Limitation and possible additive value of real world evidence in comparison with randomized clinical trials

Overview

The results of randomized clinical trials (RCTs) are the gold standard needed to approve new drugs and the new treatment indications. Their limitation is that they do not fully reflect the population of all patients in the real clinical practice due to the limitations present in the inclusion and exclusion criteria. On the other hand, the randomization process ensures an almost perfect balance between known and unknown factors that could influence the outcome of the study. In recent years, studies analyzing patients from Real World Evidence (RWE) have been increasingly published, aiming to analyze more broadly defined patient populations than those tested in RCTs. Moreover, these studies do not require a longer period of time to conduct, as they are retrospective analyses based on electronic health record databases, insurance company data and other patient registries. The randomization process is replaced by the statistical method of propensity score matching. This method has its limitations, as the factors that are taken into account when creating similar pairs of patients are defined by the investigators based on their knowledge of known confounders, i.e. those factors that may influence the outcome of the study. Thus, unlike RCTs, RWEs are also unable to stratify the study population according to previously unknown confounders. The results of several RWEs were not later confirmed in RCTs. Also, RWE-based studies conducted after RCTs, and performed on a less risky population than in RCTs, often observed a larger effect of the investigational drug than in RCTs. This finding contradicts well-known observation that the effect of a drug is greater when studied in the population with higher risk. Hence, the data obtained by such studies can be considered complementary to those obtained in RCTs. Their additive value can be used to rationalize the change of the dose of the drug or the route of administration, as well as the use of the drug in a new patient population. RWE is not suitable for demonstrating the effect or safety of a new drug, nor for comparing different groups of drugs in relation to predefined study results. These results can be considered at most hypothesis-generating and require confirmatory RCTs or indirect comparison of groups of drugs using a meta-analysis.

Keywords:

antidiabetic therapy – GLP-1 receptor agonists – randomized clinical trials – real world evidence studies – SGLT2 inhibitors