Whenever Google changes its search engine algorithm, which happens roughly 500 to 600 times a year, digital marketers and business owners tend to go into a bit of a panic over the harm it could cause to their search rankings. But in this case, Google’s newest algorithm development could actually protect you. A collaboration with Harvard University researchers resulted in a machine learned model that will reportedly flag the restaurants that are most likely to give you food poisoning.

According to the U.S. Centers for Disease Control and Prevention, approximately one in six Americans becomes ill every year due to the consumption of contaminated foods or beverages. Although many consumers are overly concerned with the possible hazards of romaine lettuce or their flawed holiday turkey preparations, the reality is that eating out is just as (if not more) dangerous than dining at home. And considering that 34% of Americans visit a casual dining restaurant once per week, the danger may be more pronounced than you’d think.

This newest algorithm development could allow you to completely bypass those dangers, however. It can apparently identify lapses in food safety in near real time, thanks to an ability to identify user search queries for phrases like “diarrhea” and “stomach cramps.” Those search queries are then cross-referenced with saved location data history from smartphones to identify potential eateries that were visited by those users in the recent past.

When the algorithm was tested in Chicago and Las Vegas, any correlations that were discovered between search terms and restaurant visits prompted official health inspections. Health inspectors were also sent to restaurants that garnered customer complaints; inspectors weren’t told which establishments were identified through the algorithm to remove bias. After researchers compared the rate of unsafe restaurant detection from routine inspections in the two cities with the rate of detection offered by the algorithm in those places, what they found was pretty astounding: while routine inspections yielded a 22.7% rate of detection, the algorithm offered a 52.3% rate of detection.

An even more illuminating aspect of the study was that the algorithm found that, 38% of the time, the potential cause of food poisoning could not be attributed to the restaurant an individual had visited most recently. In other words, the most obvious culprit of food poisoning may not actually be to blame in reality.

Study co-author and Google research scientist Evgeniy Gabrilovich explained: “In this study, we have just scratched the surface of what is possible in the realm of machine-learned epidemiology… Today, we can use online data to make epidemiological observations in near real-time, with the potential for significantly improving public health in a timely and cost-efficient manner.”

The CDC reports that of the 48 million people who contract food poisoning every year, 128,000 of them are hospitalized. And although 75% of urgent care patients rated the quality of their treatment as good or excellent in 2016, most people would prefer to not have to seek out medical attention (emergency or otherwise) for foodborne illness at all. With any luck, using Google’s new algorithm as a supplemental measure to other preventative methods could help prevent this possibility for many Americans.