Dear Respiratory Therapy Colleagues

We write this letter to help ease some of the concerns that may be raised regarding how to properly treat patients with COVID-19 as they arrive in our ICUs. Please note, that while we will not address infection control practice (this will be determined by your institutional guidelines), we would like to summarize what is currently being considered in rapidly published guidelines that particularly apply to the clinical practice of Respiratory Therapists.  We will also describe some findings from our global colleagues, as well as our approach to managing these patients.

What is currently recommended

Two guidelines have now been rapidly published by Australia and New Zealand Intensive Care Society, and the Society of Critical Care Medicine. Here is a general summary of recommendations related to Respiratory Therapy Practice.

  1. For patients not requiring immediate intubation supplemental oxygen is ok, for high oxygen needs a filtered NRB mask would be ideal.
  2. The use of high flow nasal cannula (HFNC) over other oxygen therapies is recommended when available.
  3. Non-invasive ventilation (NIV) is not recommended due to high failure rates and potential to delay intubation. HFNC is recommended over NIV. A trial of NIV is acceptable with very close monitoring for lack of improvements.
  4. Do not delay intubation when patients are showing no signs of improvement
  5. For mechanical ventilation use ARDSNet approach of 4-8 ml/kg of predicted body weight
  6. During mechanical ventilation target plateau pressure < 30 cm H2O
  7. In general, higher PEEP is recommended, but please see our comments below.
  8. Prone positioning, see our comments below.

Some Global Observations

We have been gathering information about the respiratory profile of patients with COVID-19 (Wuhan China, Italy, France) in centers able to do some measurements of mechanics.

It is early to say for sure, but it seems that there are different profiles, which means very different management. Using the same approach for every patient could be problematic.

Non-recruiters: Some patients are non or poorly recruitable. We are not sure if this is due to prolonged pre-intubation periods such as delayed intubation after receiving NIV or HFNC. For these patients low PEEP (10 or even lower) and prone positioning if PaO2/FiO2 is ≤ 150 mmHg is the best solution. Oxygenation may not always drastically improve with prone positioning, but one needs to be patient as it is mainly for lung protection and you may need hours to see an improvement in oxygenation. If the PaO2/FiO2 is consistently below 80 that would be a discussion for ECMO if resources are available, if not any other rescue strategy such as inhaled nitric oxide could buy more time.

Try to avoid 100% FiO2, which favors de-nitrogenation atelectasis. 90%, 80%, or 70% would be better. Please understand that the difference between 100% and 70%-90% may not drastically change oxygenation in the blood due to a high shunt fraction.

Compliance: Some patients seem to be very hypoxemic with preserved compliance and are presumably also non-recruitable. Similar to the patients described above, if the lung is compliant there may be nothing much to recruit within the lung, and the role of high levels of PEEP is uncertain, and the above suggestions of low PEEP (10 or lower) should be considered. We must consider that improvement in PaO2 after increasing PEEP could be related to other mechanisms related to shunt redistribution due to a reduction in cardiac output. This, of course, is not desirable and hemodynamic status should take priority. Prone position is still advised if PaO2/FiO2 is ≤ 150 mmHg.

Recruiters: Patients that are recruitable should be managed with high levels of PEEP, but would still benefit from prone positioning if the PaO2/FiO2 ratio remains ≤ 150 mmHg.

What will we be doing at the Center of Excellence

When the numbers of patients are manageable it is quite reasonable to perform measurements of respiratory system mechanics using esophageal pressure and using valuable tools such as EIT. However, in light of the number of patients you may be faced with we suggest learning and understanding three simple things that we do as our routine standard of care at St. Michael’s Hospital in Toronto; Prone Positioning, assessment of airway closure, and assessing recruitment potential using a method we recently published. All three of these things can be done by anyone, with any ICU ventilator.

Prone Positioning – Good evidence

If your team is not already experienced with placing patients in the prone position, RIGHT NOW is the time to learn and practice. Experience globally is confirming that this is working to improve oxygenation of these patients. Here is a video demonstrating how this is done with ANY ICU bed, and with very few people (understand that we may be short on staff, so know how to do it with only a few people when necessary).

Assessing Airway Closure – Our standard of practice – no current evidence

Airway closure has been identified for years (and sometimes misinterpreted) as the lower inflection point during a slow-flow (quasi-static) pressure-volume curve. Some ventilators have built in tools for performing these maneuvers (pressure-volume curves). We do it in a simpler way, using a pressure-time curve (no need for a loop). To properly assess – use 5L/min of flow and after the maneuver look for a clear inflection point. For years, one of the issues is the uncertainty of inflection point. We feel that patient where it is very clear, likely had airway closure and that this clear inflection may not reflect alveolar ‘recruitment’ so much as a point that the airways open after complete collapse. If you do not have this tool on your ventilator do not worry, it can be done on ANY ICU ventilator and assessing the pressure-time curve. The proper technique is described in this video.

Interpretation of the Inflection point

  • The absence of an inflection point means that the patient does not have airway closure above the PEEP level you start from.
  • When performing the R/I ratio (described next) it is important to understand if there is airway closure as it can change the results, so it should be determined if you are performing the R/I ratio maneuver. This value is put into the calculator to properly compensate (calculator link below)
  • The ‘High PEEP’ of the R/I ratio maneuver should be ideally AT LEAST 5 cm H2O above AOP provided Pplat is < 30 cm H2O.  We prefer to use a PEEP drop of 10 cm H2O (use the same expiratory driving pressure used in the study) and make sure to input the AOP level in the online calculator for compensation.

Recruitment to Inflation Ratio (R/I) – Our standard of practice – no current evidence

Assessing recruitment potential has been done in research with either CT scans, or with lung volume measurements (nitrogen wash-in/wash-out). An article published recently by our group validated a method for assessment recruitment potential at the bedside using a single breath PEEP reduction maneuver. The calculation takes into account what was already being exhaled + what is predicted to be lost by dropping PEEP, and compares it to the total volume exhaled. Excess release volume is considered to be from alveolar units that were recruited by the use of higher PEEP. The compliance of that recruited lung (excess volume divided by drop in pressure) is then compared to the compliance of the lung at low PEEP (when overdistension is at its least) and the result is the R/I ratio. The recruited compliance divided by the respiratory system compliance. I have developed an online calculator that does these calculations for you (requires a mobile browser, Chrome, Firefox, Safari….does not work with Internet Explorer). It is important to assess airway closure (method above) prior to assessing the R/I ratio as the airway closing pressure can affect the interpretation. The calculator includes both.  If the patient has a R/I > 0.5 we use higher PEEP (15 cmH2O) and aim to keep plateau pressure < 30 cm H2O.  We do not hesitate to use prone positioning if the PaO2/FiO2 is ≤ 150 mmHg.

Here is the video:

Here is an online respiratory mechanics calculator (with both airway closure and R/I ratio videos at the bottom)


Other ways to assess Recruitment: set PEEP at two different levels separated by at least 5 cmH20

For those who are not comfortable in learning the simple technique above, the minimal thing to do is asking the question: knowing that patients with COVID-19 can be hypoxemic BUT non recruitable should my patient receive high or low PEEP independently of the degree of hypoxemia? So please test two levels of PEEP separated at least by 5 cmH2O without changing anything else than PEEP (not FiO2!) and do ABG, hemodynamics, driving pressure and plateau pressure measurements before and after. Non recruiters usually have limited benefit in oxygenation, lower mean arterial pressure at high PEEP, higher PaCO2 and higher driving pressure. Recruiters usually have significant improvement in oxygenation, minimal change in PaCO2, minimal change in driving pressure or even a decrease. None of these criteria works perfectly however, all patients will likely have an increase plateau pressure with higher PEEP. Going higher than 30 cmH2O of plateau pressure may be risky and you may need to reduce Vt if you are using high PEEP.

Please everyone, stay safe, and protect yourself so that you can protect your patients.

Thomas Piraino, RRT, FCSRT, FAARC
Dr. Laurent Brochard, MD

Reference for R/I ratio
SCCM Recommendations
ROX Index Reference – a score to consider for HFNC failure risk, it is not perfect and does not replace clinical judgment.