TORONTO - Five years after the avian influenza strain H5N1 started killing poultry and people in Southeast Asia, researchers still don't know what to make of the dangerous and unpredictable virus.

After cutting an ever-widening swath through poultry flocks and infecting -- and killing -- mounting numbers of people in Asia, the Middle East and Africa, the virus seemed to almost take a breather in 2008. The year that just ended saw fewer recorded human cases than any since 2003, when this cycle of H5N1 activity began.

It begs the question: Is H5N1 on the wane?

Sadly, science knows too little about how flu viruses emerge, spread and jump -- or don't jump -- from one species to the next to answer that question. Given the knowledge gap, influenza scientists are still pushing for pandemic preparedness.

"Whether or not H5N1 virus is going to cause a human pandemic -- nobody can predict that," says Dr. Tim Uyeki, deputy chief of influenza surveillance and prevention for the U.S. Centers for Disease Control.

"I believe it's still a threat. But it's not the only threat," he adds, noting a two-month-old Chinese girl was hospitalized in Hong Kong in late December with an H9N2 avian flu infection.

Whatever the uncertainty about H5N1, one thing is clear. A fog of exhaustion has settled over the influenza science community as well as the public health officials who have been slaving over pandemic plans. A healthy portion of the broader public is probably sick to death of the subject too.

"I think flu fatigue is certainly a phrase which is thrown around a fair amount in the past several months or past year or so," admits Dr. Keiji Fukuda, head of the World Health Organization's global influenza program.

"In stepping back after going through three or four years of working really hard, I think that there is a genuine sense of `Wow! We have been pushing so hard and we are tired of that."'

Fukuda says the fear associated with the initial re-emergence of H5N1 -- which infected 18 people and killed six in 1997 in its first known foray into humans -- fuelled an urgent drive to prepare for what was feared to be an emerging pandemic.

"There was truly a sense that we simply do not understand what this virus is going to do and it could just change at any moment into something. And I think that that really drove people to work incredibly hard," Fukuda says.

But as time as gone on, it has become apparent the virus isn't working on a discernible timetable.

After a peak of 115 human cases and 79 deaths in nine countries in 2006, human infections declined to 88 cases and 59 deaths in 2007. In 2008, only 40 human cases and 30 deaths were reported, from six countries.

Experts can only guess at why that is and whether it signals a long-term change or is merely a short-term blip.

"It could be just a cyclical thing," says Dr. Maria Zambon, head of the respiratory viruses unit of Britain's Health Protection Agency. "I would be cautious, I think about ... inferring long-term trends from actually reasonably limited data."

Zambon notes that for years in the recent past two separate families of influenza B viruses circulated concurrently in Southeast Asia. The phenomenon was only seen in Southeast Asia; in the rest of the world, only one of the lineages of B viruses spread and caused infections.

And then after the better part of a decade, the second lineage of B viruses broke free and started spreading all over. Scientists don't know why it took so long to happen or what the tipping point was.

"So my general point is flu is highly mutable. ... It does all sorts of unpredictable things. And I just think it's something that one should not underestimate," Zambon says.

Some pathogens do work in waves or cycles, says Dr. Michael Osterholm, head of the University of Minnesota's Center for infectious diseases research and prevention.

So it is technically possible that a primary wave has worked its way through bird populations and the viruses now circulating are less likely to infect humans or cause detectable human cases if they do. Osterholm is quick to add, though, that he's seen no evidence to support that kind of change, which typically happens over a longer period of time.

"People want to play this out on a year-to-year basis. And unfortunately, this is going to be a many-year-to-many-year basis kind of evaluation," he warns.

Fukuda says studies haven't shown that the virus has fundamentally changed, so the best guess is that the reason for the decline in cases probably rests with human behaviour. Efforts to eradicate infected poultry have improved, and countries that use poultry vaccine may be lowering the number of times people come in contact with the viruses as a result.

As well, affected countries have made strides at educating the public about the risk of contact with sick and dying poultry, he says -- though he admits poor people are probably still putting the carcasses of infected birds into the pot rather than the pyre.

Uyeki suggests other possibilities.

As the problem has seemed to wane, so has attention on it. That could be translating into more lax surveillance for new cases. Doctors could be less likely to suspect and test for H5N1 infection, attributing illness to myriad other potential causes.

And the virus has divided up into many different lines or "clades" and "sub-clades" -- a development that creates challenges for testing, he notes. If the sample genetic material laboratories used to look for viruses is out of date, a test that should be positive could come back negative.

As the scientists watch and wonder, public health is struggling with its own questions.

"I think really the question for a lot of them is whether enough has been done, whether such sustained effort is needed," Fukuda admits. "They know that work has to continue and they are not certain at what level it needs to continue."

It's clear the threat of a flu pandemic -- caused by H5N1 or H9N2 or one of a multitude of other influenza viruses -- remains. It's clear the preparatory work is not finished. But it's also clear there are a vast number of other health problems competing for funding in highly challenging economic times.

It is a concern, Fukuda admits.

"These are really long-term building efforts," he says of efforts to prepare the world to weather the next pandemic.

"And if we begin to withdraw our attention and move our attention to something else which is completely different, then we really stand to lose a lot of the work which has been built up over the past four years."

"To do this over and over again is truly ... it's like being Sisyphus," he says, referring to the Greek myth of a figure condemned to a task that can never be completed.