乌鸦传媒

YouTube videos about prostate cancer often miss the mark in terms of balanced and reliable information, according to a review of the most popular videos on screening and treatment.

Videos about prostate cancer were almost 50% more likely to mention the potential benefits vs harms associated with screening or treatment, and half of the videos did not promote shared decision-making, as recommended by current clinical guidelines. Three fourths of the videos contained "potentially misinformative and/or biased content" in the video itself or in viewer comments.

The analysis revealed a significant negative correlation between the quality of a video and audience interest, Stacy Loeb, MD, of NYU Langone Medical Center in New York City, and co-authors reported in .

"I guess the main point, at least from this sample of videos, is that you are more likely to find a rosy picture," Loeb told 乌鸦传媒. "It is important for patients to be aware of both the pros and cons of screening and treatment decisions. Online sources do not necessarily make you better poised to participate in the shared decision-making process."

The message for physicians is that "we should be directing our patients to high-quality content," she added. "A lot of what's out there does not provide a balanced view or even recommend treatments that are guideline supported. Also, to create more content, to be involved in social media and support the high-quality content that is out there and in the process, help raise the signal above the noise."

YouTube's position as a leader among social media platforms speaks for itself: More than seven billion videos with a linked social networking interface that permits viewers to add feedback to each video, more than one billion registered users, and more than five billion videos watched daily. The website has a vast potential for obtaining and sharing health information, Loeb and coauthors noted in the introduction to their study.

Studies examining health-related content on YouTube often demonstrated poor quality, biased content, and commercially supported content, with an associated potential for harmful consequences, the authors continued. For example, a review of YouTube videos that had a showed high ratings among YouTube users, even though almost half of the videos contained misinformation.

YouTube's health information collection includes >600,000 videos about prostate cancer, many of which have >100,000 views. Few studies have examined the quality and accuracy of YouTube videos related to prostate cancer.

A decade-old analysis of showed that 73% of the videos provided fair to poor information, and 69% exhibited a clear bias in favor of screening or treatment. A of 100 YouTube videos about prostate cancer identified commercial and/or biased information in many of them. Previous studies of prostate cancer videos did not use validated instruments to evaluate the content, and none examined user interactions with the videos, according to .

Using default search results, Loeb and co-authors sought to perform a comprehensive contemporary review of YouTube content related to prostate cancer. They reviewed the top 150 English-language videos (according to audience engagement) about prostate cancer screening (75 of ~173,000) and treatment (75 of ~444,000). The review included user comments related to each video.

As part of the review, the researchers applied the quality criteria for consumer health information. Using the DISCERN questionnaire, the investigators rated each video with regard to quality of information presented, discussion of both risks and benefits (or harms) of a treatment or test, evidence of bias in the presentation, definitions/explanations of medical terms, references to information provided in the video, and other quality-related factors. The team also calculated two engagement metrics for each video: number of views per month and the ratio of total viewers to the number of viewers who gave the video a "thumbs up" rating.

For a subset of 50 videos, Loeb and colleagues evaluated the reading level of written transcripts and computed scores for videos by means of the Agency for Healthcare Research and Quality Patient Education Materials Tool, which addresses understandability and actionability.

The 150 prostate cancer videos had an average of 45,000 views, and total views for individual videos went as high as 1.3 million. Key findings from the analysis included the following:

• The overall quality of the information in the videos had a rating of 3 on a 5-point scale
• 75% of the videos described benefits
• 53% of the videos discussed potential harms
• 50% of the videos promoted shared decision-making between patients and their physicians
• 54% of the videos defined the medical terms used in the videos

Loeb and colleagues determined that 77% of the videos or user comments included potential misinformation or biased content. Using engagement metrics, the team estimated that the videos had a total potential reach of 600,000 viewers.

Analysis of engagement metrics and video quality measures also showed that as the quality of video content increased, audience engagement decreased, and as the quality decreased, engagement improved (P=0.004 for views/month, P=0.0015 for thumbs up/views).

"Another key takeaway from this study is the need to be aware of the source of the content and the date of it," Loeb said. "Treatment and screening guidelines change over time. Sometimes when you do a search, you might find content that is highly viewed but that is out of date. These things don't necessarily get taken down automatically if guidelines change."

Comment