Shared ventilation: how to do it if you have to
Author: Simran Kaur Matta, MD
Peer reviewer/ Editor: A/Prof Chris Nickson
Editor’s note: This is a guest post by Dr Simran Kaur Matta, MD, a US-based Critical Care Physician with a passion for physics, mechanical ventilators and POCUS. She believes that ‘proning saves lives’.
– Chris Nickson
The SARS-COV-2 pandemic is changing the way we practice medicine today. In some settings, the pandemic has led us to ration supplies severely, and to rethink and repurpose existing resources. Given the serious discrepancy between resources available and resources needed, there has been global interest in sharing one ventilator between multiple patients.
In an ideal world we would not be entertaining this idea. However, especially in regions of the world that have been hit the hardest, the situations we are facing during this pandemic are far from ideal.
Hospitals in some major cities have already started to split ventilators in rare cases. With the rising surge of the pandemic, more hospitals may follow suit. If sharing ventilators becomes a necessity, the question is: how do we support multiple lung ventilation and what is the safest, most effective way to do so?
Mode of ventilation
An important drawback of shared ventilation is that patients must be sedated and/or paralyzed to prevent ventilator triggering. This is important because if one patient is tachypneic, it will cause the other patient(s) to hyperventilate as well. This can be avoided by ensuring adequate sedation and paralysis and by increasing the trigger threshold on the ventilator in a continuous mandatory ventilation (CMV) mode.
Pressure control ventilation (PCV) is the most practical way to share ventilators. PCV allows the clinician to control over the mean airway pressure and driving pressure. Inverse ratio pressure limited ventilation (IRV) uses open lung ventilation strategy and may be a suitable alternative in difficult-to-oxygenate patients. Inspiratory pressures can be titrated to optimize tidal volumes and avoid overdistension. This mode is usually undesirable in obstructive lung disease as the short release time is problematic for patients who require a prolonged expiratory time.
The main drawback of volume control ventilation (VCV) modes is the interdependency of the patients with respect to lung mechanics. In other words,, the less affected, more compliant, lungs will receive over 50% of the total delivered tidal volume in a two patient setup. This process is dynamic and any ongoing change in the compliance of one lung will affect the other lung as well. On the other hand, in a PCV strategy, there is no patient interdependency – if lung mechanics change in one patient it will affect that patient’s delivered volume, but as the pressure is maintained in the system, the volumes delivered to other patients are not affected.
In our test model, simulation of endotracheal tube (ETT) obstruction in one test lung did not affect ventilation in the second test lung when using PCV. On the other hand, VCV is challenging and potentially deleterious given the inter-patient dependency. At a set tidal volume, simulation of blockage in one patient circuit led to diversion of volume in the second open patient circuit. This was measured using a gas flow meter and was visualized by over inflation of non-obstructed test lung.
The need for a better and safer shared ventilation circuit prototype
In shared ventilation, the ventilator circuits connected to different patients are in parallel and so the pressure in each circuit remains the same in PCV mode. Thus, the variable parameters are the flow and the volume of gas in each circuit. These parameters depend on the compliance of the individual lung. The lung with better compliance will be delivered a higher tidal volume (VT) leading to overdistension of especially the non-dependent alveoli inducing stress, shear strain and biotrauma causing ventilator induced lung injury (VILI). The lung with lower compliance, on the other hand, may be hypoventilated at the given pressure settings. Thus it is important to be able to measure and regulate VT and pressure in each circuit.
Designing a shared circuit prototype
We designed a prototype that will decrease inter-patient dependency and allow to redress for compliance mismatch between lungs. See the video below for the basic circuit set up:
The best approach is to “keep it simple”. We designed our prototype using minimal equipment that is cheap and readily available. The availability and cost of supplies was an important consideration in designing this prototype.
For each patient that will be connected to the shared ventilator, the following are needed for basic set up:
- Y piece splitters x 2 (one is included in your circuit, and one more will have to be ordered or 3D printed)
- One way valves x 2(one each for inspiratory and expiratory circuit to prevent back flow)
- HEPA filter x 1
With common pressure settings on the ventilator, regulation of flow in a circuit can be achieved by adding a resistor in series in the inspiratory limb. There are several ways to add resistance such as creating a bend, constriction or adding an obstruction or valve in the circuit. We experimented with each of the methods. Creating a bend or constriction externally did not alter the flow significantly because of the material and corrugations of the tubing. If circuits of varying diameter would be made available in the future, that could be an effective mechanism. Adding a valve, however allowed us to regulate the flow. We were able to obtain the most precise flow regulation with a globe valve. If you cannot find one, a quarter turn ball valve can be used, although this will allow for less fine tuning. The valve system also enables measurement of tidal volume in each individual lung. One valve has to be completely open at all times without resistance (except during the brief volume maneuver) for the ventilator to work properly.
The “open” circuit should be connected to the lung with least compliance (i.e. with the most disease) and lower VT. The peak inspiratory pressure (PIP) should be set to obtain the desired VT in this lung. The volume in the circuit with better lung compliance (and higher VT) can be regulated with the flow control valve.
Another way to reduce tidal volume in the “good” lung might be to place saline bags/CRRT bags on the patient’s chest. This will reduce the chest wall compliance, hence the overall distending pressure of the alveoli, thus lowering tidal volume.
We added an in line pressure monitor distal to the flow regulator near the terminal end of the inspiratory limb. Because the flow of gas is not continuous (time cycled), there will be a pressure differential before and after the valve.
If in-line Positive End-Expiratory Pressure (PEEP) valves are available, these can be used to add extra PEEP in a given circuit. Some old ventilator circuits are equipped with them, however the tubing may not be compatible with new ventilators. PEEP valves from an “Ambu bag” (bag-valve apparatus) will not work as they are not designed for use with closed circuit ventilators.
Unfortunately, FiO2 cannot be individualized. Inspiratory:expiratory (I:E) ratio and respiratory rate (RR) cannot be individualized either.
We were not able to find a flowmeter that was cheap and readily available in the hospital. There were also concerns about creating too many connections in the circuit and the potential risk of circuit breakages and disconnections.
Note that all the circuits need to be connected to the ventilator prior to calibrating the ventilator. This is necessary to take into account the compliance of individual circuits. Machine calibration feedback from the initial compliance check of all circuits is required to allow for proper pressure readings.
Grouping of patients depends on the presence or absence of underlying pulmonary disease. If used in a mass casualty incident involving otherwise healthy patients with normal lungs, I would cohort by body size, unless there is a reason for varying lung compliance such as blast lung.
In a situation involving mass causalities with respiratory disease, like we are facing with the COVID-19 pandemic, consideration of these factors helps mitigate against mismatched lung compliance between patients:
- Group by disease severity. Cohorting patients with similar disease severity and compliance would allow shared Fraction of inspired oxygen (FiO2) and PEEP settings that could work for both lungs . Such cohorting will also minimize the variation in VT between two patients and may allow pressure settings applicable to both. Nonetheless, there will still be variability in tidal volumes delivered to both lungs and see above ways to individualize ventilation.
- Group by lung pathology. Patients with obstructive lung disease (e.g. asthma, chronic obstructive pulmonary disease (COPD) should not be grouped with patients with non-obstructive lung disease (e.g. pneumonia, acute respiratory distress syndrome (ARDS)) since they will have different Time constants (Ti). That being said, ARDS lungs typically exhibit heterogeneity and different lung units may have different Ti. This problem, however, may be further compounded by cohorting with a COPD patient. Only one single I:E ratio can be set on the ventilator and this may lead to AutoPEEP (“gas trapping”) and hypercapnia in the patient with COPD. Based on the ventilator settings, an inspiratory flow bias is generated and while it may be beneficial for one lung, it may be counterproductive for the other lung.
- Haemodynamic instability. Patients with severe hemodynamic instability should not be placed on a shared ventilator. Having a code on a shared ventilator is likely to be disastrous. Patients with stable vasopressor requirements might be appropriate for shared ventilation on case-by-case basis.
In an ideal world, patients would be grouped by body size as well for lung pathology. But using so many criteria to group patients may not be practical, and in an ideal world we would not be sharing ventilators among multiple patients at all!
Alarms, monitoring, and safety checks
This is the one of the most important aspects of shared ventilation, and an area of great concern.
Low tidal volume alarm: Now that we are able to measure individual tidal volumes, low volume alarm will have to carefully set based on the expired tidal volume (VTe) of the least compliant lung. The higher the threshold VTe for the low tidal volume alarm, the higher the sensitivity. However, this comes at the cost of more false alarms.
Monitoring: Every patient should be attached to an end-tidal CO2 (ETCO2) monitor and pulse oximetry (SpO2) to assess ongoing ventilation and oxygenation. The alarm on the capnometer must be carefully set as well.
These represent the minimum required alarms and monitors for patients on shared ventilation. An ideal monitor would be connected to each patient’s ETT and display continuous tidal volumes, pressures, ETCO2, and SpO2 among other parameters. This would allow dynamic individual monitoring of both oxygenation and ventilation. Few hospitals have such a monitor, though there are some devices available capable of this functionality.
We also have to explore noninvasive methods to assess ventilation such as respiratory impedance plethysmography (RIP). RIP is used outpatient during the sleep study. RIP continuously monitors thoracic and abdominal movements and displays chest rise. During shared ventilation, RIP could allow assessment of individual parameters rather than relying on the combined parameters displayed on the ventilator. Some software also display tidal volumes (which are approximate at best, but at least to give a rough idea of TV and follow trends). Our RIP software unfortunately does not allow alarms to be set in the event of no chest rise, which could indicate an emergency such as obstruction of ETT. In the context of a pandemic it is infeasible to bring these machines to the bedside, but these approaches have a role in the future development of shared ventilation strategies.
Disadvantages of shared ventilation
Numerous disadvantages of shared ventilation remain:
- FiO2, RR and I:E ratio cannot be individualized.
- Even though the initial FiO2 and PEEP requirements may be similar, the lungs may recover at different rates and dynamic adjustments to each individual lung is difficult. This may be mitigated to some extent if you have an inline PEEP valve available.
- Patients must remain sedated +/- paralyzed during shared ventilation.
- Patients cannot be weaned from shared ventilation and must be transferred to a single ventilator when ready.
- The number of days and cumulative dose of sedation and neuromuscular blockade are likely to be large and have important consequences for the longterm recovery of patients.
- The alarms systems are limited since the ventilator is reading combined VTe thus delaying recognition of events like obstruction of ETT (hence it is important to carefully set the tidal volume alarm)
- Requirement for monitoring of patients, which may be difficult given the shortage of staff with ventilation expertise.
- Even thought viral filters are attached to both inspiratory and expiratory circuits, the risk of cross-contamination may not be completely eliminated.
Overall, there is a potential role for shared ventilation of multiple patients in the context of a pandemic if done correctly through the use of viable methods to individualize and monitor oxygenation and ventilation. However, shared ventilation can be a recipe for disaster if done poorly. VCV should be avoided in favor of PCV or IRV, at least given the equipment currently available. Shared ventilation should be considered only as a last resort once all available resources have been exhausted, when appropriate goals of care and options for palliation have been discussed with family , and the probability of survival is reasonable if shared ventilation is provided.
References and further reading
- Iwashyna, J. Should we put multiple COVID-19 patients on a single ventilator?. litfl.com. 30 March 2020. [Accessed 1 April 2020]. Available at: https://litfl.com/should-we-put-multiple-covid-19-patients-on-a-single-ventilator/
- Neyman G. & Irvin, CB (2006). A single ventilator for multiple simulated patients to meet disaster surge. Academic Emergency Medicine. 13(11), 1246–1249.[Accessed 29 March 2020]. Available at URL: https://onlinelibrary.wiley.com/doi/pdf/10.1197/j.aem.2006.05.009
- Rosenthal BM, Pinkowski J, Goldstein J. ‘The Other Option Is Death’: New York Starts Sharing of Ventilators. The New York Times. [Accessed 29 March 2020]. Available at URL: https://www.nytimes.com/2020/03/26/health/coronavirus-ventilator-sharing.html
- Paladino L, Silverberg M, Charchaflieh, JG, et al (2008a). Reply to Letter: One ventilator multiple patients — What the data really supports. In Resuscitation (Vol. 79, Issue 1, pp. 172–173). https://doi.org/10.1016/j.resuscitation.2008.06.010
FOAM and web resources
- Farkas J. PulmCrit – Splitting ventilators to provide titrated support to a large group of patients. EMCrit. 25 March 2020. [Accessed 29 March 2020]. Available at URL: https://emcrit.org/pulmcrit/split-ventilators/
- Farkas J. PulmCrit wee – Why the SCCM/AARC/ASA/APSF/AACN/CHEST joint statement on split ventilators is wrong. EMCrit. 29 March 2020. [Accessed 30 March 2020]. Available at URL: https://emcrit.org/pulmcrit/pulmcrit-wee-why-the-sccm-aarc-asa-apsf-aacn-chest-joint-statement-on-split-ventilators-is-wrong/
- SCCM/ AARC/ ASA/ APSF/AACN/ ACCP Consensus Statement on Multiple Patients Per Ventilator. [Accessed 29 March 2020]. Available at URL: https://www.sccm.org/Disaster/Joint-Statement-on-Multiple-Patients-Per-Ventilato
novel coronavirus of COVID-19
Simran Kaur Matta, MD is a US-based Critical Care Physician with a passion for physics, mechanical ventilators and POCUS. She believes that ‘proning saves lives’