Matthew W. Reynolds, Vice President, Real World Solutions, IQVIA
03.09.22
Recent FDA guidance on using real world data (RWD) to study the safety and effectiveness of products and treatments represents an exciting advancement for clinical research. Opportunities for RWD usage have existed for many years now and has grown more rapidly during the pandemic, driven by a need for immediate public health knowledge about COVID-19 from reliable near-real time data sources. The drafted guidance confirms the FDA’s interest in RWD and sets the stage for guiding principles on good practices for assessment of RWD acceptability and usage for future regulatory opportunities.
A big step forward
Issuing this framework fills a longstanding need in health and epidemiology research for how best to assure quality construction and usage of RWD. This data can play a critical part in generating evidence, as long as the data sources are thoroughly vetted, fit-for-purpose, and the datasets match both the specific research question and the specific regulatory need.
The global use of RWD has expanded over the past several decades and datasets have improved in detail, linkage, and timeliness. RWD is routinely being leveraged by the FDA, the pharmaceutical industry, and the larger public health community to gain insights and make critical decisions, and the framework is a big step towards doing so effectively.
The guidance lays out guardrails around the minimum expectation for supporting new initiatives, as well as how best to strategically think through researchers’ approaches to assessment and implementation of RWD studies. It also helps researchers identify core criteria to assess their data sources – including relevance, completeness, accuracy, and the validation of key elements like patient diagnoses, exposures, and outcomes. This suggests ways to improve data through linkage and supplementation, as well as ensuring that the dataset selected is the best fit-for-purpose for the specific research question.
Challenges to consider
The administration’s decision to issue this guidance is positive. However, it will not be without its challenges for researchers; the biggest being that the guidance refers to many study considerations but does not clearly specify the potential limitations involved in employing them. In the case of data validation, the FDA guidance mentions using unstructured data to get more information, define clinical context, and help validate exposures and outcomes. However, this is often not possible for many of the RWD sources that are currently available. It also outlines how to understand the potential misclassification of the variables of interest and statistically assess them. While these steps are important to acknowledge when characterizing a RWD approach and solution , they are not always possible for many RWD studies to execute.
Another challenge to consider is measurement of coverage continuity. The framework notes that this should be addressed in electronic health records (EHRs) and medical claims. However, there is typically no reliable measure of coverage continuity in EHRs At the same time, patient privacy issues are a standard consideration when linking a registry to another data source for supplemental information. Researchers should refrain from interpreting the guidance as purely black and white and remain flexible when it comes to selecting RWD sources and evaluating fitness-for-purpose.
Determining fit-for-purpose
No “perfect” database exists for research in general, and a database shouldn’t be broadly categorized as “good” or “bad.” A dataset is either fit for the purpose of a research question, or it’s not. While some data sources may be right for the majority of potential questions, others may be an exceptional fit for just one or two very specific questions.
Researchers should consider what context is most critical in a data source. As an example, a general practice EHR source without inpatient data would not be suitable for a study observing the risk of serious events that typically result in hospitalization. However, the very same dataset could be a relevant in a study of the factors that drive general practitioners to prescribe certain medications. Ultimately, the thoroughness of the vetting process will allow researchers to feel confident that they’ve chosen the best option for the research question at hand.
What comes next?
As the FDA implements this guidance and opportunities arise with pharma and medical device companies to evaluate RWD fit-for-purpose, new considerations will likely come to light. These points will drive expansions and further steps that complement the core guidance. Every study is different, making it challenging for the research community to set precedents.
The primary takeaway is this: after much anticipation, we have a regulator-approved blueprint on which to base decisions for RWD use. It can help with evaluating whether the evidence generated is reliable enough to complement randomized clinical trial data and offers potential for practitioners in the field to improve the quality of their research. This guidance is a stepping stone toward further refinement in future guidance. This is not the end of determining best practices for RWD use, but rather just the beginning.
Dr. Reynolds is a member of the Center for Advanced Evidence Generation, Real World Solutions at IQVIA. He designs innovative solutions for real world evidence on effectiveness and safety. Dr. Reynolds brings more than 20 years of diverse experience in non-interventional research and has a deep expertise in the usage of real world data to address questions of effectiveness, safety, and value of pharmaceutical products. He founded and led the successful Database Special Interest Group for the International Society of Pharmacoepidemiology where he served as an elected executive leader. Dr. Reynolds earned his graduate and doctoral degrees in epidemiology and preventive medicine from the University of Maryland at Baltimore in January 2000.
Source: https://www.contractpharma.com/contents/view_online-exclusives/2022-03-09/what-new-fda-guidance-means-for-the-future-of-real-world-data-use/