Real-world data (RWD) is an umbrella term for different types of data that are not collected in conventional randomised controlled trials. In the healthcare sector, RWD can be obtained from many sources and includes patient data, data from clinicians, hospital data, data from payers and social data. In order to account for these varied datasets in this study, and to analyse their collective promise, we offer the following working definition:
RWD is any data not collected in conventional randomised controlled trials. It includes data from existing secondary sources (e.g., databases of national health services) and the collection of new data, both retrospectively and prospectively.
In the view of many analysts and researchers, RWD has significant potential to improve the ways drugs are discovered and developed. Moreover, the assessment of the value of medicines and treatments in real-world settings may be made less resource intensive with RWD-based methodologies. Whilst the health and healthcare sector is an early adopter of such methodologies, the rate, direction and use of data generation is influenced by a series of factors: technological advances, for example, or data protection policy. The continuing impact of such factors will occur in ways that are difficult to predict and sometimes contradictory. The pace of technological change and the pace of change in governance arrangements, capabilities and in the building of relationships necessary to allow for successful generation and use of RWD do not necessarily move in tandem and this can create significant disappointment and frustration for all those involved. The computing technology underpinning the collection and use of data is advancing fast. Whilst necessary, these developments alone are insufficient for the successful use of data. To be of use to health researchers and innovators, and to be acceptable as evidence, data needs to be processed, analysed and presented in a coherent form. Its use and form also needs to be acceptable to the broad range of stakeholders involved in health innovation. The complex interaction of computing technology, the practice of health research and innovation, and governance and standards can make it difficult to understand what is happening in the field and what constitute enabling and constraining forces.
In that context, this study provides an overview of the use of RWD in health research and innovation in Europe and attempts to develop learning and understanding about its future potential in European research. This study was commissioned by Pfizer in October 2013 and is based on research carried out by RAND Europe and IBM over six weeks in the autumn/winter of 2013. The objectives of the work were the following:
- construct an evidence base on the establishment and evolution of standards governing the collection and use of RWD and identify the different ways in which standards have been applied
- understand the factors that have enabled or limited access to and use of RWD
- learn lessons from the use of RWD in different contexts
- identify opportunities for increasing access to RWD and contributing constructively to standard setting in Europe.
The study used a combination of approaches to survey and understand the current use of RWD and the potential opportunities for using this type of data going forwards. These approaches included a review of the academic and grey literature and also a small number of in-depth interviews with stakeholders from a range of public and private organisations in Europe. In addition, case study examples of the use of RWD in the healthcare sector were compiled. Finally, a workshop, engaging both the study team and individuals from Pfizer, brought together insights from these three streams of research to explore the options for the future and provide contextual information as to the feasibility of these opportunities.
By investigating the current forms and uses of RWD in Europe, this study has highlighted their significant potential for assessing the (short- or long-term) impact of different drugs or medical treatments and for informing and improving healthcare service delivery. Although the potential of RWD use seems quite clear, this research reveals barriers that restrict further development towards its full exploitation:
- the absence of common standards for defining the content and quality of RWD (absence of common terminology, incomplete datasets, lack of data quality assurance systems)
- methodological barriers (absence of standards for RWD analysis and for data linkage) that may limit the potential benefits of RWD analysis
- governance issues underlying the absence of standards for collaboration between stakeholders active in the field of RWD, and limitations of incentives for data sharing
- privacy concerns expressed predominantly by clinicians and patients and binding data protection legislation which can be seen to restrict access and use of data.
These issues are being addressed—although in a somewhat uneven fashion—by current initiatives from both public and private stakeholders at the regional, national and European scale. For example, the issues of data quality are being tackled through European and international initiatives aiming to improve the standardisation of terminology. Elsewhere, the development of international research coalitions is facilitating knowledge and best practice sharing and accelerating the development of common frameworks that guide RWD collection and use. This contributes both to the improvement of data quality and to researchers' analytical capabilities. In addition, a strong push towards the development of electronic health records—and eHealth infrastructures more broadly—has been observed in some European countries (Nordic countries, France, Belgium, the UK) and has been actively supported by various EU funding programmes. Such initiatives offer great potential for the automated and routine collection of patient data. Finally, access barriers—which are related to both governance and data protection issues and mostly faced by private companies—are being overcome through the implementation of strategic partnerships among stakeholders and the development of online consent management architecture. Engaging directly with academics on specific research projects, with physicians in exchange of technological and analytical services, but also with data vendors are examples of access strategies that have been extensively explored by private stakeholders in the healthcare sector.
The insights gained from the evidence can be summarised along the lines of a PESTL analysis, which explores the Policy, Economic, Social, Technological and Legal domains in relation to RWD. In each category we outline the main drivers, enablers, barriers and alternative approaches to use of RWD that have emerged from the analysis (Table 1).
Table 1. PESTL Analysis of the RWD Landscape in Europe
European Commission (EC)'s push for the development of eHealth infrastructures and use of EHR
EC's drive for the creation of Pan- European datasets and improved interoperability
National healthcare reforms aiming to greater efficiency in service management and provision.
EU funding instruments
Regional and National data infrastructure.
EC's data protection regulation
Fragmentation of national approaches to health reform
Disparities between national eHealth systems
Governance issues regarding the design and implementation of RWD standards.
Reliance on data collected in countries with easiest rules for access
Involvement in EU-funded research projects in partnership with relevant public and private stakeholders.
Resources constraints and need to develop efficient pathways to analysis
Incentives for collaboration to pool resources
Development of a market for data.
New synergies within the data value chain (e.g., with insurance companies)
National authorities encouraging data input.
Fragmented markets presenting different characteristics
Issues surrounding cost sharing for data access and use
Conflicts of interest.
Routine collection of publicly available data
Funding to academia for research in databases
Participation in research-minded consortia to spread the cost of data access and analysis
Engagement in disease specific research projects with direct access to self-reported patient data.
Increased familiarity with sharing data
Increased attention to the burden of a chronically ill and ageing society
Enthusiasm for new cures for illnesses
Willingness to access personalised health services.
Positive media coverage
Interaction with stakeholders (e.g., rare disease groups)
Practitioners care about improving outcomes for patients.
Increased suspicions about data use and potential breaches
Privacy risks due to linking different datasets
Regulation surrounding consent management
Image problem of pharmaceutical companies or insurers.
Development of personalised and stratified health services offer
Communication around the positive effects of RWD-based research.
Increased technological capabilities for data storage and analysis
Increasing capacity to link distinct datasets
Push towards standardisation of terminologies.
Machine learning, including natural language processing
National/patient identifier systems
Social media and apps for self-reported data collection.
Limits of analytical capabilities for the treatment of data
Inconsistency of existing databases and limited development of data quality insurance standards.
Leveraging methods and tools developed in other sectors
Exploration of the potential of apps/partnerships with device manufacturers.
EU level and national level debate on data protection, use and access.
Potential of using RWD to improve health services efficiency might influence existing regulation to facilitate data access
Technological advances reduce the burden of work for consent documentation collection.
Privacy and data protection likely to be strengthened
Ethical standards for research
Fragmented standards for access to databases.
Efforts on transparency and ethical commitments
Publication of RWD-based research results.
In this landscape, strategies that seek to optimise RWD access and use have to align interests of the three parties typically involved in healthcare: the organisations (payer, provider), the professionals (clinicians) and the patients. Indeed, a disproportionate advantage, or disadvantage between the three generally leads to slow adoption or refusal to change custom and practice. We therefore think that strategic partnerships between those stakeholder groups are key to defining better routes to access and improved use of data. A variety of collaborations can be developed to overcome existing barriers and facilitate RWD access and use, depending on the kind of data that is needed and the scope of their use. Those partnerships would rest on both non-monetary and monetary agreements and leverage a broad range of incentives at different levels of the health systems, from patient-level initiatives to collaborations with national health organisations.
The research described in this article was prepared for Pfizer Pharmaceuticals Ltd. and conducted by RAND Europe.