In what may come as a surprise to no one, researchers from NYU’s School of Medicine and its Perlmutter Cancer Center have found that it’s a bad idea to get your medical advice from YouTube. Social media can be a source of bad medical advice. Their article, published in the latest issue of European Urology, looked specifically at YouTube videos made to educate people about prostate cancer. The researchers evaluated videos based on accuracy, level of misinformation, and commercial bias. Of the 150 most-viewed videos assessed, 77% were found to be misleading in some significant way – either because they had factual errors or biased content in either the video or its commentary. YouTube currently houses about 600,000 videos about prostate cancer – this number represents a 1000-fold increase over the last 10 years. When it came to biased content, 75% of the videos the researchers assessed laid out benefits of various treatment options, but only 53% mentioned any potential harms and side effects. 19% of these videos recommended unproven alternative therapies. Since these videos have a viewership ranging from 45,000 to 1.3 million, it’s clear that people are relying on social media for medical advice that could potentially harm them – or at the very least leave them in the dark about potential dangers of some treatments. According to the National Cancer Institute, approximately 11.2% of American men will be diagnosed with prostate cancer at some point during their lifetime and as of 2015, there were roughly 3.1 million men living with prostate cancer in the United States. But for those who lack access to medical care or timely communication with their physicians, the Internet can be helpful – and sometimes the only – source of information. Lead author and urologist Stacy Loeb noted that these videos aren’t all bad: There is valuable information available in them, but people need to check the source to make sure it's credible and to beware of how quickly videos become outdated as care guidelines constantly evolve with the science. But the sheer number of videos available on YouTube makes it impossible for experts to watch and assess them all. And misleading medical videos on the platform are not limited to prostate cancer. Virtually anyone can post a video masquerading as medical advice and there’s no one in charge of vetting the content. Earlier this year, researchers from the McGovern Medical School at UT Austin and Georgetown University’s School of Medicine published a large-scale review on YouTube videos for medical education and found that there was no consistent way of analyzing the usefulness of videos. Of course, the problem is much bigger than YouTube. Even 20 years ago, when the Internet was in a nascent state, a study found that more than 22 million Americans used it to find medical information and 70% consulted online sources before seeing a doctor. Not much has changed since then as we’re still making the same recommendations to physician organizations, academic institutions, and medical societies to produce and promote more high-quality content, educate consumers on how to assess the validity of online information, and develop consistent criteria for judging the value of online medical information. It’s clear we need to both educate health care providers on how to steer patients towards reliable sources as well as provide non-experts looking for medical information with a reliable and relatively simple way of assessing information. Patients can’t be tasked with keeping up with the latest medical research since much of it is inaccessible (either because it is written for a specialist audience or behind a paywall). However, they can be armed with information that allows them to ask the right questions about an article or video they find online. Some questions will be easier than others to answer. We must not forget that there aren’t always straightforward answers, and trying to determine who funded the creation of the information and for what purpose can be burdensome. Take, for example, the tutorial from the National Library of Medicine – while it provides excellent information on how to research who is running a website, how they are funded, why they are creating the content, whether they are providing current information, and more, the checklist one has to go through is cumbersome. In the end, the checklist is boiled down to a short and useful list: Is the site run by experts? Does it have a clearly stated purpose? Does it label advertising as such? Is information reviewed before posting? Are the sources of data and research clear? Is the information up-to-date? Does it ensure the privacy of your information? Some of these questions are easily answered, but others require reading quite a bit of fine print as well as understanding the process of review and the nature of appropriate expertise. To make matters worse, as more and more researchers are drawn in by financial and publishing opportunities and tainted by politics, even academic and government sources can be biased. And with the rise of pay-for-play journals, sources that once sounded like legitimate publications are merely receptacles for whatever research someone wants to pay to publish. Earlier this year, science writer Michael Shulson pointed out that social media algorithms are making bad information even easier to find: Facebook and Youtube are not just hosting medical quackery. Their algorithms are helping people to find it, and to view more of it. He used an example of an Indian engineer who created a YouTube video claiming that HIV is a hoax and that anti-retroviral medication used in its treatment causes AIDS. Nearly 600,000 people have now viewed his video (certainly many just to see how ridiculous it was after Shulson’s article in UnDark was published) but undoubtedly some viewers were taken in and even convinced that his credentials (an honorary Ph.D. in diabetes science from a school in Zambia and his own nutrition clinic near Delhi) made him an expert. The video has around 14,000 “thumbs up” votes. Shulson asks a deeply important question in his article, one that we would all do well to consider in this age of misinformation: … when content on social media and similar platforms nudges people toward dangerous medical decisions, do those websites bear any responsibility? What we do know is that we can’t rely on social media platforms to police misinformation, even when it’s dangerous. UnDark’s appeal to Facebook over AIDS denialism content was met with the following statement from spokeswoman Lauren Svensson: Counter-speech in the form of accurate information and alternative viewpoints can help create a safer and more respectful environment, and we have long believed that simply removing provocative thinking does little to build awareness around facts and approaches to health. While YouTube has banned one online quack, it has the power to do more since its terms allow it to remove content “that intends to i ncite violence or encourage dangerous or illegal activities that have an inherent risk of serious physical harm or death.” While some claim that paternalistic policies risk giving a small group of people the power to decide what content is valuable, there are some pretty clear lines when it comes to content that might cause physical harm or death. But there’s also a large grey area when it comes to policing merely questionable content. That leaves us back where we started in the quest to become knowledgeable consumers of information. And, again, it brings up the problem of marketers and nefarious entities finding ways around our most probing questions. We’d do well to consider that, even while social media platforms claim that their hands are tied when it comes to the content they put in front of you, they are already employing algorithms that make those decisions, and these involve some value judgment. If Facebook can make an algorithm to sell you a particular kind of toothpaste, there’s no reason they can’t tweak one to promote information from the best sources available. Source
