Integrating Real-World Evidence in the Regulatory Decision-Making Process: A Systematic Analysis of Experiences in the US, EU, and China Using a Logic Model

Abstract

Real world evidence (RWE) and real-world data (RWD) are drawing ever-increasing attention in the pharmaceutical industry and drug regulatory authorities (DRAs) all over the world due to their paramount role in supporting drug development and regulatory decision making. However, there is little systematic documentary analysis about how RWE was integrated for the use by the DRAs in evaluating new treatment approaches and monitoring post-market safety. This study aimed to analyze and discuss the integration of RWE into regulatory decision-making process from the perspective of DRAs. Different development strategies to develop and adopt RWE by the DRAs in the US, Europe, and China were reviewed and compared, and the challenges encountered were discussed. It was found that different strategies on development of RWE were applied by FDA, EMA, and NMPA. The extent to which RWE was adopted in China was relatively limited compared to that in the US and EU, which was highly related to the national pharmaceutical environment and development stages. A better understanding of the overall goals, inputs, activities, outputs, and outcomes in developing RWE will help inform actions to harness RWD and leverage RWE for better health care decisions.

The effect of adding real-world evidence to regulatory submissions on the breadth of population indicated for rare disease medicine treatment by the European Medicines Agency

“…An approach could entail enrolling such patients accessing treatment via early access schemes, compassionate use programs [ 18 ] and named patient requests. Although organisational aspects of the inclusion of RWE at the regulatory stage have been less well-studied, initiating its generation in conjunction with early access schemes could exploit lead times with minimal effect on overall duration of regulatory engagement [ 19 ]. With the exclusion of heterogenous groups with multiple comorbidities from rare disease RCTs to ensure the internal validity of findings, RWE generated from such groups could be a complementary addition during the regulatory stage [ 19 ].…”

“…Although organisational aspects of the inclusion of RWE at the regulatory stage have been less well-studied, initiating its generation in conjunction with early access schemes could exploit lead times with minimal effect on overall duration of regulatory engagement [ 19 ]. With the exclusion of heterogenous groups with multiple comorbidities from rare disease RCTs to ensure the internal validity of findings, RWE generated from such groups could be a complementary addition during the regulatory stage [ 19 ]. However, selection bias and type I errors arising from small, non-representative cohorts in RWE are risks that must be addressed when determining the suitability of RWE for regulatory use and the use of statistical methods to address patient population differences should be considered [ 20 ].…”

Advancing the Regulation of Traditional and Complementary Medicine Products: A Comparison of Five Regulatory Systems on Traditional Medicines with a Long History of Use

“…Sentinel is a classic example of postmarketing monitoring system developed by the U.S. Food and Drug Administration in 2008 which collects and pools together data about drug usage by more than 300 million people for high-level analytics [ 87 ]. With the advent of the big data era, the innovative strategies of combining real-world study with clinical trials may present as new paths for clinical research and development, as well as clinical effectiveness and safety evaluation, thus supporting drug regulatory decision-making about new product applications as well as indication expansion [ 88 ]. In China, the importance of real-world evidence was to fill the research gaps due to the limitations of traditional TCM clinical trials and to address the need for additional evidence to inform policy decisions [ 89 ].…”

Real-World and Regulatory Perspectives of Artificial Intelligence in Cardiovascular Imaging

“…There is an intersection with the vertical MDR, though. 69 Mandatory requirements for high-risk AI applications are intended. Due to the new classification rules 11 (Medical device software) and 22 (closed-loop systems) in Annex VIII, MDR 70 medical devices based on or integrating MLMD are likely to be classified as II a or even higher.…”

“…There are no specific, legally binding requirements for MLMD in Europe at the moment, but notified bodies will 68 https://eur-lex.europa.eu/resource.html?uri=cellar:e0649735-a372-11eb-9585-01aa75ed71a1.0001.02/DOC_1&format=PDF and https://eur-lex.europa.eu/ resource.html?uri~=~cellar:e0649735-a372-11eb-9585-01aa75ed71a1.0001.02/ DOC_2&format~=~PDF accessed 22.2.2022. 69 https://portolano.it/en/blog/life-sciences/artificial-intelligence-and-medicaldevices-the-potential-impact-of-the-proposed-artificial-intelligence-regulationon-medical-device-software accessed 27.4.2022. 70 Rule 11 of Annex VIII MDR.…”