“The real world doesn’t reward perfectionists. It rewards people who get things done.”
Ziad K. Agdelnour, Lebanese-American financier and author
Real world data (RWD) consist of data sources that are not in conventional randomized controlled trials; the myriad of data sources therefore can include: patient data, data from clinicians, hospital or health system data, data from payors (including claims and billing data), data from social media, data from public health agencies and disease registries, and more currently data from health apps, wearable devices and sensors. In addition, the 2016 US Cures Act that is designed to accelerate medical product development opened the possibility of utilizing these sources of data in clinical decision making. Lastly, RWD can be used for evidence generation regarding safety and effectiveness. Just like healthcare data in general, however, there are challenges of RWD that include: lack of standardization of quality and type of RWD, fragmentation of RWD sources that can lead to access issues, accuracy and reliability of RWD, not to mention privacy and security concerns of RWD (that may be even more than clinical trial data).
Real world evidence (RWE) then is defined as the clinical evidence regarding the usage and potential benefits or risks of a medical project derived from analytics (including machine and deep learning) on the aforementioned RWD. RWE can have a myriad of trial designs (randomized trials, pragmatic trials, observational studies, etc) and be part of a hybrid design of a clinical trial that partly depends on RWD and RWE. Of note, the FD&C Act Section 505(c) encourages use of RWE for product effectiveness to help gain approval of new indications for drugs.
The future mandates a higher plane of significance for RWD and RWE for several reasons. Among these is that regulatory agencies such as the FDA (Real-World Evidence Program) have been in favor of using RWD and RWE more strategically in order to increase postmarket surveillance of medical devices and drugs for safety, especially adverse events and negative outcomes. In addition, the next decades will see the emergence of internet of everything (IoE) with embedding of artificial intelligence tools into wearable devices as well as sensors (“tinyML”). The health data from these relatively newer sources will continue to increase exponentially in volume as well as in sophistication (data basis for “multimodal” AI). Lastly, the surge of interest in population health and in particular, the focus on health equity, will emphasize the importance of RWD and RWE as proportionally more data will be coming directly from sources outside of clinical trials and health systems, with special attention to social determinants of health.