Case Reports

Prehospital management of patients with COVID-19 and respiratory failure

By April 13, 2020 No Comments

This report was prepared by the Air Ambulance Department of Oslo University Hospital. Over the past several days, we have acquired valuable knowledge, which we believe is important to share with all ambulance services in Norway.


COVID-19 patients make up a heterogeneous population, with differences in medical histories and disease progression. Yet there are some important observations being made regularly. We have seen several patients admitted to hospital from the Oslo district with an aggressive progression of the disease, involving rapid decompensation and life-threatening respiratory failure. This also includes relatively young patients, ages 30 to 60, who were previously healthy with minimal comorbidities. We will illustrate this with a case, followed by a few learning points and recommendations for treatment.


An air ambulance physician meets the ambulance to assist in the treatment of a critically ill COVID-19 patient. This is a female, age 67, previously healthy, but with hypertension. Fever and gradually increasing respiratory distress over the past 24 hours. The ambulance has requested assistance because the patient’s condition is rapidly deteriorating and she can now barely manage to breathe. The patient has RR 45, SpO2 50% with 6 litres of oxygen from a dual nasal cannula. We begin supportive care with CPAP (Flow-Safe II with HEPA filter), gradually increasing the flow from 12 to 25 liters of O2. The patient tolerates treatment well, and shows clear signs of reduced stress and respiratory improvement. SaO2 rises over 2-3 minutes to above spO2>93%


The pathophysiological picture is common among many patients. They have primary respiratory failure and are usually hemodynamically stable. The patients then develop what is referred to as “silent hypoxemia”, with a gradual decrease in oxygen saturation. A compensatory rapid respiratory rate maintains CO2 tension at a low level, so that patients do not have a sense of air hunger, contrary to what we often see in patients with pulmonary oedema, heart failure, sepsis or pneumonia.
Hypoxia with a compensatory low PaCO2 does not involve a subjective sense of distress. On the contrary, it may feel pleasant, and patients may lose their sense of self-awareness and awareness of their surroundings. Confusion is a common sign. Some, though not all, develop speaking-related dyspnoea. The patient may not sound very distressed when talking, yet may still have severe respiratory failure. Although the clinical picture will vary to a large extent, it is often different from that seen in patients with other conditions (pneumonia, heart failure, pulmonary embolism, or asthma). In addition, most patients who were previously healthy will have normal heart function and maintain adequate blood pressure despite severe hypoxia. Despite severe respiratory failure, the patient will often have relatively well-preserved lung compliance (meaning that the lungs are pliable), so that the work of breathing is relatively easy, and the patient will not necessarily be in distress. Nevertheless, work of breathing will vary based on the level of respiratory failure, and certain patients will become fatigued from the increased work of breathing (high RR) over the course of several days.
There may be an absence of pathological lung sounds upon auscultation.

The pathophysiological mechanisms behind oxygenation failure in COVID-19 patients with critical respiratory failure are likely multifactorial. These patients typically have relatively well-preserved lung compliance, with specific oxygenation failure. A hypoxic-driven respiration with tachypnoea and high minute volume results in pronounced hypocapnia (low PaCO2). It is plausible that diffusion failure is due to interstitial oedema. Oedema in the alveolar-capillary membrane increases the distance for oxygen to diffuse, leading to a fall in PaO2. Also, parts of the lungs will be unventilated, and a pulmonary shunt will be present, meaning that parts of the lungs are not participating in gas exchange. This may be caused by a combination of atelectases and interstitial oedema. The unventilated sections of the lungs will continue to have blood flow (perfusion), however the blood will then flow around (be shunted past) the healthy part of the lung and will not be oxygenated. Instead, it will flow out to the arterial side with unchanged and low saturation. This means that the oxygenated blood from the ventilated parts of the lung will mix with the nonoxygenated blood from the unventilated section of the lung (shunt). It is not possible to compensate for shunting by solely providing oxygen therapy, even with the administration of 100% oxygen, as the patient will continue to have low SpO2. By applying positive airway pressure, we can recruit (open) the unventilated section of the lung, which will increase the area that participates in gas exchange and reduce the degree of shunting. We refer to this an improvement in the ventilation-perfusion relationship (V/Q).

Oxygen therapy

Liberal oxygen therapy is essential, and a high flow 12 – 25 l/min is often necessary, regardless of the method of supply (dual nasal cannula, mask, CPAP). These patients usually respond well to positive airway pressure, and CPAP therapy is effective, even though it delivers FiO2 (fraction of inspired oxygen) which is somewhat lower than that of a mask with a reservoir bag. We performed measurements of FiO2 and airway pressure with the Flow-Safe II CPAP system. We used HEPA filters to prevent the leakage of aerosols into the environment. The tables below show that CPAP delivers a lower FiO2 than a mask with a reservoir bag. However, this might be outweighed by the fact that positive airway pressure improves lung function by recruiting areas of the lungs that are unventilated. Therefore, we recommend attempting CPAP therapy with a 15 – 25 liter flow if it is not possible to achieve SpO2>93% with a mask or dual nasal cannula. This would give a FiO2 of 0.45 – 0.55 and an airway pressure of 8 – 12 cm H20, which in most cases would be more effective than a mask with a reservoir bag of FiO2 0.6. CPAP therapy is physiologically plausible and is supported by our measurements. Furthermore, several physicians from the Air Ambulance Department have reported good results with CPAP therapy for these patients, and similar results have been reported from other countries. We do not believe that aerosol generation would be a problem, as long as a tight-fitting mask and HEPA filter are used. An oxygen mask with a reservoir bag does not have a sufficiently tight seal, and airborne infections would likely be a bigger problem with this method. We recommend the use of safety glasses and a P3 mask for all airway management.


  1. The patient should sit upright, or be placed in a steep semi-sitting position. Oxygen therapy should be liberal, i.e. up to 12 — 15 l/min with a mask and reservoir bag in order to achieve SpO2>93%. For patients who do not respond adequately to this therapy, or who are fatigued, it is best to quickly switch to CPAP therapy. When changing to CPAP, remember to turn off flow in order to avoid aerosol spread.
  2. Begin CPAP with 15 l/min and continue up to 25 l/min after response. Ensure that the mask is tight-fitting. The HEPA filter should be secured as shown in the photo. Ensure that the attachment fits tightly and securely. Place the patient in a steep semi-sitting position. Calm the patient and help them to understand the benefit of this procedure. If the patient does not respond to CPAP therapy, notify the Air Ambulance physician.
  3. If CPAP is unsuccessful or unavailable, a ventilation bag may be an option. NB! Remember to turn off the oxygen to the CPAP mask before removing it, to minimise aerosol generation. Sit behind the patient and hold a ventilation bag with high flow (12 — 25 l/min with oxygen) tightly around the patient’s nose and mouth. As long as the reservoir bag expands, there is a sufficient flow. If there is inadequate self-respiration, provide ventilatory support with caution. A PEEP valve on a ventilator bag would be an advantage. Try PEEP 5 — 10 after response. Ensure that all attachments are securely fastened.
  4. When using CPAP or a ventilator bag with high flow, it is important to be monitor the oxygen level, keep an eye on the manometer, so that you can switch to a new tank before it empties.
  5. Patients with physiological symptoms should be handled with care. These patients may show significant improvement in their physiological parameters with oxygen therapy, but will still be critically ill. They must be moved with caution and must not try to walk by themselves or exert themselves in any way.

Other aspects

  1. We have seen patients progress from being physiologically unremarkable to decompensation in the span of a few hours. If family members report that the situation has worsened, this should be taken very seriously. SpO2, RR, HR and blood pressure must always be measured, regardless of the clinical symptoms.
  2. RR > 20 or SpO2 <93% in otherwise healthy persons with no prior lung disease is very serious.
  3. You must observe the patient’s work of breathing (frequency, depth, inspiration and use of accessory muscles). Does the patient appear fatigued? If the patient is tired, exhausted or dehydrated, the situation can rapidly deteriorate. Patients often develop respiratory failure 5 to 14 days after the initial symptoms appear, with “silent hypoxemia” and rapid RR.
  4. When reporting to the Air Ambulance physician or hospital, always state the physiological parameters: RR, SpO2, HR and BP. Patients with persistently low saturation or high respiratory rate despite oxygen therapy should be received by an emergency team at hospital. Early notification to the hospital is essential.

Important information regarding SpO2 measurements in COVID-19 patients

The sickest patient will have a compensatory ventilation response with high minute volume and hypocapnia. There are reports of arterial CO2 tension (PaCO2) in the range of 1.3 — 2.0 kPa. This type of extreme hypocapnia involves an oxygen haemoglobin dissociation curve shift to the left. In a left shift, there is an increase in the affinity between oxygen and haemoglobin (oxygen binds more easily to blood). This causes oxygen saturation for a certain oxygen pressure to rise higher in a situation with hypocapnia than with hypercapnia. This means that a COVID-19 patient with only moderately reduced saturation, as measured by pulse oximetry, may have a relatively low arterial oxygen tension (PaO2). This will have many implications. For instance, a patient with previously healthy lungs who has only a slightly reduced SpO2 may be more ill than assumed. Measurements of EtCO2 would therefore be very useful. In cases of low EtCO2 and reduced oxygen saturation, we can expect a lower arterial oxygen tension (PaCO2) than the SpO2 measurement tells us intuitively. If the SpO2 is 95% and the EtCO2 is 2.0 kPa, this would indicate that the patient is more ill than the SpO2 value implies. The Norwegian Directorate of Health has stated that a SpO2 < 95% should be a cut-off value in patients with previously healthy lungs. This seems reasonable. A low EtCO2 is itself a cause for concern, as it indicates a compensatory response to progressively failing diffusion. The physiology we have described may be illustrated by what occurs during hypobaric hypoxia. In a low-pressure chamber experiments at an altitude of 22,000 feet, measurements were taken for both PaCO2 and SaO2. One subject with PaCO2 = 4.5 had SaO2 = 50%, while another subject with PaCO2 = 2.5 had SaO2 = 90%. The effect of hypocapnia is therefore significant. Read the attached article that describes this physiology.

Differential diagnostics

At a time when our attention is focused on COVID-19, we must not forget important differential diagnoses such as a pulmonary embolism, heart attack, heart failure or bacterial sepsis. All of these conditions may arise in the course of a lengthy viral infection and can complicate the clinical picture. An ECG is a useful supplementary instrument for differential diagnostics. In patients with diabetes, we must not forget to measure blood glucose levels. Ketoacidosis is a life-threatening complication that is associated with infections and may have a similar clinical picture with a high RR.

Kind regards, Senior physician, William Ottestad, Jens Otto Mæhlen, Per Olav Berve, Air Ambulance Department, Oslo University Hospital