Real-world data (RWD) has been heralded as a treasure trove of information to generate real-world evidence (RWE) which seeks to judge the comparative effectiveness of healthcare interventions when delivered through the course of routine care. RWD/RWE are already used in the US healthcare system in a variety of ways. Artificial intelligence systems comprised of machine–learning algorithms are used to predict the likelihood of an event for an individual patient and recommend treatment based on innumerable sets of archetype patient profiles created from RWD. At the same time, the slow, yet seismic, shift from a volume– to value–based reimbursement environment is using RWD to feed the reimbursement calculations for doctors and hospitals. Providers are recognizing that RWD and RWE are rapidly entering their clinical practice workflows and starting to affect compensation.
Our research explores how RWD and RWE are understood by providers and their perceptions of its utility. We have found that a vast majority (76%) of providers somewhat, or strongly, agree that RWE is necessary to augment findings from randomized controlled clinical trials (RCTs) given the well-publicized limitations of RCTs in oncology, wherein participants are younger, healthier, and less diverse. However, only 31% indicate that they somewhat or fairly often refer to RWE when making treatment decisions and only 2% indicated that real-world studies of efficacy or safety guide any clinical decisions, with more than 73% listing the NCCN Guidelines as the driver behind their recommendations.
Our research is not the first to explore these patterns and is supported by prior work. In a survey of 866 US physicians representing 27 specialties across all 50 states practicing in large multispecialty medical groups, physician hospital organizations or integrated health networks, physicians ranked RWE as the most important data for informing treatment decisions, followed by clinical trial data.1 Here we explore several potential explanations for the observed disconnect between the providers’ perception of the utility of RWD/RWE and its use in practice.
First, community oncologists may have a poor understanding of what exactly RWD and RWE comprise. A total of 41% of providers indicated that they used their personal clinical experience when deciding between treatment alternatives and 22% stated that they consult with their peers. Both their experience and that of their peers are what is captured in claims or transcribed into the electronic health records of patients and form the basis of RWD. More work is necessary to inform providers how observational, de–identified, and aggregate data analysis can be a methodologically sound approach to understanding patterns in outcomes and determining a treatment approach which may minimize bias.
Second, the setting in which physicians practice may not be conducive to the utilization of data. Nearly all of the surveyed participants practiced in community oncology clinics. This setting can be very different from an oncologist who is practicing in a closed integrated system, such as a managed care organization, or in an academic/teaching hospital facility. Within a staff model, managed care environment or integrated delivery network, oncologists and other practicing physicians have a role in making formulary decisions that impact treatment paradigms, which are heavily vetted using not only clinical trial data, but RWE studies as well. Essentially, all evidence is evaluated and sometimes, internal research is conducted to support treatment decisions.
Third, oncology providers may perceive a lack of robust standards and methods and absences of guidance from regulatory bodies regarding how RWD and RWE are gathered, analyzed and used, as a major downfall. Providers may not view RWD/RWE as a credible resource since they may feel that the data collected is in an uncontrolled manner with inherent biases and gaps. Additionally, they may perceive clinical, financial and litigious risk from utilizing these data. Together these perceptions create a significant barrier to the use or even acknowledgment of this data and related research. While clinical trials play a critical role in evaluating the safety and efficacy of drugs, rarely do these trials incorporate head to head comparisons of recent market entrants and even if they did, they would be limited by the highly selected participant pool. Comparative effectiveness can often only be judged outside of the trial in a real-world clinic setting. Thus, it is important to educate oncologists within the community setting that the ability to track more patients over a longer period of time is valuable and that RWE plays an important role in capturing evidence to demonstrate how a drug may affect patients and healthcare costs in real-world settings. For example, the expansion of the number of novel oral agents in the treatment of cancer has introduced the need for physicians to monitor adherence to therapy, a process unknown when greater than 90% of cytotoxic treatment was administered in the infusion center. Analyses of RWE, including pharmacy and administrative claims data, may help identify potential problem areas and enable pharmacists, manufacturers, and providers to support adherence by offering education, resources, and behavioral coaching.
Overall, consistent education related to RWD/RWE and a consistent message from organizations and institutions such as the FDA and professional societies may help to advance the science of RWD/RWE in healthcare and in particular for the delivery of care to cancer patents. First, robust training during medical school on theories of causal inference, bias, and data interpretation may aid providers in creating their own personal litmus test for the types of information they personally may use. At the same time, both independent scientists, professional societies, drug and device manufacturers and the U.S. Food and Drug Administration need to work collaboratively to finalize guidance on methodological approaches and intended uses of RWD/RWE. In August 2018 it will have been two years since the promulgation of the 21st Century Cures Act; and while the FDA has issued guidance on the use of RWE in regulatory decision making for medical devices, the pharmaceutical segment has yet to receive any such guidance. Finally, education campaigns and buy-in from professional societies such as ASCO will be needed. The CancerLinQ project initiated by ASCO to leverage real-world data should be further developed and implemented.
RWD and RWE promise to deliver timely and cost-effective insights when compared to a clinical trial–we should all hold ourselves accountable to that viewpoint by delivering robust methods and recommendations in a time-sensitive manner to truly leverage RWD/RWE for patients today and tomorrow.