Unless you're very old, very young or very sick, your chances of dying from an average bout of the flu are extremely low. But avian flu seems to be a disease of an entirely different sort—143 of the 244 WHO-confirmed cases since 2003 have been fatal.

Though it hasn't been clear exactly why the H5N1 avian flu virus seems to be so much more lethal than the normal human variety, a new study by doctors in Vietnam is helping to clarify the difference.

Menno de Jong of the Oxford University Clinical Research Unit at the Hospital for Tropical Diseases in Ho Chi Minh City, Vietnam, reports in the latest issue of Nature Medicine that the H5N1 virus seems to replicate in much higher numbers in victims than normal flu does, and that bird flu seems to trigger a massive immune reaction. The body's own defenses literally rip apart the lungs and other internal organs.

Doctors have suspected this for some time, but de Jong's research, which compares 18 people with bird flu and 8 people with regular flu, puts some solid numbers behind the hypothesis. The study is also a timely reminder of just how important scientists based in Southeast Asia are in the battle against bird flu.

Getting solid data on H5N1 in countries like Vietnam and Indonesia takes time and effort. Scientists need to understand the local culture and build trust with patients on the ground and officials in power. A CDC investigation team from Atlanta that descends on an outbreak site won't be around long enough to build that kind of trust.

The Oxford Clinical Research Unit at Ho Chi Minh City—a 550-bed hospital in Vietnam's second-biggest city—has produced groundbreaking research in avian flu again and again. I've interviewed the hospital's director, Dr. Jeremy Farrar, several times. The problem is that institutions like the Oxford Unit are far too rare throughout the bird flu-hit countries of Southeast Asia. (Another is the CDC's innovative International Emerging Infections Program in Bangkok.) International medical science needs to establish—and fund—more beachheads like this one if we're going to gather the data we really need to understand avian flu.

—By Bryan Walsh/Tokyo