Discussion
This is the first study to evaluate parent and caregiver utilization of educational YouTube videos in pediatric HSCT. We found, that on average, 40-70% of each video is viewed. The average watch percentage can provide insight into what happens once viewing begins. The difference in watch percentage between the languages concerned as well as non-English videos (Spanish and Arabic) had a lower mean watch percentage than their English equivalents. Non-English versions of videos are typically longer, and the correlation between video length and average watch time is negative, but also weak (-0.352), suggesting that other factors are likely contributing to their poor performance.
We can ask important questions about our system by using views per patient data as a proxy metric for our education delivery processes. Are we seeing the number of views we would expect given the number of transplants we perform? Is there a system-level reason why video topic or language differences exist?
In next steps, our team will examine our video delivery processes to ensure system reliability and survey our patients and families to identify barriers to watching the current videos. We anticipate that these efforts will present change opportunities both in the system and content areas.
Limitations:
Exactly what YouTube records as a “view” isn’t technically known to the public to discourage attempts to inflate view counts. Estimating viewers from view counts is inexact. A video’s view count would be the same if 50 people watched once or 25 people watched twice. We also do not know if the person watching the video is a patient or family member. While the videos are not searchable or recommended on YouTube, links to the videos are posted on our public website enabling access to those other than our patients and families.