Behind the CDC's Soaring H1N1 Death Totals

  • Share
  • Read Later
Spencer Platt / Getty

Jessica Forde receives an H1N1 shot at the Delany Medical Center on Nov. 11, 2009, in the Bronx borough of New York City

Correction Appended: Nov. 13, 2009

The Centers for Disease Control and Prevention (CDC) released on Thursday updated estimates of the number of H1N1 infections and deaths in the U.S. According to the new figures, about 4,000 Americans, including 540 children, have died of H1N1 flu, and 22 million people have been infected since April, when the novel flu virus first surfaced. The new death toll, which encompasses data through Oct. 17, represents a tripling of CDC estimates issued just last week; the number of deaths in children was quadruple last week's figures. But the increase does not mean that the disease has suddenly become more deadly or severe, according to health officials, who say they are not surprised by the higher numbers.

Until now, the CDC's weekly updates on the number of new cases, hospitalizations and deaths from the disease have included only laboratory-confirmed cases of H1N1 — a figure that agency officials were well aware captured only a sliver of the actual population of affected Americans. Many patients who come down with flu never go to a hospital or see a doctor and never get an official diagnosis. Many other flu patients who are admitted to the hospital may not be tested for H1N1 and may be treated under a different diagnosis. They may die from a complication, such as pneumonia, which is not reported as a case of influenza.

The CDC's new figures are based on an algorithm for estimating the true impact of H1N1 on the U.S. population; it takes into account the patients left out of the official lab-confirmed tally. "We know that a number of deaths that we're seeing are occurring outside of the hospital where testing is not possible," Dr. Anne Schuchat, director of the National Center for Immunization and Respiratory Diseases, told reporters at a briefing on Thursday. "We also know that not every patient with influenza gets a diagnosis of flu. For influenza it's virtually impossible to find every case with a lab test. So the estimation method we are using now we believe gives a bigger picture, a probably more accurate picture of the full scope of the pandemic."

Schuchat stressed that the numbers do not reflect any change in the spread of the virus or the course of the disease. The researchers simply applied a unique multiplier to each set of data in order to come up with the estimates. With respect to the number of cases of H1N1 infection, for example, CDC scientists believe that for every one case that is reported and confirmed with a laboratory test, there are 79 additional ones that go unrecorded. For every documented case of H1N1 hospitalization, there are an estimated 2.7 that are missed.

The CDC's multiplier formula comes from analysis of data from its Emerging Infectious Program (EIP), a survey of 62 counties in 10 states across the U.S. The data from these sentinel sites are then extrapolated to the general U.S. population.

"These estimates give us a better [sense] of how much disease, hospitalization and death there is than we would get by just counting individual laboratory confirmed cases," says Schuchat. "It isn't a switch or a change from the way we've been counting cases so far. What we're really trying to do is give you a bigger picture."

The numbers underscore the importance of vaccination in the effort to curb the spread of the disease, says Schuchat. They also support health officials' decision to target high-risk populations for the first immunizations. Of all hospitalized Americans, more than half were between ages 18 and 64, while only 9% were 65 or older. That distribution is the opposite of the way seasonal flu usually affects a community; in that case, 90% of hospitalizations are usually among the elderly.

Due to an editing error, the original version of this article misstated that 2 million Americans have been infected with H1N1. The number is 22 million.