The Food and Drug Administration (FDA) has announced a new strategic framework to advance the use of ‘real-world evidence’ (RWE), using ‘real-world data’ (RWD) to support the development of drugs and biologics. The framework is described in greater detail here. The framework is not (yet) a new set of policies. But it is an important step toward the formulation of policies for the inclusion of observational data and research designs in the FDA’s evaluations. This is entirely welcome.
By RWD, the FDA means
data relating to patient health status and/or the delivery of health care routinely collected from a variety of sources.
By RWE, they mean
clinical evidence about the usage and potential benefits or risks of a medical product derived from analysis of RWD.
The re-engineering of the health care system to promote continuous learning from routinely-collected RWD is the heart of the National Academy of Medicine’s proposals for a Learning Health System. To this end, the FDA framework proposes to develop standards for RWD:
To work with RWD across multiple sources, data may need to be put into a common format, sometimes referred to as a common data model (CDM), with common representation (terminologies, vocabularies, coding schemes). FDA recognizes the importance of developing data standards to maximize the utility of RWD and is working on identifying relevant standards and methodologies for collection and analysis of RWD.
The framework will also develop standards for research designs that use RWD to generate RWE supporting clinical practice.
Under FDA’s RWE Program, evidence from traditional clinical trials will not be considered RWE. However, various hybrid or pragmatic trial designs and observational studies could generate RWE. FDA’s RWE Program will cover clinical trials that generate RWE in some capacity (i.e., sources other than traditional clinical trials) and observational studies.
TIE readers know that many researchers, clinicians, and engineers have invested decades of work in developing robust data standards and observational designs. So what can the FDA contribute here?
Implementing a Learning Health System requires that we — the health care community — solve a collective action problem. To get everyone in the system collecting data, sharing it, and using the evidence derived from those data, we need to agree on technical standards for representing data and principles for drawing inferences from it. Because the FDA is a gatekeeper of access to healthcare markets, its policies have great influence. By codifying data standards and observational research designs, the FDA can define and legitimize a template for how RWD should be collected and how RWE can be generated. It could solve one of the several collective action problems requisite to building an LHS.
It’s a distinct and — how should one put it? — an unusual pleasure to write a post praising a health policy initiative from the US government.
