Why you should be wary of using 'Dr Google'

A quick trip to your favourite search engine illustrates how easily the process of self-diagnosis can go off the rails. Photo / Getty Images

Everyone occasionally experiences new symptoms that can't be readily explained or ignored. In most cases, people seek initial medical guidance from the one source they consider authoritative, available, trustworthy and nonjudgmental.

We are, of course, talking about Google - which, for many adults, has become a de facto primary care doctor.

Who can blame them? Seeing a real doctor is often too inconvenient, time-consuming and expensive to justify unless you believe your health is in serious danger. As insurance deductibles continue to rise, putting formal medical care further out of reach for most families on a budget, the need for free and reliable health information has become especially urgent. But is the advice available online actually practical - or even accurate? And how well do people without medical degrees navigate the seemingly endless, often contradictory stream of online health information?

In our experience, the answer is: not so well. By the time they reach our offices, most patients have already researched their symptoms online and reached some conclusions about the likely cause. They're often completely wrong.

A quick trip to your favourite search engine illustrates how easily the process of self-diagnosis can go off the rails. Most people seeking information about their symptoms aren't trying to become experts but just want practical guidance and advice. A top Google search result for "headache," however, begins with the unhelpful (and questionable) assertion that there are "over 150 types of headaches" - then attempts to catalogue many of them.

Medicine is also complicated, and online articles often can't account for the unique context in which each symptom occurs. Are you a healthy 25-year-old experiencing a cough after traveling abroad? Or a 65-year-old smoker with a cough who just started a new blood pressure medication? The evaluations are completely different for these two patients - though the search results may be the same. Some popular sites are even more misdirected. A Top 10 search result for "cough" offers lung cancer and cystic fibrosis as the first two explanations. Though one should always consider these possibilities, they're almost always much further down the list.

People also don't always know the exact symptom or keyword they should be searching for. A simple Web search for "bloating," for example, yields plenty of articles that recommend eating more slowly and avoiding milk-based foods. This advice is helpful if your bloating is from too much air in your guts, as it often is, but much less so if it's from fluid in your abdomen, which can be a sign of liver disease or ovarian cancer. Only after seeing a doctor and undergoing a physical examination would you know that what you're experiencing is actually called "ascites".

Most alarmingly, we've noted a growing amount of medical misinformation online. Many articles are written not to inform and guide readers but rather to generate revenue or promote certain causes. Unfortunately, it's not always easy to discern a website's intentions.

Numerous health-care-based websites appear to exist primarily to attract clicks and then show ads for medications or supplements. Since these platforms depend on advertising revenue, they've been carefully optimised to obtain high rankings in search engines - but not necessarily to provide good information.

For example, the top search result for "headache" currently features multiple banner and video advertisements for Emgality, a new medication for migraine that is recommended only for those who have not improved with other treatments. It also costs nearly $7000 a year. Of course, there is no equivalent ad campaign for rest and hydration, which are more likely to alleviate a typical headache.

Other articles are hosted by businesses and organisations striving to advance awareness and usage of their own specific treatments. In particular, some popular websites promote the idea that common symptoms can be traced to nutritional or lifestyle problems that are optimally treated by the supplements, "cleanses" and other products that are (of course) only available on those same platforms. (Paging Alex Jones and Gwyneth Paltrow!)

The bottom line is that only a small fraction of available health information online is written in the best interests of the reader. Unfortunately, a still-smaller fraction of these articles are accurate and well-organised enough to afford the layperson a decent chance of obtaining the correct diagnosis and a reasonable plan.

A few strategies, however, greatly increase the odds of finding unbiased and helpful information.

First, pay close attention to author and platform. The best medical information is written by academic physicians and/or hosted on the websites of medical schools or government agencies (such as the National Institutes of Health). If an article is on a commercial platform and simply "reviewed" by a physician, it is more likely geared toward sales or marketing goals than health ones.

Second, be suspicious whenever a website recommends a specific medication or procedure, especially when the site also offers, or contains obvious advertisements, for that treatment. These pages are frequently created by hospitals trying to increase their procedure volumes, pharmaceutical companies aiming to boost sales of their drugs, or private companies trying to sell supplements and other products.

Finally, be careful to distinguish true search results from sponsored links. For example, a Google search for "knee pain" will likely produce, depending on your location, multiple sponsored links to hospitals promoting knee replacement surgery. On the search results page, these links can be nearly indistinguishable from the non-sponsored results that correctly describe joint replacement as a last resort. Look carefully for the word "ad" next to the link.

Once you've reached a tentative diagnosis, you'll next have to decide whether to seek actual medical attention. This can be a challenging decision that depends, in part, on your comfort with uncertainty, as well as the convenience and cost of seeing a physician.

If you do decide to see a physician, we strongly recommend starting with an objective description of your symptoms, rather than relating what you have concluded based on your own research. You're much likelier to receive the correct diagnosis if your doctor can obtain a complete history without being biased toward one possible explanation.

If you're unsatisfied with your doctor's diagnosis or plan, however, or you're still convinced your own assessment is the right one, it's completely appropriate to ask follow-up questions. You should never leave a doctor's office with lingering doubts about the plan. If necessary, you can always seek a second opinion.

If you're fairly confident you don't need a doctor, we have two recommendations. First, don't skip real medical care based on the advice from any single website. Consult with multiple reputable sources to get the most comprehensive and nuanced perspective possible. To paraphrase Mark Twain: Be careful when reading medical websites, because you don't want to die of a typo. If you have any doubts, make a doctor's appointment.

Second, don't stop any medications on your own. The internet can link almost any symptom to almost any prescription drug. It's easy to convince yourself - incorrectly - that your fatigue, headache, nausea or other symptom is a side effect of one of your medications. Though you might be tempted to skip some pills, the unintended consequences can be severe. We have both seen heart attacks and strokes that occurred because a medication was stopped without a physician's guidance.

No website can replace a detailed assessment from a qualified medical provider. The rise of online misinformation that has drawn so much attention in politics also now confronts consumers of health-related content.

As physicians, we urge people to take responsibility for their own health, and we understand the desire to research symptoms and diagnoses before scheduling an office visit. Indeed, our new book provides an accurate and comprehensive framework for doing just that. If you're planning to get information from the web, however, we recommend simply exercising caution and skepticism. The stakes are just so high.

• Kelly is a senior clinical fellow and Eisenberg is an associate professor of medicine at the Columbia University Irving Medical Center. They are co-authors of "Am I Dying?!" Kelly is completing his cardiology fellowship and Eisenberg is a clinical cardiologist.