Skip to Content

Shoddy Medication? Search Engines May Already Know

People voice concerns about drugs via search queries—and that data that can predict when one might be recalled.
December 6, 2016

“Google, should I be worried about the quality of my pills?”

It’s not a question anybody wants to ask, but if you regularly take medication and things don’t feel quite right, it would be an obvious one to type into a search engine. That’s certainly what Elad Yom-Tov from Microsoft Research thought when he set out to investigate whether search queries could be used to predict when a drug might be recalled.

Turns out, they can. To work that out, New Scientist reports that Yom-Tov trained an AI to establish connections between Bing queries made in each U.S. state relating to 300 different drugs and their recall over the course of eight months. Then, for another four months, he tested the AI to see if it could predict when a drug would be recalled by its manufacturer.

His algorithms were able to predict when a drug would be recalled, but only a day or two before it actually happened. “The most indicative attributes for prediction are sudden spikes in query volume about a specific medicine in each state,” Yom-Tov writes in a paper published on arXiv. “Recalls of prescription drugs and those estimated to be of medium-risk are more likely to be identified using search query data.”

It’s not the first time that search engines have been used to investigate drugs and health. In the past, Microsoft has used queries to identify adverse effects of prescription drugs and even spot warning signs of cancer—though Google’s flu tracking famously didn’t go as well as it hoped.

While the technique can only predict recalls a couple of days in advance with accuracy, Yom-Tov tells New Scientist that it could be used as a monitoring tool much further out. Combined with other tests, it could provide drug companies and regulators with another means of detecting when something’s gone wrong with a batch of drugs.

(Read more: arXiv, New Scientist, “Can You Really Spot Cancer Through a Search Engine?”)

Keep Reading

Most Popular

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

OpenAI teases an amazing new generative video model called Sora

The firm is sharing Sora with a small group of safety testers but the rest of us will have to wait to learn more.

Google’s Gemini is now in everything. Here’s how you can try it out.

Gmail, Docs, and more will now come with Gemini baked in. But Europeans will have to wait before they can download the app.

This baby with a head camera helped teach an AI how kids learn language

A neural network trained on the experiences of a single young child managed to learn one of the core components of language: how to match words to the objects they represent.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at customer-service@technologyreview.com with a list of newsletters you’d like to receive.