Avian flu was spread from patients to health workers

Pandemic threat remains for Hong Kong strain

Transmission of avian influenza A (H5N1) virus to two health care workers caring for infected patients has been documented, increasing concerns about a global flu pandemic should the virus resurface and mutate with circulating strains, researchers warn.1 Human transmission was one of the critical questions when the outbreak first began in Hong Kong in late 1997 because there is no vaccine and little existing human immunity against an infectious H5N1 strain. (See Hospital Infection Control, March 1998, pp. 33-39.)

In 1997, 18 human cases of influenza A (H5N1) illness occurred in Hong Kong along with outbreaks of the same strain among domestic poultry, the authors report. Human H5N1 infections raised concerns about the possible pandemic, because only H1, H2, or H3 hemagglutinin subtypes had been known to cause outbreaks in humans previously. Low levels of human transmission may lead to adaptation of the H5N1 virus to humans through the mutations or through reassortment with a human influenza virus.

"The concern was that the influenza season was going to be starting in Hong Kong early that spring, and if you continued to have a high viral load of this H5 in the environment and then you also started getting human virus out in the community at high levels, that would increase the chances of developing a reassorted virus," says lead author Carolyn Bridges, MD, medical epidemiologist in the influenza branch at the Centers for Disease Control and Prevention. "The other concern was that the more human illness that you had that potentially you would have mutations that would occur within a person, and perhaps even without reassorting, you would develop some virus that was more easily transmitted person to person."

Bridges and co-authors showed for the first time — based on the epidemiologic evidence and the seroconversion of two health care workers — that human-to-human transmission of H5N1 occurred. Although no human H5N1 infections have been detected since December 1997, influenza H5 viruses, along with other influenza A subtypes, continue to circulate among bird populations. With avian viruses now having demonstrated their ability to infect humans and pass from human to human, it should be a public health priority to strengthen international surveillance for H5N1 and other influenza A subtypes among domestic and feral birds as well as human populations, they note.

To estimate the risk of person-to-person transmission, the researchers conducted a retrospective cohort study to compare the prevalence of H5N1 antibody among health care workers exposed to H5N1 case patients with the prevalence among non-exposed workers. Information on H5N1 case patient and poultry exposures and blood samples for H5N1-specific antibody testing were collected. Overall, eight workers exposed to case patients had presence of H5N1 antibodies. Of the two workers who seroconverted, one remained asymptomatic and the other reported a respiratory illness two days after exposure to a case patient. Exposed and unexposed health care workers reported no differences in poultry exposure, the other likely source of infection.

"There are these live bird markets, several on a block," Bridges says. "But we felt in the cohort study that, statistically, people had a higher likelihood of being seropositive after exposure to the patients." Bathing and changing the patients’ bed linens were activities associated with being seropositive. "I think it is a surrogate [marker] for increased intensity of exposure," Bridges says. "It has been shown with other studies [that] the sicker people are, the higher titer of virus that they are shedding."

The two workers who seroconverted were treating a very ill adult patient who eventually died. They were following standard precautions — wearing gloves and hand washing — but not using masks. Spread could have occurred via airborne or droplet routes or direct contact and self-inoculation of the nasal mucosa, Bridges says. "We didn’t do a formal study, but they did collect some serum samples from patients in neighboring beds and didn’t see any evidence of [nosocomial spread]."

Despite the finding, H5N1 has a "very low potential" to spread in humans unless it mutates further, says one of the paper’s reviewers, Arnold Monto, MD, professor of epidemiology in the school of public health at the University of Michigan in Ann Arbor. "It is very clear it doesn’t have the same potential to spread in humans as a humanized’ flu virus," he says. "The real concern at that point was not that this virus was going to spread in humans. It was that at the end of the outbreak, the regular flu viruses were beginning to circulate, and there would be a reassortment between a regular human’ virus, like an H3N2 virus, and the avian H5N1. That would be a real concern, because that is apparently how H3N2 came about — there was reassortment between an avian strain and a human strain — and that’s what made it infectious for humans. That was the 1968 [pandemic]."’

The traditional theory of pigs as the "mixing vessel" for human and bird flu strains is somewhat in question, because the H5N1 strain was transmitted directly to humans and from human to human, he adds. "Humans are assumed not to have receptors for avian viruses," Monto says. "Chickens don’t have receptors for human viruses, but pigs have receptors for both. The lesson learned from [the H5N1 infections] is that humans must have some limited numbers of receptors for the avian virus, because otherwise this couldn’t have happened."


1. Bridges CB, Katz JM, Wing HS, et al. Risk of influenza A (H5N1) infection among health care workers exposed to patients with influenza A (H5N1), Hong Kong. J Infect Dis 2000; 181:344-388.