Grocery Scanner Data Helps Contain Foodborne Illness Outbreaks
By Alex Batty, MHI Marketing Communications Coordinator |@
Foodborne illness outbreaks are always a little nerve-wracking for me because I’m a bit of a wreck, to be honest. When you hear on the news that there’s been a huge recall because it’s making people sick, the paranoid part of me freaks out a little and catalogues my entire pantry/fridge to see if there is anything that could possibly make me sick. I don’t have time for getting sick. I’ve got to write these blogs.
In cool news that will calm my panicked self and, you know, be good for the general well-being of the public, IBM researchers are using data from grocery store scanners and mapping it over data from confirmed foodborne illness cases to help speed investigation and prevent more people from getting sick.
In a very academic study (read here) that gave me flash backs to grad school days, they demonstrated that they can use as few as 10 reports of foodborne illness to narrow down the causing food product to only 12 possibilities in a few hours. Using a bunch of math and numbers that I didn’t understand because the last time I took calculus was junior year of high school (I love English degrees), they sorted all the collected data into hundreds of categories and then mapped the information to rank likelihood of contamination, thus quickly narrowing down potential offenders.
Just for a little perspective, a traditional investigation can take weeks to months and this obviously affect the economic and health impact of an outbreak. The sooner the source is discovered, the sooner a recall can be issued and the less people are consuming contaminated product. It’s also good economically because producers don’t have to recall as wide a range of product and can do it quickly, then returning to normal operations as quickly as possible.
Thought new scientific methods require lots of testing and retesting to ensure procedural soundness, the method has already been applied in real life with positive results. An E. coli outbreak in Norway was quickly managed because the method used 17 cases of infection to narrow down 2,600 possible food products to ten.
Further lab analysis was able to quickly pinpoint the source of contamination down to batch and lot numbers and the public was quickly informed – and protected.
I may not like to do math, but I still think it’s pretty awesome that there are people who do because they help my paranoid little brain sleep better about the stuff in my kitchen.
Click to read the full study, “From Farm to Fork: How Spatial-Temporal Data can Accelerate Foodborne Illness Investigation in a Global Food Supply Chain“