YouTube advice about post-stroke recovery: is it evidence based?

Problem

People with stroke are increasingly relying on internet-based resources to manage their health. YouTube is the most commonly used video-sharing site on the Internet. Previous research has shown that YouTube videos relevant to other health conditions can contain misleading or anecdotal information that is not evidence-based or in line with best practice.

Approach

The aim is to analyse the content of YouTube videos that provide advice about promoting recovery after a stroke to stroke survivors in the community, to ascertain if the advice given is evidence-based or based on best practice if evidence is not available. A search was conducted on YouTube (www.youtube.com) using the terms “stroke recovery advice”. The location was set to “United Kingdom” and language to “English (UK)”. Videos were sorted by relevance and the first 200 were title then full-video screened by two independent reviewers with a third party for arbitration of disagreements. Inclusion criteria: 1) English language; 2) videos providing advice to stroke survivors about ways to promote stroke recovery in the community e.g. change in diet, medications, physiotherapy exercises, alternative therapies; 3) videos published in the last 10 years (due to the evolving nature of stroke guidelines). Duplicate videos were removed. Video content analysis will be conducted by two independent reviewers with a third party for arbitration of disagreements. Analysis aims to ascertain: 1) Type of stroke recovery advice. 2) Sources of the videos: e.g. medical professionals, patients, charities, universities, hospitals. 3) Country of origin of the videos. 4) Whether the video content is evidence-based or based on best practice. To do this, video content will be compared to recent national guidelines. Where no evidence is available, relevant health professionals will be asked if the content is based on best practice.

Findings

After title screening of the 200 videos, 27 were excluded due to: duplication (3); publication over 10 years ago (2); non-English language (13); or irrelevant topic (9) e.g. stroke recovery in dogs. Of the remaining 173 that were full-video screened, 103 were excluded due to: being aimed at non-stroke survivors (11); no advice on promoting stroke recovery in the community (90); or poor quality of video (2). This left 70 videos included for data extraction and analysis which is currently underway.

Consequences

Potential limitations include: restriction of one search term when using YouTube; a lack of clarity about the YouTube search algorithm; no standardized method for appraising health information on the internet; difficulties in analysing video data in-depth; and the exclusion of non-English language videos. This study will provide useful information about stroke recovery information on YouTube, allowing health professionals to better advise stroke survivors on using YouTube as an information source.