Leanne Larson, MHA, Corporate Vice President and Worldwide Head, Real-World Evidence, Parexel International, USA*
Interviewed by Sally Hassan, PhD, ISMPP CMPP™, Parexel International, UK.
* The opinions expressed within are the author’s own and do not necessarily reflect the views of the author’s employer.
“What I Would Like You to Know” is an article series that shares perspectives and insights from functional area colleagues that collaborate with medical publication professionals on the planning and development of scientific publications. This article series will appear periodically in The MAP.
This article spotlights a real-world evidence expert’s perspective, presented in a question-and-answer format.
What are one or two things you want your medical publication colleagues to know from your experience of working on real-world evidence (RWE) publications?
First, research models and applications for RWE are quite different from clinical trials. Even the terminology involved is different ‒ for example, use of “patients” versus “subjects,” “studies” versus “trials,” “physicians” versus “investigators,” and “effectiveness” versus “efficacy.”
Second, RWE and real-world data (RWD) are not interchangeable with respect to their definitions. RWD reflects raw data gathered from instances of routine care, existing studies, and prospective studies, and are pragmatic to varying degrees. On the other hand, RWE reflects the transformation of data to generate evidence. For instance, when we access RWD directly from an electronic medical record, the data themselves are only “raw data” and aren’t useful to support decisions as is; they must be aggregated via appropriate medical-informatics practices, and then analyzed with specialized statistical approaches and within the context of the study objectives. Once this analysis and transformation are complete, the resulting evidence will be credible and meaningful, and able to address important questions and the goals of the study.
Third, RWD models can include different types of data: primary, secondary, or a combination of these (the hybrid model). Primary data are new data from studies, specifically designed to answer a key question (eg, prospective observational studies or pragmatic trials), while secondary data are pre-existing and have already been collected for another purpose (eg, claims databases and electronic medical records). In general, medical writers tend to think of RWE as data generated from secondary data, but historically, most RWE has been obtained from either primary or secondary data. To date, the hybrid model is arguably the most important model for RWE generation, as models based on primary or secondary data alone have important strengths and weaknesses. Secondary data are more readily accessible, but are reliant on already available information; therefore, these models may not precisely fit the purpose of the new research question. On the other hand, models based on primary data can be more relevant to the study question, and additional data can be requested for completeness. However, primary data are more complex and typically take longer to obtain, as the studies require a defined follow-up period. The hybrid model mitigates the weakness of each of these data sources, while capitalizing on their strengths, and is emerging as the most valuable model.
What is one improvement you would suggest for the development of RWE publications?
We must ensure that the analytical and strategic approach is consistent and appropriate for RWE. The statistical approach and study design need to be suitable for supporting the key study objective, increasingly for payor or regulatory audiences. Analytics in RWE differ from those for a clinical trial; statistical approaches have been developed, for instance, to address non-random populations, missing data, and other challenges inherent in RWD. The analytic approach must be described in the publication to ensure the audience appreciates that the resulting findings are relevant and accurate. Similarly, it is essential to ensure that all authors and reviewers are well-versed in the study objectives before the development of the publication begins. Given the many possible strategies, stakeholders, and goals of an RWE study, the publication must address the needs and perspectives of the appropriate stakeholder audience.
What is your biggest challenge when contributing to the development of RWE publications?
The biggest challenge is ensuring that the study objectives and design applied to generate RWE are well-described in the publication. Given the breadth of study designs and approaches in RWE, it is critical to ensure that the reader can appreciate the study’s purpose and the method used, to be able to apply the findings in their decision-making contexts.
What key recommendations would you give to colleagues working in RWE publications?
My key recommendation is first to understand who the intended audience is for the study (eg, payors have different needs than physicians) and to communicate the most relevant information for the stakeholder. It is vital to understand what the publication is trying to achieve, and that it should be about delivering value (not just data).
Secondly, there are guidelines (eg, STROBE and RECORD) that can help ensure everyone is on the same page, from an educational perspective. A useful resource is the Agency for Healthcare Research and Quality’s (AHRQ) Registries for Evaluating Patient Outcomes: A User’s Guide, third edition, which provides crucial guidance on patient registries and other RWE approaches.
Finally, I would reiterate the importance of using appropriate study designs, statistical approaches, and terminology in RWE publications.
Acronyms
RECORD, The REporting of studies Conducted using Observational Routinely-collected health Data;
STROBE, Strengthening the Reporting of Observational Studies in Epidemiology