Every year, the Centers for Disease Control and Prevention (CDC) estimates how effective last year’s flu vaccine was at preventing flu-related symptoms and hospitalizations.
But are the CDC’s estimates accurate? Can you trust them?
Dr. Eyal Shahar, a professor emeritus of epidemiology at the University of Arizona’s Mel & Enid Zuckerman College of Public Health, says no — because the CDC uses what he says is an inherently biased study design to arrive at those estimates.
In an interview with The Defender, Shahar said that for years, the CDC has insisted on using a “problematic” test-negative study design, even though better designs exist.
“There are many supporters of the test-negative design, and the CDC simply adopted that design long ago,” Shahar said. “There are CDC-associated networks, all of which exclusively use that design to estimate effectiveness. It is time for a change.”
Shahar has authored nearly 200 papers, a book chapter on biases in epidemiological research, and two books, including a textbook on epidemiology and statistics. He has taught courses in methodology and is the former associate editor of the American Journal of Epidemiology.
A test-negative study design method is relatively easy and cheap, but it is known to produce biased and incoherent results, Shahar said.
Test-negative studies on flu vaccine effectiveness enroll people who sought medical care for flu-like symptoms and were tested for the flu. The researchers compare the vaccination status between those who tested positive and those who tested negative.
But that means the study is limited to people who sought medical care and were tested. That’s a form of selection bias, because the study excluded everyone who didn’t do those things. Limiting the study in this way can also cause collider bias, which could skew the results, Shahar said.
Shahar said better study designs would likely yield more accurate results, but they require more effort and money.
