BACKGROUND
Asthma is a common disease that affects a large number of people. Improving the general public’s awareness and knowledge about asthma is crucial for the decrease of asthmatics and the prevention and cure of asthma. YouTube has become one of the major channels for asthma knowledge and information generation, sharing, seeking and use, but the content on it cannot always meet the users’ needs.
OBJECTIVE
This study illustrated the characteristics of the asthma-related videos and users’ behavioral and emotional engagements with these videos on YouTube. The factors influencing the users’ positive, negative and neutral emotional engagements were also explored.
METHODS
We collected the data of asthma-related videos from YouTube with Python. Coding method was adopted to identify the subject of each video manually. The descriptive statistical method was used to demonstrate the characteristics of the investigated videos and user engagements. Sentiment analysis helped reveal the emotional tendency of every video and comment. Negative binomial regression tests were conducted to find the factors influencing the emotional engagements of users significantly.
RESULTS
A total of 354 asthma-related videos with comments were obtained. Nine subjects were identified for the investigated videos. The Treatment (n = 97), Cause & pathophysiology (n = 49), and Complication & related disease (n = 43) subjects had the most videos among the subjects. Diagnosis & test (mean = 9.33×105), Treatment (mean = 6.25×105), and Society & culture (mean = 3.92×105) received the highest average view counts. The transcript valence influenced the positive emotional engagement (P = .001, OR = 1.13) positively and influenced the negative (P = .009, OR = 0.85) and neutral (P = .040, OR = 0.91) emotional engagements negatively. Medication (P = .036, OR = 1.24) positively influenced the positive emotional engagement; Complication & related disease (P = .020, OR = 1.45) and Society & culture (P = .013, OR = 1.64) positively influenced the negative emotional engagement; Complication & related disease (P = .035, OR = 0.77), Medication (P = .024, OR = 0.72), and Prevention (P = .008, OR = 0.68) negatively influenced the neutral emotional engagement.
CONCLUSIONS
The asthma video users’ interests were not always consistent with the video creators’ interests. The creators need generate more videos about diagnoses and social and cultural influences of asthma to meet the users’ needs. Only the video attributes, transcript valence and subject, influenced the users’ emotional engagements with asthma-related videos. Generating the asthma-related videos with more optimistic words in the transcript helps evoke users’ positive emotions and positive responses, which will benefit the video creators and general users. Patients and health consumers can gain emotional support from the positive responses and the asthma-related videos with positive emotions.