to determine correct nasoenteric tube placement is a properly taken and interpreted radiograph.¹⁹ However, because of numerous factors including cost, logistics, and variability in technique, it is often not the first-line method used for verification.¹⁹ Within our hospital, the standard for verification of nasoenteric tube placement was a single lateral radiograph of the neck and thorax. We have since moved away from this and implemented two-view orthogonal radiographs in the absence of direct intraoperative palpation of the tube in the stomach as a potential method to help prevent future fatalities from radiographic misinterpretation, such as with case 2. Orthogonal views allow for more accurate triangulation of the nasoenteric tube in space and an ability to better confirm its location. In addition to this, we have implemented a rule that nasoenteric feedings may not commence until the placement radiographs have been reviewed by two radiologists—“no feeds before two reads.” Unfortunately, although this change helps us more accurately detect tracheobronchial misplacement and pulmonary perforation, if performed when the tube has been fully advanced, it can merely identify misplacement after it has occurred, rather than prevent this potentially devastating complication.

Nasoenteric placement safety would be improved by the ability to detect tracheobronchial misplacement before the tube has been advanced far enough to cause pulmonary perforation. Additional novel methods that detect tracheobronchial misplacement earlier in the placement process in veterinary patients include the use of capnography and two-checkpoint insertions.^6,18 Capnography uses the physiologic distinctions in carbon dioxide concentrations in the respiratory system as opposed to the gastrointestinal system. If the NG tube is properly placed, a capnographic waveform will be initially present when passing the NG tube through the nares, nasopharynx, and pharynx but should be absent once placed into the esophagus and further into the alimentary tract. In contrast, a capnographic waveform would be present throughout the entire insertion process if it were being placed into the tracheobronchial tree.¹⁸ Although this method has been validated in healthy dogs, it has not yet been evaluated clinically.

The use of two-checkpoint insertions has also been discussed as a novel method of checking NG tube placement. Aspiration of the tube for negative pressure is performed within specific anatomical landmarks before being fully inserted to its distal endpoint. If the tube is correctly placed in the esophagus, there should be negative pressure when aspirated, and the NG tube can be advanced distally. If incorrectly placed into the tracheobronchial tree it will result in continuous suction when aspirated and can be retracted before traveling into the smaller airways, potentially resulting in pulmonary perforation.⁶ The effectiveness of this method is yet to be prospectively studied. In addition, two previous case reports noted negative pressure with placement in the respiratory tree presumably due to the tube lying against the tracheal wall or wedged into a bronchus.^7,16

Conclusion

Enteral nutrition is a critical pillar of patient care that helps facilitate healing.¹ Nasoenteric tubes are commonly used in veterinary medicine, and although they are cost-effective and facilitate an easier administration of nutrients, they are not without risk. The fatalities discussed within this case series demonstrate that deaths as a result of nasoenteric tube misplacement into the tracheobronchial tree occur in veterinary patients. However, it is likely that this risk can be mitigated with a few simple changes to standard operating procedure.