Surgical masks not effective, study shows

IOM panel hears range of research on masks

Surgical masks do not provide protection from aerosolized viral particles, respiratory protection experts told an Institute of Medicine (IOM) panel that was considering personal protective equipment and novel H1N1.

A randomized clinical trial comparing N95 filtering facepiece respirators and surgical masks, found that "the surgical masks had no efficacy against any of the [measured] outcomes," reported C. Raina MacIntyre, MBBS, M App Epid, PhD, FRACP, FAFHM, Head of the School of Public Health and Community Medicine at the University of New South Wales in Sydney, Australia. The respirators reduced clinical respiratory illness by 65%, influenza-like illness by 75%, and laboratory- confirmed influenza by 75%, she said.

The as-yet-unpublished study involved 1,935 health care workers at 24 hospitals who wore masks, respirators, or no facial protection for four weeks during the winter. Researchers then tracked the health care workers for five weeks to detect onset of respiratory illness. Overall, respirators were 42% more protective than masks, although no statistical difference was found between fit-tested and nonfit-tested masks.

This evidence of the lack of protectiveness of surgical masks gained attention at the meeting. Yet other presenters expressed concern about confusing guidance and lack of compliance by health care workers.

"[Employees] were telling me they were having difficulties carrying out their duties wearing the respirators throughout their shifts," said Leonard Mermel, MD, professor of medicine at Brown University and medical director for the Department of Epidemiology and Infection Control at Rhode Island Hospital. "I agreed to back down to [surgical masks due to] their requests."

The IOM panel was scheduled to issue a report by Sept. 1 to Thomas Frieden, director of the Centers for Disease Control and Prevention in Atlanta. Frieden is considering a CDC shift to droplet precautions - gowns, gloves, and surgical masks - for health care workers caring for novel H1N1.

"The guidance is effectively going to apply to all individuals presenting with upper respiratory infection," cautioned Toby Merlin, MD, deputy director of the Influenza Coordination Unit at CDC. "There is no means to [immediately] distinguish people who present with novel H1N1 infection from people who present with seasonal influenza from people who present with other upper respiratory infections. Those precautions need to be able to be used everywhere that health care workers are encountering individuals with upper respiratory infections."

The IOM heard a range of research and perspectives on personal protective equipment and influenza:

• Surgical masks vary from providing almost no protection to significant protection - but it's impossible to tell the difference between masks.

Roland BerryAnn, deputy director of NIOSH's National Personal Protective Technology Laboratory, reported that a test of five randomly selected masks showed that their filtration efficiency ranged from 12% to 98%. "The problem is you don't know which one is the 11 or 12 or which one is the 90-something," he said. The Food and Drug Administration provides marketing clearance for surgical masks but doesn't certify their performance, he noted. In the tests, N95 filtering facepiece respirators had a filtration effectiveness of 98%. Studies have found that face seal leakage of surgical masks ranges from 15% to 40%, while N95s have a face seal leakage of just 3% to 5%, reported Lisa Brosseau, ScD, CIH, of the University of Minnesota.

• Even the exhalation from breathing releases viral aerosols.

Using a special device that collects respirated air from breathing or coughing, researchers at Harvard University found viral RNA in exhaled air of 14% of 28 subjects with influenza. (The RNA also was detected in the coughs of 68% of the subjects.) Influenza A was more likely to be found in the exhaled air or coughs than influenza B. "The majority of particles we were able to measure were less than 5 µm [in size]," says James McDevitt, PhD, CIH, instructor at the Harvard School of Public Health. When a study participant wore a surgical mask, there was a significantly smaller release of influenza RNA particles that were 5 µm and larger, he said. The analysis of smaller particles is ongoing.

• Aerosolized particles exist in patient rooms, but it's not clear if they're infective.

Researchers from the National Institute for Occupational Safety and Health and West Virginia University used 24 stationary aerosol samplers to test the air for an 11-day period in an urgent care clinic in Morgantown, WV. Heath care workers also wore two other personal samplers. Eighty-one percent of the exam and procedure rooms containing a confirmed influenza patient tested positive for airborne influenza A. About half of the particles (47%) were 5µm or smaller.1 Health care worker exposures were similar to those found in the room air, said William G. Lindsley, PhD, a biomedical engineer in NIOSH's Health Effects Lab. "It's pretty clear that the health care workers themselves were being exposed to the RNA levels that we were detecting," he said. The study did not address the infectivity of the particles.

• Health care workers may resist wearing respirators.

In a study of tolerability, 27 health care workers wore eight different respirator or surgical mask ensembles for an entire eight-hour shift, with two 15-minute breaks and one 30-minute lunch break. In more than half the sessions (59%), health care workers stopped wearing the masks or respirators because they could not tolerate them. Their complaints included diminished speech communication, heat, pressure, and dizziness. The powered air-purifying respirator (PAPR) and N95s with an exhalation valve were the best tolerated; the N95 worn with a surgical mask over it was the least tolerated.2 "You can have an extremely effective respirator, but if it's not worn it's not effective. Tolerability is as important as effectiveness," said Lewis J. Radonovich, MD, director, Biosecurity Programs, for the Office of Program Development at the North Florida/South Georgia Veterans Health System in Gainesville, FL. "What we have to fix is comfort." Radonovich is the principal investigator of Project BREATHE, a collaboration of NIOSH and the Veterans Health Administration to spur development of a better respirator for health care.

• Eye exposure alone did not lead to infection.

In a study of transmissibility, 10 participants were exposed to rhinovirus 39 while breathing clean air. None became sick, indicating that transocular transmission did not occur - and that eye protection is not necessary to prevent infection, reported Werner Bischoff, MD, PhD, assistant professor of infectious diseases at Wake Forest University School of Medicine in Winston Salem, NC. However, two of five participants wearing surgical masks who were exposed to the rhinovirus in the air developed cold symptoms four to seven days later, and one of four participants wearing a fit-tested N95 developed cold symptoms five to 10 days after exposure.

• Hospitals should also use other methods of hazard reduction, including isolation or cohorting and ventilation.

Several presenters emphasized that personal protective equipment is just one part of a "hierarchy of controls," and that hospitals must also focus on engineering and administrative controls. That includes placing a mask on patients with respiratory illness, cohorting patients suspected of having novel H1N1, and use of isolation rooms with proper ventilation. Other simple measures could include placing a clear plastic barrier in the registration area to protect clerks from infectious patients, said Katherine Cox, MPH, Med, director of occupational health and safety for the American Federation of State, County, and Municipal Employees. Hospitals can minimize the number of employees who have potential exposure to patients with influenza-like illness and can conduct risk assessments to target respiratory protection to those health care workers at risk, she says. "You could end up having only 10% to 30% of your workers wearing respiratory protection," she says.


1. Blachere FM, Lindsley WG, Pearce TA, et al. Measurement of airborne influenza virus in a hospital emergency department. Clin Inf Dis 2009; 48:438-440.

2. Radonovich LJ, Cheng J, Shenal BV, et al. Respirator tolerance in health care workers. JAMA 2009; 30:36-38. Available at