"Ten Commandments" for Conducting Comparative Effectiveness Research Using "Real-World Data"

argue that RWD CER results based on nonrandomized data a priori compromises their credibility. Others may argue that clinical trials that target only regulators rather than post-regulatory decision makers, including patients, consumers, payers, prescribers, and policy makers are similarly, albeit differently, flawed because they are less informative for medical decision making than pragmatic clinical trials that address patient, prescriber, and payer concerns. In both randomized trials and studies using data with nonrandom assignment, the virtues of RWD CER results are more likely to be valued by appropriately skeptical audiences if decision makers are confident that the work has been conducted and reported with a dedication to high standards. In this spirit of devotion to good research practices for CER using RWD, we offer “ten commandments” for conducting and reporting CER based on analysis of RWD, without any claim of having received them from on high. The purpose of this article is to provide the beginning-to-intermediate practitioner or decision maker with a concise list of practices that are crucial to the proper execution of this kind of work. It is not meant to replace the growing literature which, in many cases, more extensively reviews important technical aspects of RWD analysis, and we strongly recommend that readers also review other guidance documents and Task Force reports, such as those published by the International Society for Pharmacoeconomics and Outcomes Research (ISPOR). However, we believe there is merit in a brief overview of some key tenets of the RWD research process, from planning, to analysis, to reporting, that combine general good research practices with considerations specifically relevant to CER with RWD. Before beginning, we strongly urge those who conduct RWD studies to involve those who are part of the RWD data generation and decision-making processes when designing CER studies. This will maximize the usefulness of the RWD CER results.