Doctors who have treated H5N1 avian flu patients are meeting in a Turkish seaside town to try to find answers to the myriad mysteries that remain about what the brutal virus does to its human victims and how dismal survival rates might be improved.

The World Health Organization hopes that by pooling patient data, the meeting will answer critical questions such as whether all lineages of the H5N1 virus cause the same severity of disease and how best to treat pregnant women who become infected.

Based on what it learns at this meeting, the WHO will update H5N1 treatment guidelines, so that doctors who face cases in the future can benefit from the successes -- and missteps -- of those who have treated patients in the past.

"There's a real deficit there,'' says Dr. Frederick Hayden, a WHO influenza expert and a key organizer of the meeting.

"We don't have basic information at hand to try to give the best advice. We're going to address that gap.''

The meeting will also seek buy-in from doctors for a new patient data collection system the WHO hopes to get up and running.

Doctors treating H5N1 patients -- past, present and future -- would be asked to submit a couple of pages of standardized clinical and treatment information so that the global health body can track patterns of disease and treatment efficacy on an ongoing basis.

"This will give us, I think, the best available opportunity under the current circumstances to make sense out of what's happening,'' Hayden said.

The 100 or so doctors and other experts are meeting in Antalya, in southern Turkey. The meeting, which is being held Monday through Wednesday, is being hosted by the Turkish government.

It is a follow-up to a conference that was held in Hanoi in May 2005. At that point three countries -- Thailand, Vietnam and Cambodia -- had reported human cases of H5N1; there were 89 confirmed cases and 52 deaths.

Since then nine more countries -- Azerbaijan, Iraq, Indonesia, Laos, Egypt, Nigeria, Djibouti, Turkey and China -- have reported human cases. As of Sunday, the WHO had confirmed 279 cases since November 2003; 169 of those people died.

In the intervening months, small collections of case data have been published in medical journals. An account of the treatment of eight patients in Turkey. A paper describing a similar number in Indonesia. A report that revealed how patients responded -- or didn't -- to treatment with the antiviral drug oseltamivir (Tamiflu) in southern Vietnam.

But in the main, the details of the symptoms and disease progression of the vast majority of H5N1 patients, the steps their doctors took to try to save them and the outcome of those treatment choices remain locked up in the files of doctors who cared for them.

"There is a huge number of unanswered questions about human infections with highly pathogenic H5N1 viruses,'' says Dr. Tim Uyeki, an influenza expert from the U.S. Centers for Disease Control in Atlanta, who is attending the meeting.

"I think it's so important to share information so others can benefit from the experiences of those who have dealt with this.''  

Pooling data should allow patterns to come into focus in a way that is impossible when two or three doctors are looking at a handful of cases.

The meeting is taking place behind closed doors to encourage doctors who may be working on scientific articles to share their findings before publication. Hayden says he's been told some large, unpublished sets of patient data will be presented in Antalya.

"They'll tell us what's been done and what's happened. That may allow one to say: `This does not work and we shouldn't be doing this.' That's sometimes just as important as saying: `This does work,''' Hayden notes.

In addition to learning about past cases, gathering together the physicians who have treated H5N1 cases in 12 countries should enhance research collaborations -- including planned clinical trials of intravenous forms of two antiviral drugs -- zanamivir (sold as Relenza) and peramivir, a drug still in development.

(GlaxoSmithKline, which makes Relenza, has shelved plans to test an intravenous form of the drug in the United States, but is in discussions with a WHO-organized treatment network in Southeast Asia to test the new formulation there.)

Dr. Menno de Jong believes opportunities for additional research are inevitable when so many clinicians from so many countries are brought together.

Dr. Menno de Jong believes opportunities for additional research are inevitable when some many clinicians from so many countries are brought together.

"If we can make a list -- What are the questions of highest priority and what needs to be done to answers those questions? -- I think that already will be a big gain for a meeting like this,'' says de Jong, a virologist at the Tropical Medicine Hospital in Ho Chi Minh City, Vietnam, where some of the first H5N1 patients were treated.

Still, some mysteries won't be solved until more autopsies are done on H5N1 cases, de Jong insists.

It's believed there have only been about a handful of autopsies on people who succumbed to the assault of this vicious virus. To date cases have occurred in countries where, for cultural or religious reasons, autopsies are not done.

De Jong says pleas for more autopsies are made at every scientific meeting on H5N1, but they haven't produced results. He suggests a way around the autopsy barrier would be to encourage hospitals treating H5N1 cases to do post-mortem biopsies -- using special biopsy needles to take bits of key organs for study.