Background & research topic

Social media channels focused on health videos can be a reliable point of reference for its viewers if their content is verified (BMC Medical Education, 2022). However, platforms generally incentivize business that attracts more viewers and motivatesthese to stay longer while not restricting videos’ content quality. This has led to a considerable amount of misinformation that isnow circulating on the web. Studies have shown that more than half of health-related videos on platforms like YouTube containmisleading information and biases (Hou et al., ACM/ICMI-MLMI 2020). For instance, Li et al. (BMJ Global Health, 2021) foundthat 11% of YouTube’s most viewed videos on Covid-19 vaccines contradicted information from the WHO or the Centers forDisease Control and Prevention.

Different studies have started to assess underlying causes for misinformation sharing and spreading. For example, Keselman etal. (JMIR, 2021) analyzed individual’s willingness to share non-evidence-based health YouTube videos and found thatwillingness increased with friend’s recommendations, positive reviews, and other factors such as low level of information andscience literacy. Other recent studies have found that social media posts which direct to YouTube videos are one of the most effective techniques for spreading misinformation, and understanding the full picture requires a multi-modal view on diverseplatforms (ICWSM, 2022).

This research aims to (1) identify characteristics and patterns that support or discourage the development and spreading ofhealth misinformation via social media video platforms through a literature review (conceivably, specific features regarding health promises, conspiracy claims, economic incentives, anonymity and accountability, and others) (IJMS, 2022). (2)Afterwards, an analysis of misinformation development and spreading is conducted to verify the identified characteristics. Thisincludes looking at the largest online video platforms (e.g., YouTube, TikTok, Instagram, Facebook, X/Twitter) and applying the identified theoretical findings to top channels of selected platforms (e.g., using YouTube Search and Filters and YouTubeAnalytics, TikTok Hashtags and Top Creators List, Instagram Hashtags and Explore page and Reels section featuring trending content). This could include a comparison of same content provided by major producers on different platforms toassess the effectiveness of platforms for distributing certain health topics.

Research goal & question

Identification, analysis and evaluation of factors that support and discourage the development and spreading of misinformationfocused on health content via online video platforms.

“What factors influence development and sharing of misinformation of health videos across social media platforms?”

Proposed data access & analysis

Identify 5 large channels of 3 platforms of your choice promoting health misinformation on a topic of your choice, totaling 15channels (e.g., anti-vaccine content, alternative cancer/allergy/asthma… cures, weight loss myths) – use according searchterms or video tags). The following ideas are a first discussion starter and can be adapted based on your interests andexperiences. See this exemplary study that you may use as an inspiration (BMJ Global Health, 2020).

For the top 10 videos of your 15 channels, analyze titles in relation to click numbers, likes/dislikes, comments and engagementmetrics (retweets, shares) and other criteria that you believe are relevant to estimate the reach of misinformation:

1. Classify the type of misinformation presented, i.e. what is the misleading claim, and identify connections between video titles(e.g., length, buzz words) and click numbers
2. Collect and analyze most liked comments to understand audience reactions and sentiments (e.g., do users express trust orskepticism)
3. Analyze engagement metrics to understand how widely videos are spread

Bringing the insights together, synthesize your learnings on which health misinformation have a high potential for being viewed,commented and spread.

These topics may be adapted to your research focus and approach. An alternative focus could include analyzing the languageused in the videos, e.g., is the language rather formal or is it addressing viewers directly? Does this observation lead to insightsregarding a specific “tone of viewers’ manipulation” that is rather used in certain types of videos with misinformation?

Contribution

Research call: “There is a need for research to identify common features that can be used as quality indicators in health-relatedvideos. Using this information will help users make the appropriate selection. There is also need for research to identifycommon characteristics that can serve as indicators of health-related video quality. Identifying these characteristics will behelpful in selecting health-related videos.” (BMC Medical Education, 2022)

Requirements

• Ability to work independently, including independent acquisition of suitable methodologies
• Interest in data-driven, medical innovations and their ecosystem & very good academic performance
• Enrollment in the Medical Engineering, Industrial Engineering, G-TIME, data science or a comparable degree program at theTUHH.

If interested, please share your CV and overview of grades with miriam.michel@tuhh.de

Sources

• Guler, M. A., & Aydın, E. O. (2022). Development and validation of a tool for evaluating YouTube-based medical videos. IrishJournal of Medical Science (1971-), 191(5), 1985-1990.
• Hou, R., Pérez-Rosas, V., Loeb, S., & Mihalcea, R. (2019, October). Towards automatic detection of misinformation in onlinemedical videos. In 2019 International conference on multimodal interaction (pp. 235- 243).
• Keselman, A., Arnott Smith, C., Leroy, G., & Kaufman, D. R. (2021). Factors influencing willingness to share healthmisinformation videos on the internet: web-based survey. Journal of medical Internet research, 23(12), e30323.
• Li, H. O. Y., Bailey, A., Huynh, D., & Chan, J. (2020). YouTube as a source of information on COVID-19: a pandemic ofmisinformation?. BMJ global health, 5(5), e002604.
• Li, H. O. Y., Pastukhova, E., Brandts-Longtin, O., Tan, M. G., & Kirchhof, M. G. (2022). YouTube as a source of misinformation onCOVID-19 vaccination: a systematic analysis. BMJ global health, 7(3), e008334.
• Micallef, N., Sandoval-Castañeda, M., Cohen, A., Ahamad, M., Kumar, S., & Memon, N. (2022, May). Cross- platform multimodal misinformation: Taxonomy, characteristics and detection for textual posts and videos. In Proceedings of theInternational AAAI Conference on Web and Social Media (Vol. 16, pp. 651-662).
• Osman, W., Mohamed, F., Elhassan, M., & Shoufan, A. (2022). Is YouTube a reliable source of health-related information? Asystematic review. BMC Medical Education, 22(1), 382.

Hintergrund

Digitale Gesundheitsanwendungen (DiGA) wurden 2019 in Deutschland eingeführt und haben seither ein hohes nationales und internationales Interesse geweckt. Diese digitalen Medizinprodukte bieten innovative Lösungen zur Verbesserung der Gesundheitsversorgung, Diagnosestellung und Behandlung von Krankheiten (Ludewig et al. 2021).

Bis heute sind bereits mehr als 60 DiGA im DiGA-Verzeichnis des Bundesinstituts für Arzneimittel und Medizinprodukte (BfArM) gelistet und können somit von Ärztinnen und Ärzten verschrieben werden (Bundesinstitut für Arzneimittel und Medizinprodukte 2024). Die Verschreibungszahlen sind zum jetzigen Zeitpunkt allerdings noch gering und wachsen langsamer als noch im letzten Jahr (Annual reports of GKV-SV on DiGA 2024; Schmidt et al. 2024).

Als ein wichtiger Faktor für die Adoption digitaler Innovationen im klinischen Alltag von Ärztinnen und Ärzten wurde „Education and Awareness“ identifiziert (Gordon et al. 2020). Gleichzeitig hat eine Studie zum Einfluss von medialer Berichterstattung auf die medizinische Urteilsfindung von Ärztinnen und Ärzten während der COVID-19 Pandemie eine Beeinflussung der Leistungserbringenden durch die Medien festgestellt (Goidel et al. 2023).

Ziel

In dieser Arbeit möchten wir untersuchen, wie ausgewählte deutsche Medien über DiGA berichten und ob bzw. inwieweit die mediale Berichterstattung über DiGA in Deutschland, das Verschreibungsverhalten von Ärztinnen und Ärzten für DiGA beeinflusst.

Forschungsfrage

"Wie beeinflusst die mediale Berichterstattung über DiGA das Verschreibungsverhalten der Leistungserbringenden?“

Methodik

• Befragung von Ärztinnen und Ärzten zur Wahrnehmung Digitaler Gesundheitsanwendungen und zum Einfluss von Medien auf diese
• Sentiment Analyse ausgewählter Medien

Voraussetzungen

• Fähigkeit zu selbständigem Arbeiten inkl. selbstständiger Aneignung von geeigneten Methodiken
• Sehr gute Deutschkenntnisse 
• Interesse an Daten-getriebenen, medizinischen Innovationen und deren Ökosystem
• Sehr gute Studienleistungen
• Immatrikulation in den Studiengängen Mediziningenieurwesen, Wirtschaftsingenieurwesen, G-TIME oder ein vergleichbarer Studiengang der TUHH

Bei Interesse, bitte bei sara.gehder@tuhh.de per Mail melden. Bitte schicke die relevanten Unterlagen (CV, Notenübersicht etc.) mit. 

Quellen

• “Annual Reports of GKV-SV on DiGA.” 2024. www.gkv-spitzenverband.de/krankenversicherung/digitalisierung/kv_diga/diga.jsp (February 15, 2024).
• Bundesinstitut für Arzneimittel und Medizinprodukte. 2024. “DiGA-Verzeichnis Des BfArM.” diga.bfarm.de/de/verzeichnis (July 20, 2024).
• Goidel, Kirby, Timothy Callaghan, David J Washburn, Tasmiah Nuzhath, Julia Scobee, Abigail Spiegelman, and Matt Motta. 2023. “Physician Trust in the News Media and Attitudes toward COVID-19.” Journal of Health Politics, Policy and Law 48(3): 317–50. doi:10.1215/03616878-10358696.
• Gordon, William J., Adam Landman, Haipeng Zhang, and David W. Bates. 2020. “Beyond Validation: Getting Health Apps into Clinical Practice.” npj Digital Medicine 3(1). doi:10.1038/s41746-019-0212-z.
• Ludewig, Gottfried, Christian Klose, Lars Hunze, and Sophia Matenaar. 2021. “Digital Health Applications: Statutory Introduction of Patient-Centred Digital Innovations into Healthcare.” Bundesgesundheitsblatt - Gesundheitsforschung - Gesundheitsschutz 64(10): 1198–1206. doi:10.1007/s00103-021-03407-9.
• Schmidt, Linea, Marc Pawlitzki, Bernhard Y. Renard, Sven G. Meuth, and Lars Masanneck. 2024. “The Three-Year Evolution of Germany’s Digital Therapeutics Reimbursement Program and Its Path Forward.” npj Digital Medicine7(1). doi:10.1038/s41746-024-01137-1.