"Indonesian cases" rumor and bad bargains

Xinhua Net (semi-official Chinese news agency) is the sole source today of local (Jakarta) TV station reports that two Indonesians are suspected to have died of bird flu. A family of three is allegedly stricken, with a father and daughter dead and another daughter under "intensive treatment" at a hospital 20 km south of Jakarta. The family was said to have recently traveled to India and Hong Kong. Neither locality is reported to have circulating H5N1, although undetected infections are possible.

This is clearly at the rumor stage. This doesn't mean it is false, only that there is much to confirm or disconfirm and many reasons to hesitate for the short time it should take to move it up or down a notch in concern. One of the first and primary tasks in an outbreak investigation is to verify the diagnosis. Simultaneously (so as not to lose time and keep sources as fresh as possible) information on travel history and timing of onset of symptoms should be obtained.

Indonesia, working with WHO, can prevent unfounded speculation simply by doing a rapid check to see if the infection is from influenza type A and releasing the known information on timing of travel and symptom onset. Checking the sub-type of infection and sequencing would take longer, but if it is not influenza A infection, it would rule out bird flu and we can move on from there.

It is not clear whether WHO fully understands the potential for damaging rumors in this age of rapid communication. They have given mixed signals in this regard, both sponsoring and publishing work that explicitly recognizes the value of "rumor control" and also acting on occasion in ways that fail to act on this knowledge. WHO is not a single person and it is inevitable that professional PR people will on occasion act in ways neither approved nor sanctioned by others in the agency. There needs to be a clear signal from the top leadership that openness is the best strategy in the long run, even if it seems problematic in the short run.

Buying immediate peace for lasting loss in credibility is a poor bargain.

Update (how not to do it): Within an hour of my posting this, Xinhuanet reported this "denial" from the Indonesian Health Ministry:

Indonesian Minister of Health Siti Fadillah Supari denied Wednesday that bird flu has killed two persons in the town of Tangerang over the last five days.

She said that the two, a father and his one-year-old daughter, died from acute pneumonia.

"Based on a laboratory test conducted by the ministry's research and development center on July 13, the pneumonia was caused by pathogenic bacteria, and there was no indication of pathogenic virus," Supari said in a statement as published by the Detikcom online news service.

She said health office in Tangerang, some 20 km south of Jakarta, has been instructed to further investigate into the pneumonia case and the possible outbreak in the town.

What does this say? Only that the two deaths were indeed acute pneumonia, that there was bacterial infection involved (common as secondary infections in influenza) and that they were still investigating. The best that can be said for this is that they have not yet determined that it was bird flu, not that they have confirmed it wasn't. This is just how it should not be done.