Response

To the Editor:

The coronavirus 2 pandemic has created an unprecedented rapid production of new information in the scientific literature. Translating experimental data into clinical practice remains a challenge that requires care and thoughtful validation. We greatly appreciate the commentary from Madney et al and agree with their primary conclusion that a more comprehensive study is needed to test the impact of implementation. Ultimately, careful monitoring of nosocomial transmission, especially to clinical staff, must be reported and that information considered in the risk-benefit calculation of any management paradigm. Madney et al ask a very reasonable question regarding the model—the selection of a peak flow of 40 L/min for the simulation. The communication is a short format, and the mechanistic considerations of the therapy tested could not be fully articulated.

The therapy simulated in our paper¹ evaluates a therapy classified as high-velocity nasal insufflation (HVNI). HVNI is mechanistically different from high-flow nasal cannula (HFNC) therapy, while sharing some important characteristics. Both provide supraphysiological flow of conditioned, oxygen-rich gas to the patient using an open interface.² HVNI is in a separate Food and Drug Administration classification because it imparts higher velocities to the high flows of gas to facilitate the rapid clearance of the extrathoracic dead space. This provides the ability to manage respiratory distress in spontaneously breathing patients at volumetric flow rates typically well below 40 L/min. The clinical impact of this compared with other forms of HFNC has not yet been fully defined. In a trial of 204 adult patients presenting in the ED with undifferentiated respiratory failure requiring noninvasive ventilatory support, Doshi et al³ demonstrated noninferiority to noninvasive positive pressure therapy with initial flow rates of 35 L/min and final titrated flow rates of 30 ± 6 L/min. Higher flow rates are commonly seen when using conventional HFNC therapy,⁴ and such higher mass flow of gas leaving the nose and mouth may impact the overall particle dispersal and subsequent transmission. However, experimental evidence from Hui et al⁵ suggests that a surgical mask over conventional forms of high-flow therapies likely reduce the risk of aerosol dispersion from them as well. All experimental evidence to date, although limited, suggests these therapies are safe and have limited risk for nosocomial or health-care provider infection when mitigated by use of a surgical mask on the patient.

The higher mass flow quantities generated with conventional HFNC were not evaluated in this report, and as such, the impact on the ventilatory effect of dead space washout for such therapy was not determined in this study and may be an interesting topic for future study. The current analysis demonstrated that the washout of the accessible extrathoracic dead space modeled in this simulation, with the virtual mask in place, still provided significant elimination of rebreathed CO₂, and that reduction was amenable to adjustment in the flow rate of the high velocity gas entrained into the nose of the patient.

We greatly appreciate these suggestions for addition in our ongoing evaluation of mitigation methods being considered for patient management.