The Case

A 47-year-old man presented to an emergency department (ED) with altered mental status, and was believed to have a probable overdose. He received endotracheal intubation, and, after he was stabilized, the ED staff planned to place a nasogastric (NG) tube to begin charcoal gastric lavage. A nurse passed the NG tube, and an attempt to aspirate gastric contents yielded no return. At that point, the nurse injected a 50-cc syringe full of air through the tube, and the intern reported that he heard a "woosh." Convinced that the NG tube was properly placed by its ease of passage, the confirmatory sound of the air over the stomach, and the presence of the endotracheal tube (with its inflated tracheal balloon), he gave the nurse the go-ahead to inject a slurry of activated charcoal in a sorbitol vehicle through the tube.

Soon after the charcoal infusion, the patient began to cough and developed worsening oxygenation. A chest radiograph demonstrated a new infiltrate in the right mid-lung field just distal to the tip of the NG tube, which clearly was in the right mainstem bronchus. The tube was promptly removed. The patient was treated for aspiration pneumonitis (although the charcoal is relatively inert, the osmotic stress induced by the sorbitol is quite toxic to the lungs), but he developed multiple complications including progressive pneumonia and ultimately died.

The case submitter reported that, in response to this incident, the hospital changed its protocol to require a confirmatory radiograph before anything is infused through a newly-placed NG tube.

The Commentary

Experience is the worst teacher. It always gives the test first and the instruction afterwards. - Anonymous

In treating this patient for an overdose, a nurse infused the mixture of activated charcoal and sorbitol into his lung by mistake. The therapy meant to save him eventually led to his death. As with any unintended outcome, we try to understand why it happened. It is natural to want to assign blame, but it is more productive to learn how to prevent this mischance in the future.

After reading this case, I searched the literature for similar cases and found an overwhelming number of articles on NG tube misadventures. The majority concerned the well-known problems of misplacement of light, narrow-bore feeding tubes, but there were also several reports concerning large-bore NG tubes used for lavage and infusion. Most were solitary case reports, rather than collections or reviews. I could only guess at prevalence, but it does seem that disastrous outcomes of NG intubation are not rare. Large-bore nasogastric tubes find their way into the lung causing pneumonia, pneumothorax, and hemorrhage. They perforate esophagi and bronchi. They can twist around and cause intestinal volvulus. They are pushed through the cribriform plate and curl around within the cranium with unfortunate results. There was even one prior case of charcoal infusion into a lung.(1) This breadth and variety of cases is striking when one considers that many incidents almost certainly go unreported.

Although inserting an NG tube appears to be a relatively simple procedure, one must follow correct placement technique and properly confirm placement to prevent complications. In this case, the problems probably arose more from a lack of knowledge than poor technique. The intern and nurse apparently didn’t know that instillation of charcoal in an obtunded patient was a relatively high-risk procedure that can go awry. They didn’t realize that the softly inflated endotracheal cuff wouldn’t prevent lung intubation. They didn’t know that the air insufflation test was unreliable because air infused into the bronchus can be ausculted over the epigastrium. Finally, they apparently were unaware of the other tests they could have used to check placement.(2,3) I surveyed a convenience (hallway) sample of nurses and physicians at my own hospital, however, and found that few of them knew these facts either. I concluded that in the present case the intern’s and nurse’s level of knowledge was not unusual.

How do we prevent this problem from recurring? Our tools for change are education, credentialing, and change in protocol. If lack of knowledge is the problem, education is the most obvious solution. Our current non-system for learning procedures rarely teaches beyond the basics; most trainees only receive instruction when they first learn a procedure. Advanced or refresher sessions are rare. Moreover, in the training that does occur, cognitive aspects are underemphasized in favor of technical and procedural ones. Often anecdotal or personal knowledge, rather than broader experience and data, is passed on. Credentialing of even highly challenging procedures now relies on experience alone as a measure of competence and not knowledge or measured ability. For less complex procedures such as nasogastric intubation, physicians are usually automatically credentialed with no assurance they know anything about it.

Simulations could be an answer if they were sufficiently sophisticated.(4,5) Most simulations of common medical procedures focus on elementary technique rather than the management of complications and recognition and management of pathology. In contrast, flight simulators can test experienced airplane pilots on difficult and catastrophic events such as loss of control functions or emergency management. Few medical simulators for common procedures operate in a similar fashion. Simulators teach technical skills, but not what to do if you’ve biopsied a gastric varix or sliced off a catheter during thoracentesis.

A protocol aimed at high-risk situations is a practical answer. A hospital could require, for example, a checklist or algorithm of tests that must be done to be sure an NG tube is properly placed before any infusion. A checklist might read: (a) Is the tube in proper position in the posterior pharynx? (b) Can you hear injected air in the epigastrium? (c) Test the aspirate: Is the pH below 3? (d) Hold the tube in a glass of water; do bubbles appear indicating exhaled breath? Other ways of assessing correct tube placement include capnography and radiography. Capnography testing (for exhaled carbon dioxide) is commonly used in ICUs to test narrow-bore feeding tube placement (3,6) and is often available in EDs. However, the best check of NG tube placement is chest x-ray; x-ray is standard practice in narrow-bore feeding tube placement. Would the addition of a chest x-ray in this case have delayed the instillation of active charcoal too much? Many EDs are equipped to obtain x-ray evaluations quickly. The issue has not been formally studied, but the volume of cases in the literature suggest that when an x-ray can be done without undue delay, all patients with altered mental status (and probably alert patients as well) should have x-ray confirmation of NG tube positioning before infusing anything through the tube.

This case highlights several general problems. Practitioners aren’t trained well in the cognitive aspects of common procedures, and learning often stops once they become technically proficient. Those who perform the procedure do not always follow the relevant literature. Finally, practitioners may not become sensitized to complications that occur at a low frequency in their own patient population. Ultimately, these are problems of knowledge, training, and credentialing, but they can also be addressed in part by protocols or guidelines that anticipate and prevent known complications.

Robert S. Wigton, MD Professor of Internal Medicine Associate Dean for Graduate Medical Education University of Nebraska Medical Center College of Medicine

References

1. Harris CR, Filandrinos D. Accidental administration of activated charcoal into the lung: aspiration by proxy. Ann Emerg Med. 1993;22:1470-3.[ go to PubMed ]

2. Levy H. Nasogastric and nasoenteric feeding tubes. Gastrointest Endosc Clin N Am. 1998;8:529-49.[ go to PubMed ]

3. Roberts JR, Hedges JR, eds. Clinical procedures in emergency medicine. 2nd ed. Philadelphia: WB Saunders; 1991.

4. Ziv A, Wolpe PR, Small SD, Glick S. Simulation-based medical education: an ethical imperative. Acad Med. 2003;78:783-8.[ go to PubMed ]

5. Issenberg SB, McGaghie WC, Hart IR, Mayer JW, et al. Simulation technology for health care professional skills training and assessment. JAMA. 1999;282:861-6.[ go to PubMed ]

6. Kindopp AS, Drover JW, Heyland DK. Capnography confirms correct feeding tube placement in intensive care unit patients. Can J Anaesth. 2001;48:705-10.[ go to PubMed ]