To celebrate the birthday of Dr. Erin Hennessey [@ErinH_MD] – my former co-fellow and current Stanford intensivist-anesthesiologist – I will interpret a relatively recent and terrifically high-yield overview of physiologically challenging intubations. In this must-read survey, the authors highlight particularly troublesome intubations not from the classic, anatomical perspective, but from the standpoint of the – reasonably commonplace – physiologically moribund patient. The authors provide recommendations on each of the following concerning states when transitioning a patient from unassisted to assisted ventilation: hypoxemia, hypotension, severe metabolic acidosis and right ventricular failure.
Hypoxemic – or type 1 – respiratory failure is perhaps one of the most common reasons for endotracheal intubation in the intensive care unit [ICU]. The mechanism by which this type of respiratory failure occurs is that of pulmonary arterial blood moving to the pulmonary venous system without participating in gas exchange at the alveolus. Notably, as it is common to focus on the lung as the primary pathomechansim for hypoxemia, the contribution of the embarrassed circulation to hypoxemia may be overlooked. Consider that V/Q mismatch in the lung will be accentuated by a low mixed venous oxygen saturation. When left ventricular output is diminished – as is common when a patient is in extremis – peripheral tissue oxygen extraction rises to maintain a constant tissue oxygen utilization. The result is that the oxygen saturation of venous blood returning to the right heart and pulmonary artery can be profoundly low. Consequently, poorly saturated hemoglobin presented to poorly functioning lungs will find its way to the left heart excessively deplete of oxygen. Additionally, poor RV stroke volume and under-perfusion of the lungs is a mechanism in and of itself which leads to V/Q inhomogeneity. Thus, prior to focusing on the lungs, the circulation should be optimized with fluids and/or vasoactive medications [e.g. norepinephrine or epinephrine] as needed; and this needn’t be through a central line! If one does has access to a central or mixed venous oxygen saturation [e.g. even via a PICC line], a low value should cause pause.
If the patient’s circulatory status is deemed adequate to proceed, it is standard practice to maximize the arterial partial pressure of oxygen [PaO2] and therefore hemoglobin [Hb] saturation by pre-oxygenation. It is commonly inferred that the inflection point of the hemoglobin saturation curve occurs at roughly 60 mmHg, however, this value may be shifted upwards in states of acidemia, fever, etc. [i.e. the Hb dissociation curve is right-shifted]. Pre-oxygenation is classically attempted with the patient breathing for 3-5 minutes via a standard non-rebreather. However, this method is of limited utility in critically-ill patients under respiratory duress due to the entrainment of ambient air. The authors suggest that, instead, high flow nasal cannula [HFNC], which is capable of delivering up to 60 L/min be used for pre-oxygenation, despite mixed evidence. If such a device is not available, they suggest standard, wide-bore nasal prongs at 10-15 L/min.
In patients at risk of atelectasis with apnea [e.g. excessive thoraco-abdominal load] or in those with high shunt fraction, pre-oxygenation may be better achieved using non-invasive positive pressure ventilation [NIPPV]. Pre-oxygenation by this method, even in the patient with depressed mentation, can prolong safe apnea time. However, following sedation and paralysis, there is a risk that loss of muscle tone will reduce functional residual capacity [FRC] and cause atelectasis. Consequently, acute V/Q mismatch and hypoxemia may rapidly ensue. In such patients, supraglottic airways may be employed which can allow higher insufflation pressure with a better seal. If pharmacological assistance – such as ketamine – is used to facilitate preoxygenation, this has been coined delayed sequence intubation [DSI].
Once pre-oxygenated, the patient is administered sedatives and then paralytics to facilitate direct laryngoscopy and endotracheal intubation. Notably, this procedural phase needn’t be without oxygen! Apneic oxygenation is also called “diffusion respiration” or “aventilatory mass flow;” so long as blood is passing by alveoli, an oxygen gradient is formed within the respiratory tree. Apneic oxygenation can maintain saturated hemoglobin for a very prolonged time, however, this comes at the risk of hypercapnia, respiratory acidosis and the deleterious consequences thereof. Importantly, both HFNC and nasal-applied NIPPV may be continued during apnea to help preserve alveolar recruitment and oxygenation during the apneic phase of endotracheal intubation. For a more in depth discussion on the aforementioned, pick up this review or listen to this fantastic podcast.
Peri-intubation hypotension portends a worse prognosis in the critically-ill and may be predicted by a pre-intubation shock index [heart rate/systolic blood pressure] of more than 0.8. The mechanism by which intubation impairs cardiac output is not the oft-taught “diminished gradient for venous return.” While right atrial pressure is raised by the provision of mechanical ventilation, the upstream, or driving pressure [i.e. the mean systemic pressure] tends to rise by and equal amount! The real mechanism is mostly an increase in the resistance to venous return as the abdomen is placed into “West Zone II” like physiology. Additionally, when the patient starts to trigger the ventilator, the normal thoracic pump mechanism is retarded by formation of great vein Starling Resistors at the thoracic inlet; that is, venous return to the right heart will plateau at a much lower value following intubation.
There are essentially three ways to mitigate peri-intubation hypotension: fluids, inopressors, and intelligent selection of induction agent. While volume-responsiveness is the rule in the healthy heart, it is present in roughly 50% of all critically-ill patients. There are many purported ways of establishing that the heart will increase its output in response to volume infusion, but in the spontaneously breathing patient, only a direct measure of left ventricular stroke volume [e.g. doppler flow] in response to an increase in venous return [e.g. passive leg raise, mini fluid challenge, end-expiratory occlusion] has been shown to be meaningfully predictive. In patients who are not responsive to fluid infusion, the agent of choice is norepinephrine, according to the authors. The author’s appropriately note that pure alpha-agonists such as phenylephrine are cautioned-against. Pure alpha-agonists are known to increase the resistance to venous return and therefore diminish cardiac output, despite raising arterial blood pressure. This may have disastrous results in patients with high, fixed ventricular afterload [e.g. severe pulmonary arterial hypertension or aortic stenosis]. In patients without a central line, peripherally-administered vasoactive agents can be temporarily administered without significant risk to the patient. It is also noted that benzodiazepines and propofol have sympatholytic effects which can cause veno/vasodilation as well as diminish myocardial inotropic state. In such patients, the authors note that ketamine or etomidate may be attractive agents and cite a trial which compares these two drugs; the authors do not mention the risk of adrenal-insufficiency known to occur following etomidate administration.
Stay tuned for part 2, forthcoming!