COVID-19: the mystery behind “happy” hypoxemia

The epidemiology and clinical characteristics of COVID-19 are much better understood today compared to the early stage of the pandemic several months ago. However, our knowledge of the pathophysiological changes in the lung remains largely elusive. Many patients with COVID-19 present with severe hypoxemia, yet remain remarkably comfortable, a phenomenon that has been alluded to as “happy hypoxemia”. Guan et al. studied 1099 patients with COVID-19 pneumonia admitted to 552 hospitals in mainland China during the initial phase of the pandemic. Although a large number of patients had abnormal CT scans (86%) and required supplemental oxygen (41%), and low PaO2/FiO2 ratios, dyspnea was reported by relatively few (18.7%) patients.1 The severe hypoxemia with relatively little discomfort that is often observed in patients with COVID-19 pneumonia has intrigued clinicians.  The hypoxemia is often out of proportion to the limited parenchymal involvement as evident on CT imaging.2

Mechanisms of hypoxemia in COVID-19

Failure of the hypoxic pulmonary vasoconstrictor response 

A primary vascular disorder may explain severe hypoxemia in the presence of limited lung involvement. The usual compensatory mechanism to decreased ventilation is vasoconstriction of pulmonary vessels in poorly ventilated areas of the lung by the hypoxic pulmonary vasoconstrictor (HPV) response. In COVID-19, a marked failure of the HPV response with intrapulmonary shunt is observed, clearly demonstrated by dual-energy CT scan.3 This vasoplegic state within the lung may be triggered by vasodilator prostaglandins, bradykinin, and other cytokines.4

Microthrombi and hyaline membrane formation

The SARS-CoV-2 virus infects lung endothelial cells that express angiotensin-converting enzyme 2 (ACE2) receptors. Autopsy studies of subjects with severe disease have demonstrated severe endothelial injury, with microthrombi formation, leading to capillary occlusion; extensive arterial thrombosis was also observed.5 Intravascular thrombosis triggered by a hypercoagulable state also leads to impaired ventilation/perfusion (V/Q) matching and hypoxemia. The distinctive vascular changes commonly seen in severe COVID-19 has been referred to as acute vascular distress syndrome (AVDS).6

Gas transfer from the alveoli to the capillaries (diffusion capacity) is also compromised in COVID-19 infection. The SARS-CoV-2 virus infects type II alveolar cells and leads to their destruction. This results in a denuded basement membrane that becomes enveloped by a hyaline membrane comprised of fibrin and dead cells, impairing gas exchange.7 A study of 110 COVID-19 patients revealed significantly impaired diffusing-capacity in 30.4% patients with mild, 42.4% with moderate, and 84.2% of patients with severe COVID-19 pneumonia at the time of discharge from hospital.8 As the diffusion barrier becomes more pronounced, flow through the pulmonary vasculature increases with exertion. The increased flow during exertion leads to a reduction in the time available for gas exchange, with worsening hypoxemia.9

What is the evidence for intrapulmonary shunt?

The likely presence of a significant intrapulmonary shunt in COVID-19 is supported by the study by Masi et al.10  Sixty critically ill patients with COVID-19-related acute respiratory distress syndrome (ARDS) underwent bubble contrast transthoracic echocardiography. Intravenous contrast was injected using 9.5 ml gelatine solution agitated with 0.5 ml of room air. The authors defined a moderate-to-large shunt at the patent foramen ovale level by left heart opacification within three cardiac cycles of complete right heart opacification. A moderate-to-large transpulmonary shunt was defined as left heart opacification after more than three cardiac cycles. In this study, an intrapulmonary shunt was detected in 12 (20%) of patients. 

Reynolds et al. performed contrast-enhanced transcranial Doppler in both middle cerebral arteries in mechanically ventilated patients with COVID-19 by injecting agitated saline through a peripheral or central line. The number of microbubbles in the middle cerebral arteries were counted over a period of 20 seconds by the system software and manually verified. The authors observed detectable microbubbles in 15/18 (83%) patients suggesting a right to left shunt across the lung. The number of microbubbles observed correlated with the degree of hypoxemia.11 Both these studies support the likelihood of an intrapulmonary right to left shunt due to significant pulmonary vasodilatation in COVID-19. 

Mechanisms of dyspnea 

The carotid bodies, about 2 mm in diameter, are located at the bifurcation of the common carotid artery and act as sensors of hypoxemia. They are stimulated by low arterial oxygen levels, and signals are transmitted to the respiratory centers in the medulla oblongata. The respiratory centers, in turn, transmit signals to the diaphragm through the phrenic nerves, leading to increased ventilation, which is perceived as dyspnea. The severity of dyspnea is hence, directly related to the degree of ventilatory response to hypoxemia. Usually, there is little change in the ventilatory response as the PaO2 drops from 90 to 60 mm Hg. The ventilatory response increases exponentially as the PaO2 drops further. Besides, within a PaO2 range of 40–60 mm Hg, the PaCO2 levels must exceed a threshold level of 39 mm Hg to evoke a ventilatory response by carotid body stimulation.12 The SARS-COV-2 virus affects the ACE 2 (angiotensin-converting enzyme 2) receptor, expressed in the carotid bodies. This may lead to impairment of the hypoxia-sensing mechanism and attenuation of the ventilatory response.

Why do COVID-19 patients experience relative comfort in spite of severe hypoxemia?

Different phenotypes based on lung compliance

The lung volume is generally preserved during the early stage of COVID-19 pneumonia. The lung parenchymal involvement is minimal and confined to a few peripheral ground-glass opacities. During this early phase of illness, lung compliance is reasonably well-preserved, and tidal volumes generated are nearly normal, with little change in lung mechanics. This is referred to as the type L phenotype (low elastance, high compliance). During this stage of the disease, the subjective sensation of dyspnea is attenuated, although hypoxemia may be present due to vascular abnormalities in the lung, predominantly characterized by the abolition of the HPV response. As the disease progresses, parenchymal involvement may become more extensive, with more strenuous respiratory effort. The inspiratory pressures become more negative; besides, lung permeability increases resulting in worsening interstitial edema. This phase represents the type H phenotype (high elastance, low compliance). The transition from the stage of high to low compliance may be precipitated by forceful inspiratory effort and generation of increasing negative pressures during spontaneous breathing, described as patient self-inflicted lung injury (P-SILI).13

Other possible factors, not specific to COVID-19

The ventilatory response to hypoxia may be impaired by more than 50% among subjects with diabetes14 and those above 65 years old,15 which include the majority of patients with severe COVID-19 infection. These patients with COVID-19 may not experience the expected level of discomfort associated with hypoxemia. Another factor to consider is the relative inaccuracy of pulse oximetry at low levels of PaO2. The pulse oximeter may underread by up to 7% or more below an SaO2 of 80%.16  Besides, fever, often seen in COVID-19, leads to shift of the oxygen-dissociation curve to the right, resulting in lower SaO2 levels for a given PaO2.17  


  • Severe hypoxemia, with a relative lack of subjective discomfort, is often observed in patients with COVID-19.
  • In the early stage of the disease, there is relatively minimal involvement of the lung parenchyma with preserved lung mechanics; the degree of distress is minimal at this stage.
  • In spite of limited parenchymal involvement and minimal respiratory distress, the lung vasculature may show profound changes attributable to chemokine-mediated vasodilatation, and failure of the HPV response, leading to intrapulmonary right to left shunt, resulting in severe hypoxemia. 
  • Bubble contrast echocardiography and transcranial doppler studies corroborate the existence of a significant intrapulmonary right to left shunt.
  • Furthermore, the ventilatory response to hypoxia may be impaired in the older age group and among diabetic patients. The pulse oximeter usually underestimates SaOat low levels of PaO2; besides, the oxygen-dissociation curve is shifted to the right in the presence of fever. These factors contribute to the low SaO2 levels often observed among COVID-19 patients who are relatively free of distress.


1.         Guan W, Ni Z, Hu Y, et al. Clinical Characteristics of Coronavirus Disease 2019 in China. N Engl J Med. 2020;382(18):1708-1720. doi:10.1056/NEJMoa2002032

2.         Mahjoub Y, Rodenstein DO, Jounieaux V. Severe Covid-19 disease: rather AVDS than ARDS? Crit Care. 2020;24(1):327. doi:10.1186/s13054-020-02972-w

3.         Lang M, Som A, Mendoza DP, et al. Hypoxaemia related to COVID-19: vascular and perfusion abnormalities on dual-energy CT. Lancet Infect Dis. Published online April 30, 2020. doi:10.1016/S1473-3099(20)30367-4

4.         Dhont S, Derom E, Van Braeckel E, Depuydt P, Lambrecht BN. The pathophysiology of ‘happy’ hypoxemia in COVID-19. Respir Res. 2020;21(1):198. doi:10.1186/s12931-020-01462-5

5.         Ackermann M, Verleden SE, Kuehnel M, et al. Pulmonary Vascular Endothelialitis, Thrombosis, and Angiogenesis in Covid-19. N Engl J Med. 2020;383(2):120-128. doi:10.1056/NEJMoa2015432

6.         Mahjoub Y, Rodenstein DO, Jounieaux V. Severe Covid-19 disease: rather AVDS than ARDS? Crit Care. 2020;24(1):327. doi:10.1186/s13054-020-02972-w

7.         Mason RJ. Pathogenesis of COVID-19 from a cell biology perspective. Eur Respir J. 2020;55(4). doi:10.1183/13993003.00607-2020

8.         Mo X, Jian W, Su Z, et al. Abnormal pulmonary function in COVID-19 patients at time of hospital discharge. Eur Respir J. 2020;55(6):2001217. doi:10.1183/13993003.01217-2020

9.         Hopkins SR. Exercise Induced Arterial Hypoxemia: The role of Ventilation-Perfusion Inequality and Pulmonary Diffusion Limitation. In: Roach RC, Wagner PD, Hackett PH, eds. Hypoxia and Exercise. Advances in Experimental Medicine and Biology. Springer US; 2007:17-30. doi:10.1007/978-0-387-34817-9_3

10.       Masi P, Bagate F, d’Humières T, et al. Is hypoxemia explained by intracardiac or intrapulmonary shunt in COVID-19-related acute respiratory distress syndrome? Ann Intensive Care. 2020;10(1):108. doi:10.1186/s13613-020-00726-z

11.       Reynolds AS, Lee AG, Renz J, et al. Pulmonary Vascular Dilatation Detected by Automated Transcranial Doppler in COVID-19 Pneumonia. Am J Respir Crit Care Med. Published online August 6, 2020:rccm.202006-2219LE. doi:10.1164/rccm.202006-2219LE

12.       Mohan R, Duffin J. The effect of hypoxia on the ventilatory response to carbon dioxide in man. Respir Physiol. 1997;108(2):101-115. doi:10.1016/S0034-5687(97)00024-8

13.       Gattinoni L, Chiumello D, Caironi P, et al. COVID-19 pneumonia: different respiratory treatments for different phenotypes? Intensive Care Med. 2020;46(6):1099-1102. doi:10.1007/s00134-020-06033-2

14.       Weisbrod CJ, Eastwood PR, O’Driscoll G, Green DJ. Abnormal ventilatory responses to hypoxia in Type 2 diabetes. Diabet Med J Br Diabet Assoc. 2005;22(5):563-568. doi:10.1111/j.1464-5491.2005.01458.x

15.       Kronenberg RS, Drage CW. Attenuation of the ventilatory and heart rate responses to hypoxia and hypercapnia with aging in normal men. J Clin Invest. 1973;52(8):1812-1819. doi:10.1172/JCI107363

16.       Tobin MJ, Laghi F, Jubran A. Why COVID-19 Silent Hypoxemia Is Baffling to Physicians. Am J Respir Crit Care Med. 2020;202(3):356-360. doi:10.1164/rccm.202006-2157CP

17.       Kelman GR. Digital computer subroutine for the conversion of oxygen tension into saturation. J Appl Physiol. 1966;21(4):1375-1376. doi:10.1152/jappl.1966.21.4.1375

1 thought on “COVID-19: the mystery behind “happy” hypoxemia”

  1. I really enjoy your blog, and, Since you have written about the “happy hypoxemia” in COVID-19 respiratory disease, I thought you would be interested in this paper that we published yesterday, short description follows:

    “This landmark paper provides evidence in support of the globally overlooked diagnosis occurring in COVID-19, an inflammatory response called “Organizing Pneumonia” (OP), demonstrating why OP underlies early COVID19 respiratory disease in hospital patients as evidenced by a comprehensive review of the clinical, radiographic, pathologic, and therapeutic evidence to date. The presence of OP explains the much-remarked upon phenomenon of “happy hypoxia” and the dramatic response to early corticosteroid therapy that have been observed in COVID-19 trials. Finally, the paper strongly emphasizes the importance of higher doses and longer durations of corticosteroid therapy in the treatment of COVID-19 based on OP treatment protocols, suggesting that the RECOVERY trial and current WHO recommended doses are insufficient to treat OP in many given the low dose and short duration of corticosteroids used as evidenced by the high mortality rates in the treatment arm of the RECOVERY trial, namely 23.3% in patients on oxygen and 29% in patients on mechanical ventilators.

    Hope you find it of interest – thanks, Pierre Kory

    Pierre Kory, MD, MPA
    Associate Professor of Medicine
    Advocate Aurora Critical Care Service
    Aurora St. Luke’s Medical Center, Milwaukee, WI
    Founding Member, Front Line COVID-19 Critical Care Alliance,


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s