Will avian flu act more like SARS?

ID experts counter CDC recommendations

Avian influenza challenges our usual assumptions about how influenza is spread and how to protect health care workers from infection. It may be best to think of it as a potential SARS-like disease, not a version of the seasonal flu.

That is the conclusion of Eric Toner, MD, FACEP, senior associate at the Center for Biosecurity at the University of Pittsburgh Medical Center, who conducted a literature review of influenza transmission.1 His views are contrary to those of the Centers for Disease Control and Prevention (CDC) in Atlanta, which is recommending droplet precautions for both seasonal and pandemic influenza — or the use of surgical masks rather than N95 filtering face-piece respirators.

The distinction is important for hospitals, which have been urged to prepare for pandemic influenza. Providing personal protective equipment is one aspect of those preparations.

Toner doesn't actually advocate stockpiling N95s. "It's hard enough, given the limited amount of money for preparedness, to stockpile enough surgical masks to last through a pandemic," he says. "I don't think there's any way to stockpile enough N95s or PAPRs [powered air-purifying respirators] to last through a pandemic. I believe in doing what's possible."

But Toner cautions that some of the commonly held beliefs about influenza may not apply to H5N1, the avian strain now circulating through Asia and part of Europe.

"Studies that were done previously with regard to seasonal flu that showed fairly limited transmission in hospital settings may not apply in a pandemic situation," says Toner.

Transmission of H5N1 would differ from seasonal strains in one important way: Virtually everyone has partial immunity to seasonal strains. The U.S. population has no immunity to H5N1, which carries a mortality rate of about 50%.

"Transmission is different, attack rates are different, severity is different," says Toner. "It's better to think of this as a novel virus. I wish we had a better name for it.

"People [could] think of this as SARS but with two very important distinctions. It has a very short incubation period — two days vs. five to seven days for SARS. And it has pre-symptomatic and asymptomatic transmission. That fundamentally changes control strategies."

Of course, the primary difference between H5N1 and the seasonal strains (H3N2 and H1N1) is that avian influenza is not easily transmitted among humans. In all but two cases, public health authorities can trace human cases of H5N1 to contact with sick birds. It still is an avian disease.

The CDC and its infection control advisors believe that if avian influenza becomes transmissible among humans, it will behave like other influenza strains. The CDC is recommending droplet precautions for pandemic influenza, noting, though, that when patients undergo aerosolizing procedures, such as bronchoscopy, a higher level of protection might be required.2

"There is no reason to believe a pandemic strain would operate any differently than a seasonal strain in terms of the mechanism of the virus causing transmission," says Michael Bell, MD, associate director for infection control at the Division of Healthcare Quality Improvement with the National Center for Infectious Diseases. "Yes, there are small differences. But at the end of the day, it's still influenza. It makes sense to treat it that way while doing careful monitoring [of transmission]."

The CDC's interim infection control guidelines for avian influenza recommended airborne precautions "given the uncertainty about the exact modes by which avian influenza may first transmit between humans" (www.cdc.gov/flu/avian/professional/infect-control.htm). But those guidelines are being revised, and the Pandemic Influenza Response Plan released by the U.S. Department of Health and Human Services recommends using droplet precautions.

"While we initially felt there was a need for airborne precautions, a subsequent review by CDC and discussions at HICPAC [the Healthcare Infection Control Practices Advisory Committee] indicated that there does not appear to be a difference in transmission," says Patrick Brennan, MD, chief medical officer and senior vice president at the University of Pennsylvania Health System and chair of HICPAC.

Toner acknowledges that influenza is spread primarily by large droplets, but he notes that some studies have indicated airborne spread, as well. A person can become infectious as soon as 24 hours after being infected, even if they have no symptoms, he notes.

Public health authorities are most concerned about community spread, but Toner notes that early in a pandemic, nosocomial transmission could play an important role, particularly since "the people who are most symptomatic are those who are shedding most."

One positive finding: "It seems that influenza, in general, including the 1918 virus, is probably not as transmissible as was previously thought," says Toner. "The number of people infected by each victim looks to be on the order of two as opposed to 10 or 20, as had been speculated in the past."

But if H5N1 mutates and becomes transmissible among humans, its transmission characteristics will not necessarily be identical to known flu strains, particularly early in the course of a pandemic.

"There may be a period where H5N1 becomes more SARS-like, meaning a longer incubation period, little or no asymptomatic spread," he says. "In that case, SARS-like infection control measures would be critically important."

Already there are signs that H5N1 doesn't behave like the seasonal strains, he says. "H3N2 (seasonal influenza) is a disease that exclusively infects the upper respiratory tract," he says. "H5N1 affects the lower respiratory tract. You would predict [that] its transmission characteristics would be quite different. Furthermore, H5N1 appears to infect the GI tract, which is not true with other human flus. And H5N1 affects other tissues including the central nervous system."

Bell acknowledges that infectious disease experts will need to watch the patterns of H5N1 closely and make adjustments in recommendations, if necessary. But he notes that the main distinction of H5N1 is the lack of immunity — an attribute that affects spread among populations but doesn't mean person-to-person transmission will occur differently.

"There's a difference between population-based outbreak dynamics vs. infection control for individual protection," he says.

The high mortality of H5N1 so far may lead some health care facilities to choose a higher level of protection for staff, Bell says. But they may also want to have a plan to shift from airborne precautions to droplet precautions if it becomes clear during a pandemic that the higher level of protection is not necessary, he says.

In fact, once a pandemic is widespread in the community, hospital-based transmission may be of less concern, says Toner. "Then a surgical mask is probably the best you can do and probably all that makes sense at that point," he says. "It's probably even more important to put the mask on the patient than to put the mask on the health care worker."


1. Toner, E. Do public health and infection control measures work to prevent the spread of flu? CBN Weekly Bulletin Oct. 31, 2005; University of Pittsburgh Center for Biosecurity.

2. Centers for Disease Control and Prevention. Pandemic influenza toolkit: Infection control. www.hhs.gov/pandemicflu/plan/pdf/S04.pdf.