Clinical Evidence 2030

Abstract

Excellence of clinical evidence is the heart of every well-informed decision on the development, authorization, reimbursement, use, and monitoring of medicines.

While healthcare decision makers continue to be confronted with unmet medical needs burdening patients and society at large, the slow speed and high cost of medicines development hinder new treatments reaching the patients who need them.

But the healthcare landscape in Europe is evolving and the convergence of several factors now provides the opportunity for a stronger and more sustainable approach to clinical evidence generation. The COVID-19 pandemic has shown the potential of new ways of working, with better collaboration between stakeholders and different approaches for evidence generation and evaluation. The changing policy environment in Europe, including the new legislation on a European Health Data Space (EHDS)¹ and the reform of the EU pharmaceutical regulation,² offers opportunities through greater healthcare data access, innovation in study designs, and use of advanced analytics. Increasing patient involvement in all aspects of evidence planning and healthcare decision making will further strengthen medicines development.

We highlight below the six guiding principles for excellent clinical evidence generation.

Clinical evidence is generated for patients' needs and public health. Through their engagement, patients provide critical insight into their medical needs and what really matters to them at every level of healthcare decisions. Clinical evidence generation should revolve around these needs. Patients have been increasingly involved in healthcare decisions, including those related to the evaluation of the benefit–risk of medicines by regulators, where patients bring their personal experience, knowledge, and expertise both on the conditions and the available treatment options, and also on the impact of regulatory decisions on their lives.³

Efforts are ongoing to guide the generation, collection, and use of patient experience data to support decisions on the development and benefit–risk evaluation of medicines. To further build on these efforts, multi-stakeholder collaboration in this field is encouraged.

Clinical evidence generation is planned and guided by purpose, data, knowledge, and expertise. When formulating research questions and designing clinical evidence programs, existing data, information, and knowledge should be leveraged. Currently, this is not always the case, and clinical studies may be planned ignorant of previous study results or learnings from other medicinal products. To enable this informed approach to clinical research, access to data, information and knowledge, including study protocols and results, reports on suspected adverse reactions and the outcome of regulatory assessments should be made publicly available and scrutinized when designing studies. Multi-stakeholder dialogue at the planning stage will also facilitate access to existing knowledge. In this way, past successes and failures inform identification of gaps and further clinical evidence generation and may avoid unnecessary duplication.

A “research question”-driven approach, clearly articulating one or several research questions, strengthens clinical evidence generation through optimal study design, particularly if supported by scientific advice that involves different stakeholder voices.⁴ This approach should drive the selection of the suitable study method, with some questions best answered through experiments (including clinical trials) and others through observations (including non-interventional studies).

Scientific advice and regulatory support for research and development (R&D) are key to better evidence generation, and should be strengthened, further rationalized and simplified.

Different clinical research frameworks exist for the formulation of scientific research questions. These include the PICOT framework (Population-Intervention-Comparator-Outcome-Time), historically used in the field of pharmacoepidemiology research, and the Target Trial Emulation framework,⁵ which allows clear specification of the population, intervention, comparator, outcome and time horizon for the research. The increasing understanding and use of the Estimand framework,⁶ already used to optimize clinical trial design, complements existing frameworks by stimulating further specification of research questions through considerations of intercurrent events.

Research questions drive the choice of data and methods for evidence generation, and the totality of evidence is systematically considered for decision making.

Randomized clinical trials remain the core of clinical evidence, and they should be optimized to become better, smarter and faster. Several tools have recently been put in place in the EU to achieve this. One example is the Accelerating Clinical Trials in the EU (ACT EU) initiative.⁷ Building on the momentum of the Clinical Trials Regulation (CTR) and the Clinical Trials Information System (CTIS),⁸ ACT EU seeks to transform the way trials are initiated, designed, and conducted, while better integrating clinical research into European health systems.

Evidence generated through the analysis of real-world data (RWD) should have its value established across the full spectrum of research questions.⁹ To achieve this, we need to continue building processes, setting standards, enabling faster and broader access to data, and validating methods used for collecting and analyzing data. One of the available initiatives in Europe is DARWIN EU®,⁹ which delivers a high number of quality studies, by leveraging a federated network of RWD sources, transformed in a common data model, and on which a set of standardized analytics are applied.

We also need to continue training different stakeholders to increase collective knowledge, and harmonizing approaches to account for the international environment. In this context, it is important to maintain efforts to enhance EU network skills as well as ongoing international collaborations on clinical trials and RWE, notably through the work of the International Council for Harmonization (ICH).

Some additional enablers, captured in the Big Data Steering Group (BDSG) workplan 2023–2025,¹⁰ include the data quality framework and work on data discoverability through catalogues of RWD sources and studies. More European initiatives were set up to further support the use of RWD by addressing issues of data standardization, data quality, and biases. Examples include the Vaccine Monitoring Platform¹¹ and public private partnerships, such as the GetReal project¹² and the European Health Data & Evidence Network (EHDEN) work on data standardization.¹³

Artificial intelligence is another important enabler that needs to be integrated across the spectrum of clinical evidence, including in regulatory processes, thanks to validated algorithms, established processes and trainings, and internationally aligned methods.

Finally, recent progress in European healthcare legislations offers significant opportunities for an even more optimistic future in Europe. The EHDS legislation aims to favor a more rapid, wide and deep access to healthcare data in the EU, and its implementation foresees extensive technical work, including on data quality and utility label.¹⁴ Furthermore, the reform of the EU pharmaceutical legislation will make medicines more available, accessible, and affordable in the EU thanks to more agile legislation, which includes submission and analysis of clinical study data and greater use of RWE.²

Early and integrated planning and execution of a clinical evidence strategy is critical to allow different stakeholders involved in the sequence of the healthcare decisions to leverage the totality of generated clinical evidence for their decision-making processes. This requires anticipation and close collaboration to facilitate a common understanding of (1) how unmet medical needs are identified by patients and their caregivers, (2) how medicines developers translate these needs are into an R&D plan, (3) how this plan is critically reviewed via scientific advice by both regulators and health technology assessment (HTA) bodies and ultimately, (4) how this reviewed plan informs the analysis of clinical trials data, RWD and patient experience data, resulting in a wholistic evidence package for all stakeholders to leverage.

Pre-approval joint advice procedures are successful⁴ and highly valued by stakeholders. Collaboration on post-licensing evidence generation requirements shows potential, with pilots requested from different decision makers being conducted through DARWIN EU®.

On the safety side, pharmacovigilance should be based on smarter collection and reporting of suspected adverse reactions, with RWE enabling further evaluation of on-market performance of key new medicines, and improved engagement between regulators, patients and healthcare professionals allowing to optimize risk minimization measures.

Ultimately, following these principles in a transparent and collaborative manner will reinforce trust and willingness to contribute from all stakeholders, especially patients and healthcare professionals. All stakeholders have a collective responsibility to ensure that evidence reaches the expected level of standards that society needs for it to support impactful decisions. Clinical evidence then becomes a trusted, powerful, and efficient instrument for bringing better medicines to patients faster.

Applying these six principles for clinical evidence generation will result in increased efficiency and effectiveness for R&D in Europe; it will help bring medicines faster to those that benefit from them and energize the EU as a center for clinical research. Applying these principles will engage and empower patients; enable better data-driven decision making, meet the needs of multiple decision makers and ultimately result in excellent clinical evidence.

No funding was received for this work.

The authors have no competing interests as defined by the American Society for Clinical Pharmacology and Therapeutics, or other interests that might be perceived to influence the results and/or discussion reported in this paper.

The views expressed in this article are the personal views of the authors and may not be understood or quoted as being made on behalf of or reflecting the position of the regulatory agency/agencies or organizations with which the authors are employed/affiliated.