Mitigating Viral Dispersion during Respiratory Support Procedures in the ICU

Over the past year, the world has been in the grip of a pandemic caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). The coronavirus is causing an ever-increasing number of infections globally and to date is responsible for infection in more than 124 million individuals and more than 2.73 million deaths. SARS-CoV-2 infection can cause serious hypoxemia that requires hospitalization in approximately 20% of infected individuals. Depending on the severity of their illness, 10–25% of hospitalized patients need ICU admission and ventilator assistance.

Various modalities are employed for the treatment of patients hospitalized with coronavirus disease (COVID-19), the disease caused by SARS-CoV-2. Besides antiviral drugs, immune-based therapy, monoclonal antibodies, and convalescent plasma, prone positioning and supplemental oxygen are essential adjunctive measures for relief of hypoxemia. An assortment of interfaces for delivery of supplemental oxygen, including nasal prongs, facemasks of various types, high-flow nasal oxygen (HFNO), or oxygen supplementation with noninvasive ventilation (NIV), are routinely used in critically ill patients.

Aerosols are generated during many respiratory support procedures. Among the aerosol-generating procedures (AGPs) identified by the CDC (1) and the World Health Organization (2) in the ICU, endotracheal intubation, open suctioning, tracheotomy, manual ventilation, and bronchoscopy stimulate coughing and deep respirations and could increase production of bioaerosols containing pathogens from infectious patients. Other AGPs disperse aerosols to the environment (e.g., oxygen administration with nasal prongs or facemasks, HFNO, and NIV) (3). The dispersion effects of the virus in ambient air rely on the amount of virus production, particle size of patient-generated droplets, and the speed and distance of transportation (3). Aerosols generated by these latter AGPs produce “fugitive emissions,” comprising a mixture of aerosols generated by the device and bioaerosols from the patient. The role of fugitive emissions in enhancing the spread of viruses to bystanders or healthcare workers has been a matter for debate (4).

In this issue of the Journal, Avari and colleagues (pp. 1112–1118) used a mannequin that simulated the breathing pattern of spontaneously breathing patients with mild to moderate respiratory distress and exhaled a constant breath-to-breath viral load of a bacteriophage generated by a vibrating mesh nebulizer as a surrogate for SARS-CoV-2 (5). They simulated the clinical situation in a negative-pressure ICU room. Viral dispersion was quantitated at various locations in the room during invasive mechanical ventilation via a cuffed orotracheal tube and a filter placed at the expiratory port of the ventilator and was compared with noninvasive respiratory support using nasal prongs, HFNO, a nonrebreather face mask with N99 high-efficiency particulate air filter, helmet ventilation with positive end-expiratory pressure (PEEP) valve, and bilevel positive-pressure NIV (5). The lowest bacteriophage concentrations occurred during invasive mechanical ventilation, whereas the highest concentrations were recorded while using HFNO or nasal prongs. Moreover, viral concentrations were highest closer to the mouth and lower toward the foot end of the bed. The variability in bioaerosol dispersion led the investigators to conclude that the risk of transmitting infection, and appropriate infection control, differs among several respiratory procedures used in the ICU (5).

Ideally, Avari and coinvestigators would have reported on aerosol generation during intubation and extubation of the mannequin because a burst of aerosol generation during these procedures could pose a significant risk of spreading infection (6). Such a comparison would provide a balance of the risks associated with mechanical ventilation versus other noninvasive respiratory support procedures. However, such a study requires exploring aerosol generation during a whole range of clinical scenarios involving intubation and extubation of critically ill patients.

A major strength of Avari and colleagues’ investigation was to use a bacteriophage to model viral exposure, unlike previous studies that used nonviral particles (References 17–21 in Reference 5). Viruses can spread in the hospital environment by airborne transmission. Tang and colleagues generated an aerosol of live attenuated influenza virus with a jet nebulizer from a mannequin, and when a home jet nebulizer and simple mask was used, they found viral contamination in the environment that decreased with increasing distance from the mouth (7). Lednicky and coworkers used a liquid sampling system placed 2–4.8 m away from a patient with COVID-19 and demonstrated live virions of SARS-CoV-2 in the hospital room even in the absence of AGPs (8). In hospital rooms with patients with COVID-19, SARS-CoV-2 contamination should be expected, especially in the ICU environment (9). Although live virions are present at low concentrations and only in a small percentage of air samples (9), it is essential to reduce exposure to the virus and protect healthcare workers.

Avari and colleagues reported that viral aerosol dispersion was reduced to levels comparable to those seen with invasive mechanical ventilation when a helmet device fitted with a PEEP valve was used (5). Likewise, in previous investigations it was reported that connecting a filter to an oxygen mask (e.g., HiOx Oxygen mask [Novus Medical Inc] or Respan’s Tavish mask) reduced aerosol dispersion (10). In spontaneously breathing patients on HFNO, placing a surgical mask on the patient’s face or using tissue to cover the mouth or nose could reduce the dispersion distance (11) or virus load (12). Adopting appropriate personal protective measures markedly reduces the risk of transmitting SARS-CoV-2 during AGPs (3, 13). In the initial phases of the pandemic, intubation and invasive mechanical ventilation was preferred over NIV and HFNO to diminish aerosol spread. However, with a better understanding of the transmission of SARS-CoV-2 (3), many centers have advocated a more conservative approach to oxygen supplementation in hypoxemic patients with COVID-19 that relies on initiating NIV or HFNO for critically ill patients in whom invasive mechanical ventilation is not essential (14, 15). The investigation by Avari and colleagues (5) highlights the need for further study of SARS-CoV-2 contamination in ICU environments. Furthermore, when NIV or HFNO are used in patients with COVID-19, it is all the more important to develop new methods and devices and promote other control measures that reduce the transmission of SARS-CoV-2 and mitigate the risk to healthcare workers in the ICU.