Approach Offered to Cope with Lack of Data about H1N1 Flu

Uncertainty is a predictable feature of epidemics, but the current novel influenza A H1N1 pandemic has taken the data-deficit problem to an extreme. There are too many cases to count, acknowledged the World Health Organization (WHO) in June, as the number of people infected grew exponentially.

Yet every country needs good data to respond most effectively, said Marc Lipsitch, HSPH professor of epidemiology, whose viewpoint published in the August 12 online issue of the Lancet proposes an alternative to case-based surveillance (the approach that was used in most places during the early part of the epidemic.)

Marc Lipsitch

Lipsitch also is a member of the federal 2009-H1N1 Working Group of the President's Council of Advisors on Science and Technology (PCAST). He has just been named the principal investigator of a new Center for Communicable Disease Dynamics at HSPH.

"The fundamental problem is that we don't know how many people have been infected any place in the world," said Lipsitch. "The most important thing anyone wants to estimate is the severity of the infection."

The researchers propose a two-part framework: Measure flu-like illness overall or in selected sites, and devise a consistent way to gather viral samples from people with mild and severe symptoms.

"Our approach offers the possibility of making evidence-based estimates of the burden of pandemic H1N1-attributable disease without the need to test most cases," they write. Such a system can efficiently monitor the epidemic and viral evolution and be scaled according to need and resources.

The big challenge is systematic viral sample testing at different levels of disease severity for purposes of surveillance, not treatment, Lipsitch said. The capacity to do this is limited by testing capabilities and infrastructure.

The H1N1 infections raging through some countries in the Southern Hemisphere seem to be relatively mild, but the lack of data means no one knows if severe outcomes are rare or more common and who is at most risk.

Without such data, public health advisors can only make educated guesses and pose plausible scenarios about the impact of an expected widespread fall resurgence in the United States and elsewhere in the Northern Hemisphere.

"Data are essential to inform advice from WHO and decisions made by national authorities about interventions, including possible diversion of resources from other public health concerns," Lipsitch and his co-authors write in the Lancet Perspective.

Improving the amount and quality of data is a multi-year project that transcends the pandemic and that can also help us understand patterns of common seasonal flu, Lipsitch said. He noted that seasonal flu kills tens of thousands of people every year in the U.S.

"The flu is notoriously hard to study epidemiologically, because we have such a spectrum of illness from mild to severe," he said. "It's hard to tell in a normal clinical practice if someone has flu or something else."

Lipsitch acknowledged that he and his colleagues had thought cases of H1N1 flu pandemic would be easier to count because the pandemic would be more severe. "In the 1918 pandemic, almost anyone who died during the peak of flu activity died of the flu. It overwhelmed everything else," he said.  

-- Carol Cruzan Morton. Photo by Richard Chase.