Social Media Is Replete With Patient RWD, But How Can Researchers Best Use It?

A conversation among Pistoia Alliance Social Media to Produce Real-World Evidence (RWE) project leaders Thierry Escudier and Aditya Tyagi and Clinical Leader Executive Editor Abby Proch

Social media has been around for decades, allowing users to communicate with each other on any number of topics, including — and especially — their health. Until recently, that information has largely been accessed and consumed by individuals for personal benefit. But now, companies, including drug developers, are increasingly interested in tapping into those conversations to better understand patients and thereby improve their products and strategies.

But, how to do it doesn’t come easily. Questions remain around operationalizing such an effort, in addition to those about data quality, consent, and regulatory acceptance.

For the past four years, the Pistoia Alliance has been working to address those gaps through its Social Media to Produce Real-World Evidence (RWE) project, which aims to establish best practices, engage regulators, and understand patient attitudes toward data use. The result, thus far, has been its POMELO Protocol tool to help industry and academia implement a listening program.

In this interview, project leaders Thierry Escudier and Aditya Tyagi discuss how social media listening is currently being applied, where it fits alongside established evidence-generation methods, and what it will take to make it a credible, structured input for drug development.

What are the sentiments toward social media listening in clinical research, and how is it currently being used?

Thierry Escudier: Patient engagement, patient centricity, and getting the patient voice are more and more important in recent years. Even the regulators are pushing the industry to listen to patient needs with the FDA’s patient-focused drug development (PFDD) approach. And the EMA stands for the same type of approach. Some of our members asked us if we could help them define best practices for industry when listening to communications of patients on social media.

On social media, where it's public, a patient is talking to other patients. They may discuss things they don't share even with their own friends, family, and doctor. This has been considered by industry and confirmed by the regulators to be a good way to get information that you don't get from survey questionnaires and clinical trials.

It's really trying to understand what patients are telling their community about their symptoms, their expectations. It's not about pharmacovigilance because pharmacovigilance is covered by many other guidance, policies, and SOPs from the industry.

If you can better understand your patient, you can better answer their needs.

Clinical Leader: How is the Pistoia Alliance’s Social Media to Produce Real-World Evidence (RWE) project structured?

Aditya Tyagi: There are three pieces. One is setting up best practices for collecting social media data for drug development purposes. Regulators have not defined it. They want to, but they have not taken the initiative so far, and neither has any other organization.

We consulted with industry, defining how to collect the data, how to analyze the data, how to present the data, and how to make use of that data within the drug development process.

The second piece is working with regulators, which is a long process. There are some openings here and there, but not a concerted effort by the regulators themselves.

The third piece is listening to patients and whether they are okay with pharma companies collecting their data. We did a survey in Germany, Spain, and the U.S. We thought U.S. patients would be more open, and German patients would be less open to sharing that data. But surprisingly, everybody was very keen to let pharma companies use their data as long as it was used for altruistic purposes like improving the drug development process or the drugs or addressing some of the unmet patient needs.

Clinical Leader: In practice, how does social media listening happen?

Tyagi: There are companies using AI tools to collect data from social media platforms like Facebook, Reddit, online forums, that sort of thing. These AI tools are making sense of that data and collecting it in a way that doesn't reveal the identity of patients.

Now, pharma companies are either engaging with AI vendors or have developed proprietary AI tools. A lot of that data is being used by medical affairs for marketing; some of that data is also used for drug development. But it's not a structured platform, and that is the basic agenda for our project.

This is not about pharmacovigilance, but pharmacovigilance was a new field a few decades back. Now it's a very well-regulated field. When pharmacovigilance started, there were concerns about noise in the data and how to filter out that noise from adverse effects. But they got together, set up a platform, set up best practices, and now it's a heavily regulated field. We are hoping that social media listening will follow the same trajectory.

You mentioned folks’ willingness to share what they say online with pharmaceutical companies. To what extent do users have to consent to that?

Tyagi: If I provide feedback on Reddit, it becomes Reddit's property and not mine. I do have some privileges of deleting it or hiding my username from that platform, but it becomes Reddit's IP. At the end of the day, whoever is hosting these views becomes the owner of the data, and that’s not necessarily the patient.

The way I have seen it being collected by pharma companies, they do not retain any of the identifiers. But this is something we have asked in the survey: Do you feel this can threaten you in some way? We'll soon find out, once we have analyzed the data internally.

What can we gain from incorporating social media listening into clinical research?

Escudier: In using social media, you can access a large population, instead of having only patients in the neighborhood of a doctor, and that would help with diversity efforts. But this has to be considered as a complementary or additional way to collect data.

Secondly, you may find a low-level signal the more you listen to a big population, more than with just a 10-patient interview. You may get also access to patients in rare or ultra-rare disease where you sometimes have only 10 patients per country.

What are some other potential challenges?

Escudier: I would say the data collection, data analysis, ethics, and transparency. We believe it’s going to bring big opportunities, but we have to work on the quality of the data and the purpose.

Tyagi: A lot of tools are being built based on the evolution of AI. Reddit is trying to license a lot of the data that it shares with pharma companies to make money.

We’ll also have to find a way to authenticate users, which is one of the biggest challenges. How do we know what's writing that feedback, whether it's an AI agent or an actual human being?

Also, we need regulators’ acceptance of this data into the real-world evidence packages. We want them to work with us and others to make it mainstream. We want them to join now, but another challenge is regulatory reluctance.

Where are we right now with regulators?

Escudier: In the FDA PFDD documents, they state that social media can be considered a source of real-world data. So, it was a positive sign. The EMA did the same last year in a position paper on patient experience data. So, we have been working with the EMA and are hoping to release a new edited version of the patient experience data sometime this summer.

What’s next for Pistoia Alliance and social media listening?

Tyagi: Within our Steering Committee Group, we are trying to find a solution to these challenges. How do we make social media listening more robust so that it can actually be considered as a viable source for RWE?

Our current qualitative patient survey was included 57 patients across three countries — the U.S., Germany, and Spain — and three disease areas — cardiology, oncology, and rare disease. Based on the data we have collected so far, it's looking pretty positive, and we want to expand that. Next, we are doing a quantitative survey, which is for hundreds of patients, to see whether we get the same message from the group.

About The Experts:

Since October 2023, Thierry Escudier has served as a portfolio lead for the Pistoia Alliance. Prior to this, he was a strategic leader at the Pistoia Alliance for over two and a half years. Thierry brings an extensive background in clinical operations, having worked for Pierre Fabre Medicament for 25 years, with a strong focus on digital innovation and patient centricity.

Aditya Tyagi is a clinical operations expert and project manager at the Pistoia Alliance, where he leads the ClinOps, social media listening, and IDMP ontology projects. Aditya has over 19 years of experience in managing clinical trials in the U.S., the UK, and India. He has overseen preclinical and Phase 1-4 clinical trials in areas such as oncology, cardiology, neurology, and rare diseases.