Management of COVID-19 Respiratory Distress | Critical Care Medicine | JAMA | JAMA Network
[Skip to Navigation]
Sign In
JAMA Insights
Clinical Update
April 24, 2020

Management of COVID-19 Respiratory Distress

Author Affiliations
  • 1Regions Hospital, University of Minnesota, Minneapolis/St Paul
  • 2Department of Anesthesiology, Intensive Care and Emergency Medicine, Medical University of Göttingen, Göttingen, Germany
JAMA. 2020;323(22):2329-2330. doi:10.1001/jama.2020.6825

Acute respiratory distress syndrome (ARDS) can originate from either the gas or vascular side of the alveolus. Although the portal for coronavirus disease 2019 (COVID-19) is inhalational, and alveolar infiltrates are commonly found on chest x-ray or computed tomography (CT) scan, the respiratory distress appears to include an important vascular insult that potentially mandates a different treatment approach than customarily applied for ARDS. Indeed, the wide variation in mortality rates across different intensive care units raises the possibility that the approach to ventilatory management could be contributing to outcome.1-3

COVID-19 is a systemic disease that primarily injures the vascular endothelium. If not expertly and individually managed with consideration of the vasocentric features, a COVID-19 patient with ARDS (“CARDS”) may eventually develop multiorgan failure, even when not of advanced age or predisposed by preexisting comorbidity.

Standard Approaches to Ventilating ARDS

Normally, ARDS is characterized by noncardiogenic pulmonary edema, shunt-related hypoxemia, and reduced aerated lung size (“baby lung”), which accounts for low respiratory compliance.4 In such settings, increasing lung size by recruiting previously collapsed lung units is often achieved through the use of high levels of positive end-expiratory pressure (PEEP), recruiting maneuvers, and prone positioning. Because high transpulmonary pressure induces stress across the lung that is poorly tolerated in ARDS, relatively low tidal volumes, together with tolerance for modest (permissive) hypercapnia, facilitate the goal of minimizing ventilator-induced lung injury (VILI). Indeed, in the early phases of ARDS, before a patient has fatigued or been sedated, the high transpulmonary pressures associated with spontaneous vigorous inspiratory effort may contribute to damage (so-called patient self-induced lung injury [P-SILI]).5

Clinical Features of CARDS

Soon after onset of respiratory distress from COVID, patients initially retain relatively good compliance despite very poor oxygenation.1,2 Minute ventilation is characteristically high. Infiltrates are often limited in extent and, initially, are usually characterized by a ground-glass pattern on CT that signifies interstitial rather than alveolar edema. Many patients do not appear overtly dyspneic. These patients can be assigned, in a simplified model, to “type L,” characterized by low lung elastance (high compliance), lower lung weight as estimated by CT scan, and low response to PEEP.6 For many patients, the disease may stabilize at this stage without deterioration while others, either because of disease severity and host response or suboptimal management, may transition to a clinical picture more characteristic of typical ARDS. These can be defined as “type H,” with extensive CT consolidations, high elastance (low compliance), higher lung weight, and high PEEP response. Clearly, types L and H are the conceptual extremes of a spectrum that includes intermediate stages, in which their characteristics may overlap. Another feature consistently reported is a highly activated coagulation cascade, with widespread micro- and macro-thromboses in the lung and in other organs (eFigure 1 in the Supplement); very elevated serum D-dimer levels are a consistent finding associated with adverse outcomes.7

These observations indicate the fundamental roles played by disproportionate endothelial damage that disrupts pulmonary vasoregulation, promotes ventilation-perfusion mismatch (the primary cause of initial hypoxemia), and fosters thrombogenesis. In addition, remarkably increased respiratory drive may, if unchecked, intensify tidal strains and energy loads from a patient’s respiratory effort applied to highly vulnerable tissue, adding P-SILI to the mix of the lung’s inflammatory assault.5,8 When confronting such an unfamiliar and rapidly evolving environment, only certain aspects of well-accepted lung-protective approaches to ARDS remain rational at these different stages. More important, inattention to the vascular side (eg, avoidance of fluid overload, reduction of cardiac output demands) could inadvertently promote counterproductive responses (eg, edema) and iatrogenic damage.

Protecting the CARDS Lung

Patients with type L CARDS, having good lung compliance, accept larger tidal volumes (7-8 mL/kg ideal body weight) than those customarily prescribed for ARDS without worsening the risk of VILI. Actually, in a 70-kg man, with respiratory system compliance of 50 mL/cm H2O and PEEP of 10 cm H2O, a tidal volume of 8 mL/kg yields a plateau pressure of 21 cm H2O and driving pressure of 11 cm H2O, both well below the currently accepted thresholds for VILI protection (30 and 15 cm H2O, respectively). Higher VT could help avoid reabsorption atelectasis and hypercapnia due to hypoventilation with lower tidal volumes.

The key issue in this early stage is disrupted vasoregulation, where the pulmonary vasoconstriction that normally occurs in response to hypoxia fails to occur because of an endothelial assault that mismatches perfusion to ventilation and may result in profound hypoxemia. The clinician’s first response, boosting Fio2, may indeed prove effective early on. If insufficient, noninvasive support (eg, high-flow nasal O2, CPAP, Bi-PAP) may stabilize the clinical course in mild cases, provided that the patient does not exert excessive inspiratory efforts. However, if respiratory drive is not reduced by oxygen administration and noninvasive support, persistently strong spontaneous inspiratory efforts simultaneously increase tissue stresses and raise pulmonary transvascular pressures, vascular flows, and fluid leakage (ie, P-SILI).8-10 Progressive deterioration of lung function (a VILI vortex) may then rapidly ensue. Early intubation, effective sedation, and/or paralysis may interrupt this cycle. Targeting lower PEEP (8-10 cm H2O) is appropriate. Raising mean transpulmonary pressures by higher PEEP or inspiratory-expiratory ratio inversion redirects blood flow away from overstretched open airspaces, accentuating stresses on highly permeable microvessels and compromising CO2 exchange without the benefit of widespread recruitment of functional lung units.

If lung edema increases in the type L patient, either because of the disease itself and/or P-SILI, the baby lung shrinks further, and the type H phenotype progressively develops. Concentrating the entire ventilation workload on an already overtaxed baby lung increases its power exposure and blood flow, thereby accentuating its potential for progressive injury.

There are 2 major contributors to this VILI vortex of shrinking the baby lung: airspace VILI8 and intensified stresses within the vessels that perfuse it9,10 (eFigure 2 in the Supplement). Over time, superimposed VILI and unchecked viral disease incite inflammation and edema, promoting local and generalized thrombogenesis, intense cytokine release, right ventricular overload, and systemic organ dysfunction. In this advanced state, it is advisable to apply a more conventional lung-protective strategy: higher PEEP (≤15 cm H2O), lower tidal volume (6 mL/kg), and prone positioning while minimizing oxygen consumption. Whichever the disease type, weaning should be undertaken cautiously (Table).

Table.  Time Course and Treatment Approach to Ventilation Support for Patients With CARDS
Time Course and Treatment Approach to Ventilation Support for Patients With CARDS

COVID-19 causes unique lung injury. It may be helpful to categorize patients as having either type L or H phenotype. Different ventilatory approaches are needed, depending on the underlying physiology.

Back to top
Article Information

Corresponding Author: John J. Marini, MD, Regions Hospital, University of Minnesota, 640 Jackson St, MS11203B, St Paul, MN 55101 (

Published Online: April 24, 2020. doi:10.1001/jama.2020.6825

Conflict of Interest Disclosures: Dr Gattinoni reported receiving fees from General Electric for ventilator development and seminars; a grant from Estor for technical equipment and to perform a study; fees from Masimo for technical consultation and Nutrivent; and a patent licensed to Sidam Biomedical Solutions for development of an esophageal balloon for the measurement of esophageal pressure together with enteral feeding.

Grasselli  G, Zangrillo  A, Zanella  A,  et al; COVID-19 Lombardy ICU Network.  Baseline characteristics and outcomes of 1591 patients infected with SARS-CoV-2 admitted to ICUs of the Lombardy region, Italy.   JAMA. 2020. Published online April 6, 2020. doi:10.1001/jama.2020.5394PubMedGoogle Scholar
Arentz  M, Yim  E, Klaff  L,  et al.  Characteristics and outcomes of 21 critically ill patients with COVID-19 in Washington State.   JAMA. 2020. Published online March 19, 2020. doi:10.1001/jama.2020.4326PubMedGoogle Scholar
Wang  D, Hu  B, Hu  C,  et al.  Clinical characteristics of 138 hospitalized patients with 2019 novel coronavirus-infected pneumonia in Wuhan, China.   JAMA. 2020;323(11):1061-1069. doi:10.1001/jama.2020.1585PubMedGoogle ScholarCrossref
Gattinoni  L, Marini  JJ, Pesenti  A, Quintel  M, Mancebo  J, Brochard  L.  The “baby lung” became an adult.   Intensive Care Med. 2016;42(5):663-673. doi:10.1007/s00134-015-4200-8PubMedGoogle ScholarCrossref
Brochard  L, Slutsky  A, Pesenti  A.  Mechanical ventilation to minimize progression of lung injury in acute respiratory failure.   Am J Respir Crit Care Med. 2017;195(4):438-442. doi:10.1164/rccm.201605-1081CPPubMedGoogle ScholarCrossref
Gattinoni  L, Chiumello  D, Caironi  P,  et al.  COVID-19 pneumonia: different respiratory treatments for different phenotypes?   Intensive Care Med. 2020. doi:10.1007/s00134-020-06033-2PubMedGoogle Scholar
Zhou  F, Yu  T, Du  R,  et al.  Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: a retrospective cohort study.   Lancet. 2020;395(10229):1054-1062. doi:10.1016/S0140-6736(20)30566-3PubMedGoogle ScholarCrossref
Marini  JJ, Rocco  PRM, Gattinoni  L.  Static and dynamic contributors to ventilator-induced lung injury in clinical practice. pressure, energy, and power.   Am J Respir Crit Care Med. 2020;201(7):767-774. doi:10.1164/rccm.201908-1545CIPubMedGoogle ScholarCrossref
Vieillard-Baron  A, Matthay  M, Teboul  JL,  et al.  Experts’ opinion on management of hemodynamics in ARDS patients: focus on the effects of mechanical ventilation.   Intensive Care Med. 2016;42(5):739-749. doi:10.1007/s00134-016-4326-3PubMedGoogle ScholarCrossref
Marini  JJ, Hotchkiss  JR, Broccard  AF.  Bench-to-bedside review: microvascular and airspace linkage in ventilator-induced lung injury.   Crit Care. 2003;7(6):435-444. doi:10.1186/cc2392PubMedGoogle ScholarCrossref
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    6 Comments for this article
    Conditional Management of COVID-19
    Michael McAleer, PhD (Econometrics), Queen's | Asia University, Taiwan
    The illuminating and instructive paper on COVID-19 respiratory distress suggests that extremely careful ventilatory workload management of the vasocentric features is required, failure of which may lead to multiorgan deterioration and ultimate failure.

    On the face of the detailed clinical results, what is especially troubling is that organ failures do not seem to be linked to advanced age or pre-existing comorbidity.

    As different comorbidities might have markedly different impacts in assaulting specific organs, or combinations of organs, additional information on the types of pre-existing comorbidities according to age would help in understanding the cause-and-effect relationships.

    In particular,
    pre-existing lung comorbidities, such as different stages of cancer, and whether primary, secondary or tertiary, would change the control variables, and hence the conditioning set used to provide management of COVID-19 respiratory distress.

    Patient self-induced lung injury might also be considered in light of pre-existing lung comorbidities that could transition to more serious and progressive injuries.
    Compliance Loss and Death in COVID
    John R. Dykers, Jr., MD | Chatham Hospital, Siler City, NC, USA
    If interstitial edema from vascular injury begins to predominate, gaseous diffusion in the alveoli is unlikely to be restored by high PEEP. The high oxygen flow will be less effective and the high PEEP lung damage will be fatal.
    Is Prone Positioning (and high PEEP) a One-Size-Fits-All Strategy for Type 1 Patients With COVID-19 Pneumonia?
    Mohamad Abdelsalam, Master of Medicine | Suez General Hospital, Critical Care Department, Suez, Egypt
    Professor Gattinoni proposed an interesting mechanism by which prone positioning improves oxygenation in type 1 patients with COVID-19 pneumonia who have near-normal compliance with only minimal lung recruitability. According to Professor Gattinoni, prone positioning (and high PEEP) does not improve oxygenation by recruitment of collapsed areas, but through redistribution of pulmonary perfusion and improvement of V/Q matching. The physiologic mechanism underlying oxygenation improvement in prone-positioned patients who have non-recruitable lung is gravity-dependent reduction in perfusion of the basal (non-dependent) regions such that perfusion becomes more properly matched with ventilation. However, for this mechanism to act there must be a “baseline” increase in pulmonary perfusion relative to alveolar ventilation such as might result from loss of hypoxic pulmonary vasoconstriction and/or AT2 receptor-mediated vasodilatation (dysregulated pulmonary perfusion). Obviously, when the "baseline" V/Q abnormality is increased shunt due to increased pulmonary blood flow, measures that reduce blood flow such as prone positioning (and high PEEP) would be expected to improve V/Q ratio and arterial oxygenation—regardless of alveolar recruitment. However, this is unlikely to occur if dead space ventilation (resulting from pulmonary micro-thrombi, right heart failure or AT1 receptor-mediated pulmonary vasoconstriction) is the main mechanism of hypoxemia. By further decreasing pulmonary perfusion, prone positioning (and high PEEP) may actually worsen dead space ventilation and pulmonary gas exchange. The effect of prone positioning (and high PEEP) on arterial oxygenation in type 1 patients is probably determined by which type of V/Q mismatch is present at baseline—shunt vs dead space ventilation. Thus, it is likely that prone positioning (and high PEEP) is not a “one-size-fits-all” strategy for patients with type 1 disease, and that careful evaluation of the underlying physiologic abnormality is mandatory before considering prone positioning (and high PEEP) in these patients.
    Mechanisms of Right-to-Left Shunt in Type 1 COVID-19 Pneumonia
    Mohamad Abdelsalam, Master of Medicine | Suez General Hospital, Critical Care Department, Suez, Egypt
    In a recently published editorial in Critical Care, Professor Gattinoni described type 1 “non-ARDS” COVID-19 pneumonia as having no significant lung areas to recruit (1), but the right-to-left venous admixture is typically around 50%. However, in the absence of significant alveolar collapse/consolidation, such a high degree of venous admixture cannot be fully explained by intra-pulmonary shunting even after considering the increase in pulmonary blood flow due to loss of hypoxic vasoconstriction and AT2 receptor-mediated vasodilatation (dysregulated pulmonary perfusion). Certainly, increased perfusion alone can increase shunt fraction without a concomitant decrease in ventilation. However, the associated increase in cardiac output may counterbalance the increased shunt by decreasing tissue oxygen extraction and increasing mixed venous oxygen saturation. In other words, intra-pulmonary shunt resulting from increased perfusion (with normal ventilation) is not typically associated with the severe hypoxemia often seen in type 1 disease. Another explanation would be intra-cardiac shunt through a patent foramen ovale due to acute pulmonary hypertension resulting from pulmonary micro-thrombi and AT1 receptor-mediated pulmonary vasoconstriction (dysregulated perfusion).


    1. Gattinoni, L., Chiumello, D. & Rossi, S. COVID-19 pneumonia: ARDS or not?. Crit Care 24, 154 (2020).
    Clinical Phenotyping of COVID-19 into L and H Types
    Mohamad Abdelsalam, Master of Medicine | Suez General Hospital, Critical Care Department, Suez, Egypt
    It is hypothesized that COVID-19 respiratory failure has two distinct phenotypes—L and H types that have different clinical and radiological patterns. However, disease progression occurs in only one direction—from L type to H type, but not the other way round; an observation which suggests clinical staging rather than phenotyping. It is assumed that L-to-H progression is due either to the natural course of the disease or to P-SILI resulting from increased respiratory drive of spontaneously breathing patients. P-SILI is essentially a type of lung injury induced by alveolar over-distenstion secondary to high trans-pulmonary pressure. In other words, L type can progress to H type when the lung is overinflated due to large volume ventilation during spontaneous breathing. However, it is also assumed that L type, because of good lung compliance, can accept larger tidal volumes (8 mL/Kg IBW) without increasing the risk of VILI. According to this assumption, the very same L type which is susceptible to P-SILI resulting from high trans-pulmonary pressure during spontaneous ventilation is also immune from VILI during controlled ventilation even when the tidal volume (and driving pressure) is high. What makes L type susceptible to and immune from essentially the same type of lung injury—volutrauma resulting from regional over-inflation? And, what makes COVID-19 progresses from L type to H type, but not in the reverse direction, if they are not actually different stages of severity within the same clinical spectrum?
    CARDs and Coagulopathy
    Camilo Colaco, PhD | ImmunoBiology Ltd
    Could a virally induced coagulopathy be the underlying cause of CARDs and thus require high dose anticoagulant therapy (preferably early when breathing problems mild/moderate)?