The intricate physiology of veno-venous extracorporeal membrane oxygenation: an overview for clinicians.

During veno-venous extracorporeal membrane oxygenation (V-V ECMO), blood is drained from the central venous circulation to be oxygenated and decarbonated by an artificial lung. It is then reinfused into the right heart and pulmonary circulation where further gas-exchange occurs. Each of these steps is characterized by a peculiar physiology that this manuscript analyses, with the aim of providing bedside tools for clinical care: we begin by describing the factors that affect the efficiency of blood drainage, such as patient and cannulae position, fluid status, cardiac output and ventilatory strategies. We then dig into the complexity of extracorporeal gas-exchange, with particular reference to the effects of extracorporeal blood-flow (ECBF), fraction of delivered oxygen (FdO2) and sweep gas-flow (SGF) on oxygenation and decarbonation. Subsequently, we focus on the reinfusion of arterialized blood into the right heart, highlighting the effects on recirculation and, more importantly, on right ventricular function. The importance and challenges of haemodynamic monitoring during V-V ECMO are also analysed. Finally, we detail the interdependence between extracorporeal circulation, native lung function and mechanical ventilation in providing adequate arterial blood gases while allowing lung rest. In the absence of evidence-based strategies to care for this particular group of patients, clinical practice is underpinned by a sound knowledge of the intricate physiology of V-V ECMO.

Quantifying pH-induced changes in plasma strong ion difference during experimental acidosis: clinical implications for base excess interpretation.

It is commonly assumed that changes in plasma strong ion difference (SID) result in equal changes in whole blood base excess (BE). However, at varying pH, albumin ionic-binding and transerythrocyte shifts alter the SID of plasma without affecting that of whole blood (SIDwb), i.e., the BE. We hypothesize that, during acidosis, 1) an expected plasma SID (SIDexp) reflecting electrolytes redistribution can be predicted from albumin and hemoglobin's charges, and 2) only deviations in SID from SIDexp reflect changes in SIDwb, and therefore, BE. We equilibrated whole blood of 18 healthy subjects (albumin = 4.8 ± 0.2 g/dL, hemoglobin = 14.2 ± 0.9 g/dL), 18 septic patients with hypoalbuminemia and anemia (albumin = 3.1 ± 0.5 g/dL, hemoglobin = 10.4 ± 0.8 g/dL), and 10 healthy subjects after in vitro-induced isolated anemia (albumin = 5.0 ± 0.2 g/dL, hemoglobin = 7.0 ± 0.9 g/dL) with varying CO2 concentrations (2-20%). Plasma SID increased by 12.7 ± 2.1, 9.3 ± 1.7, and 7.8 ± 1.6 mEq/L, respectively (P < 0.01) and its agreement (bias[limits of agreement]) with SIDexp was strong: 0.5[-1.9; 2.8], 0.9[-0.9; 2.6], and 0.3[-1.4; 2.1] mEq/L, respectively. Separately, we added 7.5 or 15 mEq/L of lactic or hydrochloric acid to whole blood of 10 healthy subjects obtaining BE of -6.6 ± 1.7, -13.4 ± 2.2, -6.8 ± 1.8, and -13.6 ± 2.1 mEq/L, respectively. The agreement between ΔBE and ΔSID was weak (2.6[-1.1; 6.3] mEq/L), worsening with varying CO2 (2-20%): 6.3[-2.7; 15.2] mEq/L. Conversely, ΔSIDwb (the deviation of SID from SIDexp) agreed strongly with ΔBE at both constant and varying CO2: -0.1[-2.0; 1.7], and -0.5[-2.4; 1.5] mEq/L, respectively. We conclude that BE reflects only changes in plasma SID that are not expected from electrolytes redistribution, the latter being predictable from albumin and hemoglobin's charges.NEW & NOTEWORTHY This paper challenges the assumed equivalence between changes in plasma strong ion difference (SID) and whole blood base excess (BE) during in vitro acidosis. We highlight that redistribution of strong ions, in the form of albumin ionic-binding and transerythrocyte shifts, alters SID without affecting BE. We demonstrate that these expected SID alterations are predictable from albumin and hemoglobin's charges, or from the noncarbonic whole blood buffer value, allowing a better interpretation of SID and BE during in vitro acidosis.

Causes of Hypoxemia in COVID-19 Acute Respiratory Distress Syndrome: A Combined Multiple Inert Gas Elimination Technique and Dual-energy Computed Tomography Study.

BACKGROUND: Despite the fervent scientific effort, a state-of-the art assessment of the different causes of hypoxemia (shunt, ventilation-perfusion mismatch, and diffusion limitation) in COVID-19 acute respiratory distress syndrome (ARDS) is currently lacking. In this study, the authors hypothesized a multifactorial genesis of hypoxemia and aimed to measure the relative contribution of each of the different mechanism and their relationship with the distribution of tissue and blood within the lung. METHODS: In this cross-sectional study, the authors prospectively enrolled 10 patients with COVID-19 ARDS who had been intubated for less than 7 days. The multiple inert gas elimination technique (MIGET) and a dual-energy computed tomography (DECT) were performed and quantitatively analyzed for both tissue and blood volume. Variables related to the respiratory mechanics and invasive hemodynamics (PiCCO [Getinge, Sweden]) were also recorded. RESULTS: The sample (51 ± 15 yr; Pao2/Fio2, 172 ± 86 mmHg) had a mortality of 50%. The MIGET showed a shunt of 25 ± 16% and a dead space of 53 ± 11%. Ventilation and perfusion were mismatched (LogSD, Q, 0.86 ± 0.33). Unexpectedly, evidence of diffusion limitation or postpulmonary shunting was also found. In the well aerated regions, the blood volume was in excess compared to the tissue, while the opposite happened in the atelectasis. Shunt was proportional to the blood volume of the atelectasis (R2 = 0.70, P = 0.003). V˙A/Q˙T mismatch was correlated with the blood volume of the poorly aerated tissue (R2 = 0.54, P = 0.016). The overperfusion coefficient was related to Pao2/Fio2 (R2 = 0.66, P = 0.002), excess tissue mass (R2 = 0.84, P < 0.001), and Etco2/Paco2 (R2 = 0.63, P = 0.004). CONCLUSIONS: These data support the hypothesis of a highly multifactorial genesis of hypoxemia. Moreover, recent evidence from post-mortem studies (i.e., opening of intrapulmonary bronchopulmonary anastomosis) may explain the findings regarding the postpulmonary shunting. The hyperperfusion might be related to the disease severity.

State of the art: Monitoring of the respiratory system during veno-venous extracorporeal membrane oxygenation.

Monitoring the patient receiving veno-venous extracorporeal membrane oxygenation (VV ECMO) is challenging due to the complex physiological interplay between native and membrane lung. Understanding these interactions is essential to understand the utility and limitations of different approaches to respiratory monitoring during ECMO. We present a summary of the underlying physiology of native and membrane lung gas exchange and describe different tools for titrating and monitoring gas exchange during ECMO. However, the most important role of VV ECMO in severe respiratory failure is as a means of avoiding further ergotrauma. Although optimal respiratory management during ECMO has not been defined, over the last decade there have been advances in multimodal respiratory assessment which have the potential to guide care. We describe a combination of imaging, ventilator-derived or invasive lung mechanic assessments as a means to individualise management during ECMO.

Combining the Physical-Chemical Approach with Standard Base Excess to Understand the Compensation of Respiratory Acid-Base Derangements: An Individual Participant Meta-analysis Approach to Data from Multiple Canine and Human Experiments.

BACKGROUND: Several studies explored the interdependence between Paco2 and bicarbonate during respiratory acid-base derangements. The authors aimed to reframe the bicarbonate adaptation to respiratory disorders according to the physical-chemical approach, hypothesizing that (1) bicarbonate concentration during respiratory derangements is associated with strong ion difference; and (2) during acute respiratory disorders, strong ion difference changes are not associated with standard base excess. METHODS: This is an individual participant data meta-analysis from multiple canine and human experiments published up to April 29, 2021. Studies testing the effect of acute or chronic respiratory derangements and reporting the variations of Paco2, bicarbonate, and electrolytes were analyzed. Strong ion difference and standard base excess were calculated. RESULTS: Eleven studies were included. Paco2 ranged between 21 and 142 mmHg, while bicarbonate and strong ion difference ranged between 12.3 and 43.8 mM, and 32.6 and 60.0 mEq/l, respectively. Bicarbonate changes were linearly associated with the strong ion difference variation in acute and chronic respiratory derangement (ß-coefficient, 1.2; 95% CI, 1.2 to 1.3; P < 0.001). In the acute setting, sodium variations justified approximately 80% of strong ion difference change, while a similar percentage of chloride variation was responsible for chronic adaptations. In the acute setting, strong ion difference variation was not associated with standard base excess changes (ß-coefficient, -0.02; 95% CI, -0.11 to 0.07; P = 0.719), while a positive linear association was present in chronic studies (ß-coefficient, 1.04; 95% CI, 0.84 to 1.24; P < 0.001). CONCLUSIONS: The bicarbonate adaptation that follows primary respiratory alterations is associated with variations of strong ion difference. In the acute phase, the variation in strong ion difference is mainly due to sodium variations and is not paralleled by modifications of standard base excess. In the chronic setting, strong ion difference changes are due to chloride variations and are mirrored by standard base excess.

Effects of CPAP and FiO2 on respiratory effort and lung stress in early COVID-19 pneumonia: a randomized, crossover study.

BACKGROUND: in COVID-19 acute respiratory failure, the effects of CPAP and FiO2 on respiratory effort and lung stress are unclear. We hypothesize that, in the compliant lungs of early Sars-CoV-2 pneumonia, the application of positive pressure through Helmet-CPAP may not decrease respiratory effort, and rather worsen lung stress and oxygenation when compared to higher FiO2 delivered via oxygen masks. METHODS: In this single-center (S.Luigi Gonzaga University-Hospital, Turin, Italy), randomized, crossover study, we included patients receiving Helmet-CPAP for early (< 48 h) COVID-19 pneumonia without additional cardiac or respiratory disease. Healthy subjects were included as controls. Participants were equipped with an esophageal catheter, a non-invasive cardiac output monitor, and an arterial catheter. The protocol consisted of a random sequence of non-rebreather mask (NRB), Helmet-CPAP (with variable positive pressure and FiO2) and Venturi mask (FiO2 0.5), each delivered for 20 min. Study outcomes were changes in respiratory effort (esophageal swing), total lung stress (dynamic + static transpulmonary pressure), gas-exchange and hemodynamics. RESULTS: We enrolled 28 COVID-19 patients and 7 healthy controls. In all patients, respiratory effort increased from NRB to Helmet-CPAP (5.0 ± 3.7 vs 8.3 ± 3.9 cmH2O, p < 0.01). However, Helmet's pressure decreased by a comparable amount during inspiration (- 3.1 ± 1.0 cmH2O, p = 0.16), therefore dynamic stress remained stable (p = 0.97). Changes in static and total lung stress from NRB to Helmet-CPAP were overall not significant (p = 0.07 and p = 0.09, respectively), but showed high interpatient variability, ranging from - 4.5 to + 6.1 cmH2O, and from - 5.8 to + 5.7 cmH2O, respectively. All findings were confirmed in healthy subjects, except for an increase in dynamic stress (p < 0.01). PaO2 decreased from NRB to Helmet-CPAP with FiO2 0.5 (107 ± 55 vs 86 ± 30 mmHg, p < 0.01), irrespective of positive pressure levels (p = 0.64). Conversely, with Helmet's FiO2 0.9, PaO2 increased (p < 0.01), but oxygen delivery remained stable (p = 0.48) as cardiac output decreased (p = 0.02). When PaO2 fell below 60 mmHg with VM, respiratory effort increased proportionally (p < 0.01, r = 0.81). CONCLUSIONS: In early COVID-19 pneumonia, Helmet-CPAP increases respiratory effort without altering dynamic stress, while the effects upon static and total stress are variable, requiring individual assessment. Oxygen masks with higher FiO2 provide better oxygenation with lower respiratory effort. Trial registration Retrospectively registered (13-May-2021): clinicaltrials.gov (NCT04885517), https://clinicaltrials.gov/ct2/show/NCT04885517 .

Clinical impact of screening computed tomography in extracorporeal membrane oxygenation: a retrospective cohort study.

BACKGROUND: Data on the prevalence and clinical impact of extrapulmonary findings at screening computed tomography (CT) on initiation of veno-venous extracorporeal membrane oxygenation (V-V ECMO) are limited. We aimed to identify the prevalence of extrapulmonary findings on screening CT following V-V ECMO initiation. We hypothesized that extrapulmonary findings would influence clinical management and outcome. METHODS: Retrospective analysis (2011-2021) of admission screening CT including head, abdomen and pelvis with contrast of consecutive patients on initiation of V-V ECMO. CT findings identified by the attending consultant radiologist were extracted. Demographics, admission physiological and laboratory data, clinical decision-making following CT and ECMO ICU mortality were recorded from the electronic medical record. We used multivariable logistic regression and Kaplan-Meier curves to evaluate associations between extrapulmonary findings and ECMO ICU mortality. RESULTS: Of the 833 patients receiving V-V ECMO, 761 underwent routine admission CT (91.4%). ECMO ICU length of stay was 19 days (IQR 12-23); ICU mortality at the ECMO centre was 18.9%. An incidental extrapulmonary finding was reported in 227 patients (29.8%), leading to an invasive procedure in 12/227 cases (5.3%) and a change in medical management (mainly in anticoagulation strategy) in 119/227 (52.4%). Extrapulmonary findings associated with mortality were intracranial haemorrhage (OR 2.34 (95% CI 1.31-4.12), cerebral infarction (OR 3.59 (95% CI 1.26-9.86) and colitis (OR 2.80 (95% CI 1.35-5.67). CONCLUSIONS: Screening CT frequently identifies extrapulmonary findings of clinical significance. Newly detected intracranial haemorrhage, cerebral infarction and colitis were associated with increased ICU mortality.

Prevalence and Indications for Oxygenator Circuit Replacement in Patients Receiving Venovenous Extracorporeal Membrane Oxygenation.

In this retrospective observational cohort study, we aimed to describe the rate of extracorporeal membrane oxygenation (ECMO) circuit change, the associated risk factors and its relationship with patient characteristics and outcome in patients receiving venovenous (VV) ECMO at our center between January 2015 and November 2017. Twenty-seven percent of the patients receiving VV ECMO (n = 224) had at least one circuit change, which was associated with lower ICU survival (68% vs 82% p=0.032) and longer ICU stay (30 vs . 17 days p < 0.001). Circuit duration was similar when stratified by gender, clinical severity, or prior circuit change. Hematological abnormalities and increased transmembrane lung pressure (TMLP) were the most frequent indication for circuit change. The change in transmembrane lung resistance (Δ TMLR) gave better prediction of circuit change than TMLP, TMLR, or ΔTMLP. Low postoxygenator PO 2 was indicated as a reason for one-third of the circuit changes. However, the ECMO oxygen transfer was significantly higher in cases of circuit change with documented "low postoxygenator PO 2 " than those without (244 ± 62 vs. 200 ± 57 ml/min; p = 0.009). The results suggest that circuit change in VV ECMO is associated with worse outcomes, that the Δ TMLR is a better predictor of circuit change than TMLP, and that the postoxygenator PO 2 is an unreliable proxy for the oxygenator function.

Physiological adaptations during weaning from veno-venous extracorporeal membrane oxygenation.

Veno-venous extracorporeal membrane oxygenation (V-V ECMO) has an established evidence base in acute respiratory distress syndrome (ARDS) and has seen exponential growth in its use over the past decades. However, there is a paucity of evidence regarding the approach to weaning, with variation of practice and outcomes between centres. Preconditions for weaning, management of patients' sedation and mechanical ventilation during this phase, criteria defining success or failure, and the optimal duration of a trial prior to decannulation are all debated subjects. Moreover, there is no prospective evidence demonstrating the superiority of weaning the sweep gas flow (SGF), the extracorporeal blood flow (ECBF) or the fraction of oxygen of the SGF (FdO2), thereby a broad inter-centre variability exists in this regard. Accordingly, the aim of this review is to discuss the required physiological basis to interpret different weaning approaches: first, we will outline the physiological changes in blood gases which should be expected from manipulations of ECBF, SGF and FdO2. Subsequently, we will describe the resulting adaptation of patients' control of breathing, with special reference to the effects of weaning on respiratory effort. Finally, we will discuss pertinent elements of the monitoring and mechanical ventilation of passive and spontaneously breathing patients during a weaning trial. Indeed, to avoid lung injury, invasive monitoring is often required in patients making spontaneous effort, as pressures measured at the airway may not reflect the degree of lung strain. In the absence of evidence, our approach to weaning is driven largely by an understanding of physiology.

End-Tidal to Arterial Pco2 Ratio as Guide to Weaning from Venovenous Extracorporeal Membrane Oxygenation.

Rationale: Weaning from venovenous extracorporeal membrane oxygenation (VV-ECMO) is based on oxygenation and not on carbon dioxide elimination. Objectives: To predict readiness to wean from VV-ECMO. Methods: In this multicenter study of mechanically ventilated adults with severe acute respiratory distress syndrome receiving VV-ECMO, we investigated a variable based on CO2 elimination. The study included a prospective interventional study of a physiological cohort (n = 26) and a retrospective clinical cohort (n = 638). Measurements and Main Results: Weaning failure in the clinical and physiological cohorts were 37% and 42%, respectively. The main cause of failure in the physiological cohort was high inspiratory effort or respiratory rate. All patients exhaled similar amounts of CO2, but in patients who failed the weaning trial, [Formula: see text]e was higher to maintain the PaCO2 unchanged. The effort to eliminate one unit-volume of CO2, was double in patients who failed (68.9 [42.4-123] vs. 39 [20.1-57] cm H2O/[L/min]; P = 0.007), owing to the higher physiological Vd (68 [58.73] % vs. 54 [41.64] %; P = 0.012). End-tidal partial carbon dioxide pressure (PetCO2)/PaCO2 ratio was a clinical variable strongly associated with weaning outcome at baseline, with area under the receiver operating characteristic curve of 0.87 (95% confidence interval [CI], 0.71-1). Similarly, the PetCO2/PaCO2 ratio was associated with weaning outcome in the clinical cohort both before the weaning trial (odds ratio, 4.14; 95% CI, 1.32-12.2; P = 0.015) and at a sweep gas flow of zero (odds ratio, 13.1; 95% CI, 4-44.4; P < 0.001). Conclusions: The primary reason for weaning failure from VV-ECMO is high effort to eliminate CO2. A higher PetCO2/PaCO2 ratio was associated with greater likelihood of weaning from VV-ECMO.

Mechanical power thresholds during mechanical ventilation: An experimental study.

The extent of ventilator-induced lung injury may be related to the intensity of mechanical ventilation--expressed as mechanical power. In the present study, we investigated whether there is a safe threshold, below which lung damage is absent. Three groups of six healthy pigs (29.5 ± 2.5 kg) were ventilated prone for 48 h at mechanical power of 3, 7, or 12 J/min. Strain never exceeded 1.0. PEEP was set at 4 cmH2 O. Lung volumes were measured every 12 h; respiratory, hemodynamics, and gas exchange variables every 6. End-experiment histological findings were compared with a control group of eight pigs which did not undergo mechanical ventilation. Functional residual capacity decreased by 10.4% ± 10.6% and 8.1% ± 12.1% in the 7 J and 12 J groups (p = 0.017, p < 0.001) but not in the 3 J group (+1.7% ± 17.7%, p = 0.941). In 3 J group, lung elastance, PaO2 and PaCO2 were worse compared to 7 J and 12 J groups (all p < 0.001), due to lower ventilation-perfusion ratio (0.54 ± 0.13, 1.00 ± 0.25, 1.78 ± 0.36 respectively, p < 0.001). The lung weight was lower (p < 0.001) in the controls (6.56 ± 0.90 g/kg) compared to 3, 7, and 12 J groups (12.9 ± 3.0, 16.5 ± 2.9, and 15.0 ± 4.1 g/kg, respectively). The wet-to-dry ratio was 5.38 ± 0.26 in controls, 5.73 ± 0.52 in 3 J, 5.99 ± 0.38 in 7 J, and 6.13 ± 0.59 in 12 J group (p = 0.03). Vascular congestion was more extensive in the 7 J and 12 J compared to 3 J and control groups. Mechanical ventilation (with anesthesia/paralysis) increase lung weight, and worsen lung histology, regardless of the mechanical power. Ventilating at 3 J/min led to better anatomical variables than at 7 and 12 J/min but worsened the physiological values.

Albumin Oxidation Status in Sepsis Patients Treated With Albumin or Crystalloids.

Inflammation and oxidative stress characterize sepsis and determine its severity. In this study, we investigated the relationship between albumin oxidation and sepsis severity in a selected cohort of patients from the Albumin Italian Outcome Study (ALBIOS). A retrospective analysis was conducted on the oxidation forms of human albumin [human mercapto-albumin (HMA), human non-mercapto-albumin form 1 (HNA1) and human non-mercapto-albumin form 2 (HNA2)] in 60 patients with severe sepsis or septic shock and 21 healthy controls. The sepsis patients were randomized (1:1) to treatment with 20% albumin and crystalloid solution or crystalloid solution alone. The albumin oxidation forms were measured at day 1 and day 7. To assess the albumin oxidation forms as a function of oxidative stress, the 60 sepsis patients, regardless of the treatment, were grouped based on baseline sequential organ failure assessment (SOFA) score as surrogate marker of oxidative stress. At day 1, septic patients had significantly lower levels of HMA and higher levels of HNA1 and HNA2 than healthy controls. HMA and HNA1 concentrations were similar in patients treated with albumin or crystalloids at day 1, while HNA2 concentration was significantly greater in albumin-treated patients (p < 0.001). On day 7, HMA was significantly higher in albumin-treated patients, while HNA2 significantly increased only in the crystalloids-treated group, reaching values comparable with the albumin group. When pooling the septic patients regardless of treatment, albumin oxidation was similar across all SOFA groups at day 1, but at day 7 HMA was lower at higher SOFA scores. Mortality rate was independently associated with albumin oxidation levels measured at day 7 (HMA log-rank = 0.027 and HNA2 log-rank = 0.002), irrespective of treatment group. In adjusted regression analyses for 90-day mortality, this effect remained significant for HMA and HNA2. Our data suggest that the oxidation status of albumin is modified according to the time of exposure to oxidative stress (differences between day 1 and day 7). After 7 days of treatment, lower SOFA scores correlate with higher albumin antioxidant capacity. The trend toward a positive effect of albumin treatment, while not statistically significant, warrants further investigation.

Prevalence and outcome of silent hypoxemia in COVID-19.

BACKGROUND: In the early stages of COVID-19 pneumonia, hypoxemia has been described in absence of dyspnea ("silent" or "happy" hypoxemia). Our aim was to report its prevalence and outcome in a series of hypoxemic patients upon Emergency Department admission. METHODS: In this retrospective observational cohort study we enrolled a study population consisting of 213 COVID-19 patients with PaO2/FiO2 ratio <300 mmHg at hospital admission. Two groups (silent and dyspneic hypoxemia) were defined. Symptoms, blood gas analysis, chest X-ray (CXR) severity, need for intensive care and outcome were recorded. RESULTS: Silent hypoxemic patients (68-31.9%) compared to the dyspneic hypoxemic patients (145-68.1%) showed greater frequency of extra respiratory symptoms (myalgia, diarrhea and nausea) and lower plasmatic LDH. PaO2/FiO2 ratio was 225±68 mmHg and 192±78 mmHg in silent and dyspneic hypoxemia respectively (P=0.002). Eighteen percent of the patients with PaO2/FiO2 from 50 to 150 mmHg presented silent hypoxemia. Silent and dyspneic hypoxemic patients had similar PaCO2 (34.2±6.8 mmHg vs. 33.5±5.7 mmHg, P=0.47) but different respiratory rates (24.6±5.9 bpm vs. 28.6±11.3 bpm respectively, P=0.002). Even when CXR was severely abnormal, 25% of the population was silent hypoxemic. Twenty-six point five percent and 38.6% of silent and dyspneic patients were admitted to the ICU respectively (P=0.082). Mortality rate was 17.6% and 29.7% (log-rank P=0.083) in silent and dyspneic patients. CONCLUSIONS: Silent hypoxemia is remarkably present in COVID-19. The presence of dyspnea is associated with a more severe clinical condition.

The impact of ventilation-perfusion inequality in COVID-19: a computational model.

COVID-19 infection may lead to acute respiratory distress syndrome (CARDS) where severe gas exchange derangements may be associated, at least in the early stages, only with minor pulmonary infiltrates. This may suggest that the shunt associated to the gasless lung parenchyma is not sufficient to explain CARDS hypoxemia. We designed an algorithm (VentriQlar), based on the same conceptual grounds described by J.B. West in 1969. We set 498 ventilation-perfusion (VA/Q) compartments and, after calculating their blood composition (PO2, PCO2, and pH), we randomly chose 106 combinations of five parameters controlling a bimodal distribution of blood flow. The solutions were accepted if the predicted PaO2 and PaCO2 were within 10% of the patient's values. We assumed that the shunt fraction equaled the fraction of non-aerated lung tissue at the CT quantitative analysis. Five critically-ill patients later deceased were studied. The PaO2/FiO2 was 91.1 ± 18.6 mmHg and PaCO2 69.0 ± 16.1 mmHg. Cardiac output was 9.58 ± 0.99 L/min. The fraction of non-aerated tissue was 0.33 ± 0.06. The model showed that a large fraction of the blood flow was likely distributed in regions with very low VA/Q (Qmean = 0.06 ± 0.02) and a smaller fraction in regions with moderately high VA/Q. Overall LogSD, Q was 1.66 ± 0.14, suggestive of high VA/Q inequality. Our data suggest that shunt alone cannot completely account for the observed hypoxemia and a significant VA/Q inequality must be present in COVID-19. The high cardiac output and the extensive microthrombosis later found in the autopsy further support the hypothesis of a pathological perfusion of non/poorly ventilated lung tissue.NEW & NOTEWORTHY Hypothesizing that the non-aerated lung fraction as evaluated by the quantitative analysis of the lung computed tomography (CT) equals shunt (VA/Q = 0), we used a computational approach to estimate the magnitude of the ventilation-perfusion inequality in severe COVID-19. The results show that a severe hyperperfusion of poorly ventilated lung region is likely the cause of the observed hypoxemia. The extensive microthrombosis or abnormal vasodilation of the pulmonary circulation may represent the pathophysiological mechanism of such VA/Q distribution.