Randomization as an experimental control method has been widely used in human clinical trials and other biological experiments. It prevents selection bias and ensures against accidental bias. It produces comparable groups and eliminates the source of bias in treatment assignments. Finally, it allows the use of probability theory to express the probability of chance as a source of difference in the final result.
A good experiment or trial minimizes evaluation variability and provides an unbiased assessment of the intervention by avoiding the confusion of other factors, which are known and unknown. Randomization ensures that each patient has an equal chance of receiving any of the treatments under study. In addition, it generates comparable intervention groups, which are the same in all important aspects except for the intervention each group receives. It also provides a basis for the statistical methods used in data analysis. The basic benefits of randomization are: it eliminates selection bias, it balances the groups with respect to many known and unknown confounding or prognostic variables. In the same way, it forms the basis for statistical testing and for a free statistical testing assumption of equality. of treatments. In general, a randomized experiment is an essential tool for testing treatment efficacy.
Randomization in practice
In practice, randomization requires generating randomization programs, which must be reproducible. The generation of a randomization program generally includes obtaining random numbers and assigning random numbers to each subject or treatment condition. For simple experiments with a small number of subjects, randomization can be easily performed by assigning the random numbers from the random number tables to the treatment conditions. However, it all depends on whether a restricted or stratified randomization will be performed for an experiment or whether an unbalanced allocation ratio will be used. In this case, it is better to use computer programming to perform the randomization, such as SAS, R environment, etc.
Why researchers demand randomization
Life science researchers demand randomization for several reasons. First, subjects from various groups should not differ systematically. In clinical research, if treatment groups are systematically different, the research results will be biased. Assume that subjects are assigned to control and treatment groups in a study examining the effectiveness of a surgical intervention. If a greater proportion of older subjects are assigned to the treatment group, then the outcome of the surgical intervention may be influenced by this imbalance.
Secondly, adequate randomization ensures that there is no a priori knowledge of the group’s assignment, i.e., concealment of the assignment. The researchers, subject or patients or participants, and others, should not know to which group the subject will be assigned. Knowledge of group assignment creates a layer of potential selection bias that can stain the data. Schul and Grimes stated that trials with inadequate or unclear randomization tended to overestimate treatment effects by up to 40%. This compares with those using adequate randomization. The outcome of research may be negatively affected by this inadequate randomization.
Statistical techniques used
Statistical techniques such as covariance analysis (ANCOVA), are often used to adjust for covariate imbalance. However, interpretation of this place-adjustment approach is often difficult. Covariate imbalance often leads to unanticipated interaction effects, such as uneven slopes between subgroups of covariates. One of the critical assumptions in ANCOVA is that the slopes of the regression lines are the same for each group of covariates. The adjustment needed for each group of covariates may vary, which is problematic. ANCOVA uses the average slope between the groups to adjust the outcome variable. Therefore, the ideal way to balance covariates between groups is to apply robust randomization in the design stage (before the adjustment procedure) rather than after data collection. In such cases, randomization is necessary and ensures the validity of statistical tests of significance that are used to compare treatments.
Types of Randomization
Many procedures have been proposed for the random assignment of participants to treatment groups in clinical trials. Each method is described along with its advantages and disadvantages. It is very important to select a method that will produce valid and interpretable results for your study. The use of online software to generate randomization code using the block randomization procedure will be presented.
Randomization based on a single sequence of random assignments is known as simple randomization. This technique maintains a complete randomization of the assignment of a topic to a particular group. The most common and basic method of simple randomization is to flip a coin. For example, with two treatment groups (control versus treatment), the side of the coin (i.e., heads – control, tails – treatment) determines the assignment of each subject.
Other methods include using a shuffled deck of cards (e.g., even – control, odd – treatment) or rolling a die (e.g., under and equal to 3 – control, over 3 – treatment). You can also use a random number table found in a statistics book or computer-generated random numbers for simple subject randomization. This randomization approach is simple and easy to implement in a clinical investigation. In large clinical research, simple randomization can be relied upon to generate a similar number of subjects among groups. However, the results of randomization could be problematic in clinical research with a relatively small sample size. This results in an uneven number of participants between the groups.
Randomization in blocks
The block randomization method is designed to randomize subjects into groups resulting in equal sample sizes. This method is used to ensure a balance in sample size between groups over time. The block size is determined by the researcher and must be a multiple of the number of groups (i.e., with two treatment groups, block size of 4, 6, or 8). Blocks are best used in smaller increments because researchers can more easily control the balance.
Once the block size has been determined, all possible balanced allocation combinations within the block should be calculated (i.e., the same number for all groups within the block). The blocks are then chosen at random to determine the allocation of patients to the groups. Although a balance in sample size can be achieved with this method, groups can be generated that are rarely comparable in terms of certain covariates. Pocock and Simon stressed the importance of controlling for these covariates because of the serious consequences for the interpretation of results. Such an imbalance could introduce biases into the statistical analysis and reduce the power of the study. Therefore, sample size and covariates must be balanced in clinical research.
The stratified randomization method addresses the need to control and balance the influence of covariates. This method can be used to achieve a balance between groups in terms of the initial characteristics of the subjects (covariates). Specific covariates must be identified by the researcher who understands the potential influence that each covariate has on the dependent variable. Stratified randomization is achieved by generating a separate block for each combination of covariates and subjects are assigned to the appropriate block of covariates.
Once all subjects have been identified and assigned in blocks, a simple randomization is performed within each block to assign subjects to one of the groups. The stratified randomization method controls for the possible influence of covariates that would jeopardize clinical research findings. For example, a clinical investigation of different rehabilitation techniques after a surgical procedure will have a number of covariates. It is well known that the age of the subject affects the prognosis rate. Therefore, age could be a confounding variable and influence the outcome of clinical research.
Adaptive covariable randomization
A potential problem with small to moderate-sized clinical research is that simple randomization (with or without consideration of stratification of prognostic variables) may result in a significant covariate imbalance between treatment groups. Covariate imbalance is important because of its potential to influence the interpretation of research results.
Many investigators have recommended adaptive covariate randomization as a valid alternative method of randomization for clinical research. In adaptive covariate randomization, a new participant is sequentially assigned to a particular treatment group taking into account specific covariates and previous participant assignments. Adaptive covariate randomization uses the minimization method when assessing sample size imbalance among several covariates.
Available online methods
Using online randomization, the researcher can generate a randomization plan for assigning treatments to patients. This online software is very simple and easy to implement. Up to 10 treatments can be assigned to patients and repeat treatment can also be performed up to 9 times. A maximum of only 10 treatments can be assigned to patients.
By entering the web address http://www.graphpad.com/quickcalcs/index.cfm in the address bar of any browser, the graphpad page appears with several options. Select the “Random Numbers” option and then press continue, the Random Number Calculator will appear with three options. Select the “Randomly assign subjects to groups” tab and press continue. On the next page, enter the number of subjects in each group in the “Assign” tab and select the number of groups in the “Subjects for each group” tab and keep the number 1 in the repeat tab if there is no replication in the study. For example, the total number of patients in a three-group experimental study is 30 and each group will be assigned to 10 patients. Type 10 in the “Assign” tab and select 3 in the “Topics in each group” tab and then press the “do it” button.
Another online randomization software that can be used to generate a randomization plan is http://www.randomization.com. The seed for the random number generator (Wichmann and Hill, 1982, modified by McLeod, 1985) is obtained from the local computer clock and printed on the bottom of the randomization plan. If a seed is included in the application, it overrides the value obtained from the clock and can be used to reproduce or verify a particular plan. Up to 20 treatments can be specified. The randomization plan is not affected by the order in which treatments are entered or the particular boxes left blank if not all treatments are needed. The program starts by sorting the names of the treatments internally. However, the sorting is case sensitive, so the same capitalization should be used when recreating a previous plan.
The benefits of randomization are numerous. It ensures against accidental bias in the experiment and produces groups that are comparable in every respect except for the intervention each group received. Simple randomization works well for large clinical trials . For small to moderate clinical trials (n without covariates, the use of block randomization helps to achieve balance. For small to moderate clinical trials with several prognostic factors or covariates, the adaptive randomization method may be more useful in providing a means to achieve treatment balance.
Pocock SJ, Simon R. Sequential treatment assignment with balancing for prognostic factors in the controlled clinical trial. Biometrics. 1975;31:103–15.
McLeod AI. Remark AS R58. A remark on algorithm AS 183. An efficient and portable pseudo-random number generato. Appl Stat. 1985;34:198–200
Schul KF, Grimes DA. Allocation concealment in randomized trials: Defending against deciphering. Lancet. 2002;359:614–8.
You might also be interested in: What’s a theory?