Fatal flaws in a controlled clinical study of respirator usage in pregnant healthcare workers

To the Editors:We are writing regarding a fatally flawed study(1) published in Antimicrobial Resistance & Infection Control that has resulted in significant and dangerous misinformation circulating during the pandemic.The study, based on data collected under conditions wrongly described as “simulating the actual wearing of an N95 respirator”, concluded that “Breathing through N95 mask materials have been shown to impede gaseous exchange and impose an additional workload on the metabolic system of pregnant healthcare workers, and this needs to be taken into consideration in guidelines for respirator use.” Without reference to respirator surface area and the intended real-world use of N95s, these conclusions are neither meaningful nor accurate.The study and its subsequent citation as evidence for harms of N95 use appear to result from a failure to understand the basic scientific principles relevant to airflow through a respirator. Two fundamental concepts expose the critical flaw in the study. First, all porous materials (including those from which respirators are constructed) have inherent “resistances” to airflow that, in the case of a respirator, manifest in some perceived effort in breathing during use. The relationship between airflow (in, e.g., units of L/min) and this resistance depends on the flow velocity (in, e.g., units of m/s) of the inhaled and exhaled gases as they travel through the respirator material. Second, conservation of mass necessitates that for a given volumetric flow rate, the velocity is inversely proportional to the surface area – halving the surface area, for example, doubles the velocity. At relevant airflow rates for real-world respirator use, resistance is approximately linearly proportional to flow velocity, and therefore inversely proportional to surface area. Limiting breathing to a sufficiently small area of any material will “impede gaseous exchange” and increase breathing effort by greatly increasing the resistance at a given flowrate, while use of a sufficiently large area of any permeable material will not.This study was carried out with subjects wearing a Hans Rudolph mask fitted with a spirometry turbine, with or without a section of N95 material cut from an N95 respirator taped over the turbine holder (study Fig. 2). We obtained the dimensions of the turbine holder (53 mm outside diameter and 37.2 mm inside diameter) from the manufacturer (Cortex, Leipzig, Germany). Very approximate measurements of the more visible right side of the single image provided (study Fig. 2) indicate roughly 20-25% of the radius of the section of N95 material is covered with tape, leaving an exposed patch of radius ~2 cm, and thus area on the order of ~12.6 cm2. The patch used appears to have been cut from a 3M 1860 or 1860S respirator. The breathable area of these respirators will vary with face shape and contact geometry but is on the order of 150 cm2 – more than an order of magnitude larger than what was tested. For a given minute ventilation value, the flow velocity and the associated resistance to flow through the study-modified respirator will therefore be at least ten times greater than in real-world conditions.We emphasize: this entire trial rests on data collected under conditions equivalent to wearing ten or more well-sealed respirators layered atop one another. In fact, the experimental setup would fail to satisfy existing standards for respirators that effectively define maximum permitted resistances to airflow as a primary performance requirement. As a result, the stated findings and conclusions have no relevance to real-world use. Importantly, our criticisms cannot be dismissed as a simple difference of opinion – the study relies on assumptions that are objectively and demonstrably false. It is no more accurate to describe the intervention tested as “simulating the actual wearing of an N95 respirator” than it would be to have subjects perform a one-second rinse with water, and then report that as “simulating proper handwashing technique”.This is not merely a topic of academic interest. Influential authors arguing against universal use of the respirators necessary to protect healthcare workers and others against aerosol-transmitted Risk Group 3 pathogens such as SARS-CoV-2 have used it to claim(2,3) that “…impaired gas exchange and increased metabolic load with the use of respirators has been documented among healthy pregnant women”. To the extent that misinformation arising from this study has interfered with the use of appropriate PPE during a pandemic that has cost “an enormous number of lives”,(4) there is a significant possibility that injuries and death may have already resulted and that future harms will occur if this work is not clearly recognized to be invalid. We therefore request the study be immediately retracted.Further, such a fundamental error illustrates an absolute requirement for adequate mechanistic and subject matter expertise to ensure rigour in the design and execution of clinical trials. It is difficult to imagine that this study would have been carried out had someone familiar with the physics of respirator performance been involved in its conception or approval. In addition to financial costs, trials consume the time and energy of willing participants – a limited resource – and often expose them to increased risk. There is an absolute ethical imperative that trials be scientifically sound, so that these resources are not wasted. Thus, in addition to the trial team, review processes at both the funding and ethics approval stages require relevant interdisciplinary expertise as well.Finally, the review process for publication of such trials also requires this expertise. While the mechanistic understanding of PPE function may not be of interest to the intended audience of Antimicrobial Resistance & Infection Control, this may simply mean that such errors are less likely to be identified in the normal course of events. Adequate interdisciplinary peer review is essential to preventing similar flawed findings from undermining the value of the infection control literature in future.References cited1.Tong, P. S. Y. et al. Respiratory consequences of N95-type Mask usage in pregnant healthcare workers—a controlled clinical study. Antimicrobial Resistance & Infection Control 4, 48 (2015).2.Ng-Kamstra, J., Stelfox, H. T., Fiest, K., Conly, J. & Leigh, J. P. Perspectives on personal protective equipment in acute care facilities during the COVID-19 pandemic. CMAJ 192, E805–E809 (2020).3.Conly, J. et al. Use of medical face masks versus particulate respirators as a component of personal protective equipment for health care workers in the context of the COVID-19 pandemic. Antimicrobial Resistance & Infection Control 9, 126 (2020).4.Bredow, R. von & Hackenbroch, V. Epidemiologist Jeremy Farrar on the Next Viral Threat: “I Fear We Are at the Beginning of an Era of Pandemics”. Der Spiegel (2022).