A pandemic would evolve week by week, with the most likely scenario being "a cluster of cases like those we saw in Indonesia. Rather than fizzle out as this one appears to have done, it would continue to cause transmission," he said. Depending on how fast that happened, governments might respond with travel restrictions to try to contain the virus.

Such a course of action might buy some time, depending on how well the virus is transmitting in the initial phases, according to Mr. Lipsitch. During a warning period is when "people will take it seriously," he said. Any threat also would be recognized by the World Health Organization, although its reporting ability is limited, he noted.

"They are allowed to issue press reports, but they can't release data they get from governments unless they are given permission," he explained.

He noted that a warning period is hard to predict because "we don't know how the virus will be evolving, so there will be some period of time when it's not in other countries and it's not worldwide, and people will prepare and get their contingency plans in place as best they can." Once it begins to spread within a country, however, "I think it will happen very quickly, within a matter of weeks."

What is the likelihood of this happening? "Nobody will give a hard number unless they're irresponsible, because they don't know," Mr. Lipsitch told NU. In fact, he said, "we may be at an elevated risk now, because of the bird-to-human contacts, for one--but on the other hand, those factors might suggest that we've already had a lot of exposure."

The risk is hard to predict "because we don't know how to translate the warnings into numbers," he said.

Parents taking care of sick children or kids kept home from school in a pandemic will be a major workplace exposure for risk managers to consider, he noted. "This is a huge social issue. The 40 percent peak absenteeism projected by the government is based on people taking care of ill children and themselves, and people staying away from work to avoid exposure to infection."

What can public and corporate risk managers do to prepare? "They need to plan and then rehearse, because otherwise people forget things," he said, warning that without rehearsal, disaster planning becomes merely a "tabletop exercise."

Peter Ulrich, senior vice president of the RMS model management team, told the audience that influenza offers some unique risk management challenges.

"The most obvious are insurance losses to all lines of life and health and workers' compensation," he said. "But productivity is another factor if employees are not at work." He also noted that changes in interest rates and stock market values could affect insurance company assets during an outbreak.

One side effect of a pandemic to insurers, he added, could be a reduction in auto claims if people are not driving to work and stay home to avoid public exposure.

Another challenge to risk managers, he said, is the length of a pandemic. "An earthquake or hurricane are quick events," he said. "But with a pandemic you could be looking at two or three years if there is an outbreak. These highlight the need to take care of risk management at the front-end."

Mr. Lipsitch said that with the flu virus in general, people with a vigorous immune system tend to get sick, but do not experience severe disease or mortality. Despite this, about 34,000 die annually in the United States from influenza--the number may actually be higher, however, because many deaths attributed to diabetes, heart disease or pneumonia could have influenza as an underlying cause, he said.

What's different about a pandemic virus, he explained, is that it is a strain to which the population has not been exposed, "and so our immune response is not as vigorous as it would be to a normal flu."

There are two ways to get infected from the bird flu. One is when a nonhuman virus and a human virus simultaneously infect another host such as a pig, which is well suited to accept both types of virus. There is also the possibility of direct transfer to humans.

"Most evidence is saying that the 1918 flu--the most severe of our recent pandemics--went straight from birds to people, but that is not entirely agreed upon," he said. "The major concern at the moment is that the H5N1 subtype of flu is widespread in wild birds. The other is that it is able to enter humans, and when it does, it causes a remarkably severe disease."

Roughly half of those infected with this virus have died of it, he said.

Mr. Lipsitch explained that pandemics occur with a historical frequency of three-to-four times a century. Although they happen with some regularity, the intervals between pandemics are irregular. "So, when you hear people say we're overdue for a pandemic, because it's been 38 years since the last one, you should be skeptical," he said.

He added that many think the risk at the moment is above the historical average. "The reason has to do with H5N1, because it's widespread in birds, in places where they have impact with people--and the risk of death where there is infection is considerable," he said.

The other argument for why the risk may be high, he said, is "the pattern of contact we have between humans and hosts. China is perhaps the most notorious, where human contact with [bird] species is high."

On a positive note, he said most of the decline in disease mortality in the 20th century preceded or predated "our ability to do modern medicine--which happened before the introduction of antibiotics and vaccines in the 1940s."

The good news about that, he said, is that "something besides modern medicine improved our ability to withstand infectious diseases."