Dueling studies: Mask debate is not academic
Dueling studies: Mask debate is not academic
Aussie researcher defends embroiled N95 study
Talk about peer review. Rarely do medical researchers have their specific papers described to the president of the United States, but that is the case for two studies playing a central role in the national furor over surgical masks vs. respirators for H1N1 protection. Both studies were mentioned in a Nov. 5 letter to President Barack Obama, as the nation's leading infectious disease groups urged him to rescind a federal requirement that health care workers wear N95 respirators while caring for H1N1 patients.
One was published and the other, as of this writing, was not. Therein hangs a convoluted tale about randomized trials, statistical significance and Chinese health care workers, but let's begin with the study that was published — in the Journal of the American Medical Association no less.1 The lead author is Mark Loeb, MD, a health care epidemiologist at McMaster University, Hamilton, Ontario, Canada.
"Based on this data, it appears that the surgical mask is not inferior to the N95 respirator for protecting health care workers against influenza, so it's basically not worse than an N95 respirator," he tells Hospital Infection Control & Prevention. "I don't think it's the end of the story; people have to do more studies. But the data suggest [surgical masks and N95s] are very comparable. The attack rate between the two groups was very similar, not only for influenza but for other respiratory viruses."
It should be noted at the onset that this was a serologic study involving blood draws, meaning the transmission could have occurred in the community in the absence of any respiratory protection. "That's a good point, but because this is a randomized trial we can infer that any differences in these groups would be related to the intervention," Loeb says. "They were extremely balanced [groups] and we also looked at household exposures [to flu-like illness]. Those were similar between the two groups as well. Hopefully, the randomization takes care of those potentially confounding factors."
In the study of 446 nurses in eight Ontario hospitals, the nurses were randomized into two groups: 225 were assigned to receive surgical masks and 221 were assigned to receive a fitted N95 respirator, which they were to wear when caring for patients with febrile respiratory illness. The primary outcome of the study was laboratory-confirmed influenza. Effectiveness of the surgical mask was assessed as "noninferiority" of the surgical mask compared with the N95 respirator. Between Sept. 23, 2008 and Dec. 8, 2008, influenza infection occurred in 50 nurses (23.6%) in the surgical mask group and in 48 (22.9%) in the N95 respirator group. "Our data show that the incidence of laboratory-confirmed influenza was similar in nurses wearing the surgical mask and those wearing the N95 respirator. Surgical masks had an estimated efficacy within 1% of N95 respirators," the authors concluded.
As a result, Loeb said he would be comfortable wearing a standard surgical mask for H1N1 patients, though public health officials in Ontario are requiring N95s like their colleagues in the United States. "We did have some H1N1 in the study — we were taking blood from nurses until May — and we didn't notice any difference between the two groups."
The subject of considerable controversy, the unpublished study showing clear efficacy of N95 respirators over surgical masks is by lead author C. Raina MacIntyre, MBBS, PhD, FRACP, FAFHM, head of the School of Public Health and Community Medicine at the University of New South Wales in Sydney, Australia. Originally presented to an Institute of Medicine panel assessing the evidence on the issue for the Centers for Disease Control and Prevention, the MacIntyre study was recently presented in Philadelphia at a meeting of the Infectious Disease Society of America (IDSA).2 The abstract submitted to IDSA described a randomly controlled trial in 24 hospitals in Beijing. "By intention to treat analysis, surgical masks had no efficacy for any of the outcomes," the authors concluded. "N95 masks were significantly more protective than control. Fit testing did not appear to improve the efficacy of N95 masks, which as a group had statistically significant efficacy of 60% against CRI, 75% against ILI [influenza-like illness], 56% against laboratory confirmed respiratory viral infection, and 75% against confirmed influenza infection."
Reports from the IDSA meeting indicate the authors were less conclusive in presenting their findings, which were subject to question in the area of the randomized trials. "When they presented their study [at IDSA] it wasn't a randomized controlled trial," says Neil Fishman, MD, director of the department of health care epidemiology and infection control at the University of Pennsylvania in Philadelphia. "So they reanalyzed their data without the control and [found] in their primary outcomes that there was no difference in protection between a fitted N95, a nonfitted N95 and a surgical mask. My take is that they had an uncontrolled, unrandomized, underpowered study and they showed no statistical difference between their primary outcomes. We are asking that OSHA enforcement be suspended until this new information can be critically
analyzed and hopefully incorporated into the CDC recommendations."
Indeed, in the letter to President Obama, the IDSA, SHEA and the Association for Professionals in Infection Control and Epidemiology note that "the MacIntyre study was initially reported to show a significant benefit associated with use of the N95 respirators and this preliminary report was both presented to the Institute of Medicine's Committee on Respiratory Protection for Healthcare Workers in the Workplace Against Novel H1N1 Influenza and cited in its final Sept. 3, 2009 Letter Report, which recommended the routine use of N95 respirators."
MacIntyre responded to the situation in an e-mail interview with HIC. "It is very common indeed for conference presentations to be preliminary, and to revise analysis of data based on reviewer comments, which is exactly what we did for complete academic honesty and integrity," she says. "The results, however, still suggest superiority of N95s, and that is what was presented at IDSA."
The main reason the research changed was that the original analysis was compared to a control group, which did not wear masks, she explained. "In this analysis, the infection outcomes were statistically significant (and still are) showing efficacy for N95s but not surgical masks," MacIntyre says. "The Chinese health workers and IRB would not accept being "randomized" to not wearing a mask (i.e., they found it unacceptable to be "told" not to wear a mask), so we had to use a convenience control group (not randomized), selected from hospitals where mask wearing was not the norm. A control group that is not randomized can be subject to confounding, so for the main analysis in the final paper, we removed almost 500 health workers (the no-mask group) which then contributed to the p values losing statistical significance.
"However, even after removing the no-mask group, rates of infection outcomes remain approximately double in the surgical mask arm compared to the N95 arm," she continues. "A bootstrapping [computer] analysis showed that this difference between the arms was about 97% likely to be a real difference, suggesting the loss of statistical significance was due to loss of study power from the removal of nearly 500 subjects. . . . The current campaign to overturn the CDC/IOM guidelines on the use of respiratory protection for health care workers asserts that they are based on [our] study. It is explicitly stated in the IOM guidelines that they were never based on this study, because it was and still is unpublished, but on the large body of published evidence on fit and filtration."
That point was reiterated by the IOM and the CDC, which posted the following comment on the studies: "The IOM panel explicitly stated that it did not rely on the MacIntyre study in issuing its recommendations, nor is the IOM Report or this study mentioned in the CDC Guidance. In presenting the results of the IOM letter report to the sponsors of the review, CDC and OSHA, Dr. Ken Shine, the panel chair, made a point of specifically saying that the two recently conducted randomized trials of respirator use, the MacIntyre and the Loeb studies, were not considered in the committee's decision."
Fair enough, but it seems to beg the reasonable question of why an unpublished study was presented if it was not going to be considered. That appears a little like the judge telling the jury to disregard the testimony after they have heard it, and the fact that MacIntyre was on the IOM panel only added to umbrage. "I'm assuming it carried a fair amount of weight because Dr. MacIntyre was on the IOM committee," says Fishman, president-elect of the Society for Health Care Epidemiology of America. "The bottom line is that the [N95] mandate is not driven by the science. The best science that we have right now of practical mask use is the Loeb paper. That has been published and gone through peer review. The use of a surgical mask in that study was equivalent to the protection provided by an N95, at least for routine care."
In response to questions about MacIntyre's membership on the IOM panel that rendered the recommendation to the CDC, the university communications department issued this statement to HIC:
"Professor MacIntyre is a medical epidemiologist who has spent 17 years as a researcher in infectious diseases. She has over 120 publications in this field in peer-reviewed medical journals, a significant track record of success in peer-reviewed grants, and an active research program in infectious diseases . . . . [MacIntyre] was invited to be on the IOM committee based on her track record in clinical facemask research. She was among the first researchers in the world to do a clinical trial of face masks, and the first to pose the research question about the comparative clinical efficacy of N95 respirators and surgical masks. Professor MacIntyre has no vested interests in policy in the U.S., nor any affiliation with stakeholder organizations in the U.S. who are currently fighting over this issue. Neither was she the only IOM committee member with a research track record relevant to face masks — in fact, expert committees generally comprise people who have a research or policy track record in the field."
References
- Loeb M, Dafoe N, Mahony J, et al. Surgical Mask vs. N95 respirator for preventing influenza among health care workers: A randomized trial. JAMA 2009; 302(17):1,865-1,871.
- MacIntyre CR, Wang Q, Cauchemez S, et al. The first randomised, controlled clinical trial of surgical masks compared to fit-tested and non-fit tested N95 masks in the prevention of respiratory virus infection in hospital health care workers in Beijing, China. Abstract 1247. Presented in Philadelphia at the Infectious Diseases Society of America Conference; Oct. 29-Nov. 1, 2009.
Subscribe Now for Access
You have reached your article limit for the month. We hope you found our articles both enjoyable and insightful. For information on new subscriptions, product trials, alternative billing arrangements or group and site discounts please call 800-688-2421. We look forward to having you as a long-term member of the Relias Media community.