Pistoia Alliance Furthers Research on Using Social Media Listening in Drug Discovery, Clinical Research

By Allison Proffitt 

April 22, 2026 | In an expansion of its advocacy work into clinical fields, the Pistoia Alliance has launched new patient research to shape the ethical use of social media listening in drug development. The work is part of its ongoing Social Media Real-World Evidence (RWE) project.  

The project launched in 2022, said Aditya Tyagi, Project Manager at the Pistoia Alliance, and he described a three-pillar approach to formalizing the use of social listening in drug development—not pharmacovigilance. The first pillar of the work was to standardize how social listening has been used in the past. The second pillar is to solicit input from patients and patient advocacy groups. The third pillar will be to share findings with regulatory groups via publications.  

While both the FDA and EMA have signaled that they are open to considering social media as a source of data, practically these findings are not yet accepted as part of real-world data, a missed opportunity Pistoia Alliance believes.  

“This approach is not to change the whole process of submitting a new drug for registration,” Thierry Escudier, Portfolio Lead at the Pistoia Alliance, explained. “What we are working with our members to confirm is that social media can be an additional source for real-world data. That doesn’t mean it’s going to replace any other means [of gathering data]: patient surveys, questionnaires, interviews with physicians, review of health records and things like that.”  

Including the Patient Voice 

The patient phase of research, approved by the German Ethics Council, began last month with “nearly 57” patients from US, Spain, and Germany across oncology, rare disease, and cardiology communities. The indications were chosen, Tyagi explained, to represent a broad patient experience: small patient populations from rare disease; a more chronic indication in cardiology; and a more acute and privacy-sensitive therapeutic area in oncology.  

They began by speaking with patient advocacy organizations for feedback on their planned survey. Recruitment was finished last week, and the patients are now being surveyed about their social media use, their understanding of how social listening might work, and their views on how helpful that might be for drug discovery.  

Although the data have not all been gathered or analyzed, Tyagi said early signals are very positive. Patients in both the US and Europe are keen to share their data for altruistic purposes.  

The goal is to ensure methods for collecting and analyzing social media data reflect real patient expectations, priorities, and concerns. Following the first cohort, the project group plans to run a larger quantitative study involving 300-400 patients and is calling on regulators and pharmaceutical companies to support this expansion. 

Framework Foundations  

This step of the work builds on the best practices the project group published in January. Experts from Bayer, Roche, Boehringer Ingelheim, Chiesi and Semalytix published a paper in Frontiers in Medicine (DOI: 10.3389/fmed.2025.1703923) that provided a framework for pharmaceutical companies seeking to use social media data in a consistent and ethically governed way.  

“Social media plays an important role for patients as they receive emotional, network and information support that helps them in coping with their condition,” the authors wrote in the paper. Because patients are using social media to discuss and learn about their conditions, it is undoubtedly a valuable source of information for researchers.  

The authors observe that social media listening is ideal to support explorative research and find new or previously disregarded concepts from subjective patient perceptions; is a cost-effective method that allows to collect data longitudinally and globally, offering high immediacy since real-time posts can be included; and allows insights into rare diseases and rare events which cannot be detected in small data groups or segmentation of patient groups. 

But there are drawbacks. Patient statements posted on social media might be inaccurate and difficult to interpret and they are not reliable reporting of the exact timing of events/symptoms, dosages, treatment schedules, causal effects, etc. There are ethical questions around gathering data online, even posts that are public and have no assumption of anonymity.  

In response, the authors outlined several best practices for using social media posts in research. Among the twelve listed:  

  • Prefer selecting different social media sources to mitigate source bias and try to prioritize collection of data from public forums.
  • Collect whole forums instead of single isolated posts to reduce bias by search terms and obtain the full context/history of patients.
  • Follow data protection regulations and principles, create a Data Protection Impact Assessment or similar document that details how harm to social media users is minimized.
  • Put safeguards to ensure that analyzers of the data do not search for quotes from the data online, as this might lead to re-identification of individuals.
  • Put measures in place to detect and remove fake accounts/users/advertisements from the data.
  • Only process data of groups that can be assumed to be aware of the sensitivity of information they are sharing publicly on social media. 

In parallel, the project team has developed a decision-support questionnaire, POMELO: Protocol for Social Media Listening Online, designed to help pharmaceutical companies assess whether and how social media listening should be used in real-world data initiatives. The structured, step-by-step questionnaire and examples covers study setup, data collection, algorithm selection, and analysis, helping organizations apply social media listening in a consistent, transparent, and scalable way. 

“We have a couple of pharma companies who are doing the project with us, but it’s really focusing on defining best practices in general,” said Escudier, “not for a specific disease, not for a specific type of drug, not for a specific pharma.”  

“This is why it’s a pre-competitive [effort] as well, because it benefits everybody—whoever is involved in this space. It piques a lot of interest and we are glad to be the first one, so to speak, to initiate the effort,” Tyagi added.

Load more comments
comment-avatar