Top searches on health topics? It may depend on where you live

Robert H. Shmerling, MD

Faculty Editor, Harvard Health Publishing

You can learn a lot about a person’s medical concerns by looking at the health topics they’ve searched for online. It’s fascinating (and a bit creepy) to take a peek at what others are searching — and to compare what you find to what sends you online.

I’ve posted before about how the health issues people report worrying about the most differ from those that are most common, deadly, or have the biggest impact on quality of life. There’s overlap, of course, but certain important conditions (such as lung disease, the third leading cause of death in 2015) did not make the top 10 list of health concerns in a 2015 survey.

In that same post, I discussed how causes of death varied among states. For example, bicycle-related deaths occurred more often in Florida, and deaths due to accidents involving machinery were highest in Iowa and North Dakota. While we can probably come up with reasonable explanations for these, the higher rate of accidental suffocation in Connecticut is perplexing.

Health concerns as reflected by searches on health topics

A new study looks at searches for health information online and how that varies between states. The results are intriguing — and in many cases, quite difficult to understand.

Analyzing health-related Google searches by state, researchers found that:

  • The top searches were for ADHD (nine states), syphilis (six states), and HIV/AIDS (four states). This makes sense considering that nearly 10% of school-age children have been diagnosed with ADHD (according to the CDC), more than a million people are living with HIV/AIDS (and even more are trying to avoid infection), and rising rates of syphilis have been reported in many parts of the country.
  • The opioid epidemic has affected many states across the country, but it was the top search in only one state (Vermont, a state among the hardest hit by opioid misuse).
  • Some of the findings likely relate to recent events in particular states. For example, Kentucky experienced a recent outbreak of hepatitis A, and Idaho had one of the highest rates of coli infection linked to contaminated romaine lettuce in 2017. So it makes sense these would be the top searches in those states.

Other findings were harder to explain. Why would “ear infection” be the most common search in North Dakota? Similarly, many common conditions were the most commonly searched in only one or two states. It’s hard to understand why the top search was for hemorrhoids in Minnesota, insomnia in South Dakota, and hypertension in Delaware and Montana. These conditions are quite common all over the country. (If you can explain these findings, let me know!)

Limitations of this study

It’s important to consider the limitations of this sort of analysis. One issue is that people search for information using a variety of terms, and that might throw off the tally. For example, eating disorders might be a common problem in your state, but if people used several different search terms to get information about them (bulimia, bulimia nervosa, eating disorder, and so on), none might register as a top health concern. In fact, two states (Washington and Nebraska) had top searches that might include eating disorders, but different terms were used most often (anorexia for Nebraska, body dysmorphia for Washington).

Another limitation is that not everyone uses Google for their searches (though estimates suggest it accounts for about two-thirds of them). Data regarding searches on non-Google sites were not included, and it’s possible that the choice of search engines varies by state.

It’s more than just interesting

Health concerns are common everywhere. And while it’s intriguing to see that different states seem to be focused on different health problems, the analysis of searches for health information can provide insights that go well beyond “interesting.” In some cases, through a review of online searches researchers have been able to identify the spread of disease before public health officials. For example, online searches for cold and flu remedies in a community may signal the spread of flu well before patients’ test results or hospital admissions confirm it.

What’s next?

You can expect to hear more about how the things we’re looking for online mirror our health concerns and illnesses. Of course, analysis of search results is nothing new. Advertising and targeted marketing helps to fuel a remarkable amount of commerce on the Internet. And it’s important to point out that in any analysis of online searches, privacy and confidentiality remain important concerns.

My guess is that researchers will refine their methods as they look at the large volume of online searches we generate every day, so that their analyses can help identify and even prevent illnesses as they develop. One can only hope that our searches can be used for more noble pursuits than just targeted advertising.

Follow me on Twitter @RobShmerling


  1. Steve B

    It would interesting to report separately the results for urban & rural areas (counties?) within each state. It also might be worth considering regional differences within states especially large ones like Alaska, Texas & California. Each island within Hawaii might have different results. Rural regions could be based on occupational considerations, e.g., petroleum-related/mining, agriculture, fishing or based on geographic/climate considerations, e.g., high desert, coastal, mountains…

Post a Comment:

This blog aims to provide reliable information as well as healthy dialog about the topics covered. We do not provide responses to personal medical concerns nor do we endorse any recommendations offered in the comments. We reserve the right to delete comments for any reason, particularly those that do not relate directly to the contents of this post, are commercial in nature, contain objectionable or inappropriate material, or otherwise violate our Privacy Policy. Promotional URLs will be removed from comments. Comments on this blog do not represent the views of our editors or Harvard University, and have not been checked for accuracy. All comments submitted to this site become the non-exclusive property of Harvard University.