Study design

In this study, we conducted a search on the YouTube platform (https://www.youtube.com) using the keywords “hereditary angioedema” in June 2023, following the ethical principles outlined in the Helsinki Declaration. To minimize search history and cookie bias, we cleared the browser history and used incognito mode with default search settings, aiming to replicate typical user behaviour. Additionally, to prevent bias related to the country or content creator, we disabled location services and set English as the default language. We aimed to replicate a simple search strategy that could be conducted by any person without using filters. Therefore, YouTube’s ranking algorithm was used to sort video results by relevance. Unlike some studies that limit their selection to the first 50 or 150 videos, we included all videos accessible through the YouTube platform at the time of the search. This approach was selected to maximize the comprehensiveness of our dataset and reflect the full range of available content on this topic. To maintain consistency, new videos uploaded during the screening process were excluded from the dataset. The 135 videos available to viewers were selected based on their relevance according to YouTube’s sorting algorithm at that specific time. As of June 2023, the oldest video was uploaded 153 months earlier, whereas the most recent video was uploaded just 1 month prior.

The identified videos were added to a spreadsheet, and their URLs were used for subsequent screening and coding. In the initial screening phase, three researchers independently reviewed the videos to apply the inclusion and exclusion criteria. For videos where discrepancies arose regarding the criteria, the researchers convened to discuss and reevaluate these videos collaboratively, ensuring consensus in the final decision. Three independent allergists, each with relevant experience, evaluated the videos in separate environments over a 4-week period to ensure unbiased analysis. An initial set of 135 videos was screened, and a flow diagram was created in accordance with PRISMA guidelines, illustrating the videos that were excluded based on the specified criteria [29, 30]. This manual process was chosen to ensure a thorough evaluation of each video's compliance with the criteria.

The exclusion criteria were as follows: videos that were non-English, lacked sound, were unrelated to HAE, were purely advertisements, were shorter than one minute, were audiovisually inappropriate, or were duplicates. The inclusion criteria were used to select videos that were in English, had sound, were related to HAE, were not purely advertisements, were at least one minute in length, were audiovisually appropriate, and were not duplicates. From an initial set of 135 videos, a detailed screening process was applied, with each video evaluated individually by three independent allergists in separate settings. Videos were excluded if they were non-English (n = 3), lacked sound (n = 3), were unrelated to hereditary angioedema (HAE) (n = 2), were purely advertisements (n = 4), were shorter than one minute (n = 2), were audio-visually inappropriate (n = 4), or were duplicates (n = 6). This process resulted in a final selection of 111 videos, each reviewed independently to ensure an unbiased analysis (Fig. 1).

Fig. 1

Flow diagram of video selection process

Evaluation and categorization of the videos

All YouTube videos were evaluated via assessment tools for general information, quality, reliability, understandability, and actionability. The videos were then compared across two origin categories and four content-purpose subgroups. The videos were categorized into two groups based on the presenter and/or YouTube channel: the 'Health Group' and the 'Nonhealth Group', following the methodology used in previous studies [31, 32]. A video was classified as part of the health group if the presenter was a medical doctor, paramedic, nurse, pharmacist, or an unspecified health care professional. Similarly, if the channel belonged to a medical doctor, paramedic, nurse, pharmacist, health care facility, training or educational centre/company, nonprofit medical association, or governmental medical organization, the video was also considered part of the health group. All other presenters and channels were classified as part of the nonhealth group.

In addition to categorizing the videos by their origin, all videos were further subdivided into four subgroups based on their content and purpose:

• Medical Professional Education (MPE): These videos are aimed primarily at health care professionals and focus on providing advanced knowledge and skills related to HAE.

• Patient Education (PE): These videos are designed to educate the general public about HAE, offering important information to enhance the understanding and management of the condition.

• Patient Experience: Videos in this category centre on the personal experiences of patients or their relatives, who share individual narratives without an educational component.

• Awareness: These videos focus solely on raising awareness about HAE without offering educational content or sharing personal experiences.

Evaluation of general information

General information included data on views, likes, time of upload (in months), and duration (in minutes) of the YouTube videos. To provide a deeper understanding of user engagement, we calculated the view-to-month ratio to assess video popularity.

Assessment of quality, reliability, understandability and actionability

The quality, reliability, understandability and actionability of the videos were assessed via several tools: the Global Quality Scale (GQS) [27], the Patient Education Materials Assessment Tool for Audiovisual Materials (PEMAT-A/V) [33], the Quality Criteria for Consumer Health Information (DISCERN) [34], and a modified version of DISCERN [35] (Supplemental Table 1). Video quality and streaming were evaluated via a 5-question GQS score, where a higher GQS score indicated superior content quality and information. To evaluate the understandability and actionability of the videos, the PEMAT-A/V score was applied. PEMAT assesses educational material in two ways: understandability, which enables individuals with varying levels of health literacy to comprehend and identify key video content, and actionability, which determines whether viewers can take appropriate actions based on the materials presented. The DISCERN scoring system, developed by Charnock et al. in 1999, consists of 16 questions designed to evaluate the quality of information. Each question is scored on a scale from 1 to 5 points. The questions are categorized into three sections: reliability (questions 1 to 8), quality of information about treatment options (questions 9 to 15), and an overall score (question 16). DISCERN scores are categorized as follows: excellent is denoted by scores of 63–75 points, good is denoted by scores of 51–62 points, fair is denoted by scores of 39–50 points, poor is denoted by scores of 27–38 points, and very poor is denoted by scores of 16–26 points. In 2012, Singh et al. employed a modified version consisting of 5 questions. According to this scoring system, each question is awarded 1 point for a 'yes' answer and 0 points for a 'no' answer, with each video receiving a total score ranging from 0 to 5 points. Each of these scoring systems was rated on a scale, with higher scores indicating greater reliability. To ensure objectivity, the average of three independent results was used.

Evaluation of specific information on hereditary angioedema

The presence of detailed information about HAE was examined in each video. Information on family history, genetic inheritance, symptoms, response to antihistamines, prodromal symptoms, and the frequency and triggers of attacks was analysed. Additionally, all videos were categorized based on whether they contained laboratory findings such as increased serum bradykinin levels, low serum levels of C1-INH and C4, and low C1-INH function. Videos were also evaluated for information on the possibility of delayed diagnosis or misdiagnosis of the disease. Information on therapeutic interventions, including bradykinin receptor antagonists, C1-INH extracts, kallikrein enzyme inhibitors, androgens, tranexamic acid, berotralstat, and lanadelumab, was analysed.