Editorial 

November 29, 2023

The Value of Not Intubating Comatose Patients With Acute Poisoning

Zaffer Qasim, Jeanmarie Perrone, M. Kit Delgado

JAMA. Published online November 29, 2023. doi:10.1001/jama.2023.22462

Imagine if a new multicenter, randomized trial of a pharmaceutical drug found that for every 3 patients treated on presentation to the hospital with a common severe illness, one intubation and an intensive care unit (ICU) admission were prevented. This would be an instant blockbuster drug. Insurers would scramble to reimburse it, and guidelines and quality metrics would be established to facilitate implementation. In the setting of worsening hospital and ICU strain globally, health systems would work rapidly to implement pathways for using this new therapy. For reference, the RECOVERY trial testing the use of dexamethasone in hospitalized patients with COVID-19, cited more than 2500 times since publication 3 years ago, reduced the need for intubation for 1 in every 50 patients.¹

In this issue of JAMA, Freund et al² report the results of a multicenter, randomized trial of a health care strategy akin to this hypothetical blockbuster drug tested among 225 patients treated in 20 emergency departments (EDs) and 1 ICU. The patients enrolled were obtunded or comatose with a Glasgow Coma Scale (GCS) score of 8 or less from a suspected poisoning, intoxication, or overdose and had a mean age of 33 years. The most common toxins were alcohol (66.7%) and benzodiazepines (39.6%). The authors found that compared with usual care, a conservative strategy withholding intubation among these patients with acute poisoning and a GCS score of 8 or less was not only safe, but had a markedly reduced rate of intubation of 57.8% vs 16.4%, rate of ICU admission of 66.1% vs 39.7%, and rate of adverse events of 14.7% vs 6.0%. It should be noted that patients who met emergency intubation criteria including respiratory distress, oxygen desaturation, vomiting, or persistent shock were excluded. While there are some caveats, this is a practice-changing trial that will affect many prehospital, emergency medicine, and ICU physicians. The trial provides rigorous evidence to counter an ingrained dogma in clinical practice that has been overextrapolated beyond its limited evidence and narrow indications.

The evaluation and management of obtunded, comatose patients can be challenging, especially if their depressed mentation puts the patient at higher risk of airway and ventilatory complications such as aspiration of gastric contents. While seemingly protecting the patient from this risk, the intubation process is not itself without harm. It can result in postintubation cardiovascular collapse, ventilator-induced lung injury, and utilization of limited intensive care resources.³ Thus, having evidence to determine which obtunded patients should be intubated can be extremely useful.

The GCS was designed by 2 Scottish neurosurgeons in the 1970s to uniformly describe the level of consciousness in comatose patients with brain injury.⁴ The GCS score is the sum of 3 subscales (verbal, eye movement, and motor components) and provides a score ranging from 3 to 15. This scale was rapidly adopted by clinicians worldwide in trauma patients and extended to nontraumatic causes of coma.

The GCS soon became a guide to clinicians about when to intubate their patients, with Gentleman et al⁵ bringing the adage of “less than 8, intubate” into common clinical use in the early 1990s. This has been supported by the Advanced Trauma Life Support course and the Eastern Association for the Surgery of Trauma’s practice-management guidelines.⁶ High-quality prospective evidence, however, is lacking to continue to promote this dogmatic approach. In their systematic review of 13 studies, Orso et al⁷ found a lack of evidence to support or refute intubation at low GCS levels to prevent aspiration events. Additional retrospective evidence in both trauma and nontrauma populations only strengthens the need to conduct a more robust prospective trial to validate the important clinical question of whether GCS alone should dictate the need to intubate.^8,9

To this end, Freund and colleagues² should be commended for designing and executing a multicenter, pragmatic randomized clinical trial to generate the level of evidence necessary to determine the safety of expectant management rather than immediate intubation in patients with acute poisoning and a GCS score of 8 or less.

The authors prespecified a primary composite end point of in-hospital death, ICU length of stay, and hospital length of stay. The prespecified statistical analysis, the Finkelstein-Schoenfeld method, prioritizes outcomes within the composite measure and enables a way to combine the categorical and continuous scales of individual outcome to determine net benefit. “Wins” and “losses” were calculated for each group based on pairwise comparisons. The conservative strategy of withholding intubation was found to be more effective based on the primary composite end point with a win ratio of 1.85 (95% CI, 1.33-2.58). Given that no patients died in the trial, the benefit was driven by reduced ICU length of stay and hospital length of stay.

The results of this study have the potential to significantly impact both patient-centered and institution-centered outcomes. Avoiding intubation can protect patients from complications associated with using a ventilator, the risk of nosocomial infections, exposure to additional sedative medications, and prolonged hospitalization. From a resource perspective, the institution can use limited critical care capacity for other patients, and the shorter length of stay allows bed capacity to be increased. While this study was conducted in a high-income country, low- and middle-income countries will especially benefit from the resource and financial savings. With the escalating challenge of ED overcrowding whereby ICU patients board in the ED and may experience worse outcomes, safely avoiding intubation has indirect benefits.¹⁰

However, there are several caveats to consider. First, while withholding intubation can avoid ICU admissions, it is unclear from this study how shifting care to the ED affects nursing and physician workload to care for other patients. Close monitoring would be required to identify markers of deterioration, and this can be an added burden to already overworked nurses and physicians.¹¹

Second, unique aspects of the emergency care system in France, where this study was performed, should be considered. Some patients are evaluated by a physician in the prehospital setting. Should that physician decide to intubate, these patients are taken directly to the ICU. This highlights 2 challenges to the external validity of these findings in other emergency services systems—the senior clinician has firsthand information on scene to make their airway management decision, and the patients arriving intubated to the ICU may undergo earlier extubation by a medical team more familiar with doing this than if they arrive in the ED initially. Nevertheless, this subset of patients do lend themselves to safe ED extubation.¹²

Having a physician at the scene also may secure a more accurate history to exclude concomitant trauma secondary to intoxication, which often prompts ED/trauma teams to default to intubation to facilitate a trauma evaluation more rapidly in patients with intoxication. Given that modern computed tomography (CT) scanners can complete a head CT in less than a minute, future research should investigate protocols using physical examination, and mechanism of injury to withhold intubation, and proceed with rapid CT if feasible. There has been a progressive shift away from the use of gastric decontamination procedures such as orogastric lavage and activated charcoal administration, which may have previously prompted teams toward prophylactic intubation in an earlier era.¹³

This study was also successful in the intervention group because of the underlying toxicology of the poisoning. The predominant causes of altered mental status were alcohol, benzodiazepine, or γ-hydroxybutyric acid (GHB)/γ-butyrolactone (GBL) intoxications. Benzodiazepine and GHB/GBL exposures are associated with less frequent need for ventilatory support, and GHB/GBL is noteworthy for ultra-short duration of effect, which supports a brief course of observation and supportive care rather than default to intubation. In the US, amidst a profound polysubstance overdose crisis, naloxone reliably restores sufficient ventilatory status in patients with opioid co-intoxication. An opportunity to evolve clinical practice to support expectant management to those with persistent depressed mental status but improved ventilation can confirm the suspected diagnosis of intoxication.

There continues to exist much dogmatic practice in medicine grounded in poor-quality evidence. This study demonstrates the ability of thoughtful, well-conducted trials to challenge that dogma, provide improved quality of care to patients, and improve resource utilization in the health care system. In the case of a patient with acute poisoning when another cause of altered consciousness has been excluded, doing less is actually doing more.