Googling your symptoms when you're feeling sick might lead you towards some pretty unreliable medical information most of the time, but that doesn't mean it's an entirely useless exercise.

In a recent study, researchers measured the popularity of medical symptom searches on Google and discovered that the volume of such searches can subsequently help predict the incidence of COVID-19 cases arising weeks later in the area.

While the most common symptoms associated with coronavirus might be things like cough, fever, and difficulty breathing, in this case, researchers wanted to examine whether there was correlation between COVID-19 cases and surges in searches for a more distinct subset of gastrointestinal (GI) symptoms caused by the disease, such as abdominal pain and diarrhoea.

"We identified common GI symptoms attributed to COVID-19 from previous studies as search terms, which included ageusia [loss of taste], abdominal pain, loss of appetite, anorexia, diarrhoea, and vomiting," the team, led by first author and gastroenterologist Imama Ahmad from North Shore Medical Centre in Salem, Massachusetts, explains in their paper.

Using Google Trends, the researchers compared the volume of anonymised searches for these terms in 15 US states against the reported incidence of COVID-19 cases, in the period between January and April of this year.

They found Google searches for specific, common GI symptoms were indeed linked with subsequent coronavirus cases in most of the states studied, with the strongest relationship being evident about three to four weeks after the searches were made.

While this is an important and potentially helpful insight, it's not an entirely surprising link. For several years, it's been well known that search engine queries can help alert us to things like influenza outbreaks.

So the main takeaway here is that – as experts have suggested – the same technique really can also help inform us on the spread of COVID-19, potentially signalling which suburbs might be about to become hotspots.

"Searches for GI symptoms preceded the rise in reported COVID-19 in a predictable fashion, slightly longer than the one to two-week lag time observed in prior studies on influenza," the authors write.

"The observed time difference could be related to differences in testing availability, reporting, or longer incubation period of COVID-19 compared with influenza."

The tool is most useful in identifying correlations in areas that are already experiencing a high burden of disease: in this study, the states with the highest incidence of cases during the study period were New York, New Jersey, California, Massachusetts, and Illinois.

Not all related GI symptoms correlated strongly with increases in COVID-19 diagnoses, with ageusia, loss of appetite, and diarrhoea demonstrating the firmest links.

Another limitation of the study, the researchers acknowledge, is that the largely anonymised nature of the search information available in Google Trends makes it hard to filter out confounding variables that could have an effect on the data.

With a view to helping health researchers as much as they can, though, Google this month announced it was making these kinds of search data more widely available for scientists trying to ascertain COVID-19 incidence from search queries.

The hope is that, with elevated access to search interest and trends for more than 400 medical symptoms, signs, and conditions, it will be easier for health professionals to visualise and forecast ahead of time likely areas of coronavirus impact in the US.

For Ahmad and her team, it's already clear the technology can be of great aid.

"Our data underscore the importance of GI symptoms as a potential harbinger of COVID-19 infection and suggests that Google Trends may be a valuable tool for prediction of pandemics with GI manifestations," the researchers explain.

The findings are reported in Clinical Gastroenterology and Hepatology.