Enhancing patient voice: the European Medicines Agency keen on integrating patient experience data from drug R&D to reimbursement

The importance of patient experience data (PED)

Patient experience data (PED) plays a vital role in the regulatory framework for drugs and medical devices, serving as a bridge between the clinical objectives of healthcare providers and the lived experiences of patients and their quality of life. By capturing the perspective of individuals undergoing treatment, regulatory bodies such as the European Medicines Agency (EMA) or the Food and Drug Administration (FDA) in the US can gain insight into how therapies impact patients' daily lives and into the effectiveness and tolerability of treatments from a patient's viewpoint, as well as their needs and expectations. This understanding is essential for driving improvements in healthcare practices and outcomes. Healthcare systems are (slowly) understanding that emotional well-being is equally important, not just physical health, as patient satisfaction is closely linked to medication adherence and overall health outcomes. There is an increasing number of research case studies and papers supporting this - A 2017 broad systematic review of 55 studies published in the BMJ found consistent positive associations between patient experience and objective clinical outcomes including adherence to treatment, preventive care uptake, reduced hospital use and costs, and some safety measures (1).

Incorporating patient experience data into regulatory frameworks not only enhances the understanding of various diseases and treatment impact but is also the foundation of a patient-centric healthcare system, that becomes therefore more responsive to patient needs and more efficient.

European Medicines Agency recent initiatives

The EMA has undertaken significant initiatives aimed at the integration of PED into its regulatory frameworks. These initiatives fundamentally prioritise patient voices in the drug development and regulatory processes, seeking to enhance the relevance, safety and effectiveness of medicines delivered to the public.

The EMA devised its first formal framework for interaction with patients and consumers in December 2011 with a significant update to its Engagement Framework in February 2022.

The Agency had another notable initiative in the form of a Reflection Paper (2) for the regulatory integration of PED published at the end of September 2025, thus giving patient voice more weight. Regulators are formally recognising patient evidence in benefit - risk (BR) assessments and asking for it in dossiers, whilst industry associations are pushing for transparency and structured patient evidence because there is now a formal recognition that patients may value different aspects of their disease and available treatments compared to what tech or drug developers typically think, including particular outcome measures of treatment. The aim is to also inject more innovation and competitiveness in drug research and drug development in the EU.

The paper identifies three principal types of PED, which is usually collected in clinical trials, particularly phase III studies (3):

📌 patient-reported outcomes

📌 patient preference studies (not yet fully established in drug development in many companies) and

📌 data obtained through patient engagement activities (such as surveys, interviews, workshops, scientific advisory groups etc).

This is the actual EU consultation paper: https://www.ema.europa.eu/en/patient-experience-data-ped-reflection-paper

As a side note, the FDA in the US already publishes structured BR frameworks and has formal pathways to collect PED. So the EMA has also been developing tools and guidelines aimed at encouraging pharmaceutical companies to embrace PED in their clinical trial designs and post-marketing studies.

Challenges and gaps in data collection and validation

The collection and strategic analysis of PED represent a crucial aspect of the EMA's mission to enhance patient engagement in the pharmaceutical development process and to do that in an equitable and ethical way for all parties involved, which is not without challenges given some risks associated to AI technologies widely used in data collection and analysis nowadays.

In fact the timing of the EMA consultation paper is likely also linked to the explosion of new digital and AI data collection and analysis solutions over the last couple of years. These technologies are more and more successfully used for unstructured data collection methods (such as conducting unstructured interviews but also increasingly using NoSQL databases, scraping websites and using application program interfaces (API) to access and collect unstructured data from social media platforms and online forums) - so they make spontaneous data generated in (patient) forums and social media platforms such as Facebook or Instagram easier to collect and analyse. This type of data can indeed be considered PED and included but the validation requirements of such data scraping algorithms are not discussed in the EMA reflection paper.

Unstructured data collection methods lack standardised frameworks for validation which can be a risk since such social media data can often suffer from validity errors, missing key data points, suffering from limited representativeness and bias (exaggeration, sarcasm etc.), leading to inconsistencies in data interpretation, ultimately compromising the reliability of the findings.

Additionally, from an ethical standpoint, patient privacy and informed consent can be at risk as some public online posts may contain patient sensitive information and though geo location and metadata individuals can become re-identifiable.

Addressing ethical implications and future directions

As the EMA looks to enhance patient voice through the integration of patient experience data, ethical challenges and ethical standards (particularly regarding informed consent, data security, transparency and privacy protection) surrounding this data collection become increasingly prominent.

In terms of structured data collection, one of the primary ethical challenges lies in ensuring that patients fully understand what their participation entails. Improving the informed consent process is essential; patients should not only be aware of how their data will be used but also understand the potential risks and benefits associated with their involvement - now and in the medium and long term future. Developing standardised, clear and accessible consent documents can aid in achieving this goal. Moreover, the EMA can promote frameworks that emphasise patient education, enabling individuals to make informed choices.

Another concern is related to transparency and privacy protection. In unstructured data collection, patient privacy can be at risk as some public online posts may contain patient sensitive information and though geo-location and metadata individuals can become re-identifiable. As more and in-depth data is collected, ensuring both the protection and anonymisation of personal information becomes paramount and ideally, information from certain social media platforms should never be included - both because it is highly personal and sensitive and because it can also be highly biased and invalid. The EMA should collaborate with stakeholders to establish robust guidelines that enforce transparency regarding data storage, handling, and sharing practices and perhaps set clear validation standards and boundaries for which type of data should and shouldn't be collected. Addressing these privacy concerns is essential for building trust among patient communities.

Also in connection to the patient privacy is the data cybersecurity aspect as AI, like any new technology, is less robust in its security, more data rich (so more valuable) and hence more prone to cyber attacks (4, 5). This is a fundamental security aspect for patients since (a) larger volumes of really valuable personal (medical) data in PED databases can be stolen compared to pre-AI times but also (b) that it can render clinics and hospitals more "hackable" since their (not always very performant) IT systems hold increasing patient data and these cybersecurity incidents can mean some patients lose access to lifesaving treatments and surgery for days or weeks, creating even more backlogs. Notable examples include a ransomware incident in June 2024 in the UK that crippled pathology services for many London hospitals (6), forcing cancellation/postponement of thousands of appointments and elective procedures and contributing to reported patient harms and a cyberattack in April 2024 which forced SYNLAB Italia to disable nationwide computer and telephone systems, suspend sampling points and laboratory activity and investigate data theft (7).

Looking ahead, the EMA has the opportunity to solidify its commitment to ethical data integration by adopting innovative strategies. This includes leveraging technology to enhance data security while ensuring seamless communication with patients regarding the management of their data. A collaborative approach, involving regulatory bodies, healthcare organizations, and patient advocates, can also highlight the importance of ethical practices throughout the research and regulatory landscape.

Sources:

(1) https://bmjopen.bmj.com/content/3/1/e001570

(2) https://www.ema.europa.eu/en/patient-experience-data-ped-reflection-paper

(3) https://www.osborneclarke.com/insights/european-medicines-agency-encourages-regulatory-integration-patient-experience-data

(4) https://securitybuzz.com/cybersecurity-news/71-of-hackers-say-ai-boosts-hacking-efficiency-but-introduces-new-risks/

(5) https://www.mckinsey.com/about-us/new-at-mckinsey-blog/ai-is-the-greatest-threat-and-defense-in-cybersecurity-today

(6) https://www.insurancejournal.com/news/international/2025/01/15/808371.htm

(7) https://securityaffairs.com/162172/hacking/synlab-italis-cyberattack.html