3

Exploring the State of the Science and Potential Priorities for Research and Standards Development

Over the course of the next two panel sessions, seven speakers discussed various aspects of the state of the science and potential priorities for research and standards development in the areas of filtration performance, fluid resistance, flammability, and biocompatibility/usability. Prior to the workshop, the panelists were asked to address the following questions in their remarks:

• What improvements are needed to the tests and test methods?
• What efforts are under way to revise the standards?
• What are the research gaps and priorities?
• What are the priorities for research, test method development and refinement, and post-market surveillance of N95s to improve health care workers safety and health?
• What are the priorities to be considered in integrating FDA and NIOSH evaluation processes for N95s used in health care settings?

TESTING THE FILTER PENETRATION PERFORMANCE OF RESPIRATORY PROTECTION DEVICES

Robert Eninger, Air Force Institute of Technology

Numerous factors influence filtration efficiency (see Box 3-1), and as a result, a filter’s measured filtration efficiency will depend on the test aerosol, the particle size, the way in which aerosol is measured inside and outside of the filter, and other test parameters. “If you want to compare apples to apples, all of these [different factors] must be similar between the two different test regimes to have results that are comparable,” said Eninger.

TABLE 3-1 Filtration Test Methods

NOTE: lpm = liters per minute; MMAD = mass median aerodynamic diameter; PSL = polystyrene latex spheres.

SOURCE: Presented by Robert Eninger, August 1, 2016.

With regard to filtration test methods, Eninger considers the NIOSH certification standards (see Table 3-1) as a more conservative set of filtration test parameters as compared to the ASTM F2299 standard included in FDA’s regulatory requirements. Starting with particle size, he explained that the NIOSH certification uses sodium chloride particles with a mean particle size corresponding to the most penetrating particle size based on the principles of mechanical filtration.

The electrical charge characteristics of an aerosol can significantly influence its ability to be filtered depending on the specific filtering medium. The NIOSH test uses a charge-neutralized aerosol, creating a worst-case scenario for a filtration medium, while the ASTM method uses an unconditioned aerosol, which means that the aerosol’s particles are charged to some extent. The NIOSH test uses an entire respirator that has been preconditioned for 24 hours at 38°C and 85 percent relative humidity, which represents a challenging high-humidity environment. The ASTM method tests a sample of the filtration material with no preconditioning. In addition, the NIOSH method is run until a mass of approximately 200 milligrams is deposited on the filter—a relatively large amount of mass—while there is no specific requirement for a specific mass to be tested in the ASTM method.

Flow rate is important because people respire at different rates under different workloads. The flow rate influences the residence time of particles as they pass through the filtration medium. The NIOSH method uses a constant flow rate of 85 liters per minute, which is an estimate of what a high peak inspiratory flow might be for a very active workplace,

while the ASTM method uses a constant flow rate of 28 liters per minute that may be more typical of the respiratory rate of a surgeon or other health care worker. Finally, the NIOSH method assesses the mass of particles passing through the respirator using light scattering, while the ASTM method uses light scattering to count the number of particles passing through the filter material sample.

Another ASTM test method listed in the FDA notice for N95 respirators, ASTM F2101, assesses biological filtration efficiency. This test changes two parameters from the ASTM particle filtration efficiency test: using aerosolized Staphylococcus aureus particles with mean diameter of approximately 3 microns and collecting the particles that pass through the filtration medium on a six-stage aerosol impactor. The collected particles are then plated out for a defined time and the number of bacterial colonies that develop on the plates are counted to produce a measure of colony-forming units per volume of air.

Eninger said he considers NIOSH certification to be rigorous, repeatable, and a near worst-case scenario for a respirator. It does not use a biological aerosol for certification of N-type respirators. Over the past decade, researchers have revised the parameters regarding the understanding of the most penetrating particle size for a respirator. Many respirators manufactured today use electrostatically charged filtration media to increase filtration efficiency while reducing breathing resistance, creating respirators that place a lower physiological burden on users. Research has shown that the most penetrating particle size for respirators using these electrostatically charged materials shifts to 100 nanometers or less and even as low as 30 nanometers (Eninger et al., 2008). Work from NIOSH (Rengasamy et al., 2013) has shown that removing the charge from particles increases the most penetrating particle size. Other research (Harnish et al., 2013, 2016) has shown that biological particles follow the same filtration physics as do inert particles, and so the capture mechanism does not discriminate in favor of or against biological particles. These results suggest that adding the FDA requirements regarding testing for filtration performance to the NIOSH requirements would not improve the assessment of N95 filtration efficacy (Rengasamy et al., 2016).

PARTICLE PENETRATION PATHWAYS

Sergey Grinshpun, University of Cincinnati

For someone wearing a respirator there are two potential pathways for aerosol particle penetration: through the filter medium or via faceseal

leakage. The current NIOSH certification process, which measures particle concentrations inside and outside of the filter used by a respirator and assesses the filtration efficiency, addresses the first of these pathways. Over the past several decades, said Grinshpun, “industry has done a wonderful job improving the physical collection efficiency of filters, but faceseal leakage has been largely ignored or overlooked in designing new respirators or considering new configurations of facepieces.”

To measure faceseal leakage, the respirator undergoes performance testing while fitted on a human subject. Measuring particle concentration inside and outside of the respirator on the user allows the determination of what is termed “total inward leakage”—the sum of the two types of particle penetration (through the filter and through the faceseal). Given the improvements made in filtration media, leakage through the faceseal can be comparable to or exceed penetration through the filter material. Thus, said Grinshpun, it becomes important to look at the relative contributions of each pathway. Assessing the relative contribution in a laboratory setting is done by fitting a manikin with a respirator and having the manikin “breathe” according to exercise-specific breathing patterns, which have been recorded when the same type of respirator was fitted on a human subject following the same set of exercises. Subtracting the filter material’s contribution as determined using the NIOSH test procedure for total inward leakage yields a value that quantifies the faceseal leakage (Grinshpun et al., 2009). Studies in Grinshpun’s laboratory have shown that penetration through faceseal leakage is, in fact, a major contributor to total particle penetration into the respirator and in some instances can exceed penetration through the filter material by 10-fold. Other studies (Chen and Willeke, 1992; Rengasamy and Eimer, 2012) produced similar results, as has computational modeling of total inward leakage. The latter has shown that amount of faceseal leakage is related to the ratio of the area of a leak in the faceseal to the area of the filter.

Having identified faceseal leakage as a critical factor in a respirator’s particle filtration performance, Grinshpun and others have been developing novel faceseal designs to address this problem. One such effort, which aimed to improve respiratory protection against surgical smoke, came up with an elastomeric material with a shape that was designed based on the anatomy of the human face. Comparing the performance of an N95 respirator with this new type of faceseal with that of a standard N95 respirator on 10 human subjects (who were conducting simulated electric cautery surgery) showed that this new faceseal afforded greater protection against particle exposure at a statistically significant level.

Grinshpun noted that this is not the only design that companies have recently developed to address the faceseal leakage problem.

In conclusion, Grinshpun suggested that because faceseal leakage may represent a major pathway for aerosol particle penetration into respirators, particularly for those made with highly efficient filter materials, particle penetration tests for certification should include total inward leakage measurements and quantification of faceseal leakage. While there are experimental and computational methods for determining the efficiency of filtering facepiece respirators, he said that priorities for research and standard development should include the development of test methods capable of accounting for faceseal leakage. Existing NIOSH and FDA evaluation processes, Grinshpun indicated, might benefit from integrating a leakage test into the existing protocols.

FLUID RESISTANCE TESTING

Brandon Williams, Nelson Laboratories

Testing of respirators to meet the FDA criteria for fluid resistance is usually done using the ASTM F1862 test method standard, which assesses whether the synthetic blood penetrates into the layers of the respirator. The test involves dispensing synthetic blood using a pressurized cannula for a distance of 12 inches through a small hole in a target plate to which the respirator is attached. A visual inspection of the inside of the respirator for fluid penetration is then conducted. As Williams explained, the multiple layers of filtering material in some respirators can make visual identification of blood penetration difficult and sometimes requires lightly brushing a cotton swab on the inside of the device to see if there was penetration. This human factor—the force of brushing the swab against the inside of the product can be too hard or too soft—is one aspect of the test method that is subjective.

Several features of the test method approximate potential surgical exposures to blood. The pressure levels used to shoot synthetic blood at the respirator range from 80 millimeters of mercury (mm Hg) pressure to 120 to 160 mm Hg, which spans the typical human blood pressure. Williams noted that the 12-inch distance between the cannula and the respirator mimics the scenario of how close a surgeon can come to the patient during surgery. He also explained that the surface tension of the synthetic blood specified in ASTM F1862 is at the low end of the range for human blood, which mimics human blood at its most penetrating surface tension. The standard specifies that three or fewer failures are al-

lowed out of a set of 32 samples in order to pass the test. One advantage of this test aside from its simplicity, he added, is that it has been used for more than 15 years and as a result there is a significant body of data against which to compare various respirator materials and products.

ASTM F1862 is not a perfect method, however. One challenge, said Williams, is measuring surface tension accurately, and he noted that researchers at CDC have been trying to standardize surface tension measurements so that the test will produce consistent results from one laboratory to the next. “Most people purchase the test blood from the same manufacturer, yet sometimes they get different results using the same test method, so that is something that is being improved on right now,” said Williams. Another challenge is that the visual inspection must occur within 10 seconds after the mask or respirator has been sprayed. He questioned whether that is a reasonable amount of time for a surgeon to remove a mask or respirator after getting sprayed with blood or if it really takes longer than that. “When the standard comes up for revision, I think that should be looked at to see if that is a reasonable expectation,” said Williams. Two other disadvantages are that this is merely a penetration test and not a microbiological one, and that respirators with seams or the newer duck-billed products are difficult to test because of the low locational accuracy of the fluid-dispensing apparatus. The standard does not address how to test different areas of a respirator.

In Williams’s opinion, blood penetration testing is important for surgical respirators because filtration materials have to be porous so the user can breathe, but porous products will allow for some liquid penetration. He noted that they are not aware of any other test result that could be used as an indicator for how the product will perform in the blood penetration test.

FLUID RESISTANCE TESTING AND PROTOCOLS

Steven Elliott, Food and Drug Administration

Fluid resistance is one of the elements that is assessed in FDA’s 510(k) approval process for surgical N95 respirators, said Elliott. The 510(k) approval process, he added, compares new devices against those that have already been granted clearance. As discussed above, FDA recognizes consensus standard ASTM F1862 as the specific methodology to be used to test fluid resistance. The agency also recognizes ASTM F2100 as the consensus standard for performance specifications of the materials used in respirators and medical face masks.

Elliott said FDA recognizes both the strengths and limitations of the

current fluid resistance tests that the previous speaker discussed. Using the ASTM standards is voluntary—manufacturers can develop their own comparable tests—but because several stakeholders from industry, the health care sector, and FDA have had input into the development of these standards, it often turns out to be easier for manufacturers to use the ASTM methods and performance specifications. He also pointed out that FDA only requires a blood resistance test, but it recognizes that other fluids, such as caustic chemotherapy drugs, could have different properties that may require additional testing. “I am unaware of any specific clearances we have along those lines, but that is something that would be evaluated and require additional challenge conditions in the testing and possibly an entirely new test,” said Elliott. Other performance claims might also result in the need for a manufacturer to use additional tests.

Elliott said that while FDA is open to changes and modifications to make the test methodology more appropriate for N95 respirators, any new procedure would have to take into account the regulatory history of these standards. He noted that any new procedure that would be incorporated into the agency’s regulatory framework would need to look at performance expectations with regard to fluid resistance. “You can increase fluid resistance on any mask or respirator, but there will be some sort of tradeoff in terms of breathability and we want to make sure that we would be moving in the appropriate direction and taking in the concerns from all fronts on that,” said Elliott.

FLAMMABILITY TESTING FOR RESPIRATORS

Samy Rengasamy, National Personal Protective Technology Laboratory

At first glance, fire hazards would not seem to be a risk to which health care workers are exposed, but in fact, said Rengasamy, when surgical fires do happen, they can be catastrophic. He noted that the Emergency Care Research Institute reports that about 600 surgical fires occur in the United States each year. Operating room fires have many sources (including electrical surgical equipment, alcohol-based agents, surgical drapes, and gases such as oxygen and nitrous oxide) and are most likely to occur during procedures such as endoscopic airway surgery, oropharyngeal surgery, tracheostomy, and cutaneous surgery.

In 2014, when NIOSH published a notice on respiratory protective devices used in health care in the Federal Register, issues were raised as to whether NIOSH should consider adding tests and requirements on splash and spray protection, protection against flammability hazards, and

bacterial filtration efficiency to the 42 CFR 84 conformity assessment process for respirators (HHS, 2014). To address these questions, NIOSH initiated a research project on fluid resistance and flammability of respirators and other head and face personal protective equipment. The equipment tested in this research included N95 respirators, PAPRs, hoods, surgical head covers, surgical N95 respirators, and surgical masks. The project evaluated fluid resistance for N95 and surgical N95 respirators, as well as surgical masks, using the ASTM F1862 standard method (Rengasamy et al., 2015; see description of the method above). Filtration efficiency of these devices was assessed using the NIOSH sodium chloride method and the results were compared with the FDA-required particulate filtration efficiency and bacterial filtration efficiency test methods (Rengasamy et al., 2016). The flammability testing is under way for this project and is using the Consumer Product Safety Commission (CPSC) method described in 16 CFR 1610. This test measures the time that it takes for fire to traverse across a sample of material (positioned at a 45 degree angle) once it has been ignited. The average burn time for a set of five samples is used to assign a material to a flammability class. Class 1 materials (normal flammability) are less likely to burn and take longer than 3.5 seconds to burn.¹ The average burn time for Class 3 materials is less than 3.5 seconds for plain surface textile fabrics. FDA recommends that surgical masks and respirators be made from Class 1 and Class 2 materials and requires a flammability warning notice if Class 3 materials are used.

At the time of this workshop, Rengasamy and his colleagues had tested 11 N95 models and 8 surgical N95s, and all of the models passed the flammability test. Many of the samples, he said, did not ignite at all and some ignited but self-extinguished. Others did burn but with an average burn time exceeding 3.5 seconds. Comparable results were obtained and confirmed by a third-party independent laboratory. Rengasamy noted that 7 out of 11 N95 models met the FDA requirements for fluid resistance and flammability testing, as did all 8 surgical N95s. The take-home messages, he said, are that NIOSH has the capacity to perform the 16 CFR 1610 flammability testing and that there may be several N95

___________________

¹16 CFR 1610 states that plain surface textile fabrics that have a burn time of 3.5 seconds or more are classified as Class 1, normal flammability; for raised surface textile fabrics, Class 1 is used for those with a burn time of more than 7 seconds. For raised surface textile fabrics, intermediate flammability (Class 2) is defined as a burn time of 4 to 7 seconds. Class 2 does not apply to plain surface textile fabrics. Class 3 (rapid and intense burning) materials are those that “exhibit rapid and intense burning, are dangerously flammable and shall not be used for clothing.”

models on the market that meet FDA requirements but that have not been submitted to FDA for clearance. This could provide more models of N95s that meet FDA criteria that are available for emergency use during the infectious disease seasons, said Rengasamy.

FLAMMABILITY TESTING: TEST METHODS AND STANDARDS

Roger L. Barker, North Carolina State University

FDA cites three standards with respect to the flammability of surgical N95 respirators, said Barker. The first, developed originally in the 1950s by CPSC² and described above by Rengasamy, is used in testing apparel worn in the United States. This standard stipulates a test method that has been used for more than 60 years so there are extensive data. CPSC has established three classes of flammability based on the rate of burn propagation in that test. FDA also cites National Fire Protection Association standard 702, which is similar to the CPSC standard and was actually withdrawn in 1986 in deference to the CPSC standard, and Underwriters Laboratory 2154, which measures the level of atmospheric oxygen required to propagate flame when the ignition is caused by electrical surgical laser. As far as Barker could determine, this latter test is not readily available. He noted that all clothing materials, even those treated to be flame-resistant, will burn if a high-intensity heat source is applied to them in the presence of sufficiently elevated oxygen levels.

Having spent much of his academic career developing test methods to look at thermal protection for protective clothing, Barker said one thing he knows to be true is that there are many variables, such as the ignition source and oxygen level, that can affect and determine flammability beyond the material itself. He also pointed out that all of the recognized test methods look solely at the filtration material, not the entire respirator. Many different metrics can be used to describe flammability, including ease of ignition, burning rate, heat release, thermal stability, and others. Barker said that the tests focus on the flammability of the product’s materials and not on the way in which the material is put into use (e.g., the form of the products such as respirators, masks, and gowns), which can affect the outcome of a material’s burning behavior.

With regard to what he believes to be the most important considerations for assessing a flammability test method or whether it is even nec-

___________________

²This standard is alternatively referred to as 16 CFR 1610 and CPSC CS-191-53.

essary, Barker said the first is to understand the flammability hazards in surgical and nonsurgical environments as they relate to the exposures that could occur to the user wearing a respirator. “When you do that, what you are able to do is more accurately match the flammability requirement with the potential hazard in a reasonable and practical type of way with a goal of ensuring safety,” said Barker. Two other considerations are the need for uniformity in reporting test results so that there is clarity about the class to which a material belongs, and the importance of risk assessment that accounts for the entire landscape of the factors that might affect the functional performance of the respirator or other product. As noted above, FDA has stated that surgical N95s should use materials that meet Class 1 or Class 2 flammability standards.

EVALUATING BIOCOMPATIBILITY

BiFeng Qian, Food and Drug Administration

As had already been discussed, surgical N95 respirators are Class II medical devices subject to FDA’s 510(k) review to ensure they address the performance and safety issues described in a 2004 FDA guidance document for surgical masks (FDA, 2004). According to this document, biocompatibility is an expected element of surgical N95 respirators (as it is with many other medical devices), and in September 2016, FDA issued a new guidance document on biocompatibility evaluation (FDA, 2016).

Biocompatibility evaluation of a medical device, said Qian, should first consider how the device will contact the body and how long it will remain in contact with the body. Based on their intended clinical use and the FDA 2004 guidance document for surgical masks (FDA, 2004), surgical N95 respirators are classified as surface devices that contact intact skin for a limited duration. Biocompatibility evaluation also needs to consider the primary material used as well as all of the other ingredients, including plasticizers, additives, crosslinkers, reagents, colorants, inks, adhesives, surfactants, detergents, antimicrobial coatings, process contaminants, and sterilant residues. All surgical N95 respirators seeking 510(k) clearance, Qian added, should address biocompatibility concerns for cytotoxicity, skin irritation, and dermal sensitization as specified in the FDA-recognized test standard for a surface device, intact skin contact, and limited duration use (International Organization for Standardization [ISO] standard 10993-1). In addition, if the surgical N95 respirator is to be sterilized, the sterilization residues should be examined to ensure

that requirements are met for the ISO 10993-7 acceptance criteria for limited exposure devices.

Qian provided an overview of the four consensus standards that FDA recognizes for testing biocompatibility (see Box 3-2). These evaluations are supposed to be conducted on the final product, representative samples from the final product, or materials processed in the same manner as the final product, including sterilization. Biocompatibility testing based on manufacturing raw materials or unfinished device parts may have limitations and is generally not accepted by FDA without solid justification. Evaluations should cover all device components with the potential to contact patients or users and not be limited to the inner layer. “In the worst clinical use condition, chemical residues may leach out from other layers and come into contact with the patient or user,” Qian explained. FDA, she added, considers that addition of a color additive to a medical device is a significant change to the device. If a surgical N95 respirator has more than one color type, each color type needs to be assessed for biocompatibility.

In some cases, sponsors may conduct biocompatibility testing on a different device. If so, the sponsor will need to justify the switch and include a certification stating that the test article is identical to the proposed medical device in its final finished form in formulation,

processing, sterilization, and geometry, and that no other chemicals, such as plasticizers, fillers, additives, cleaning agents, and mold release agents, have been added. In other instances, the surgical N95 respirator may be identical to a previously cleared device in terms of the materials, chemicals, and processes, including sterilization. If this cleared respirator has established a history of safe use in the intended application and user population, the sponsor may provide a material certification statement for comparison to the previously cleared device in lieu of new biocompatibility testing.

FDA may have additional concerns about biocompatibility, Qian noted, if a surgical N95 respirator has additional specific issues, including having materials known to be associated with significant health problems, having antimicrobial or other specific coatings, if it is intended to be used in biologically vulnerable populations, or if it contains special labeling claims. Based on those concerns, FDA may request additional testing or information.

As a final note, Qian said biocompatibility of a medical device may need to be reevaluated if changes are made to the device (including changing the source or specification of the materials used in the product or how it is manufactured, packaged, or sterilized), if there is a change in product shelf life or intended use, or if there is any evidence that the product may produce adverse effects when used in humans. “Based on the assessment of potential impact of the changes, new biocompatibility testing may or may not be warranted,” said Qian.

DISCUSSION

The discussion following the presentations focused on the range of tests outline in FDA and NIOSH requirements and criteria.

Particle Filtration Testing

James Zeigler, from J.P. Zeigler, LLC, asked the panelists if they knew of any examples where respirators were tested for viral penetration in the same way that personal protective clothing fabrics have been tested. Williams said he was unaware of anyone conducting such a test on an N95 respirator, although it has been done on surgical masks.

In response to a question about priorities for integrating the FDA and NIOSH certification processes, Eninger replied that if the goal is to look at filtration for purposes of determining filtration efficiency, then testing

with an inert particle is adequate. He said that one drawback of the ASTM test is that it does not utilize test conditions that are well controlled in other areas, such as the electrostatic condition of the aerosol or the loading of the filter. He noted that additional research is not warranted because methods already exist for dealing with these variables and implementing those methods should be the next step.

Elliott clarified that the ASTM specifications regarding filtration efficiency are not FDA methods. Rather, when FDA clears an N95 surgical respirator, it works with NIOSH in terms of certification. Terrell Cunningham from FDA added that FDA only clears NIOSH-certified N95 respirators. FDA confirms that the respirator has a valid NIOSH certification number and therefore has met the particle filtration efficiency standards and the other NIOSH requirements. Then FDA looks at whether the product also meets the biocompatibility, flammability, and fluid resistance standards and, if so, the product could be approved as a surgical N95 respirator. If the surgical N95 respirator is proposed to have an antimicrobial additive, then FDA imposes additional requirements because of the concern that chemically attaching or embedding an antimicrobial could affect a filter’s barrier performance. Biocompatibility would also be a concern with an antimicrobial respirator.

Williams noted that NIOSH will be publishing a paper showing that if an N95 respirator passes NIOSH testing, it is close to guaranteed that it will pass bacterial and particle filtering efficiency testing. However, passing the bacterial and particle filtering efficiency tests does not guarantee passing the NIOSH testing protocol. Elliott then explained that the expectation at FDA is that an N95 respirator submitted for FDA clearance as a surgical N95 will have been certified by NIOSH to pass the differential pressure and particle filtration efficiency tests, but with regard to bacterial and viral filtration efficiency, data from those tests will be necessary for FDA’s review.

Elizabeth Claverie-Williams from FDA pointed out that whether a device requires bacterial and viral testing has to do with its intended use. “Whatever claims the sponsor makes as related to the performance and effectiveness of the device will drive the types of tests that we will require in order to support those claims,” she explained. In response to a question about who certifies a nonsurgical N95 that someone would wear in a tuberculosis isolation room, for example, Claverie-Williams said that if the product is not designated as a surgical N95, then it is solely approved by NIOSH; FDA is not involved in the approval process for standard N95s.

Grinshpun noted that “a particle is a particle” and that the important

characteristic of particles with regard to testing is their aerodynamic, not actual, size. Aerodynamic size, he said, accounts for the fact that bacterial and viral particles may not be perfect compact spheres such as the test particles used for testing filtration efficiency.

Jeffrey Peterson from NIOSH’s NPPTL explained that when NIOSH receives products that have an antimicrobial or infection control claim, they talk with the manufacturer and coordinate with FDA to ensure that the product is undergoing the appropriate tests to meet the FDA criteria.

Fluid Resistance Testing

James Chang noted that blood and body fluid splash is a significant challenge in protecting health care workers. David Prezant agreed and noted that fluid exposures are also a major occupational health issue for paramedics and emergency medical technicians and these exposures raise the potential need for prophylactic therapy for HIV or other infections. Prezant also raised an issue having to do with the fact that the ASTM test protocol uses a narrow stream of fluid directed at the respirator, but first responders and health care workers wearing a respirator are more likely to be struck by a fine mist from coughing or a large fluid mix from vomiting rather than a jet of blood. If that is the case, he wondered why such testing is required when the best protection would be afforded by a face shield. In Prezant’s opinion, face shields, not fluid-resistant N95s, are the solution.

Elliott agreed this was a good point, and said FDA would never argue that a respirator alone would be the appropriate choice for personal protection from workplace hazards across the entire range of possible uses or in every contamination situation possible. Williams said in his opinion, testing fluid resistance using a high velocity stream would be a worst-case scenario.

Howard Cohen from the Yale School of Public Health wondered if the test for fluid resistance is putting up an unnecessary barrier that limits the supply of respirators that would afford particle protection—the predominant role for these respirators—for the majority of health care workers, both on a routine and emergency basis. Elliott agreed this was a fair point and suggested that one solution would be to require a different device for fluid protection in situations where fluid protection is important. However, the current regulations do not make that distinction because surgical N95 respirators are regulated as surgical apparel, where the expectation is that there would also be barrier protection against bodily fluids. Prezant noted that any health care worker concerned about

fluid exposure would don a face shield and not depend on a surgical N95 respirator because the respirator only covers a small portion of the face.

Zeigler noted that in his opinion having a requirement for fluid resistance is creating a false sense of security and could result in breathability and air flow problems given that fluid resistance is tied to air flow. James Johnson said Zeigler’s comment points to the need for research on how these respirators are used in practice in operating suites and other areas of the hospital.

Flammability Testing

In clarifying the need for flammability testing of N95s for use in health care settings, Cohen mentioned that the specific occupational hazards need to be more carefully assessed.

Nesbitt and Chang noted that they view flammability resistance as a low-priority issue for health care respirators. Prezant also questioned the extent to which fires related to respirators are happening in operating rooms or endoscopy suites. These discussions led to comments about the need for increased hazard assessment efforts in the health care environment to more fully examine the hazards that are present where N95 respirators are used. Mark Shirley from Sutter Health noted that he and other workshop planning committee members had done some exploration of the literature and experiences regarding surgical fires and had not found evidence of filtering facepiece respirators being involved with fires. He emphasized the risk-assessment needs regarding the flammability of respirators and noted that is an area of opportunity for further research and guidance for health care professionals.

In response to questions about the methodology of the flammability test, Barker noted that this is a long-standing and reproducible methodology.

Biocompatibility Testing

Responding to questions about whether biocompatibility is an attribute that FDA and NIOSH need to address as part of the MOU, Chang and Sood both said that biocompatibility should be considered, particularly for respirators that have specific additions, such as antimicrobial properties. Craig Colton said that biocompatibility is an important attribute that manufacturers assess for respirator products in all industries, including health care. Biocompatibility testing, as noted by Qian, is not limited to respirators used in the operating room if the respirator is claimed to be a medical device. Peterson added that even though nonsur-

gical N95s do not require FDA clearance, they still need to meet NIOSH’s requirement that a device worn by a wearer cannot cause any harm to the wearer. “We deem that as the manufacturers’ responsibility and that when they submit the application they have done their due diligence in actually performing that work” said Peterson. Colton said one reason 3M has a toxicology department is to ensure that none of its products with intended human use has a biocompatibility issue regardless of whether or not the product will be submitted for FDA clearance.

Maryann D’Alessandro pointed out that NIOSH has not had any reports of biocompatibility issues with nonsurgical N95s and that the agency has a certified product investigation process and audit process that would follow up on any reported issues. Jennifer Goode from FDA added that the tests used to assess biocompatibility are well understood and consistently conducted. “When we see differences in results, it is usually not because it is at one test lab or another,” said Goode. “It is because there is something that is in the final product that somebody did not anticipate. When we see a toxicity for a normally used material, it is because something happened either from the supply or in the manufacturing.”

Cunningham then explained that FDA’s expectation is that NIOSH would not issue a certification of a surgical N95 respirator until it had conferred with FDA and confirmed that FDA was also reviewing the product. The important point, he added, is that NIOSH and FDA do coordinate the review and labeling of these products. Similarly, Peterson pointed out that when NPPTL receives an application for a respirator that does have an impregnated antimicrobial agent, it notifies the company filing for certification that it needs to notify FDA as well and submit a 510(k) application. “We would not take action on issuing approval until that request has been received by FDA and validated against their requirements,” said Peterson. He added that NIOSH and FDA keep each other informed about deficiencies that turn up during the certification process and that the two agencies work together on product labeling.

Cecile Rose from the University of Colorado Denver said it appears based on the presentations and discussion that it makes sense for manufacturers to do biocompatibility testing as opposed to having it done by NIOSH or FDA. Cohen, commenting on the importance of biocompatibility and flammability testing, agreed with Rose and said, “it appears to me that these are two issues that could be well handled by the manufacturer by saying these are standards that you must meet, which is really what FDA does.” If there is an issue, he said, FDA or NIOSH would investigate as opposed to adding a requirement for biocompatibility and flammability testing to the NIOSH certification test.

Qian pointed out that biocompatibility applies to both the patient and the health care workers. Goode added details on this, noting that the ISO 10993-1 standard is currently undergoing revision and that FDA does have regulatory authority for medical devices that are primarily in contact with the clinician as well as those in contact with the patient. Joyce Lee, a toxicologist at Halyard Health, agreed that examining irritation and sensitization are important both for the patient and the clinician.

Overarching Comments

Andrew Levinson from OSHA noted that the protective value of an unworn respirator is zero and that this is “really the first time that NIOSH is looking at an industry-specific respirator standard that would be just for health care usage.” Given that, he wondered if there was a need to have standards for user acceptability criteria such as heat and moisture buildup and breathing resistance that would strike a balance between adequately protecting people but not overly protecting them with a respirator that is uncomfortable and less likely to be worn. Mark Shirley agreed that is an important concern from an end-user’s perspective and balancing protection. Usability, said Shirley, should be part of a risk assessment that each organization needs to conduct to ensure the appropriate equipment is worn and that employees understand how to put it on and take it off. Geeta Sood agreed with this last idea and said, “From a user’s point of view, I would much rather have a simple, easy-to-use mask that would not be fluid-resistant but that would allow people to use it and not have as much user error.”

With regard to expiration dates, Prezant asked if anyone has ever tested these products to see if they have the same flame resistance, fluid resistance, and biocompatibility over time. Barker noted the need for this type of research and emphasized that the degradation of flammability performance in textiles can occur over time. Qian said devices with materials that may degrade over time need more testing to evaluate the expiration date. Johnson said the biggest factor in the degradation of performance would likely be how the products are stored. D’Alessandro added that research is under way at NPPTL to look at these issues and the main finding so far has been that filtration performance is not affected over time, but the bands that secure the respirator on the face and the foam around nose bridges do degrade.