What can real-world evidence contribute to regulatory science in pre- and post- marketing setting?
Historically, information on drug safety and efficacy were derived from randomized clinical trials (RCT), although RCTs have several limitations. The need to generate evidence in a real-world setting has led to a massive growth in the field of observational research using real world data (RWD). Real word evidence (RWE) is already filling the gaps of pre-marketing clinical evidence through evidence generated in the post-marketing phase, becoming increasingly recognized as an essential element to inform healthcare decisions, both at the regulatory and clinical and organizational level. In the past decade, the role of RWE in regulatory setting has evolved significantly, consequently, several regulatory agencies such as Food and Drug (FDA) and the European Medicines Agency (EMA), have set up specific task forces to better define how RWD can be leveraged for regulatory purposes. RWE is currently recognized as a relevant source of information for conducting studies aimed at identifying and characterizing the risk of medicines in the post-marketing phase but, there is still a great debate on how to exploit this information in the pre-authorization phase. The role of RWE is of importance in settings where it may be difficult to identify a sufficiently large population to perform robust analyses, such as in pediatric drug safety or rare diseases, as well as in chronic diseases. The usefulness of RWE in both the pre- and post-marketing setting would justify the creation of national centers of excellence which can store and analyze RWD in a timely and appropriate manner. However, the creation of such centres of excellence must be accompanied by a greater of receptiveness of Health Technology Assessment and regulatory agencies. However, RWE also has its limitations. For example, RWD is usually generated for purposes other than research, as a result of which the quality and completeness of RWD is not always appropriate to address some research questions and there are limitations related to data accessibility.