The Physiologically Difficult Airway – Part 2
Jon-Emile S. Kenny [@heart_lung]
In part 2, I continue my commentary on this excellent review; part 1 may be found here. In this post I will consider patients with severe metabolic acidosis and those with right ventricular [RV] dysfunction and/or failure.
Severe Metabolic Acidosis
In patients with severe metabolic acidosis, alveolar ventilation tends to be maximal as a compensatory mechanism. The relationship between alveolar ventilation [on the abscissa] and PaCO2 [on the ordinate] is an inverse, asymptote [figure 1]. Thus a fall in alveolar ventilation – even during a very brief apneic period – during intubation can result in a dramatic increase in PaCO2 and a profound worsening of acidemia. Patients with intense lactic acidosis, and especially Kussmaul respirations during diabetic ketoacidosis [DKA] may have minute ventilation approaching 40 liters/minute! The authors note that intubating such a patient and placing him or her on volume-control with a tidal volume of 1 liter and 30 breaths per minute will still not adequately compensate for the patient’s metabolic acidosis. Accordingly, they recommend avoiding intubation when possible in such patients. If work of breathing must be attenuated, the authors suggest using NIPPV to assist the patient and use the ventilator to estimate minute ventilation. If endotracheal intubation is required, the authors recommend against using a paralytic and opting for a sedative which maintains respiratory drive. Consequently, if successfully intubated, these patients are at very high risk for ventilator asynchrony secondary to flow starvation. Pressure control may, therefore, be a better choice as flow and minute ventilation are not preset by the clinician. Ostensibly, this will make the patient more synchronous with the ventilator.
Patients with severe metabolic acidosis and concomitant acute respiratory distress syndrome [ARDS] are exceptionally challenging. Once intubated, lung protective ventilation with small tidal volumes is essentially impossible in severe metabolic acidosis. Firstly, adequate tidal volumes are simply not conceivable; there will be undue stress placed upon the ‘baby lung.’ Secondly, it is often neglected that by lowering the tidal volume, the dead space fraction necessarily rises. As anatomic dead space remains constant, the fraction of total dead space [Vd] relative to tidal volume [Vt] – the Vd/Vt – must rise when Vt is lowered. In totality, patients with ARDS and severe metabolic acidosis should raise consternation when considering endotracheal intubation.
Equally, patients with significant airway obstruction within the milieu of severe metabolic acidosis are physiologically troubling. A high respiratory rate coupled with long pulmonary time constants [lung unit resistance x compliance] results in air-trapping and auto-PEEP. The clinician should be especially cognizant of this complication not only for its hemodynamic effects, but also its ability to cause trigger asynchrony. Excessive auto-PEEP requires that the patient overcome a greater load to trigger the ventilator. If this begins to happen, the patient’s breathing work and carbon dioxide production will rise, while the patient concurrently loses minute ventilation; this is a deadly combination. Treatment requires recognition of the problem and augmentation of ventilator-applied PEEP to minimize the trigger threshold load [see part E].
Right Ventricular Failure
A right ventricle under duress is an unwanted houseguest for the intensivist. Impending intubation raises concern over cardiovascular collapse secondary to depressed RV preload and increased RV afterload. Tackling each of these principles in turn, prior to intubation, may be fruitful. Diminution of RV afterload is favored by ensuring good pre-oxygenation and apneic oxygenation [as mentioned in part 1] as well as avoiding acidosis. Additionally, consideration for inhaled nitric oxide or inhaled epoprostenol should be given as these agents will vasodilate the pulmonary vasculature in communication with alveoli – such physiology will lower impedance to RV forward flow [figure 2]. Secondly, consideration should be given to improving RV contractility with inopressors [e.g. norepinephrine and epinephrine] or – while not mentioned by the authors – a myocardial sensitzer [e.g. levosimendin]. The use of vasopressin was not specifically addressed.
Figure 2: Optimizing RV function first will move the operating point from A to B. Provision of fluids at point B will shift the venous return curve [curved red arrow] and RV forward flow will rise from point B to C. If volume is given prior to optimization of RV function, the operating point will move from A to D. This will simply raise right atrial pressure without augmenting flow. RV is right ventricle, LV is left ventricle, RCA is right coronary arteryAccording to the authors, the issue of fluid administration, should be evaluated using echocardiography. While this is intuitively appealing, I have concerns that respiratory variation in lung volume [and therefore RV afterload] will induce variation in the RV outflow tract velocity time integral. To the ultrasonographer, this may imply preload variation when, in fact, it represents afterload sensitivity. Similarly, because of systolic ventricular interdependence, isovolumic contraction velocity of the RV may vary not as a function of RV characteristics, but because of respiratory variation of left ventricular features [consider that roughly 60% of systolic RV pressure generation is a consequence of the left ventricle [LV]]. It is my opinion, that assessment of the LV in response to fluid challenge remains the intensivist’s ultimate test of cardiac function, even in the setting of a failing RV. Eventually, it is LV output which is of most hemodynamic importance, not only for the brain and kidneys, but also the right coronary artery and the RV itself.
During intubation, the authors support using hemodynamically neutral induction agents such as etomidate, IV fentanyl to prevent the hypertensive response to laryngoscopy and a low mean airway pressure to prevent excessive RV afterload. With respect to the latter recommendation, the most important physiological principle is that of the ‘vascular waterfall’ first termed by Permutt in the 1960s. Increasing alveolar pressure will increase the impedance to RV outflow only to the extent that West Zone I and II are generated. Such physiology is of less concern in WHO type II pulmonary hypertension because pulmonary venous pressure is high; that is, most of the lung is in West Zone III. One means of assessing alveolar over or under-distention is to utilize the stress index. I find this method particularly appealing because it allows for a qualitative assessment of atelectasis [terminal alveolar recruitment or stress index < 1] which should be aggressively avoided in patients with pulmonary arterial hypertension.