How we’re pursuing health equity at internet scale

Today’s world has more internet users than people with access to essential health services such as primary care, dental care, or surgery. We’ve reached 700,000 of them across 18 countries, but so much more remains to be done.

Meedan’s Digital Health Lab team have created a B2B fact-checking model for health content based on syndication. We are testing a hypothesis that a rapid-response process where a team of health experts distill scientific evidence based on requests made by fact-checkers—requests that other communciators around the world may also have— we can meet the needs of audiences searching for health information online more efficiently, at larger scales, across regions, and directly where searching takes place.

Here are our top-level highlights:

• Our incredible team of experts successfully responds to requests for contextualizing the scientific research in as little as 3 hours
• Our process has successfully reached an audience of over 700,000 people across Asia, Latin America, Africa, Europe and North America, with localization of both text and accompanying visuals (thanks, to Students Against COVID-19 and Localization Lab) in 8 languages prioritized by our partners
• The public health community, including health literacy practitioners, is eager and excited to get more involved in supporting local misinformation response efforts

Reporting done by Reuters about herd immunity and COVID-19 using Meedan Digital Health Lab's COVID-19 Expert Database content.

Through the COVID-19 pandemic, online information searches serve as a supplement to reduced availability and accessibility of in-person healthcare. This coincides with public health authorities realizing the important role that trusted fact-checking organizations with strong community ties play in minimizing the impact of health misinformation, especially during a pandemic. At a presentation our team participated in about health misinformation hosted by the National Academies of Science, Engineering and Medicine, health fact-checking was identified as "the next frontier of health literacy", and, thanks to generous support from the Omidyar Foundation, Google, and the Robert Wood Johnson Foundation, Meedan’s Digital Health Lab gets to contribute to the development of this new frontier.

Over the last year and a half Meedan’s Digital Health Lab has been developing and deploying a rapid-response framework for supporting fact-checking organizations, journalists and other community information leaders that are working to address health misinformation online across literacy levels and languages. In addition to rapid response, we collaborate with behavior scientists, fact-checking organizations to design workshops and educational modules to strengthen the efficaciousness of health misinformation response efforts.

Our COVID-19 rapid response, the COVID-19 Expert Database, is focused on:

• Collecting critical pandemic-related queries directly from fact-checking organizations, community newsrooms and other community information leaders
• Triaging COVID-19 content to send to team members with specific public health expertise to provide standardized responses
• Designing visuals to address disparities in health literacy
• Conducting behavior science, machine learning and rigorous user research to strengthen this rapid-response process as a replicable and scalable framework

Through three 48-hour weekly editorial sprints, as well as 3-hour turnarounds for urgent requests, the content produced by the Digital Health Lab’s COVID-19 Expert DB & Toolkit has served three major purposes: as a reliable source of information relating to rapidly evolving science; providing explanations surrounding conflicting evidence and treatment recommendations; sharing recommended additional resources for fact-checking organizations and journalists to review as they produce the information their communities need.

This content reaches global audiences through newsletters (Berkman Klein Center for Internet & Society, Splice Media, Speak Up Africa, Suno India Daily Updates), podcasts (Suno India, Misinfodemia), online forums such as Misinfodemia on Instagram Live, National Academies of Science, Engineering and Medicine, RightsCon, SwitchPoint, Global Fact, and in content produced by fact-checkers including Students Against COVID-19, VERA Files, Africa Check, Nigeria Health Watch, BOOM Live, and Reuters.

Fact-check conducted by BOOM Live following a request through their WhatsApp tipline using the Digital Health Lab's COVID-19 Expert Database content.

B2B Fact-Checking: Helping journalists check facts with health expertise

In the digital health information ecosystem, building direct communication channels involves developing interventions where target audiences spend time finding answers to the health questions they have, which today means content easily encountered through searches online and through social media. Of course, as any one organization, this is really challenging to achieve.

During this project, we developed a process we call B2B fact-checking: we help the fact-checkers fact-check. It was our hope that by building a model based on syndication, fact-checkers, journalists and other community information leaders can use the content we create directly for them as well as the content we have developed in response to questions shared by other communicators. With this model, we can provide the needed health expertise to reduce health misinformation more efficiently, at larger scales, and directly where a users’ searching takes place.

Fact-checking organization VERA Files uses Digital Health Lab's COVID-19 Expert Database content to respond to dangerous misinformation spread in the Phillippines.

Resources from the COVID-19 Expert Database influencing conversations about complex COVID-19 health misinformation topics as early search results on Google.

Some examples include our contextualization of misinformation about COVID-19 and alkalinity requested by VERA Files, which was also used by India Today in responding to misinformation through their fact-checking tipline. The three responses including information that we contextualized for Nigeria Health Watch related to mask safety and use (here, here and here), in addition to being a top search result, has been shared by Speak Up Africa in their newsletter, was referenced the Misinfodemia Instagram Live series, and has been used by Suno India in their podcasts and daily updates.

Health communications organization Nigeria Health Watch uses Digital Health Lab's COVID-19 Expert Database content to respond to health misinformation topics that emerge in countries all over the world, with a localized response focus.

Our research with these users has demonstrated how having access to experts can increase the subject matter that fact-checkers can cover, and therefore the content available online to the fact-checkers’ audiences. The examples of where our collaboration with partners in different parts of the world have led to first item or first-page search results on Google highlight the impact that smaller health organizations can have on influencing the discourse online when we can support the work of fact-checkers directly.

Data sprints for the triple bottom line: Substance, Syndication and Speed

One of the biggest challenges of achieving impact at the scale of the internet is moving at the speed of the internet. We studied design sprint methodologies from Agile, Google Design, and others to create what we call a Data Sprint: a rapid-response framework that ensures the triple bottom line of substantive, quality content; syndicated, networked impact; and speedy, responsive turnaround times.

Substance: To date, a small team of scientists with expertise across infectious disease, nutritional epidemiology, pandemic preparedness, vaccine adoption and health behavior change have created over 110, 500-word science articles over 20 weeks. Each article or database ‘entry’ responds to questions submitted by fact-checkers and journalists, with additional information to further distill each topic’s historical and scientific significance for health communicators.

This impact will soon be magnified through a collaboration with Localization Lab, targeting 8 languages prioritized by our current partners. We are currently working with SciLine to expand the global reach of the resources that they have created by including them in the short articles we produce for our partners.

Syndication: Through the syndication model developed for this project, our partners have identified an audience of over 700,000 people in more than 18 countries across Asia, Latin America, Africa, Europe and North America. This audience includes podcast listeners, website visitors and newsletter subscribers, and does not include the social media followings of our partner organizations. This level of impact was made possible through our syndication model for content dissemination, enabling a small team of scientists, researchers and communicators to influence diverse information spaces.

Speed: The turnaround time for the content we produce is 48 hours or less, following the tight editorial cycles of our journalism partners. Responses are often created in as few as 5 hours by team members on-call to meet urgent partner deadlines. This rapid response directly targets midinformation: misunderstandings, misinterpretations and misrepresentations based on scant or emerging scientific evidence. Midinformation happens often when expert consensus changes over time, and it is distinct from more straightforward misinformation.

We apply principles from Google’s Design Sprint methodology into three weekly sprints, including problem mapping and prioritization, solution sketching across text and visual responses, and feedback integration from partners. Through consistent use of this model, we are able to optimize for the number of requests we can address weekly, and can account for responding to a diverse set of topics that require reviewing extensive scientific literature across a variety of disciplines.

Key lessons our team has learned over the last 6 months

As a team, we have distilled four key lessons to carry our project strategy forward:

1. It is important to educate health professionals about the value of existing approaches being taken by platforms to reduce the impact of misinformation. Influencing impacts of health misinformation at scale requires closer connections between health scientists and communicators using tools already optimized for easy online encounters with factual health information, such as Google’s ClaimReview and MediaReview annotation frameworks,, and partnerships with third-party fact-checking organizations. The health community is ready and excited to contribute to projects like this one, and needs guidance about where time and effort is best spent.
2. Syndication models can maximize impact at the scale of the internet. Our partnerships with e-newsletters, podcasts, communicators with their own syndication models, and collaborations with fact-checking organizations and newsrooms have reached over 700,000 people in just three months of collaboration. Our goal is to continue supporting communicators with their own content development, know that while we can distill the ever-changing science successfully, community information leaders know their audiences best, and that maximizes the impact of any public health message.
3. A focus is needed on misinformation related to both vaccines and treatment. While these represent the majority of the questions shared with us by fact-checkers and journalists, the most common questions are not reflected in resources provided by intergovernmental or government health institutions. Further, while current public health officials estimate that a successful vaccine will be developed by the end of 2020 and that vaccination campaigns will start by early 2021, a vaccination campaigns’ success is determined by people’s willingness to get vaccinated. Patterns in uneven distribution of vaccines globally, and thus uneven eradication of certain illnesses, will likely yield inequalities in the elimination of COVID-19 in different parts of the world. For this reason, it is essential to not only address vaccine misinformation, but also misinformation associated with COVID-19 treatments, which may emerge in successful forms earlier than the release of a vaccine.
4. Localization conducted by various stakeholders is required to reach & be accessible to community information leaders at a global scale. Our partners have requested that local health experts, as well as users, combine efforts to translate the database content and accompanying visuals into local languages. We’re grateful to the Localization Lab, and to our public health expert translator team, for making this novel approach at localization possible! This optimizes for accessibility (including factors related to availability, relevance, and literacy), and ensures that fact-checkers and journalists operating in diverse languages can all benefit from, and contribute to, improved health literacy where their audiences already seek content.

What’s next: expanding and diversifying our impact

With these lessons in mind, here’s what’s next for our team:

Expanding scope of response effort: Our experts have provided contextualization for over 120 topics related to COVID-19, moving at a rate of, on average, 10 database articles or "entries" per week. Our collaborators are increasingly asking questions about public health topics significant with respect to the pandemic, but are distinctly relevant outside of COVID-19. These include questions about nutrition, non-communicable diseases, antibiotic resistance and reproductive health. Our goal is to provide additional public health expert contextualizations for topics outside of COVID-19.

Engaging with community-based communications networks: Our goal is to expand our offerings with a focus on rapid responses for community newsrooms, freelance journalists, community nonprofits and other information leaders within communities. In collaboration with our partners, we’re working with the product and development teams at Meedan to create more efficient pathways of information-sharing between our experts and community information leaders, using dedicated helplines through SMS, WhatsApp, Facebook and Twitter. We’re also sharing customizable bot infrastructure so community newsrooms can more easily connect their audiences with the health information that is directly relevant to their needs. We hope to support these groups directly by providing the information that they can tailor to the needs of their own communities that they know best.

Strengthening relationships between public health experts, the health literacy community, local newsrooms and fact-checking organizations: Given the important role community newsrooms and fact-checking organizations play in responding to health misinformation via social media and online search platforms, integrating more closely with the public health community can serve to strengthen health literacy at scales larger than ever before.

To learn more about work, please contact health@meedan.com.