Official newspaper of The University of Texas at Austin

The Daily Texan

Official newspaper of The University of Texas at Austin

The Daily Texan

Official newspaper of The University of Texas at Austin

The Daily Texan

Advertise in our classifieds section
Your classified listing could be here!
October 4, 2022

UT, Columbia researchers predict this year’s flu vaccine to be 36 percent effective

Chelsea Purgahn

Getting the flu vaccine for 2016–2017 will reduce patients’ risk of contracting the flu by 36 percent, according to research by UT and Columbia scientists. This data means this year’s vaccine is 23 percent less effective than last year’s, which the Centers for Disease Control estimated as 59 percent effective.

However, these numbers shouldn’t discourage people from getting the flu vaccine, said researcher Andrew Blumberg, an associate math professor at UT and collaborator on the study. Blumberg said that because researchers can’t know for certain which flu strains will be prevalent in the coming year, it is difficult to make completely accurate predictions about vaccine effectiveness.  

Blumberg and his collaborators’ predictions are based on analysis of the genetic diversity of flu strains in New York City and the effectiveness of flu vaccines in the past two decades.

The team developed two methods of predicting the rate of effectiveness of the flu vaccine. Unlike previous methods, which require directly obtaining data from the strains of the virus in a given year, the new methods require only a diagram that shows the biological relationships between the different flu strains, called a phylogenetic tree. 

Blumberg said the first method, called tree dimensionality reduction, allows for the division of a phylogenetic tree into smaller sections. 

“What I developed with my coauthors was a method [that cuts] up a giant phylogenetic tree into little pieces and looks at the distributions of the little trees rather than the big tree in order to better understand the structure of the flu virus,” Blumberg said. 

The other method involves the geometric structure of the space distribution of the trees. 

“We developed a way to understand [the implications of] when collections of trees were spread out in a space that didn’t necessarily have the geometry of the set of trees [from the preceding years],” Blumberg said. 

Researchers used this information to estimate the genetic diversity among prevalent flu strains in a year, according to Blumberg. The team determined a negative correlation between the flu vaccine’s effectiveness during a given season and the genetic diversity of viruses from
previous seasons. 

“What we found is that if you look at the distribution of the little trees that were made in previous years, and you can predict what type of vaccine is effective in any given year by understanding how much genetic diversity existed in the past year,” Blumberg said. 

The World Health Organization creates flu vaccines each year that target the most common strains from the previous year. 

According to Blumberg, if there is a wide range of genetic diversity of flu viruses in the past year, then the effectiveness of the vaccine is lower. He said this is probably due to a decreased probability of selecting the correct strains of flu viruses to target.

“If you look at the data from the last couple of years, including last year, you’ll notice there’s more spread and more genetic diversity, and so that’ll make it harder to predict which strains will be active this year,” Blumberg said. 

Although the increased genetic diversity may account for the decreased effectiveness, Blumberg said that “noisy” data on flu strains means the statistics are uncertain.

“The estimates from the CDC aren’t that great, and although the [graphs] in our paper are very suggestive, they’re not a precise prediction,” Blumberg said. “I think from the standard of how accurate these numbers are, I’m not sure there’s a huge difference between 60 percent and 36 percent.”

Still, Blumberg said that the methods and techniques developed in this study can be used to better understand disease populations, particularly different types of cancer, and to tailor treatment methods for patients.

“The hope is to use these techniques to develop ways to figure out which treatments are most effective for patients, see what tumors are susceptible to which kinds of interventions, and predict what patient mortality is likely to be,” Blumberg said. 

More to Discover
Activate Search
UT, Columbia researchers predict this year’s flu vaccine to be 36 percent effective