As pediatricians, we can guide patients and their families to learn how to look up medical questions online properly. Why is it important for people to learn how to do this? There are multiple reasons:

1. We are likely to instantly get a practically infinite number of answers to any question we pose to a search engine. It is not the quantity of answers that is lacking but often the quality. The ‘signal-to-noise’ ratio decreases as we get greater and faster access to more detailed information about the world (1).

2. Our minds are not adapted to handling such large amounts of data especially when we are emotionally invested. Our brains are apt to make quick, safe, somewhat-right judgements based on limited information (2), e.g., deciding if there is a lion in the nearby rustling bushes at night.

3. Few people get advice on how to find and interpret relevant information online. Although, most of it is common sense, patients and families may be emotional and vulnerable to misinformation. Facts obvious to clinicians may not be so to non-medical people and may go undiscussed.

4. The solution is not to ask people to refrain from ‘googling’. Virtually everyone will look up something online: a weird symptom, an unfamiliar disease given as the diagnosis to a close relative, or the side effects of a medication.

5. The internet will always be a source of both true and false information. Our knowledge is fallible, and science needs alternative points of view to move forward (3).

6. In medical ethics we accept that the principle of autonomy supersedes beneficence. This means that we believe that to truly help patients and families is to empower them to be able to make sound medical decisions, not to make the decisions for them.

Recommendations for discussing potential sources of confusion:

• The problem of anecdotal “evidence”: e.g., “So-and-so tried this alternative treatment for her disease, and it worked for her so it might work for me”. An example can illustrate the difference between anecdotal and scientific evidence to patients. In a randomized controlled trial, a large group of patients is given the novel treatment A and a control group that is similar is given placebo or the standard treatment B. Then, data is collected from both groups and analysed using statistics to find a significant difference. Only then, we can say treatment A works or does not work. This should demonstrate why it is so difficult to draw any conclusions from anecdotes. Even if we assume the anecdote of someone’s success with an alternative treatment was not falsified, many factors could distinguish them from another patient, none of which can be controlled for. Even if all other factors were the same, any positive or negative effect seen from the alternative treatment will likely be due to random chance.
• About social media posts: Remind patients and families that people do not have any obligation to tell the truth on social media, may have different motivations than expected, or may be unwittingly misleading by leaving out pertinent facts. Even if the social media post is citing seemingly scientific studies, try to verify that a) the studies are real, b) the studies are coming from a peer-reviewed scientific journal, and c) the post is not misrepresenting the study or misinterpreting its conclusions.
• Why searching a disease could be better than a symptom:  usually a constellation of symptoms and physical exam findings make clinicians suspect one disease or another. A symptom by itself can be remarkably non-specific and lead to excessive worry or delay in seeking help. It’s probably a good idea to suggest sticking to reputable medical websites of institutions such as the American Cancer Society or the CDC for basic information.
• The distinction between scientific articles published on medical journals and news articles or articles on science blogs: a) Scientific articles are peer-reviewed, b) data is gathered, analyzed, and interpreted by a professional who knows the context of previous studies and standards of practice, and c) any potential conflict of interest must be declared eliminating a potential source of bias. None of these can be guaranteed for non-scientific articles.
• Documentaries, or videos online as a source: Even if they seem to have a high production cost, these usually conflate anecdotal evidence with real scientific evidence, embellish facts for dramatic effect, and do not have to meet any scientific standards.

Today’s families are surrounded by differing sources of knowledge. Thus, part of the doctor-patient relationship should more than ever be helping families sift through the information and make the right decisions for themselves and their children.

Can Ozlu, MD

References:

1. Taleb NN. Fooled by Randomness. New York: Random House; 2016.
2. Kahneman D. Thinking, Fast and Slow. Penguin Books; 2012.
3. Popper KR. The Open Society and Its Enemies. Princeton, NJ: Princeton University Press; 2011.