Many have now added the Internet to their personal health toolbox, helping themselves and their loved ones better understand what might be ailing them.
— Pew Trust

Nearly 60% of all Americans sought health information online in 2013 (Pew Internet Project), and roughly 50% of these 'online diagnosers' subsequently consulted a clinician.

This isn't surprising.

Now, vast amounts of medical / scientific information find their way onto sites indexed by search engines (i.e., Google) and eventually filtered by sites like Wikipedia, WebMD, and others.

The real question is the filter. How accurate is the information that appears on these sites?

Take FDA's drug safety warnings, for example. It is not until new drugs and devices are used in the real world that we begin to be able to detect rare safety events. When these unanticipated side effects do emerge, it is an important job of the regulator to let patients taking the medicine and their doctors know; and we would hope that this info is available online and readily accessible.

So, along with co-authors Drs. Bourgeois and Seeger, we designed a simple study to investigate the effect (or digital imprint, if you will) of these safety warnings on Google search and Wikipedia page traffic as well as Wikipedia content. As you might have guessed, there was a surge in search traffic and page views immediately following FDA warnings.

Overall, 41% of Wikipedia pages pertaining to the drugs with new safety warnings were updated within 2 weeks after the warning was issued with information provided in the FDA announcements. Yet, 23% of Wikipedia pages were updated more than 2 weeks after the FDA warning was issued (average, 42 days), and 36% of pages remained unchanged more than 1 year later (as of January 2014).

There were other differences in how quickly the Wikipedia pages were updated. Pages for drugs treating 'common' diseases (affecting more than 1 million people in the US) were more likely to be updated quickly (58% were updated within 2 weeks) than were those for drugs designed to treat less-prevalent conditions (20% were updated within 2 weeks, P=0.03 for log-rank χ²). 

These disparities are troubling, because this information is important. Knowledge of a drug's side effects are necessary for patients to make informed decisions that weight the risks and benefits of a particular treatment.

We offer a few recommendations in our NEJM article. There are likely many, many more ways that we can improve the quality of the health information online. 

Ultimately, I suspect we canand shoulddo (much) better.