Randomization is a key feature of clinical trials aiming for a valuable study outcome. In this article we will explain the concept, its purpose and how it can be applied in practice.
What is randomization in clinical trials and why do we use it?
Clinical trials are performed to provide insights about the efficacy and effectiveness of an intervention (e.g. treatments or changes in behavior). Usually a trial aims to compare two or more groups of patients on a specific outcome, (for example the survival rate) given two treatment options, or a treatment option versus a control (no treatment).
“The RCT is considered the golden standard of clinical research”
The randomized controlled trial (RCT) is universally considered to be the gold standard when conducting clinical research. In an RCT, the allocation of patients to a specific intervention group is done by way of randomization. The groups of participants are kept as similar as possible, with one exception: the intervention itself. By keeping the differences between the intervention groups as small as possible, you reduce the chance that a measured effect on the selected outcome is due to bias; for instance, only very sick patients received the intervention.
What is blinding in clinical trials?
“Double-blind randomization ensures researchers and patients do not influence the trial”
An extension to the normal RCT, often used in research along with placebos (a non-therapeutic way to mimic the process of administering treatment, like a sugar pill), is the double-blind randomized controlled trial. The term double-blind refers to the fact that both the researcher and the patient do not know which randomization group the patient was allocated to. The advantage of this blinding is twofold: the researcher cannot affect the outcome of the study by treating patients in the intervention group differently and the patient is not influenced by knowing whether he or she received the placebo.
What is the purpose of randomization and blinding in research studies?
Both randomization and blinding are common methods to guarantee higher-quality outcomes of clinical trials by preventing any subjective biases as well as maximizing the study result’s validity. The methodology of randomization ensures that participants will be randomly assigned to different treatment groups to prevent the systematic arrangement of intervention and to avoid predictability. The other main purpose of randomization is to achieve comparability between the intervention and the control group to explain differences in results or outcome by treatment.
During the study, neither the participants nor the assigned physicians should know to which group they belong. Hence the process of blinding aims to make investigators as well as patients unaware which treatment the patients are going to receive. This prevents selection bias and increases objectivity among researchers. It also helps to ensure participants’ behavior and reporting will not be affected by their group affiliation.
How does randomization work?
The random allocation of study participants to intervention groups can be described by using a classic probability case: the vase with marbles everyone remembers from high school. By putting in an equal number of marbles for each intervention group (red or blue) and (blindly) taking out a marble for each patient, you can ensure that the intervention allocation is performed at random.
How Castor uses randomization for clinical trials.
A potential problem with the retrieval of marbles from the vase is that the researcher must know how many patients are going to be included (how many marbles should be put in the vase). And with that information, the researcher can infer what the colour of the last marble will be, potentially breaching a double-blind randomization procedure and influencing the research results.
To overcome this problem, Castor uses a variable block randomization method.
With this method, Castor puts a small but equal number of marbles in the vase (a ‘block’) for each intervention group. These blocks consist of either 2, 3, or 4 marbles. This way the researcher does not know how many marbles are in the vase at any given moment, and thus cannot predict the next randomization.
Online randomization with Castor
To guarantee a validated variable block randomization for your clinical trial, Castor EDC uses a carefully curated randomization algorithm. The special algorithm is constructed to divide randomized inclusions across groups in variable block sizes to ensure true randomness. The groups as well as the block sizes can independently be defined in the study settings. As various factors, such as number of inclusions and strata, influence the block size, we highly recommend consulting a statistician to ensure you are selecting the right block size.
In Castor EDC, the success of double randomization is performed with two different randomization rights that can be assigned individually to every user. The randomization right allows investigators to distribute participants into different intervention groups. Users with the randomization right will be able to see the date of randomization and who performed it. The actual randomization number and allocation will remain hidden. Only users with the ‘view randomization’ right can actually see to which group a participant (record) has been allocated. This separation between people who need to be blinded and persons involved who allocate the randomization is inevitable when conducting a RTC. Hence, researchers should pay special attention to set up the randomization rights correctly to ensure that all users who need to be blinded don’t have access to the allocation data.