Medical Misinformation on Social Media
2019; Lippincott Williams & Wilkins; Volume: 140; Issue: 14 Linguagem: Inglês
10.1161/circulationaha.119.041719
ISSN1524-4539
Autores Tópico(s)Vaccine Coverage and Hesitancy
ResumoHomeCirculationVol. 140, No. 14Medical Misinformation on Social Media Free AccessArticle CommentaryPDF/EPUBAboutView PDFView EPUBSections ToolsAdd to favoritesDownload citationsTrack citationsPermissions ShareShare onFacebookTwitterLinked InMendeleyReddit Jump toFree AccessArticle CommentaryPDF/EPUBMedical Misinformation on Social MediaCognitive Bias, Pseudo-Peer Review, and the Good Intentions Hypothesis Samuel P. Trethewey, MBChB, BSc Samuel P. TretheweySamuel P. Trethewey Samuel P. Trethewey, MBChB, BSc, Birmingham Heartlands Hospital, Bordesley Green East, Birmingham, B95SS United Kingdom. Email E-mail Address: [email protected] University Hospitals Birmingham NHS Foundation Trust, United Kingdom. Originally published30 Sep 2019https://doi.org/10.1161/CIRCULATIONAHA.119.041719Circulation. 2019;140:1131–1133The importance of understanding factors that drive the spread of medical misinformation has been highlighted by recent discourse.1,2 In response to the growing and pervasive threat of medical misinformation, editors of leading cardiology journals collaborated to shine a spotlight on this issue.3 In their commentary, Hill et al3 highlight the responsibility of purveyors of web-based media content to help address medical misinformation.But what motivates an individual or organization to share a specific scientific or healthcare narrative? Motivations are likely to be multifactorial and may be influenced by individual or organizational biases and competing interests. A news organization may be motivated by the need to generate readership, and therefore income, using sensationalist, inaccurate headlines regarding new medical treatments. An additional contributing factor is that individuals may be motivated to share information that they perceive to be both accurate and of social value. For example, misinformation surrounding Cough CPR has been perpetuated for almost 20 years, which, in part, has been facilitated by social media.4 People who share articles endorsing Cough CPR may do so with good intentions because of a perception that the medical advice may benefit others. This is a powerful idea because the perceived validity of information shared online is likely related to the level of trust the reader has in the source. Information shared by friends and family may be more likely to be perceived as accurate by a layperson than an article shared by an independent organization. Moreover, social media can create echo chambers of misinformation where ideas are not challenged and users of social media fall foul to confirmation bias and polarized viewpoints within their social networks.2Historically, the onus has been on clinicians and academics to moderate the dissemination of medical information. However, dissemination goes beyond conference presentations and publications in peer-reviewed journals; the general public plays a key role. This second stage of dissemination is unregulated and unfiltered and is therefore subject to a high risk of bias. There is no 90% rejection rate after peer review for comments posted on social media platforms such as Twitter, Facebook, or Instagram; comments are available instantly without restriction and are accessible to all indefinitely. In some ways, comments on social media platforms are like a sort of pseudo-peer review: unfiltered comments and unsubstantiated claims or critiques, often made by nonexperts. Anyone can comment on a social media post without giving their credentials. Social media platforms therefore act as an open access publishing forum, with users having instant access to preprints with no mechanism in place to formally moderate claims. This can result in information overload from many sources of variable quality, with Internet search engines such as Google facilitating dissemination. The sheer volume of medical information shared on social media can make it increasingly challenging for laypeople to distinguish fact from fake news.The commentary by Hill et al3 shifts the burden of responsibility of addressing these issues to those involved in the spread of information via social media. However, it is unclear exactly how purveyors of web-based media content should combat the spread of medical misinformation. It is likely that a multimodal approach, using a wide range of interventions, will be required. Interventions based on a form of fact-checking of popular social media posts may be a promising avenue to explore. For example, Twitter could assign a Reviewed Content status to Tweets that have been reviewed by a verified expert in a relevant field (for example, a professor of immunology in the case of a Tweet regarding vaccination). The expert would be tasked with reviewing the content of the Tweet and would undergo a fact-checking, critical appraisal process, leading to the production of a written response to the Tweet. This written response could have 2 key components: a lay summary of the evidence along with a detailed, referenced review of the evidence. Once a Tweet has been reviewed, a dated Reviewed Content label could be shown with the Tweet. The written expert review would then be directly linked to the original Tweet for users to access. For practical purposes, only Tweets containing specific medical/healthcare terms that have received a sufficient number of interactions (for example, >5000 shares/likes/comments) would be considered for content review and critical appraisal. These Tweets could be identified using algorithms and screened to determine if they warrant formal review. This could help to target popular issues that have received more widespread attention and are of public health concern. Of course, the threshold at which a Tweet is reviewed would require careful consideration. Furthermore, users of Twitter could be assigned a simple, cumulative Fact-Sharing score based on the number of shares of unique expert reviews. The purpose of this metric would be to encourage users to share fact-checked information originating from expert sources via the Reviewed Content process. Such a score could be displayed on a user profile as a reflection of their engagement with the fact-checking process. Other possible interventions include displaying targeted educational material regarding important public health topics to social media users using automated algorithms and recruiting trusted social media influencers to share fact-checked information on their profiles.Medical journals are engaging more with social media; many journals have active Twitter profiles sharing articles and hosting Twitter chats and journal clubs. It is interesting to note that the Journal of Cardiovascular Electrophysiology has recently begun peer reviewing Tweets for consideration of publication. In their #JCETweet2Press initiative, Journal of Cardiovascular Electrophysiology offers Tweeters the opportunity to share interesting cases, with a chance to be published in the journal following peer review. Initiatives like this are promising and encourage a proactive approach to reviewing content shared on social media.The foundation of evidence-based medicine is peer review. Peer review enables scientists to critique medical research and question the validity of conclusions that can be drawn from that research. However, laypeople may not understand the importance of the peer review process, nor may they appreciate the significance of the publishing source as an indirect marker of scientific validity (ie, a reputable medical journal versus an online blog post). Moreover, it is possible that the general public lack identifiable figureheads from the medical profession whom they trust regarding health information. A possible solution would be to identify and allocate roles to experts (for example, an experienced senior cardiologist) whose specific role would be to advise the public regarding treatments and to help dispel misinformation regarding their specialty. We need more science communicators who can take complex ideas and share them accurately and in an understandable format with the general public. We cannot expect everyone to understand the scientific method or be able to critically appraise literature to identify misinformation. Despite good intentions, many people may unknowingly share misinformation via social media. Although there is a role for medical journals in addressing these issues,5 targeted public health education and social media campaigns may also be required to help prevent the spread of medical misinformation and develop public health literacy. It is likely that proactive interventions, such as promoting scientific literacy in school, will be crucial and may have a greater long-term impact. Furthermore, proactive engagement of the scientific and clinical community with news and social media providers is essential to facilitate accurate communication of medical information to laypeople.DisclosuresNone.FootnotesThe opinions expressed in this article are not necessarily those of the editors or of the American Heart Association.https://www.ahajournals.org/journal/circSamuel P. Trethewey, MBChB, BSc, Birmingham Heartlands Hospital, Bordesley Green East, Birmingham, B95SS United Kingdom. Email [email protected]org.ukReferences1. Merchant RM, Asch DA. Protecting the value of medical science in the age of social media and "fake news."JAMA. 2018; 320:2415–2416. doi: 10.1001/jama.2018.18416CrossrefMedlineGoogle Scholar2. Chou WS, Oh A, Klein WMP. Addressing health-related misinformation on social media.JAMA. 2018; 320:2417–2418. doi: 10.1001/jama.2018.16865CrossrefMedlineGoogle Scholar3. Hill JA, Agewall S, Baranchuk A, Booz GW, Borer JS, Camici PG, Chen PS, Dominiczak AF, Erol C, Grines C, et al. Medical misinformation: vet the message!Circulation. 2019; 139:571–572. doi: 10.1161/CIRCULATIONAHA.118.039193LinkGoogle Scholar4. Trethewey SP. 'Cough CPR': misinformation perpetuated by social media.Resuscitation. 2018; 133:e7–e8. doi: 10.1016/j.resuscitation.2018.10.003CrossrefMedlineGoogle Scholar5. Armstrong PW, Naylor CD. Counteracting health misinformation: a role for medical journals?JAMA. 2019; 321:1863–1864. doi: 10.1001/jama.2019.5168CrossrefMedlineGoogle Scholar Previous Back to top Next FiguresReferencesRelatedDetailsCited By Patrick M, Venkatesh R and Stukus D (2022) Social media and its impact on health care, Annals of Allergy, Asthma & Immunology, 10.1016/j.anai.2021.09.014, 128:2, (139-145), Online publication date: 1-Feb-2022. Sitto K and Lubinga E (2021) "My Birth Control Makes Me Emotionally Psycho": Online Female Narratives about Contraceptives, Communicatio, 10.1080/02500167.2021.1987945, 47:2, (96-121), Online publication date: 3-Apr-2021. Trethewey S (2019) Strategies to combat medical misinformation on social media, Postgraduate Medical Journal, 10.1136/postgradmedj-2019-137201, 96:1131, (4-6), Online publication date: 1-Jan-2020. St.Amant K (2020) Communicating About COVID-19: Practices for Today, Planning for Tomorrow, Journal of Technical Writing and Communication, 10.1177/0047281620923589, 50:3, (211-223), Online publication date: 1-Jul-2020. Carius B and Schauer S (2020) Ibuprofen During the COVID-19 Pandemic: Social Media Precautions and Implications, Western Journal of Emergency Medicine, 10.5811/westjem.2020.4.47686, 21:3 Trethewey S, Beck K and Symonds R (2020) Experience and perspectives of primary care practitioners on the credibility assessment of health-related information online, Postgraduate Medical Journal, 10.1136/postgradmedj-2020-138111, (postgradmedj-2020-138111) October 1, 2019Vol 140, Issue 14 Advertisement Article InformationMetrics © 2019 American Heart Association, Inc.https://doi.org/10.1161/CIRCULATIONAHA.119.041719PMID: 31567022 Originally publishedSeptember 30, 2019 Keywordsbiassocial mediapublic healthpublication biaspeer reviewhealth literacycommunicationPDF download Advertisement SubjectsCardiovascular DiseaseEthics and PolicyHealth Services
Referência(s)