Home > All articles > Registry data to support clinical trials

Registry data to support clinical trials

Randomized clinical trials (RCT) have traditionally been held up as the gold standard for studying the efficacy of a treatment. Recently, real-world evidence (RWE) has been pushed alongside the RCT. Both of the methods have their strengths, which is why RCT and RWE have their place in the life cycle of a drug. Those who utilize both RCT and RWE are surfing on the wave’s crest.

Rekisteritiedosta täydennystä kliinisille kokeille - Terveyttä datasta | Registry data to support clinical trials - Health from Data

RCT and RWE are methods utilized in medical research. Both can be used as an example to demonstrate the efficacy or safety of a medicine. However, both methods have pros and cons, which are the topic of this posting. In addition, I will blog on how RWE can be used to supplement the data obtained from an RCT.


Clinical trials provide information on the efficacy or other intervention of the drug in relation to control therapy. The set-up of an RCT is experimental, and the research design is controlled in many ways. This ensures the safety of patients, the scientific quality, and the reproducibility of the research. However, the generalization of the results can be questioned.

The real-world data (RWD) utilized in the RWE study has been routinely recorded during the patient’s treatment. As the name “real-life evidence” indicates, RWE describes the treatment in real-world clinical practice. The data used for the RWE study has been collected retrospectively, so it cannot be controlled. The sex, age, medication, stage of disease, and comorbidity of the study subjects vary.

The patients participating in an RWE-study may have several ongoing treatments. Thus, it is impossible to analyze a specific treatment and a control therapy, as in RCT. Furthermore, the study’s retrospective design prevents the study of the effect of a drug that has not yet been in clinical use.

The recording practices of an RCT are agreed upon in advance, which makes them consistent throughout the study. On the other hand, RWD is recorded by numerous professionals at different times and sites, which is why the recording practices may differ or contain errors. Therefore, quality control plays an essential role in RWE research.

An RCT could be compared to an experiment carried out by a zoologist in a laboratory, where factors affecting animals – such as the length of the day, nutrition, and temperature – are controlled. Using the same analogy, an RWE study would be an experiment in the animal’s natural habitat, where variables cannot be influenced.

Big data

In Finland, data recorded in social and health registries can be used for research purposes with a data permit, which is based on the Act on the Secondary Use of Social Welfare and Health Care Data. Thus, there is no need to recruit participants for an RWE study. Furthermore, the RWE study covers all individuals stored in the register retrospectively, meaning that RWD exists already at the beginning of the study – without the need to record new information.

However, recruitment and follow-up of the study subjects may be a challenge for a prospective RCT. The patient may be wondering whether they should participate in the RCT or not. As an example, they may not want to end up in a placebo-treated group.

A large amount of data corrects uncertainties in the RWE study, such as errors in RWD recording. Thanks to the number and the origin of the data, machine learning can be utilized for RWD.

Let’s work together
RWE can supplement an RCT which is focusing on the efficacy of medicine. Here, RWE may be provided on the epidemiology, mortality, disease progression, or treatment costs of the patient group targeted, to mention a few possibilities.

“A virtual clinical trial combines RCT and RWE research.”

Sometimes an RWE study is conducted because an RCT would be an unethical option. In cancer treatments, the control arm of a clinical trial cannot usually be carried out because patients cannot be left without treatment.

In this case, a virtual control arm is created for RCT. Here, control patients are extracted from a retrospective patient mass using the patient selection criteria of the corresponding RCT. In this case, the RCT and RWE studies shake hands, and RWD offers something that can not be collected utilizing RCT. Furthermore, it is possible to use RWD to peek at how the patient population in RCT would be doing in real life.


The relationship between an RCT and an RWE study is not competitive but supplementary. The data stored in the social and healthcare registry were not originally created for research purposes. However, the extensive utilization of RWD is not only sensible but also cost-effective.

The outcome of an RCT or an RWE study is more uncertain alone than when these two are presented together. The authorities and decision-makers value both RCT and RWE.

Virtual control arms are a hot topic you can read more about in future World of Methods guest blogs.