Coronavirus Disease 2019 Acute Respiratory Distress Syndrome: Guideline-Driven Care Should Be Our Natural Reflex

In December 2019, a novel virus, the severe acute respiratory syndrome coronavirus 2, arose and spread globally over the ensuing months (¹). This new human pathogen causes coronavirus disease 2019 (COVID-19) associated respiratory failure and acute respiratory distress syndrome (ARDS), creating a substantial challenge to both public health and critical care medicine as a specialty. Patients with severe COVID-19 respiratory disease present with bilateral patchy opacities and hypoxemia, meeting criteria for ARDS (^1–3). Early reports have, at times, proposed a variety of disease phenotypes, suggesting that the clinical manifestations of COVID-19 ARDS are unique when compared with traditional ARDS (^4,5). Some patients are noted to have only minor limitations in respiratory compliance yet significant hypoxemia, sometimes considered out of proportion to the degree of alveolar infiltrates and overall severity of illness, raising concerns about novel pathophysiologic mechanisms specific to COVID-19 (⁶). Others have noted significant ventilation/perfusion mismatch and significant pulmonary vascular involvement (⁷). Questions have been raised as to whether the classical description of ARDS, as a disease of low compliance and severe alveolar injury with shunt-related hypoxemia, even applies to COVID-19. This has led to some calls for the reevaluation of mechanical ventilation strategies and proposals for unconventional treatment approaches for patients with ARDS from COVID-19 (⁸) that focus on treating a “disparate physiology” as compared with traditional ARDS.

While the variation in severity of lung injury and respiratory mechanics among patients with COVID-19 ARDS is notable and should prompt further investigation, similar degrees of compliance and lung injury are seen in previously described ARDS cohorts (^9,10). Bhatraju et al (³), in a case series of 24 patients with COVID-19 ARDS in the Seattle region of the United States found that the median compliance on day 1 of mechanical ventilation was 29 cm H₂O (interquartile range [IQR], 25–36 cm H₂O) with a median lowest Pao₂:Fio₂ of 142 (IQR, 94–177). All patients in the case series also had bilateral infiltrates on imaging, meeting criteria per the Berlin definition of ARDS. Similar severity of lung injury has been described in larger cohorts of patients with COVID-19 ARDS (¹¹). In the landmark trial of low tidal volume ventilation in ARDS, degree of lung injury defined by Pao₂:Fio₂ was similar to that seen thus far in descriptions of COVID-19 ARDS. Similar wide ranges of respiratory compliance were also described (¹²). In this seminal ARDS Network study, low tidal volume ventilation improved mortality regardless of compliance. Additionally, disordered coagulation, endothelial function, and thrombosis are well-known in ARDS prior to the COVID-19 pandemic (¹³). Increased pulmonary dead space fraction (i.e., ventilation of nonperfused regions of lung) early in ARDS is also common and associated with mortality (¹⁴). Maldistribution of pulmonary blood flow to areas with limited participation in gas exchange has also been described in ARDS (¹⁵) with attenuated hypoxic vasoconstriction being a potential mechanism (¹⁶). In each of these respects, the pathophysiology of COVID-19 ARDS disease varies, but thus far, it has not been shown to differ qualitatively enough from already established scientific literature describing traditional ARDS to justify abandoning established guidelines and protocols.

Based on these early clinical observations and physiologic reasoning, new phenotypes for COVID-19 ARDS have been proposed along with guidance recommending alternative approaches to mechanical ventilation based on the proffered phenotypes (¹⁷). Interest in phenotyping ARDS even prior to COVID has been substantial, particularly as we have come to understand the heterogeneity of ARDS. Research by Calfee et al (¹⁸) has suggested that differential phenotypes may have different outcomes and clinical features. In a secondary analysis of two large randomized clinical trials of patients with ARDS, they identified two subphenotypes, a hypo-inflammatory and a hyper-inflammatory phenotype, noting that the latter was associated greater mortality and that outcomes differed based on positive end-expiratory pressure strategy used between the two subphenotypes. Additional research has strengthened the evidence for these phenotypes and they hold significant potential for predictive enrichment in future ARDS interventional trials and in elucidating the pathology behind the heterogeneous feature of ARDS. The field of critical care medicine has thus far, however, taken a careful approach in adopting these phenotypes to alter clinical management, patiently awaiting further data from prospective, randomized trials. In contrast, the existence of specific phenotypes of COVID-19 ARDS remains speculative and currently lacks substantial scientific evidence beyond the handful of preliminary reports authored by leaders in the field of critical care.

Thus, implementing strategies such as higher tidal volume ventilation, particularly in lieu of data showing benefit of low tidal volume ventilation across a range of respiratory mechanics and in non-ARDS populations (^19,20), is not currently warranted. No randomized clinical trial has yet been performed studying the effect of an intervention with stratification based on ARDS phenotype from any etiology, and many interventions within critical care medicine have fallen to the refiner’s fire of well-designed, prospective randomized controlled clinical trials despite prior positive observational data and rational physiologic reasoning.

The question thus remains: how should the intensivist approach caring for critically ill patients with COVID-19 ARDS? We recommend, given the knowledge surrounding both COVID-19 and the well-established management principles for ARDS, that the natural reflex of the intensivist should be to follow current evidence-based guidelines. Advances within the field have improved patient outcomes over the last several decades, built on the foundation of well-designed scientific studies that have shown the benefit of interventions such as low tidal volume ventilation, paired spontaneous awakening and spontaneous breathing trials, and prone position ventilation when needed (^9,12,21). Focusing on these and other well-established, evidence-based management strategies is of particular importance given that the vast majority of studies in critical care, even those based on physiologic mechanism and reasoning, demonstrate no mortality benefit when studied in a robust manner (²²). Likewise, conservative fluid management, limiting sedation, managing delirium, and mobilizing patients early improve outcomes across a range of critical illness, from reducing days of mechanical ventilation to improving functional status (^23–25). The COVID-19 public health crisis has also highlighted the importance of reducing the impact of isolation on patients and families whenever possible, utilizing both technology and family centered care approaches to facilitate the very important need for human connection between patients and their loved ones to overcome the burdens of delirium, post-traumatic stress disorder, and other sequelae of critical illness. These are the bedrocks of modern critical care practice, and they should remain so, especially in the face of a pandemic.

Bedside assessment of physiology is integral to the job of the intensivist faced with complex diseases and organ failure, and its application to critically ill patients remains vital. Physiology can and must teach us, guiding both treatments and the design of our scientific inquiries. Yet, when our bedside assessment of physiology is either incomplete or stands in contrast to the best scientific data available, these observations should be used to generate testable hypotheses prior to the rejection of the most current evidence-based, guideline-driven critical care strategies. Such an approach allows our patients to not only survive but also thrive in survivorship. During this pandemic, when we provide such care, we can take comfort in knowing we are providing the best treatment possible at this point in medical history.