What does FiO2

Pathophysiology and ventilation principles in COVID-19 (as of April 26, 2020)


At the beginning of the global spread of SARS-CoV-2 it was postulated that the development of a classic ARDS in severe forms underlies the frequent respiratory insufficiencies in COVID-19 patients, there is increasing uncertainty about the generalizability of this assumption.Rather, a distinction must probably be made between different etiologies or forms of respiratory insufficiency.

According to current theories and knowledge, possible causes can be virus pneumonia, ARDS (-like picture) and possibly also (micro) thromboembolism of the pulmonary vessels. Endotheliitis also seems to play a prominent role. While the diagnosis of pneumonia can be made through fever or a suspected respiratory infection, high respiratory rate, low saturation values ​​and dyspnea, the diagnosis of ARDS according to the Berlin definition is based on the oxygenation index (paO2 / FiO2).

As a result, the therapy recommendations with regard to ventilation therapy may have to be adapted and individual and differentiated considerations for ventilation must be made for each patient. However, this is difficult at this early point in the pandemic, as a systematic classification of patients is currently not yet possible with certainty and therefore guidelines of guideline quality cannot be made, nor can the likelihood of a response to a specific therapy be estimated.

At this point we try to give an overview of the various explanatory and therapeutic approaches and to summarize possible strategies for ventilation in order to be able to derive possible conclusions for practice. A statement about the clinical relevance and correctness of these theories is not yet fully possible at this early stage. The transferability to the specific clinical case must therefore be carefully checked. The purpose of this article is to find out about possible factors that lead to respiratory insufficiency and how these can be treated so that they are not disregarded in everyday clinical practice.

Certainly there is not only black and white with COVID-19 and overlapping or merging symptoms and manifestations of the clinical picture must be expected.

With every published new theory, critics who want to refute the newly established theses also become loud.

The similarity to HAPE (High Altitude Lung Edema) postulated by some authors does not seem to exist and therefore COVID-19 should not be treated based on this clinical picture.

Non-indicated or pathophysiological therapies should be refrained from, even if the level of knowledge about COVID-19 is not yet extensive.

The top priority in every phase of treatment is the protection of personnel by means of suitable personal protective equipment. Fear of infection should not be the primary reason for intubation.


The primary replication of SARS-CoV-2 takes place in the upper bronchial epithelial cells and the mucosal cells of the gastrointestinal tract. If the infection is not immunologically controlled at this early stage of the disease, the virus replicates in the alveolar epithelial cells and thus in the lungs as the primary target organ. This happens in the early infection phase. An inflammatory reaction occurs with local vasodilation, increased endothelial permeability and invasion of leukocytes. The pneumonia caused by this leads to hypoxemia and cardiovascular stress reactions. The strong local inflammation causes severe damage to the lung parenchyma with subsequent respiratory insufficiency. The severity of the respiratory insufficiency depends on the severity of the infection, the immune response and the general constitution of the patient, but also on the patient's own compensation mechanisms - i.e. the respiratory drive - and the time span between the appearance of the first clinical symptoms and the start of clinical treatment. On the one hand, there is pronounced hypoxemia, but also recurrent, considerable increases in CO2 in patients with increased alveolo-arterial oxygen differences and a high difference in arterial and end-tidal CO2. In addition, hyperinflammation with a cytokine storm can occur in a late phase. This means that COVID-19 can be divided into three phases.

  • Phase I: early infection
  • Phase II: pulmonary disease
  • IIa: without hypoxia (paO2 / FiO2> 300 or paO2> 63 mmHg for room air)
  • IIb: with hypoxia (paO2 / FiO2 <300 or paO2 <63 mmHg for room air)
  • Phase III: hyperinflammatory disease

Cardiac consequences of pulmonary pathology

In addition to pulmonary problems, cardiac comorbidity must also be taken into account in every phase of treatment. In addition to possible myocarditis, other causes of cardiac complications also play a role in COVID-19: Hypoxemia leads to increased cardiac work to ensure tissue perfusion by increasing cardiac output. As a result of the compensatory hyperventilation, there is an increase in afterload for the left ventricle and due to the increased negative intrathoracic pressures, possibly due to an increase in the preload, an increased load on the right heart. Cardiac arrhythmias, cardiac insufficiency and coronary events as well as (micro) embolism of the pulmonary arteries are also favored. Regular determinations of troponin in serum to evaluate a cardiac component are recommended.

Clinical course

Typically, dyspnea as a symptom of respiratory failure in COVID-19 does not appear until 6.5 days after the onset of the first clinical symptoms (phase II). This is the point in time from which the patients are mainly treated as inpatients.

This mainly affects older or chronically ill patients, but young patients can also develop pronounced and life-threatening symptoms and should therefore be closely monitored and, if the symptoms worsen, be treated immediately by intensive care. The fact that many COVID-19 patients remain asymptomatic for a long time despite already poor arterial oxygen saturation values ​​and deteriorate within a very short time with decompensation is deceptive. A median ARDS occurs 2.5 days after the onset of dyspnea.

Regular measurement of vital parameters such as respiratory rate and saturation, but also pulse, blood pressure and temperature are therefore essential in order to identify high-risk patients and to recognize the transition from phase IIa to phase IIb. In addition to the (q) SOFA score, further instruments for this are the laboratory determination of D-dimers, LDH, CRP, IL-6, ferritin and (differential) blood counts with special attention to the lymphocytes. Arterial or capillary blood gas analyzes in room air should also be carried out to assess the extent of the respiratory insufficiency. Another possible predictor for an impending respiratory decompensation could be a decreased CO2 in the exhaled air. British authors explain this with the fact that hyperventilation to compensate for the hypoxemia leads to hypocapnia. Due to its high diffusivity, CO2 can still be exhaled while oxygen uptake is already disturbed. It is noteworthy that these patients do not feel and present themselves to be airless for a long time ("happy hypoxic"), since the respiratory alkalosis may lead to a left shift in the oxygen binding curve of the hemoglobin and thus to an increased oxygen affinity. Relatively good oxygen saturation values ​​can be measured here despite a low oxygen partial pressure. If the CO2 increases in the course of the aggravation of the respiratory insufficiency, the patient's condition deteriorates rapidly. This should also be taken into account during intubation and subsequent ventilation, since an increase in paCO2 leads to rapid desaturation due to a shift in the oxygen binding curve to the right.

pathophysiological theories of respiratory failure and corresponding therapeutic consequences

L and H type

In contrast to the initial hypothesis that the pulmonary pathology in COVID-19 is a classic ARDS, Gattoni et al. developed the theory that a distinction must be made between two main phenotypes with different treatment responses, since the ARDS definition cannot be applied to all patients with respiratory failure. This model is based on CT examinations and still requires clinical and histopathological validation.

The L-type (low = low -> L-type;> 50% of patients) is characterized by low elastance (= good compliance), a low ventilation-perfusion ratio, a low lung weight and low recruitment.

Overall, the lungs are well ventilated, the typical frosted glass-like opacities can be found subpleurally or in the fissures of the lungs. Severe pulmonary edema is absent; in contrast to the atelectasis-related mechanism in ARDS, the hypoxemia can a.e. can be explained by a loss of pulmonary autoregulation of the vessels, which - due to pronounced vasodilation in the diseased, poorly ventilated lung areas - leads to an increased right-left shunt. To compensate for this, there is an increase in the MV, mainly via an increase in the tidal volume due to more negative intrathoracic pressures. This results in poor recruitment and the frequent lack of subjective dyspnea. The difference to the classic ARDS (see below) becomes clear in the preserved lung mechanics with good compliance. If the patient is clinically inconspicuous, low saturation values ​​are possibly tolerable and intubation can be dispensed with.

The H type (high = high -> H-type; 20-30% of the patients) are characterized by high elastance (= low compliance), a high ventilation-perfusion ratio, a high lung weight and high recruitment.

These patients therefore have a typical ARDS (see below) with a relevant right-left shunt in the context of viral pneumonia, which CT-morphologically is characterized by extensive consolidations typical of ARDS. Invasive ventilation is often unavoidable, although the response to prone positioning (proning) may depend on the aetiology of the ARDS: the response is better in the secondary than in the primary ARDS.

As a result of high negative inspiratory pressures and the resulting larger intrathoracic pressure fluctuations (approx. 15 cm H2O) to generate higher tidal volumes combined with increased inflammation-related permeability of the pulmonary vessels, there could be a transition from the L to the H type with the development of severe interstitial edema. Early intubation should probably be favored here. However, one should not ignore the fact that too early intubation and aggressive ventilation strategies can possibly also promote a transition from L- to H-type iatrogenically.

In contrast to the H-type, which should be treated like a typical ARDS and with invasive ventilation, with the L-type a more cautious approach in the form of a step therapy is probably the means of choice:
initially oxygen administration via mask in case of hypoxemia (less aerosol formation than nasal cannula)
non-invasive procedures such as HFNC and NIV in patients with dyspnea, as long as the patient benefits from them, taking into account self-protection (aerosol formation!)
Critical indication of intubation, but avoid emergency intubation
Avoid aggressive ventilation parameters if possible (high PEEP & peak pressure), as compliance is usually good, possibly also tolerating lower oxygenation values ​​(lactate?)
in hypercapnia, higher tidal volumes (up to 8 ml / kg body weight) with good compliance may be less critical than ventilation in ARDS
Prone positioning of intubated patients only in the case of Profit, selfproning in spontaneously breathing patients (even with non-invasive procedures) but worthy of support

To differentiate between L and H types see also https://m4mvscovid.de/de/intensivstation/warum-covid-19-kein-klassisches-ards-ist/


The term ARDS (Acute Respiratory Distress Syndrome or Acute Respiratory Distress Syndrome) encompasses acute pulmonary dysfunction in the sense of an oxygenation disorder of various causes, but excluding purely cardiac or hydropic genesis, in which radiological condensation of the lung structure occurs. Hyperinflammation leads to increased endo- and epithelial permeability and, consecutively, to pulmonary, intraalveolar edema with difficult gas exchange and thus to oxygenation disorders. Functional right-left shunt occurs as a result of atelectasis, which is favored by reduced surfactant formation. Pulmonary fibrosis is a complication in the further course of ARDS. Such a clinical picture can be caused by SARS-CoV-2.

Radiologically, many COVID-19 patients have ARDS-typical bilateral, disseminated, but above all subpleurally located shadows. (Ro-thorax images). Due to the high infectivity of the pathogen, however, transport and repeated radiological examinations should be avoided if possible. For this reason, too, lung ultrasound is increasingly becoming the focus of imaging diagnostics for making diagnoses and monitoring the progress. According to current studies, ultrasound of the lungs is equivalent to CT diagnostics in ARDS and can provide indications of the presence of lung damage by the virus and the resulting inflammatory reaction even before clinical deterioration.

In ventilated patients, the severity of the ARDS can be divided into, using BGA and ventilation parameters by calculating the so-called Horovitz index

  • mild: paO2 / FiO2> 200 but ≤ 300 mmHg
  • moderate: paO2 / FiO2> 100 but ≤ 200 mmHg
  • difficult: paO2 / FiO2 ≤ 100 mmHg

Note: The paO2 value is expressed here in mmHg, FiO2 is a dimensionless number (fraction). For calculation purposes, an oxygen content of 35% in the inspiratory air means, for example, an FiO2 of 0.35. 100% therefore corresponds to an FiO2 of 1.

A calculation example:
paO2 in the BGA 114 mmHg, 65% oxygen in the breathing air on the device:
114 mmHg: 0.65 = 175.38 mmHg -> falls into the “moderate ARDS” category

Ventilation therapy for ARDS

While avoiding invasive ventilation was previously the primary goal and the use of high-flow O2 therapy and NIV was preferred in ARDS, the ventilation management of ARDS is increasingly being discussed.

In mild ARDS, oxygenation via high-flow nasal cannula (HFNC) and non-invasive ventilation (NIV) can be attempted, taking into account the criteria for lung-protective ventilation (see Possible oxygenation routes). It is important to recognize failure of these procedures at an early stage. The failure rates for the NIV are 50%, with mortality rates of again 50%. In the absence of an early response (FiO2 <150 or 175 mmHg after one to two hours) to NIV, intubation should therefore be carried out, since intubation in these patients can be delayed but ultimately not avoided.

The S3 guideline on invasive ventilation recommends that patients with a Horovitz index <100 mmHg should be intubated and invasively ventilated, although one study showed increased mortality from an index of <150 mmHg (for information on intubation, see Sect. Training: Intubation & Intubation Algorithm). Further criteria for intubation are increasing O2 demand, steadily or rapidly decreasing SaO2, increasing breathing rate> 30 / min and increasing work of breathing. Since COVID-19 can lead to acute aggravation of the hypoxemia and a rapid deterioration in condition, intensive monitoring of the patient is essential in order to identify the need for intubation in good time.

So far, there are no good data on the ventilation modes to be selected in the initial phase after intubation (0-48h). Overall, however, a pressure-controlled ventilation mode (e.g. BIPAP) could be beneficial from a physiological point of view. Spontaneous breathing activity is also possible in this ventilation mode. To optimize ventilation, deeper sedation and, in individual cases, relaxation can be considered in the initial phase in order to ensure better compliance and to improve the oxygenation index while at the same time lowering the ventilation pressures required.

If the clinical outcome is good, the aim should be to achieve rapid spontaneous breathing, e.g. in CPAP mode, by reducing the analgesic sedation. As with controlled ventilation, care should be taken to protect the lungs (see below). If longer ventilation therapy is necessary for more severe forms of ARDS with episodes of prone positioning, this requires deep analgesic sedation up to around RASS -4 with possible relaxation.

In order to reduce atelectasis and the resulting right-left shunt as well as pulmonary edema, an increased PEEP is recommended, which can be further adjusted with the help of the FiO2-PEEP tables of the ARDS network (see below) as an orientation aid if the FiO2 requirement increases can. If possible, ventilation should initially be based on the low PEEP table.

  • initial setting according to low PEEP table according to ARDS network
  • possibly.In the case of severe ARDS, initial setting according to the high PEEP table (should be discussed with an experienced intensive care doctor)

As a rule of thumb, patients with critical hemodynamics or good compliance should preferably be ventilated according to the low PEEP table; for patients with low compliance (e.g. obese patients, severe ARDS), the high PEEP table is more recommended.

In general, measurement maneuvers (e.g. Best PEEP Trial) should be carried out to check whether the patient benefits from an increase in PEEP.

In the case of respiratory improvement, the FiO2 should first be reduced before the PEEP is reduced.

Lung protection is the focus of ARDS ventilation. In order to avoid further damage to the lungs from ventilator-induced mechanisms (VILI) such as excessive tidal volumes, high shear forces and high transpulmonary pressures, a few principles must be observed. First and foremost, the tidal volumes should be selected as low as possible and a driving pressure (difference between peak pressure and PEEP) higher than 15 mbar should be avoided.

The target parameters for ventilation are SaO2 = 90-94% and paO2 = 60-80 mmHg (with previously known COPD, possibly also> 55 mmHg). Permissive hypercapnia up to a paCO2 <80 mmHg and pH values ​​up to> 7.2 is possible; a buffer attempt can be made if necessary.

Possible parameters for entering into lung protective ventilation are as follows:

  • low tidal volume (4-) 5-6 ml / kg ideal body weight (measure patient, size -100)
  • PEEP ≥ 10 mbar
  • Peak pressure ≤ 30 mbar
  • FiO2 0.5 (depending on the value in the BGA)
  • Breathing rate 16-20 / min
  • I: E = 1: 1.5

Apply as needed

  • PEEP increase to 16-20 mbar
  • Increase in MV by increasing AF to 30 / min
  • Adaptation of the I: E to
    • 1: 1 with leading oxygenation disorder
    • 1: 2 with paCO2 / etCO2 increase or air trapping / intrinsic PEEP - recognizable by the termination of expiration in the flow curve of the ventilator
    • CAVE: an inverse I: E should not be used (e.g. 1.5: 1 or 2: 1)

If these adjustments do not lead to an improvement in oxygenation (paO2 / FiO2 <150 mmHg) or if pulmonary sonography indicates the presence of atelectasis, a consistent prone position for 16 hours alternating with supine position (e.g. for 8 hours). This should be done in several cycles, recommendations speak of up to 7 repetitions. There are contraindications to be observed here! If necessary, lateral positions of 90/135 ° with emphasis on one side of the lung and improved oxygenation when lying on the "healthier side" (= good lung down) are possible. The radiological / sonographic findings should also be taken into account here.

If the above measures do not achieve a sufficient improvement in oxygenation, NO inhalation can be considered in individual cases. The indication for ECMO therapy must be considered early and broadly (Horovitz index <60 mmHg).

In the case of other primary diseases as a cause of ARDS, the mortality of the severe form is up to 60%, which is why a high case fatality rate must also be expected with ARDS due to COVID-19.

Microembolism & Endotheliitis

Experience from intensive therapy in British patients suggests that in the initial phase of respiratory failure in some patients, microembolic pulmonary vascular occlusion may be at the fore of the oxygenation disorder. There are both observations that can possibly be explained by an increase in procoagulatory factors. The compliance of the lungs is not restricted. ARDS and bacterial superinfections as well as the resulting problems are only relevant in the later course.

The British colleagues conclude from this that ventilation should not be made too aggressive in this phase of the disease in order to avoid negative effects in later stages of the disease due to fan-associated damage. A low initial PEEP (approx. 10 mbar) seems to be sufficient for many patients.

In this phase of the disease, too, early prone positioning is recommended by improving the perfusion ratio (from paO2 / FiO2 <120 mmHg). NO inhalation or the inhalation or intravenous administration of prostacyclin could possibly also improve the perfusion situation and thus the gas exchange.

More recent pathological evaluations of tissue samples from COVID-19 patients have shown that virus-related endothelial damage in the context of endotheliitis is a serious pathomechanism in severe COVID-19 courses. This means that COVID-19 is not just a lung disease. The clinical consequences that can be drawn from this are still unclear.

There are currently no well-founded recommendations for anticoagulation in COVID-19 (unfractionated heparin vs. LMWH, therapeutic vs. prophylactic dosage).

Practical consequences for ventilation therapy in COVID-19

In the synopsis of the previous knowledge and conclusive theories on pulmonary pathophysiology in COVID-19, a step-by-step approach, weighing the advantages and disadvantages of non-invasive oxygenation routes compared to invasive ventilation, appears to be useful. At the beginning of this step therapy, O2 should be administered via an oxygen mask. If this is no longer sufficient, you should switch to HFNC and NIV.

In the initial stage of the pandemic, non-invasive oxygenation routes such as (O2 nasal cannula / mask with higher flow rates, HFNC and NIV) were not recommended due to the high potential for aerosol formation and the possibly poorer outcome of patients with late intubation. With increasing experience with COVID-19, according to the current state of knowledge, under certain conditions (NIV using a helmet instead of a mask, non-vented masks with virus-proof filters, HFNC with low flow), these procedures could possibly be safe for the staff and, depending on the etiology of the respiratory insufficiency, beneficial be for the patient in order to avoid intubation and possible fan-associated damage and the negative influence on the course of the disease caused by this. Therefore, these procedures are also moving back into the focus of therapy. Inhalation with saline solution reduces the aerosol release.

It must be taken into account here that patients with increased respiratory drive due to excessive tidal volumes and intrathoracic pressure fluctuations, even with spontaneous breathing (with or without NIV), can develop lung damage (patient-self-inflicted lung injury = P-SILI).

With all etiologies of respiratory failure in COVID-19, the question arises as to how the optimal time for intubation can be determined. A sole indication based on hypoxemia without further aggravating factors such as subjective dyspnea or signs of insufficient tissue oxygenation is currently being questioned by many authors and clinicians.

Whether intubation is advantageous or disadvantageous for the patient can certainly not be generalized and should be decided on the individual case.

The German Society for Pneumology and Respiratory Medicine eV (DGP) recommends in the absence of an early response (FiO2 <150 or 175 mmHg after one hour) to NIV, increasing O2 requirement, steadily or rapidly falling SaO2, increasing respiratory rate> 30 / min and an increase in Breathing work to check the indication for intubation. Signs of impending respiratory exhaustion are rapid shallow breathing, clear activity of the inspiratory auxiliary breathing muscles, retraction of the upper or lower thoracic aperture and / or visible muscular activity of the muscles of the shoulder girdle, agitation and reduced vigilance due to respiratory insufficiency and Development or intensification of respiratory acidosis with hypercapnia. It should be noted here that COVID-19 patients are often asymptomatic due to silent atelectasis and, despite impending respiratory deterioration, do not yet show any warning signs such as dyspnoea, tachypnea or increased work of breathing.

No matter what stage of the disease the patient is in and what form of ventilation he is being treated with, the principles of lung protective ventilation must always be observed.

modified according to DGP position paper, Fig. 2

Initial data show positive effects of NO inhalation, but tachyphylaxis must be expected after a few days. Prostacyclin therapy could also be a useful option.

ECMO therapy must be considered (Horovitz index <60 mmHg).

Overall, intensive stays are prolonged and weaning is described as delayed. Processes with secondary deterioration after initial improvement have also been described, so weaning from ventilation should not be started too quickly. Up to 60% of patients require re-intubation within 24-48 hours. This may be due to severe upper airway swelling. The use of dexamethasone and nebulization of adrenaline are recommended here. A secondary air test can be carried out with the tube unblocked (CAVE: aerosol formation!) After weighing up the benefits and risks. Material for the management of the difficult airway including surgical instruments should be available quickly (intubation (algorithm) & difficult airway (algorithm)).

Since the tracheotomy is also one of the high-risk procedures with regard to the risk of infection for staff with COVID-19, the indication for this should be made strictly and not too early (tracheotomy).

In the case of COVID-19, the tube should always be clamped off before disconnecting the tube because of the possible exposure to infectious aerosols. This also prevents a loss of PEEP, which can lead to the immediate formation of atelectasis with subsequent respiratory deterioration. In order to prevent contamination from the ventilator, which may continue to run, after disconnection, disconnection must always take place between the filter and the hose system of the device. In addition, the device should be switched to stand-by mode or a stop immediately beforehand. As far as possible, closed suction systems should be used for suction and sample collection.


  • AMBOSS chapter Acute Respiratory Distress Syndrome (ARDS)
  • S3 guideline for invasive ventilation and the use of extracorporeal procedures in acute respiratory insufficiency
  • Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China, Huang et al.
  • Findings of lung ultrasonography of novel corona virus pneumonia during the 2019-2020 epidemic, Peng et al.
  • Respiratory support for patients with COVID-19 infection, Namendys-Silva
  • Recommendations for intensive care therapy for patients with COVID-19, Kluge et al.
  • COVID-19: a synthesis of clinical experience in UK intensive care settings, Intensive Care Society as part of the National Emergency Critical Care Committee
  • COVID-19: Are we actually treating it correctly ?, dasFOAM.de (JUWO)
  • COVID-19 Lung Injury is Not High Altitude Pulmonary Edema, Luks et al.
  • Position paper on the practical implementation of apparatus-based differential therapy for acute respiratory insufficiency in COVID-19, German Society for Pneumology and Respiratory Medicine e.V. (DGP)
  • COVID-19 patients with respiratory failure: what can we learn from aviation medicine? Ottestad et al.
  • COVID-19 is also systemic vascular inflammation

Version from: 04/26/2020

This page was printed out on May 23, 2021 at 1:55 pm.

Due to the dynamic information situation, we ask you to closely check the latest version.

Address of this page: m4mvscovid.de/de/beatmung/beatmungsprinzipien-bei-ards-speziell-bei-covid-19/

Due to the dynamic nature of the situation, it is impossible to accept responsibility for the correctness of the information. Under certain circumstances, guidelines and recommendations from various specialist societies that appear every day require an independent reflection of the content offered.

All contents offered on this page are non-binding recommendations. All information on dosages, application forms or recommendations for action must be checked for correctness in each individual case. All information was written to the best of our knowledge.

The authors and the responsible editors assume no liability for the correctness of the information.