More on Iraq bird flu

The WHO official reference lab in the UK has confirmed what we already knew. The 14 year old girl, Tijan Abdel Qader, died of bird flu. As we have done a couple of times we are posting her picture to remind ourselves and everyone else that this was a real person, someone's little girl, sister or best friend (pic courtesy Holland to Kurdistan)

Tijan's uncle also died from a severe respiratory disease ten days later and there are currently 13 more Kurds being treated in Sulaimaniya's main hospital. Medical News Today is reporting a serious shortage of the antivirals Tamiflu and Relenza, saying there are currently only 30 tablets of Tamiflu in the entire area (enough for three or four people). To be effective these drugs must be given shortly after onset of the illness so they are not likely to do much good now for those under treatment. But it is hard to understand why US authorities have not sent these urgently needed supplies as yet.

As in Turkey, the first reports of the disease are in humans not birds, suggesting either a change in the virus or, more likely, longterm undetected or unreported disease in poultry. In addition there was confusion about the diagnosis, with local doctors initially reporting negative tests, results that were eagerly seized upon by WHO. But wishing won't make it so. Indeed it is well known that false negatives are now common in the bird flu tests. New Scientist has an excellent story, noting some of the reasons:

Tijan Abdel-Qader died on 17 January and two days later the WHO declared Iraqi tests had shown she had not died of bird flu. “There were also no reports of dead chickens in the area, and she seemed to have another illness,” Dick Thompson, a WHO spokesman, told New Scientist.

Fortunately, he says, her doctors asked for a second opinion. On Monday 30 January, the US Navy Medical Research Unit in Cairo, Egypt, confirmed H5N1 in samples from the girl. On Tuesday evening, the National Institute for Medical Research at Mill Hill in London, one of the WHO’s official collaborating centres for flu, also confirmed the virus.

“When we get a positive, we’re sure,” Alan Hay at the NIMR told New Scientist. “But when we get a negative, we aren’t.” One problem is getting a sample with virus in it. The amount of virus present during the course of bird flu in humans varies more than with human flu. And test samples are usually mucus from the nose or throat. But because H5N1 is a bird virus, it prefers the higher temperatures – and the more bird-like cell-surface molecules – of the lower lungs.

But sometimes positive diagnoses can be also false. The London lab has re-tested samples from 21 people that Turkey’s central laboratory had declared positive for bird flu. Only 12 were confirmed, including all four known fatalities.

The now-negative samples might have degraded between tests. “We have to look carefully at the victims' symptoms and circumstances before we decide,” says Hay. But if it turns out there were only 12 cases in Turkey, that makes the death rate – one in three – very similar to the most recent wave of cases in Vietnam, suggesting the virus has not yet become better-adapted to humans.

The problem is that one common test for the virus – amplifying any viral genes using a process called PCR – is tricky and notoriously subject to contamination. Another, looking for antibodies to H5N1 in the blood, does not work until well into an infection, and the standard test kits are designed for human flu, sometimes miss bird flu.

But leaving the diagnostic tests to the world’s scattered experts takes time, for local approval, transportation and then testing. “In the early days of an epidemic, when it is not clear whether the virus is in a country or not, it is important that tests be completely reliable,” says leading flu expert Albert Österhaus of Erasmus University in Rotterdam in the Netherlands. “Better a little delay than a result you aren’t sure of.” (New Scientist)

I have some hesitation about Österhaus's demand for absolute certainty, given the difficulties here. When the stakes are high and delay is costly, waiting for an overburdened lab thousands of miles away to do a test on a sample that may or may not be valid seems problematic.

But one thing for sure. WHO's eagerness to embrace negative results is inappropriate. Even casual observers (see here, here, here, here and here) could see the Iraq diagnosis was likely bird flu. It wasn't just a "lucky" request for confirmation but a mandatory one (and an unlucky one for Dick Thompson, one of WHO's chief spinmeisters).

Thompson needs to be reminded of the old saying: if it quacks like a duck, looks like a duck and walks like a duck, it probably infects like a duck.