Can you trust YouTube videos on mammography?

2017 01 03 16 38 30 662 Youtube Button 400

With more and more patients turning to digital media, particularly YouTube, for the answers to their mammography-related questions, are the videos made by knowledgeable people rising to the top? Or are they getting buried? A new study suggests the latter: Quality mammography videos are not reaching the majority of YouTube viewers.

Sharon Lee from the Western University of Health Sciences in Pomona, CA, and colleagues searched YouTube using the term "mammogram" and analyzed the first 100 videos. They specifically evaluated the academic quality of the YouTube videos and found the most viewed videos came from nonmedical professionals, at 722,145 views versus 23,211 views (the Breast Journal, September 13, 2020).

"[T]here is no association between the number of views and the quality of the video," they wrote.

Who's presenting on YouTube?

Numerous studies over the years have asked the question, "How well is YouTube educating patients?" Past research has looked at general characteristics such as the information provided in the video and how well it addressed a patient's concerns.

However, to date, Lee and colleagues haven't found a study evaluating the academic quality of the people presenting the information, which they sought to rectify with this study.

In their study specifically focusing on mammograms, the researchers looked at physicians' characteristics including mammography-associated publications, h-index (a metric that measures both the productivity and citation impact of an author's publications), academic affiliation, and specialty.

Lee and colleagues used the search term "mammogram" in May 2020 with YouTube's default setting, which sorts videos by relevance. They included the first 100 videos and categorized them into four groups: academic videos, vlogs, patient interviews, and other.

Constituting the top 100 videos were 62 academic videos, 11 vlogs, six patient interviews, and 21 "other" videos. In the academic videos, Lee and colleagues identified 34 physicians. To analyze each physician's characteristics, the researchers scoured Scopus and Google.

Physicians who trained in diagnostic radiology were more likely to have a greater average view count than physicians who trained in other residencies (p < 0.5). In addition, more often than not, academic videos contained physicians with mammography-related publications, academic affiliations, and diagnostic radiology residency.

However, vlogs -- or videos uploaded by people chronicling daily life activities and not necessarily people with reliable medical information -- were viewed more frequently by about 31%.

"The discrepancy in views may be due to factors including how the YouTube algorithm ranks videos and the clickthrough rate of a video's thumbnail and title," Lee and her co-authors wrote.

The YouTube algorithm makes a big difference in terms of what people watch and what they don't. For instance, each user's search result will be ranked differently based on their watch history. In the future, it would be beneficial for patients if YouTube developed an algorithm that preferentially ranked videos uploaded by a reliable medical center or healthcare professional, the researchers added.

For now though, Lee and her colleagues advise patients to cross-reference YouTube videos with other sources of medical information given the inconsistent nature of the videos that currently appear at the top of search results.

Page 1 of 569
Next Page