Ethics statement
This study was exempt from Institutional Review Board approval of the study institution since it involved the use of public access data only.
Search strategy
Our methodology is based on previous study[16]. The study population was composed of all YouTubeTM videos containing information about Candida auris infection on September 21, 2019. The follow search key terms were used: “Candida auris infection” and “Candida auris”. And the YouTubeTM search was sorted by the ‘‘Relevance” option of videos, which is probably the most common option for users. The first 100 videos (20 videos/page, first 5 page) of each search result were selected, because users usually screen within the first 5 pages of a search result. We ignored all the advertisements in the search results and in the beginning of video.
Inclusion and exclusion criteria
The inclusion criteria were: (1) available on August 21, 2019; (2) related to Candida auris in content; (3) in English.
The exclusion criteria were: (1) non-English; (2) videos had no accompanying audio; (3) irrelevant videos; (4) duplicate videos; (5) advertisements.
In addition, Videos with multiple sub-parts were counted as one video.
Variables
The follow information was extracted for each of the videos: the title; length of the video (in minutes); total numbers of views, comments, “dislike”, “like”; days since upload. If information was missing because of the video publisher restrictions, they were not considered in the corresponding analysis. For each video, we recorded the source of upload, categorized as government/news agencies, Universities/professional organizations/non-profit, physician/physician groups, Stand-alone health information websites, Medical advertisement/for profit companies, Individual and other. Regarding the characteristics of the individuals appearing in a video as well as the primary protagonist(s) in each video, we categorized it as patients, patient's family or caregiver, physicians, nurses, reporter, social individual and others. In addition, Video style was categorized as follows: “education”, “entertainment”, “news report”, “politics”, “personal experience and blog”, “interview” and “others”. We also determined percentage positivity (defined as the number of likes divided by the total number of likes or dislikes of that video); likebility (like per day); viewing rate (view per day) and viewer’s interactions (defined as the number of likes minus the number of dislikes divided by total number of views of that video).
Scoring system
Similar to other studies, a point-based rating tool was constructed to evaluate video quality and specific content[17]. The overall quality of all selected videos was assessed using The Global Quality Scale (GQS) which is a five-point scale that was used to assess the educational value of each video (Table 1)[18]. Seven specific contents of the videos were systematically evaluated, including introduction, epidemiology, risk factors, symptoms, susceptible population, treatment and prevention of Candida auris infection. According to whether each content was specifically discussed, the content was given 0 point (Not mentioned), 1 point (Briefly introduced) and 2 points (Introduced in detail) as described previously[19]. A total score of 19 was available and a qualitative rating was given based on the reviewer’s score: “poor”(0-6), “good”(7-13), or “excellent”(14-19).
Each video was evaluated independently by two viewers to determine video eligibility for study inclusion (Jiangqing Huang and Shengcen Zhang). All viewers were blinded to each other’s result. Disagreements (differ by three or more points) between viewers regarding the content scores or GQS scores were resolved by an arbitrator (Qirong xiao) who given the final scores.
Jiangqing Huang and Shengcen Zhang are microbiology fellows with an interest in Candida auris infection. Qirong Xiao is a graduate student in clinical medicine with subspecialty training in hematology department who is knowledgeable in treatment of bacterial infections. They were trained before assessing the quality of videos. They received a document with URL of videos and scoring criterion from Bin Li respectively. And they didn’t discuss any detail during the assessment process.
In addition to the scores given by the arbitrator, the scores given by the two viewers were averaged for the final results and statistical analysis.
Assessment of reliability
The DISCERN tool was used to assess the reliability of the YouTubeTM videos[20]. This is a a five-point scale based on five questions, and each question is answered as yes or no. Each yes was given 1 point, for a total possible score of 5 points. As shown in Table 2.
Statistical analysis
Statistical analysis was performed using SAS 9.4 Statistical Software. The Cohen’s kappa coefficient (κ) and the interclass correlation coefficient (ICC) were calculated to assess the interobserver agreement. Fisher’s exact test (two tailed), Kruskal-Wallis test, Pearson test and Mann-Whitney test were performed for data comparison. Only p< 0.05 was considered statistically significant.