Everyone is scrambling to put together their own “protected code blue” or COVID airway plans. We are creating these with limited evidence and lots of strong opinion. One of the incredible benefits of FOAM is the ability to rapidly share information and collaborate. Below is a modified version of the COVID-19 resuscitation principles document we created for my department. It is the first draft. I expect it to change with time (especially as we test it through more simulation). However, in case anyone is still working on their own document, I thought I would share it here. I would love for people to pick this apart. This isn’t about me being right in the first draft, but about finding a way to keep both staff and patients safe in trying times. (With luck, we might get a look at similar documents designed by people much smarter than me – like Chris Hicks and Andrew Petrosoniak down at St Mike’s hospital. I will share other guidance as it becomes available.)
Note: These principles are brought together from a number of different documents and sources. (They rely heavily on the Safe Airway Society consensus statement.) They have also been peer reviewed through our local group. However, evidence in this area is sparse, and these should in no way be considered standard. Practices will vary significantly from hospital to hospital with good reason (you need to adapt to your local resources and skills). You should read them, like you do everything, with a critical eye.
- No staff member should become infected by COVID as a consequence of caring for patients.
- We should aim to provide the same high-quality critical care for our patients during this crisis as we would provide at any other time, understanding that changes are necessary during a crisis.
- We should presume that any critically ill patient is COVID positive until proven otherwise.
- Good communication is essential through any crisis.
- No guideline or protocol can account for every possible scenario. Team members need to be flexible in their approach, while ensuring the safety of both staff and patients.
Donning appropriate PPE is essential before seeing any patient with a respiratory complaint. Full airborne PPE should be worn for any undifferentiated critically ill patients. As information becomes available, the physician can change the level of precaution as appropriate.
Appropriate doffing of PPE is also essential. We suggest that teams members doff in pairs to ensure best practice is followed after managing these high risk patients. Ideally, every resuscitation will be assigned a “safety officer” whose only responsibility is overseeing the safety of the staff.
General respiratory support
All hypoxic patients will be assumed to be high risk for COVID and therefore should be managed with full PPE.
A rapid assessment from outside the room should place the patient in one of three categories:
- Mild hypoxia/ respiratory distress: Can be managed with nasal prongs, targeting an oxygen saturation greater than or equal to 88%. We suggest applying a surgical mask over the nasal prongs to limit viral spread.
- Moderate hypoxia/ respiratory distress: Patients that don’t look like they will need to be intubated in the first 4 hours of their hospital stay. Apply the HiOx non-rebreather oxygen mask and titrate to an oxygen saturation greater than or equal to 88%. Any patients on HiOx or requiring oxygen flows of greater than 6 L/min should be placed on airborne precautions.
- We suggest an early phone call between the ED physician and the anesthesiologist and intensivist on call to facilitate safest timing of intubation.
- If there is any doubt about how sick the patient is, err on the side of intubating early.
- If the HiOx mask is not available, a standard non-rebreather can be used as an alternative. However, a standard non-rebreather is an open system without any filters. If using a standard non-rebreather, we suggest placing a surgical mask over the nonrebreather to cover the exhalation ports.
- Severe hypoxia/ respiratory distress: If the patient is in severe distress and it appears that they will need intubation within the next few hours, do not delay intubation. If there is a strong indication to move the patient before intubation (more controlled situation, more experienced staff, better equipment), that may be reasonable, but consider the risks of transferring a patient with an uncontrolled airway.
- In general, we recommend against delaying intubation once it appears necessary. The suggested best practices for airway management are below.
- For these patients, anesthesia should be paged stat overhead.
Update: As we get more experience with managing COVID-19 patients, we are recognizing multiple distinct presentations, which could contradict this simple breakdown. Critically, people are noticing patients with significant hypoxia but no other distress, who may be better managed without early intubation. You can hear more about this from Scott Weingart:
- Any patient with a cough or respiratory complaints should have a mask on at all times.
- As a general principle, the higher the oxygen flow rate, the more aerosolization that will occur, so use the lowest rate of oxygen possible. Titrate patients to an oxygen saturation equal to or greater than 88%, unless there is a specific reason for an alternative target.
- Noninvasive ventilation and optiflow should be avoided. (Update: This was always a controversial statement. I included it because many hospitals were making this statement, but at the same time, these treatment modalities were being used widely in China and Italy. In the video above, Scott suggests using both modalities. I think that the risk benefit probably favours using these, assuming you are treating the patient in a negative pressure room with full airborne PPE.)
- Nebulizers should be avoided.
- The only exception would be the need to nebulize epinephrine for severe croup, in which case it should occur in a negative pressure room with staff in full airborne PPE.
- Puffers via MDI can be given using droplet precaution (although the exact level of PPE required for procedures is unclear).
- Oxygen flows greater than 6/L per minute should be considered aerosolizing and need an airborne room. COVID patients requiring high oxygen flow (greater than 6-8 L/min) or fiO2 greater than 50% have a very high failure rate, and intubation should be considered early to prevent higher risk exposure to staff.
Airway management for patients requiring intubations
Full PPE is mandatory for any airway management. If at all possible, airway management should occur in a negative pressure room. If a negative pressure room is not available, a normal pressure room with the door closed is acceptable.
Limit the number of people in the room to just those necessary.
The most skilled person should be performing the intubation. This should be the anesthesiologist whenever possible.
Inside the airborne room (all with full PPE):
- Note: Different hospitals use very different teams. If all you have to do is intubate, I would limit the team in the room to 1 MD, 1 RN, and 1 RT. However, this document was written for emergency department resuscitations, where patients are undifferentiated and lots needs to get done. In that setting, I think you want the full team.
- 1 RT who is responsible for gathering and setting up equipment.
- 1 MD as the resuscitation lead
- 1 airway doctor. (If the anesthesiologist is not available, the most experienced person in the room should intubate.)
- I will note, many protocols are only putting 1 doctor in the room. I have mixed feelings about this. Lowering exposure to staff is good, but these are going to be high stress intubations, and I really think there needs to be a lead resuscitation MD who isn’t task saturated by the airway.
- 2 RNs – with critical care background
Inside the anteroom:
- 1 RN/RPN – with full PPE to pass items from outside to inside airborne room
- 1 RT – with full PPE to pass items from outside of room to inside as needed or assist in room as needed
Outside the room
- 1 RN – to act as a “runner” and obtain supplies
- 1 RN – to prepare medications and communicate with team inside
- 1 MD – if available to help manage the case, anticipate the needs of the doctors in the room, and provide backup as needed.
In the room
- HiOX Mask
- BVM (with viral filter and PEEP valve attached)
- Tracheal tubes (2 sizes), stylet, 10 mL syringe, lube, and ETT securing device
- Oropharyngeal airway
- Video laryngoscope with both standard geometry and hyperangulated blades.
- End-tidal CO2 monitor
- NG tube
- Current suggested RSI medications: ketamine 1.5 mg/kg and rocuronium 1.5 mg/kg
- We suggest ordering a norepinephrine drip be prepared in any patient with potential hemodynamic compromise.
In the anteroom
- ETT size above and below
- Direct laryngoscope with multiple blade
- Second generation LMA, with appropriate size selected before starting intubation
- Surgical airway kit (scalpel, bougie, and 6.0 ETT)
In the hallway
- Crash cart (Do not bring the carts into the room. There are pre-prepared bags with essential medications to bring in the room.)
- Difficult airway cart
- ACLS medication kit
Before entering the room, perform a prebrief. Ensuring that the team knows their roles, the plan for intubation, and the plan for failure.
- Noninvasive ventilation, bag valve mask ventilation, and high flow oxygen are all considered to increase the risk to providers, and should be avoided if possible.
- We think the best technique for ensuring limited exposure to the staff is to apply a BVM using a two hand seal. Use the V-E or vice grip method. (See picture)
- If the patient is breathing spontaneously DO NOT PROVIDE RESPIRATIONS. (DO NOT SQUEEZE THE BAG). The mask is for oxygenation only, not for ventilation.
- If the patient is apneic, they will not get any oxygen from this set up. Ensure you have an excellent 2 hand seal, and provide gentle respirations just until you see minimal chest rise, at the slowest rate you can (6 breaths / minute).
- If unable to secure a tight seal and a leak is evident – the physician will decide what is the most effective and least risk for aerosolization.
- The BVM must have a viral filter between the mask and the bag, and the PEEP valve should be applied. (See picture)
- The BVM should be attached to wall oxygen like normal.
- The BVM should be attached to waveform capnography. A low end tidal CO2 number, or a triangular waveform (rather than rectangular) indicates a mask leak, and should prompt interventions to improve the seal.
- PEEP may be used, up to 15 cm H2O.
- The airway doctor will be responsible for holding the two hand mask seal, allowing the RT to prepare equipment, and the nurses to prepare medications and vascular access. However, if the physician is needed for procedures (such as placing an IO) or to otherwise manage the complexities of the case, the RT can temporarily assume this role, or the second RT can be called into the room.
- Preoxygenation using the HiOx mask is an alternative, but increases exposure to staff and provides inferior preoxygenation, so it is discouraged. If it is used, oxygen flow rates should not exceed 15 L/minute, and the team needs to be careful about leaks from around the mask.
- We are currently recommending against the use of CPAP, BiPAP, and high flow humidified nasal oxygenation for these patients.
- Hypoxia frequently causes agitation. It will make preoxygenation difficult, and increases risk to staff members. Consider early use of ketamine to control agitation and allow preoxygenation to proceed.
- Assess the patient for markers of anatomical and physiological difficulties with intubation, and clearly state any alterations to the standard plan based on that assessment.
- Intubation should proceed by rapid sequence intubation. Paralysis decreases coughing, and increases first pass success. Ensure that you wait for the patient to be fully paralyzed prior to inserting the laryngoscope, as any coughing will put the staff at increased risk.
- Do not use nasal prongs for apneic oxygenation.
- The first attempt at intubation should be by the most experienced intubator available. The first option should be anesthesia.
- Video laryngoscopy is suggested for the first approach, but the airway doctor will use their judgement.
- Be careful about tube depth. Watch the cuff pass through the cords and don’t advance any further (you will not have easy access to x ray to confirm placement, and adjusting tube position exposes staff to additional risk.)
- If the intubation is successful, a viral filter should be attached immediately to the endotracheal tube, and the patient should be rapidly attached to the ventilator. Use WAVEFORM capnography to confirm tube placement (this prevents us from having to detach the tube to take out the colour based CO2 monitor, and also allows for ongoing monitoring of the patient.)
- Place an NG tube after intubation (prevents another staff member from being exposed later).
- Delay portable chest x ray by at least 15 minutes unless there is a clinically emergent indication.
We don’t have the appropriate adaptors, and don’t want to complicate things, but if you want to achieve CPAP and also have oxygen delivery in an apneic patient without providing ventilations, this is the suggested set-up by Scott Weingart:
Some videos on optimizing video laryngoscopy
- If the first intubation attempt fails – management will depend on the patient, and must proceed at the discretion of the lead physician in the room. These patients are at very high risk for rapid desaturation, so anticipate that attempts at re-oxygenation may be required between intubation attempts.
- Our priority is not contaminating clinical staff, while maintaining adequate oxygenation for the patient. Remember, the intubation itself is a high risk procedure, and is made higher risk if rushed, complicated by trauma, or secretions. A controlled attempt to reoxygenate the patient is preferable to a rushed second intubation attempt.
- If the patient’s oxygen saturation allows, make another intubation attempt, with an adjustment based on the first look (different blade, different size tube).
- If the patient requires re-oxygenation between attempts, the use of an LMA using a BVM with a viral filter attached, is probably the safest approach. BVM ventilations are also reasonable, focusing on a perfect two hand seal with the V-E grip, and very gentle, slow ventilations performed by a second provider.
- If the first intubation attempt was made by someone other than anesthesia, and the LMA is oxygenating the patient without significant leak, it is reasonable to leave the LMA in place until anesthesia arrives for the second attempt.
- The intubation procedure is the highest risk moment for transmission of COVID to staff members. Attempts at intubation should be limited to 3, unless there is an obvious way to optimize a further attempt by changing to different equipment or a different intubator. If intubation is not possible, we would suggest placing an LMA and waiting for help, or performing a bougie assisted cricothyrotomy if you are in a can’t intubate, can’t oxygenate scenario.
Can’t oxygenate, Can’t ventilate
- If there is a can’t oxygenate, can’t ventilate situation, a cricothyrotomy is the recommended procedure.
- The one major modification is that when the decision to proceed with cricothyrotomy is made, attempts to ventilate through the mouth and nose should stop, because ventilations will increase the risk of aerosolization through the incision.
- There are no modifications to the procedure itself. We recommend scalpel, bougie technique, as described here. If you perform a cricothyrotomy, place the patient on the ventilator, and cover the patient’s mouth and nose with a mask.
The primary problem for most COVID patients has been hypoxemia from ARDS. However, these patients also develop septic shock. We need to be prepared for peri-intubation deterioration. CPR is a very high risk procedure, so we would prefer to be somewhat more liberal than usual in our use of vasopressors if you think they might be needed in order to prevent peri-intubation cardiac arrest. Norepinephrine is the first line agent. We suggest having this prepared and in the room for any patient with a shock index greater than or equal to 1 (the heart rate is greater than or equal to their systolic blood pressure).
Cardiac arrest management
- All patients arriving in cardiac arrest should be assumed to be high risk for COVID until we can assess otherwise. Full PPE is recommended. Precautions can be altered as more information is gathered, but we should err on the side of caution when faced with uncertainty.
- Standard ACLS algorithms should be followed.
- For airway management, intubation is greatly preferred over any other option. Unlike our normal ACLS teaching, we are willing to accept a brief pause in CPR to achieve intubation. Because CPR is aerosol producing, chest compressions should not continue during the intubation attempt.
- Ventricular fibrillation and ventricular tachycardia arrests are presumed to be cardiac in nature, and therefore don’t require ventilation. For these patients, we suggest chest compression only CPR for the first 2 rounds of CPR. To prevent aerosolization, a BVM should be applied over the face with a tight 2 hand seal held by the RT. After 2 unsuccessful defibrillations, if the physician believes ventilations are important, intubation is the preferred method, again with a pause in chest compressions during the procedure.
- PEA and asystolic arrests should be presumed to be hypoxic, and intubation is the safest procedure. If for some reason there is a delay to intubation equipment being available, chest compression only CPR is reasonable, with the airway covered with a BVM using a tight 2 hand seal, but not providing ventilations. An LMA is also reasonable.
- If the patient arrives with a King LT in place, the act of removing the device is likely to be higher risk than leaving it in place, so you can continue the resuscitation using the SGA.
Update: I think the biggest change to this is a discussion of what to do if you are in a patient’s room in droplet precautions, and they arrest. I still believe that no chest compressions or airway management should be done until everyone in the room is in airborne precautions (although some people are getting away from that by allowing chest compressions as long as the patient’s mouth and nose are covered.) However, that doesn’t mean the provider should immediately leave the room. The first step is to call for help (or activate the protected code blue). Then, while waiting for the team to get into airborne PPE, they should apply the pads, and if the patient is in a shockable rythm, they should shock up to 3 times in a row. If the team is still not in the room, time can be used to prepare for the code (place an IV, get medications out, etc). As soon as the full team arrives, the person in droplet precautions leaves the room, and the team takes over running standard ACLS, modified as above. The UK ACLS algorithm is below.
We are dedicated to ensuring that every member of the emergency department team feels completely comfortable with the management of critically ill COVID patients. In addition to in-situ simulations, we will be providing specific hands on training to review the use of:
- The Glidescope
- BVM (both the set up with the viral filter and the proper use)
- How to set up capnography properly (on both the ETT and the BVM)
- How to interpret CO2 waveform to assess for leak
If there is any equipment that you are unfamiliar with or want some refresher training, please ask.
References and further reading
The information about COVID is rapidly evolving, and the above document was adapted from a number of sources. It will change as more information becomes available, and in response to feedback from our staff.
Version one of this document was based largely on the following guidelines, which are freely available, and staff are encouraged to read on their own:
Morgenstern, J. COVID Resuscitation Principles, First10EM, March 21, 2020. Available at: