The process of rapid sequence induction (RSI) often forces the clinician to choose between two conflicting priorities. Torn between maintaining a safe level of oxygen, rendering the patient apneic, and limiting the amount of positive pressure ventilatory support given prior to obtaining a secure airway in the hopes of minimizing the risk of aspiration. Traditionally this has translated into administering vast quantities of supplemental oxygen to preoxygenate the patient, with the intent of producing a sufficient oxygen reservoir in the hopes of avoiding positive pressure ventilation altogether.
But some have suggested that our priorities are misaligned. Our overblown concern over the risk of aspiration is out of proportion to the true rate of clinically important events, leading to undertreatment of the far more prevalent risk of hypoxia. To empirically test this assertion Casey et al conducted, the PreVenT trial, an RCT examining the use of positive pressure ventilation during the apneic period in critically ill patients undergoing RSI1. The authors enrolled adult patients admitted to 7 ICUs undergoing RSI, and randomized them to a traditional approach where the patients was left apneic throughout the intubation process or one where bag mask ventilation was used to support them until the initial attempt at laryngoscopy was made. Patients were excluded if the treating clinicians determined that ventilation between induction and laryngoscopy was either required or contraindicated.
For patients randomized to the bag mask ventilation group, ventilation was provided during the interval from induction until the initiation of laryngoscopy. The authors encouraged clinicians to use oxygen flow rates of at least 15 liters per minute, a PEEP valve set to 5 to 10 cm of water, an oropharyngeal airway, a two-handed mask seal using a head-tilt and chin-lift maneuver, and ventilating at 10 breaths per minute with the smallest volume required to generate a visible chest rise. Patients randomized to the no ventilation group received no ventilatory support between induction and securing an ETT, except following a failed attempt at laryngoscopy, as treatment for hypoxemia, or if the treating clinician determined it was necessary. Methods to pre-oxygenate patients and the use of apneic oxygenation was left to the judgment of the treating physician. The authors enrolled a total of 401 patients, nearly 50% had sepsis or septic shock, with nearly 60% having hypoxic respiratory failure as the reason for intubation.
The authors report that patients randomized to the bag mask ventilation group experienced significantly less episodes of desaturation during the intubation process compared to their no ventilatory controls. The median lowest oxygen saturation was 96% in the bag mask ventilation group and 93% in the no-ventilation group (P=0.01). 29.5% of patients in the bag mask ventilation group experienced oxygen saturation values of less than 90%, compared 40.1% in the no-ventilation group. 10.9% of patients in the bag mask ventilation group had an oxygen saturation of less than 80%, compared to 22.8% of controls. 4.1% had an oxygen saturation of less than 70%, as compared with 10.1%. The authors found no increase in the rates of aspiration (2.5% vs 4.1%), new opacity of CXR (16.4% vs 14.8%), new PTX (1.1% vs 3.1%), need for new vasopressor following induction (19.9% vs 23.1%), or episodes of peri-intubation hypotension (4.1% vs 8.6%) in the bag mask ventilation and no ventilation groups, respectively.
While this is a fairly positive trial, it does present a number of issues that limit its broad generalization. A large portion of patients randomized to the control group did not receive optimal alternative oxygenation strategies in either the pre-oxygenation and apneic phases of the intubation process. Would the authors have observed significant difference between the two groups if optimal preoxygenation and apneic oxygenation strategies were utilized in the no ventilation group? 46.1% of the patients in the control group were preoxygenated with either a NRB mask, a standard nasal cannula alone, or nothing, none of which historically provide adequate preoxygenation conditions. 15.2% and 8.6% of the cohort had an oxygen saturation of less than 95% and 92% at the time of induction. While 77.7% of the patients in the control group had supplemental O2 in line during the apneic period, only 34.1% had O2 supplementation in the form of HFNC or standard nasal prongs.
But more important than the less than ideal comparator is the fact the study likely asked the wrong question. I doubt much equipoise exists as to whether the use of positive pressure ventilation during the apneic period decreased the rate of desaturation. Especially when compared to a group that of patients in which imperfect pre-oxygenation strategies are employed. What was unclear was whether this risk of hypoxia should be prioritized above the risk of aspiration. And this study did very little to answer this question. They planned for a sample size which could demonstrate an absolute between-group difference of 5% in the lowest oxygen saturation with a power of 90%. This left the trial vastly underpowered to assess the bag mask ventilation strategy’s effect on the rate of aspiration, especially the catastrophic aspiration events that greatly inhibit one’s ability to secure a definitive airway. The authors themselves admit to this shortcoming in their discussion stating, “determining whether bag-mask ventilation increases the relative risk of aspiration by 50% would require a trial enrolling approximately 4000 patients.”
I suspect the reality lies somewhere between the two extremes of always and never providing positive pressure ventilation when endeavoring to secure an airway in the critically ill. There is a subset of patients who have very little lung pathology limiting their pre-oxygenation potential, and/or have an extremely high-risk of aspiration, (the authors actively excluded this subset of patients), who likely can tolerate the apneic period without the assistance of positive pressure ventilation. Equally there is a subset of patients in whom the pulmonary insult is severe enough that safe oxygen saturation levels cannot be maintained through pre-oxygenation strategies alone, and thus the risk of hypoxia far outweighs the risk of aspiration. In this case gentle positive pressure ventilatory assistance seems like a reasonable alternative to the more traditional approach (See here for a great PulmCrit post on a safe and effective method). Despite its methodological rigor I fear Casey et al have provided us with little novel information. Patients are complicated. Critically ill patients undergoing endotracheal intubation are more complicated still. As clinicians we are tasked with balancing the multitudes of ways our patients can continue to decompensate and how best to prevent it.
Sources Cited:
- EM Nerd-The Case of the Partial Cohort - May 24, 2020
- EM Nerd: The Case of the Sour Remedy Continues - January 20, 2020
- EM Nerd-The Case of the Adjacent Contradictions - December 23, 2019
In addition to the points you raise, in this post the PreVent trial is described as being “in critically ill patients undergoing RSI”. While the authors imply this in the introduction, a careful reading of the methods section and trial protocol will reveal that, actually, they never even tried to select patients undergoing RSI! The trial allowed basically any intubation method to be chosen that included any intravenous sedative/anaesthetic. We do not know what proportion of patients received RSI, but examining the results in detail (including the supplementary data) suggests that most patients did not undergo RSI. We can draw… Read more »