Introduction

Nasogastric tubes (NGTs) are widely used in human and veterinary medicine for enteral nutrition. NGTs are an inexpensive, convenient option for enteral feeding and usually do not require anesthesia for placement.¹ NGTs allow for gastric decompression, measurement of residual gastric volume to assess for gastrointestinal motility, reduced nausea (reported in humans), reduced risk of aspiration pneumonia, and administration of enteral nutrition and medications.^2,3 Despite these benefits, NGTs are associated with complications.³ The incidence of complications secondary to NGT placement in veterinary patients has been reported at 57%, including sneezing, discomfort, inability to advance the tube, pneumothorax, pneumonitis, and isocalothorax.^3–7 Isocalothorax is accumulation of pleural fluid caused by enteral feeding into the pleural space.⁸ To the authors’ knowledge, there has been one reported case of isocalothorax resulting from NGT misplacement in a dog, which proved fatal.⁶ Literature is sparse regarding isocalothorax in both human and veterinary medicine.^4–7,9–13

This case report describes three cases of isocalothorax and reviews surgical and medical management strategies. To the authors’ knowledge, this is the first veterinary report describing successful management of isocalothorax. The contemporary standard of care was provided to each animal described in this case report. All animal treatment was consistent with acceptable practices as described in AAHA policy statements.

Case 1

A 7 yr old, 3.56 kg, neutered male Pomeranian was hospitalized at an academic teaching hospital for daily wound care that necessitated sedation and had a styletted NGT placed on the third day of hospitalization because of anorexia. A styletted NGT^a was placed on day 3 because of anorexia. The patient was sedated with 11 mcg/kg medetomidine^b IV and 2 mg/kg propofol^c IV and a styletted NGT was placed by the attending veterinarian. The NGT was premeasured to the last rib and advanced to the premeasured mark. Three-view placement radiographs showed the NGT traversing the left bronchial tree before retroflexing at the diaphragm, but this was not noticed by the veterinarian (Figure 1A). Bolus feeding^d was initiated at 4 mL/kg every 4 hr (50% resting energy requirement per day). Approximately 15 hr later, the patient developed dyspnea and tachypnea. Thoracic radiographs revealed NGT misplacement, bilateral pleural effusion, and bilateral pneumothorax. There was concern that the fluid would cause sepsis, so a thoracostomy catheter^e was placed in the right hemithorax, and 40 mL of serosanguinous fluid and 20 mL of air were aspirated. Thoracotomy was then performed at the left sixth intercostal space. The pleural cavity contained a large amount of liquid diet and fibrinous pleuritis. The NGT was located exiting the caudal aspect of the left caudal lung lobe. Lavage and gentle debridement of the left hemithorax and a partial left caudal lung lobectomy were performed. The right hemithorax was mostly unaffected besides minimal pleural effusion, which was lavaged and suctioned. A second thoracostomy catheter was placed into the left hemithorax. Postoperative radiographs confirmed appropriate placement of the thoracostomy tubes and absence of pneumothorax or pleural effusion. The catheters were aspirated every 4 hr and left in place until negative suction was obtained from the left side and there was 2.5 mL/kg/day of fluid production on the right. No further lavage was performed. Septic pleuritis was suspected in the right hemithorax, as inflammatory cells and intracellular cocci were observed on cytology of aspirated fluid on the first day after surgery by the attending veterinarian. However, culture samples taken during surgery were negative, possibly owing to early initiation of parenteral antibiotics. The patient was treated supportively for 5 days and discharged. On recheck examination at 7 days, thoracic radiographs were normal, and he was clinically well.

View Figure

• Download as Powerpoint
• Download Figure

Case 2

A 7 yr old, 3.2 kg, female, spayed poodle was referred to an academic teaching hospital after a gastrotomy at a general practice clinic. The patient presented in shock. The left nostril was desensitized with proparacaine drops and an 8-Fr styletted NGT^a was premeasured to the last rib, lubricated, and advanced through the left nostril. It was met with resistance, removed, and replaced with a 5-Fr styletted NGT^f. The 5-Fr NGT was advanced through the left nostril to the premarked level, and the patient swallowed during placement. A small amount of fluid was aspirated before encountering negative pressure. A right lateral radiograph was taken, and the veterinarian mistakenly believed the tip of the NGT was within the gastric lumen (Figure 1B). Bolus feeding^g was initiated at 2 mL/kg every 4 hr (25% resting energy requirement) on the first day. After 16 hr in hospital, the patient was dull and developed moderate dyspnea with a normal respiratory rate. The patient was placed in an oxygen cage. NGT malpositioning was suspected. The patient also developed hypotension (80 mm Hg systolic arterial pressure, Doppler^h) and hypoglycemia (3.1 mmol/L, reference range 4.11–8.84 mmol/L, glucometerⁱ). An automated hemogram^j showed severe neutropenia (0.49 × 10⁹/L, reference range 2.95–11.64 × 10⁹/L) with a left shift, moderate lymphocytosis (7.52 × 10⁹/L, reference range 1.05–5.10 × 10⁹/L), mild monocytosis (1.31 × 10⁹/L, reference range 0.16–1.12 × 10⁹/L). A blood chemistry analysis^k revealed mild hypoalbuminemia (22 g/L, reference range 23–40 g/L). Mild hypothermia (36.2°C) was noted. These clinical findings indicated sepsis. Thoracic radiographs showed moderate pleural effusion, multilobar alveolar pattern, and bilateral interstitial pattern in the caudal lung lobes. The NGT was positioned in the mediastinum, with the esophagus superimposed over the trachea owing to obliquity (Figure 1C). Approximately 3 mL of opaque pink fluid was removed by thoracocentesis. On cytology, the aspirated fluid contained red blood cells, and the background of the smear was similar to a smear prepared of the liquid diet, bolstering suspicion of isocalothorax. Median sternotomy was elected to decontaminate the thorax. The thoracic cavity contained a large amount of thick, white material. The mediastinum was ruptured, and the mediastinal space was filled with white liquid. Perforations were found in the left caudal lung lobe and the caudal part of the left cranial lung lobe. Total left-sided pneumonectomy was performed. During debridement and thoracotomy, asystole developed four times and the patient was resuscitated with epinephrine, atropine, and direct cardiac massage. During closure of the surgical site, asystole developed again, and resuscitation was ultimately unsuccessful. No culture was sent out.

Case 3

A 1 yr old, 3.4 kg, female, intact British shorthair presented to an academic teaching hospital for suspected NGT misplacement by another clinic, where the patient was being treated for seizures secondary to suspected intoxication. The NGT had reportedly been placed without sedation, as the patient was obtunded. The NGT was premeasured to the last rib and blindly inserted, and a swallowing reflex was noted. A lateral radiograph showed the tip of the NGT at the level of the last rib. This was misinterpreted as correct placement. A total of 109 mL (32 mL/kg) enteral liquid^d was delivered over 36 hr q 4 hr at the referring clinic. Intermittent pyrexia, tachypnea, and dyspnea were noted. Thoracic radiographs revealed pleural effusion with right caudal lung lobe consolidation. The patient was referred for ongoing management. On presentation, the patient was dull, and an NGT was present in the left nostril. Oxygen saturation was 94% with supplemental oxygen. Reduced lung sounds were auscultated bilaterally with severe dyspnea and tachypnea. The patient was sedated with 0.3 mg/kg butorphanol^l IV and 1 mg/kg alfaxalone^m IV. Thoracic focused assessment with sonography for trauma showed pleural effusion, and 140 mL of fluid was drained by bilateral thoracocentesis. Radiographs taken after thoracocentesis showed residual pleural effusion and consolidation of the right middle and caudal lung lobes. The NGT passed through the trachea and entered the right bronchus and right caudal lung lobe. An automated hemogram^j and blood chemistry analysis^k showed normal neutrophil count (9.84 × 10⁹, reference range 2.95–11.64 × 10⁹/L) with left shift, eosinopenia, thrombocytopenia (125 K/µL, reference range 151–600 K/µL), and moderately elevated alanine aminotransferase (422 U/L, reference range 12–130 U/L). A blood gas analysisⁿ showed a respiratory acidosis (pH 7.27, partial pressure of CO₂ 50.4 mm Hg, HCO₃ 19.9 mmol/L). A thoracostomy tube was placed in the right hemithorax and ∼40 mL of serosanguinous fluid was removed. Cytological analysis showed many degenerative neutrophils, nucleated giant cells, and foamy macrophages with no bacteria present. Culture of the pleural fluid was negative, although antibiotics had been administered at the referring clinic. A computed tomography scan to evaluate the precise position of the tube and extent of damage showed the NGT in the right caudal lobar bronchus, right caudal lung consolidation and mineral opacity, bilateral pleural effusion with the right side more affected, and atelectasis of the right lung lobes, likely secondary to NGT misplacement. The chest tube was aspirated and the pleural cavity was lavaged with 20 mL of lactated Ringer’s solution every 4 hr. Fluid therapy, analgesia, ampicillin-sulbactam^o (30 mg/kg IV q 8 hr), and bronchodilators were initiated. Chest tube aspiration and lavage frequency were gradually decreased based on drain production and clinical signs. The chest tube was removed on day 5 following no overnight production of fluid. The patient saturated well on room air with normal respiratory rate and effort. Thoracic radiographs showed improvement of the interstitial alveolar pattern of the right caudal lung lobe. The patient was discharged the following day and continued to improve at home. Thoracic radiographs 1 wk and 1 mo after discharge demonstrated continued improvement.

Discussion

In human medicine, misplaced NGTs are considered “never events”—patient safety incidents that are preventable and so serious that they should never happen.¹⁴ Predisposing factors for NGT misplacement are altered mental status with decreased cough or gag reflex, a preexisting endotracheal tube, and severe illness.⁹ Gag reflex is often reduced if the patient is sedated for NGT placement, a likely contributing factor in case 1.

Radiographs are the gold standard for confirmation of NGT placement.² However, radiographic misinterpretation is the most common cause of undetected NGT misplacement in humans (46.2% of cases).¹⁵ Radiographs were taken of the three aforementioned cases, but the NGT position was incorrectly assessed. Additional confirmatory methods, such as additional orthogonal radiographs, capnography, pH testing of aspirate, and ultrasound-guided placement, should be used if there is doubt regarding radiographic positioning.¹⁶

Isocalothorax is rarely reported in human literature, with the majority of cases resulting in death.^9–13 In most cases, the NGT was removed upon discovery of tachypnea or dyspnea, and the patients deteriorated rapidly before any treatment could be implemented.^9–12 In one case, the patient recovered successfully following chest tube drainage, physical therapy, and antibiotics.¹³ To the authors’ knowledge, there is only one case of isocalothorax reported in veterinary literature.⁶ In the case report by Rodriguez-Diaz et al., NGT misplacement in a dog was discovered 36 hr after initiation of feeding when the patient developed tachypnea and dyspnea. The patient deteriorated rapidly, and bilateral thoracostomy tubes were placed. Removal of the NGT caused a continuous pneumothorax, which was managed with vacuum-assisted thoracostomy tube drainage. The patient died 4 days later. It is difficult to compare this case with those reported here because of the difference in species and circumstances. More cases are required to determine the predisposing factors and best treatment strategies in isocalothorax cases.

In cases 1 and 2, puncture of the NGT through the respiratory tract resulted in food material deposited in the pleural space and subsequent atelectasis, ventilation-perfusion mismatch, and hypoxemia. In case 3, it was unclear whether the NGT punctured into the pleural cavity, but pleural effusion was still present. It is likely that there was a small perforation into the pleural cavity undetected by computed tomography. Alternative hypotheses were proposed in human cases where there was isocalothorax without clear perforation, including disruption of endothelial permeability via calcium in enteral nutrition, or secondary pneumonitis resulting from production of proinflammatory cytokines and increased capillary permeability.^10,17 Whether bacterial infection is a component in cases of isocalothorax is unknown. Two of the cases had negative cultures of the pleural fluid, but antimicrobials were administered before sampling, which may have influenced the yield.

In case 3, the patient developed tachypnea and dyspnea 36 hr after initiation of feeding. The NGT was not removed until after diagnostics and thoracostomy tube placement. The treatment protocol was based on treatment of pyothorax. Although there is no consensus on the treatment of pyothorax, broad-spectrum antimicrobial therapy and tube thoracostomy drainage are well described in literature.¹⁸ Thoracostomy tube lavage with buffered balanced crystalloid solution and drainage was performed frequently initially, and frequency was reduced based on the clinical presentation of the patient and amount of fluid collected. The tube was removed when pleural fluid volume was less than 2 mL/kg/day and respiratory signs resolved. No major complications resulted from pleural lavage and suction.

Surgical management of isocalothorax is not well described in veterinary or human literature. Should surgical treatment be elected, the preferred surgical treatment of pyothorax is thoracotomy via median sternotomy, which allows adequate visualization, exploration, and thoracic lavage.¹⁸ In case 1, surgical intervention was successful and the patient was clinically normal 1 mo later, but it was unsuccessful in case 2. The patient in case 2 was already septic and cardiovascularly unstable before surgery, which likely contributed to the different outcome. The surgery performed in case 2 was also more radical, involving a total left-sided pneumonectomy. The time from NGT misplacement to surgical intervention was similar in both cases (14 hr and 16 hr). More information is needed to elucidate the mortality risk factors in such cases.

Ultimately, medical and surgical interventions can successfully treat isocalothorax in small animals. The incidence of NGT misplacement in veterinary patients is unknown, and the consequences can be catastrophic. After case 2, our hospital adopted a modified version of the Roubenoff and Ravich two-step protocol, as described by Gladden.⁵ The NGT is advanced to the level of the fourth to fifth intercostal space, and a lateral oropharyngeal radiograph is taken to ensure the NGT is in the esophagus before advancing into the stomach. Two-view orthogonal radiographs are taken after full insertion to confirm appropriate placement. Our previous protocol was to advance the NGT to the premeasured level of the 13th rib before taking two-view orthogonal radiographs. This has eliminated NGT misplacement in our hospital. However, this protocol potentially increases labor, cost, and time as additional radiographs are needed. Validation of accessible point-of-care methods, such as pH paper testing, capnography, and point-of-care ultrasound, is justified to reduce the incidence of NGT misplacement in general practice clinics.

Conclusion

Although NGT placement is common, clinicians should be aware of the potential risks. This case report aims to raise awareness of the risk of isocalothorax secondary to NGT misplacement so that clinicians are prepared to prevent and address such complications. However, more studies are needed to investigate the best way of managing cases of isocalothorax. Our hospital requires two-view orthogonal thoracic radiographs to confirm appropriate positioning before enteral nutrition is started. In patients who are sedated and at a higher risk for inadvertent placement into the trachea, a cervical radiograph is obtained before advancement of the tube. For cases requiring bedside placement, the decision is both clinician and patient dependent and typically consists of aspiration of gastric contents, laryngoscopic visualization of the tube by bypassing the trachea via the oropharynx, and visualization of the distal tip of the tube with point-of-care ultrasound in the cervical esophageal region as well as gastric lumen.