Methods
This study was reviewed by the Cincinnati Children’s Hospital Medical
Center (CCHMC) Institutional Review Board and deemed to be non-human
subjects IRB 2021-0468.
CCHMC is a large, urban pediatric medical center; the transplant service
performs approximately 100-120 transplants annually. In 2016, the
division of Bone Marrow Transplant and Immune Deficiency at CCHMC
partnered with the internal marketing department and created educational
videos for patients and caregivers undergoing HSCT. Families are
informed of the YouTube-hosted videos through the Bone Marrow Transplant
Education page on the main hospital website, printed education
materials, and one-on-one communications with the care manager.
Fifteen videos were selected for analysis, representing the five
transplant topics, each delivered as separate videos in the three most
common patient languages spoken at CCHMC: English, Spanish, and Arabic
(Table 1). These educational videos were selected due to their
“unlisted” status on YouTube, meaning that the videos can only be
viewed if someone had a direct link and are not available through
YouTube searches or recommendations, supporting a better comparison
between view counts and patients transplanted.
YouTube offers aggregated video performance statistics for videos. The
available data include the aggregated overall view count and overall
average watch percentage for each video over a specified time range.
These data were obtained for each of the videos on the YouTube Analytics
site9. It is important to note that individualized
view data are not provided.
We compared each video’s views to the number of patients going through
the transplant process during each video’s lifetime (from the video
upload date to January 2022). We identified relevant patients from the
CCHMC HSCT database using the infusion date of the first transplant.
Patients who underwent their first transplant concurrent to or after
each video’s release date were selected. The patient’s language was
extracted from the electronic health record (EHR).
Mean views per patient was calculated using the total video views of
each topic and language divided by the number of patients undergoing the
first transplant during the video’s release. Overall topic and languages
watch percentages were calculated using weighted means with topic and
language performance weighted by view count. Pearson correlation
coefficient was computed to explore the relationship between video
length and average watch percentage.