What regulators expect of patient experience data
Patient experience data is the form in which patients' own experience and preferences reach a regulator or a health technology assessment body, and the European Medicines Agency's reflection paper sets out what it may be drawn from and what level of evidence it has to meet.
What patient experience data is
Patient experience data are data directly reflecting patients' experience or preferences on treatments or outcomes, without any interpretation by a clinician or anyone else.
Patients' perspectives on medicines and their benefits and risks are of great value to the European Medicines Agency (EMA). Patient experience data give regulators and other decision-makers in healthcare systems insight into what matters most to patients, beyond other well-established scientific outcomes. Cancer medicines are an example, since patients may in some cases prioritise quality of life over traditional clinical endpoints such as overall survival. Patients' lived experiences, reflected in quantitative or qualitative patient experience data, can complement the other data currently collected, notably through clinical trials, and further inform regulators' decisions on medicines.
The reflection paper
The EMA published its draft reflection paper on patient experience data for public consultation on 29 September 2025. The consultation was open until 31 January 2026, and comments were to be sent using the template provided to PED_RP@ema.europa.eu. The paper is intended for medicine developers, patient groups, researchers and other decision-makers. It was produced by a multidisciplinary group of experts from the EU regulatory network, including representatives of patient and consumer organisations. Feedback from the consultation will help shape the EU approach to patient experience data, with the ultimate goal of further improving evidence generation and optimising future medicine development towards the outcomes that matter most to patients.
The patient experience data initiative
The European Union's initiative on patient experience data has involved multiple stakeholders from its beginning. It takes a holistic view, aiming to include patients in every step of the development cycle, even after a product is approved, when a further indication may be registered. Patient experience data is meant to serve both regulatory purposes and health technology assessment, and the reflection paper confirms that it can be relevant to the decision-making of both kinds of body. In one regulatory professional's view, the current cultural and societal environment in Europe would support its introduction. The same professional holds that the initiative should leverage the scientific benefits already established from patient engagement in clinical trial design, such as patient-reported outcomes and, more recently, patient preference data.
What the paper asks of developers
The paper encourages developers to gather and include data reflecting patients' real-life perspectives and preferences throughout the lifecycle of a medicine, during pre-authorisation, benefit-risk evaluation and post-authorisation, for regulatory and health technology assessment purposes, and describes general principles for generating, collecting and analysing them. It states clearly that patient experience data is not a firm requirement for any marketing authorisation, so a dossier can still be submitted without it. One regulatory professional finds this somewhat disappointing, though possibly only because the paper is the EMA's first step in the area. To support developers, the EMA offers early interaction platforms, such as scientific advice and the qualification of novel methodologies, to discuss development plans and proposals for regulatory submissions.
Where patient experience data comes from
Patient experience data can be provided through patient-reported outcomes, patient preference studies and other information obtained through patient engagement activities. The paper repeats the familiar concept of patient-reported outcomes but also opens the door to accepting patient preference studies, although the EMA states that it has little experience of handling such studies in a regulatory context. The data can be obtained from clinical trials, from real-world data such as safety surveillance systems, and from sources yet to be fully validated and used, including mobile health technologies and social media. Social media is viewed especially critically as a source, because the quality of its data can be very uneven and it still carries a significant degree of bias.
The level of evidence a regulator expects
Patient experience data and patient preference studies have been familiar terms for years, but, in one regulatory professional's experience, clients struggle to understand what level of data they need to bring to a regulator. They ask whether a patient preference study must reach the quality of a randomised clinical trial, whether it resembles a real-world evidence study, and what quality and level of detail its write-up needs. In that professional's view, the reflection paper gives some, though not all, of the answers. The EMA acknowledges that it has limited experience in handling such data, so the paper is an invitation to industry and academia to learn together how to handle the field. It provides only first steps towards clarifying some aspects.
The paper states that patient experience data can have certain limitations and is, in the EMA's view, associated with relevant missing data. From a statistical point of view, one regulatory professional holds, this limits the value of the data unless the study design is very thorough. The paper clarifies that the quality criteria for the data are the same as for a traditional randomised clinical trial. It states that the best case is a randomised clinical trial in the pivotal programme that includes elements of patient experience data, because that gives the best quality of data. It also states clearly that patient experience data reaches a product label by no abbreviated route: as with any other data, it must be statistically compelling.