As regulators and payers demand stronger evidence of product value, there is increased pressure on companies to consider more innovative approaches for drug development. The abundance of new data sources and advances in technology has created an opportunity for increased speed and efficiency in data collection and analysis, allowing for new and valuable insights for decision-making and communication of product value.
pharmaphorum spoke with Dr Radek Wasiak, VP and global head of peri- and post-approval studies at Evidera, a PPD business unit, to get his perspective on the changing dynamics and impact of data generation and analysis in the world of healthcare.
I have a different perspective on this issue. Although there may be organisational disruption due to development of new ways of working, the explosion of big data is most likely enabling the pharma industry rather than causing a disruption. The routine use of and access to electronic medical records, as well as technological advances allowing for the safe storage and sharing of data, have created an opportunity to inform the drug development process. While clinical trials provide the ‘gold standard’ for establishing treatment efficacy, they are conducted in small populations under tightly controlled conditions, and alone are insufficient to guide clinical practice. As a result, real-world data (RWD) have become important components for healthcare decision-making. By operating outside the confines of clinical trials, RWD can be used to understand treatment effectiveness and provide insight into post-launch drug safety and outcomes.
However, working with these kinds of ‘big data’ requires an understanding of how information is collected and coded, so effective strategies can be developed for analysis and insight generation. For example, variation in disease coding (or failure to code) requires focused consideration on the part of researchers using these data. These are just a few elements that will change how the pharmaceutical industry operates.
It’s safe to say the Information Age has a large and diverse impact on the healthcare system, including various dimensions of drug development and patient care, and has enabled the possibility of more real-world evidence (RWE)-based decisions on the part of drug developers, regulatory agencies, clinicians, health plans, and patients. The rapid and continuous development of information infrastructures and capabilities have resulted in an explosion in the amount and quality of RWD, and linkages that have expanded the possibilities for how RWD can be built into RWE to inform decisions across the healthcare system.
Although conventional randomised controlled trials (RCT) remain the gold standard for regulatory submissions for marketing authorisations across the globe, they come with a number of limitations. These issues have raised the need for research with a more pragmatic focus designed to answer a somewhat different set of questions, directed at the real-world effectiveness and safety of interventions. The goal has shifted to not only bringing to market safe and efficacious interventions, but those for which enough evidence exists that patients will ask, providers will prescribe, and payers will pay. To meet this goal effectively, RWE is needed throughout the development cycle.
Importantly, regulators including the US Food and Drug Administration (FDA) and the European Medicines Agency (EMA), as well as other stakeholders such as the National Institutes of Health (NIH) and the Patient-Centered Outcomes Research Institute (PCORI), are active participants in a number of efforts aimed at the incorporation of RWE from pragmatic trials into regulatory decision-making. Multi-stakeholder bodies like MIT’s New Drug Development Paradigms (NEWDIGS) or non-profit organisations such as the International Consortium for Health Outcomes Measurement (ICHOM) are also pushing the envelope, with thinking about evidence-generation ecosystems (NEWDIGS) and standardisation of outcomes (ICHOM).
Solving the puzzle of how automation impacts the scientific rigour related to studies using RWD. In particular, the use of machine learning (ML) in biomedicine is increasing at a rapid rate; however, there is a substantial and well-documented division of opinion on its value. Critics argue that the methodology is not transparent and too often leads to mistaken inference, calling it the new alchemy. Supporters argue that ML is a potential game-changer and enables powerful insights when traditional approaches fall short.
Although common in other industries and widely used in the drug discovery process, the application of ML to healthcare data presents several specific challenges, particularly when it is used as a diagnostic tool. Considering the current RWD explosion, the benefit of ML is potentially substantial; however, we also need to acknowledge the potential for incorrect application and the subsequent likelihood of mistaken insights. Additionally, the 2018 European legislation on data protection (the General Data Protection Regulation or GDPR) will potentially prohibit some current uses of ML (e.g. automated individual decision-making and profiling).
There are a few areas where we expect to see further innovation, and they are consistent with our areas of focus. First, data access – we are moving beyond accessing only structured data sets toward also working with unstructured healthcare data. Our advances in ML focus specifically on establishing a credible way of extracting value from social media, which allows access to constantly updated and unfiltered patient opinions, and clinical notes, providing rich clinical data. Secondly, we believe virtualisation is the next wave of clinical innovation, with technology enabling movement away from site-based studies. We have seen an increase in interest from our research partners and have several recent examples of studies that have approached data generation in this way. Finally, how the data are used will change. For instance, we believe patient e-cohorts, bringing different data types into data repositories, will enable RWD to support clinical development operations. Our work with MIT NEWDIGS and the Innovative Medicines Initiative’s (IMI) IMPACT-HTA initiative are examples of leading the way to redefine the evidence generation environment and infrastructure.
There are myriad impacts that can be seen with better, faster, more complete data, such as better investment decisions, reduced timelines, and cost savings. One example is regenerative medicine, where there is greater scrutiny by regulators and payers, including the requirement to show evidence of unmet need and the potential to demonstrate a transformative impact of treatments. RWE techniques have never been more important in painting a complete picture to get regenerative designation, fast-track status and a strong value story that is sufficient for all stakeholders.
In response to the growing demand for RWE, many pharma companies have restructured and created cross-functional RWE teams focused on optimising RWD use. Those who have not will need to rely heavily on the expertise of external partners to design and conduct RWE studies.
In general, companies should begin their evidence planning early, and identify and engage all relevant stakeholders to ensure that their evidence needs are considered. Thoughtful attention should be given to understanding what data exists, what data needs to be collected and/or supplemented, and ensuring mechanisms are in place to effectively capture the patient voice throughout the process.
While RWE can provide value at all stages of the product lifecycle, if not used properly it can also misinform, leading to loss of credibility and negatively impact commercial success of the drug.
It is important to consider that not all RWE is created equally, and defining the strategy that can support a strong case for a company’s product is dependent on a number of factors, including, but not limited to:
Reliance on existing data collection methods (claims data or electronic health records), confidentiality protection or data ownership laws, and the rules of research funding are all constraints to broader adoption. However, they are completely understandable given the investment needed to generate patient-level information in a longitudinal way.
Are we therefore stuck with an imperfect system? Not necessarily, as there is a growing recognition that improvement is needed. IMI’s GetReal was one of the first initiatives that recognised the need to do better. It aims to show how robust new methods of RWD collection and synthesis could be created and considered for adoption earlier in pharmaceutical research and development, and the decision-making process. Previously mentioned initiatives at NEWDIGS or newer IMI initiatives (e.g. IMPACT-HTA) are taking this work further. It is the collaboration of multiple stakeholders involved in RWE generation that is the first of the necessary conditions for improved RWD. Development of research standards, making more complex and robust approaches part of these standards, and ensuring comparability of findings through the use of tools, such as a common data model, should be the desired outcomes of increased collaboration.
The second barrier is an extension of the first one – the lack of knowledge required to interpret findings from RWD. RWE should produce stakeholder-relevant, and in particular, clinically-relevant, outputs. The drug development process has become increasingly multidisciplinary, and issues related to market access and HEOR (health economics and outcomes research) supporting market access are now discussed early as part of multifunctional brand teams. This increased visibility, which involves RWE generation, requires going beyond technical delivery. For instance, RWE translational research should help clinicians understand what to expect from RWD studies, what constitutes good research practices, and how clinicians can engage in the design and interpretation of these studies. These activities would only improve the value of the generated evidence.
Another barrier is time. We need faster research outputs. In many cases, it takes more than a year to get access to data and this inevitably has a negative effect on research quality and impact. Analyses of real-world treatment patterns or outcomes are often outdated by the time they are made public. Post-authorisation safety studies, focused on confirming the drug is safe in actual clinical practice, are prone not only to challenges associated with a new drug gaining the necessary market share but also with data availability delays, causing them to run for several additional years. This also extends beyond studies involving existing data sources, as data collection studies typically do not produce results for several years, with frequent interim analyses often being cost prohibitive.
Many of these challenges cannot be easily overcome and will require non-research solutions, but there are some steps that pharma can take, as we have discussed: an increase in the use of technology and automation to speed up data management and analysis.
We recently had a need to gain rapid and repeatable access to RWE to support the use of immuno-oncology (IO) products in the UK. We developed tools and a framework, which was validated against a traditional chart review approach, to extract and analyse data directly from an oncology prescription platform. The results identified 30% more patients compared to the traditional methodology and reduced the time until we received the data from more than three months to 15 days.
This is just one example of how embracing the new technology and using available resources in more productive and innovative ways can lead to real savings.
Radoslaw (Radek) Wasiak, PhD, MA, MSc, is a member of Evidera’s executive leadership team and the vice president and global head of Peri- and Post-Approval Studies – a global team of scientists and research operations professionals responsible for planning and executing real-world evidence (RWE) and interventional studies. Prior to this role, Dr. Wasiak served as general manager of the RWE and Meta Research practice areas, focused on generating and synthesising real-world and clinical evidence. He also formerly held the position of EU director, with operational and business responsibilities for the growth of European service offerings. Before joining Evidera, Dr. Wasiak was a research scientist at the Liberty Mutual Research Institute for Safety, where he developed an extensive expertise with multi-level claim and secondary databases as well as with different aspects of research project quality review. Dr. Wasiak has presented at numerous conferences and his work has been published in multiple scientific journals. Dr. Wasiak holds a PhD and an MA in economics from the University of Connecticut, as well as an MSc in finance and banking from the University of Economics, Poznan, Poland. He has held adjunct appointments at Trinity College (USA) and the Harvard School of Public Health.