What is meant by a double-blind, randomized controlled trial?
Blinding represents a very rich history that spans over two centuries. Very recent investigators and psychological researchers internationally comprehend the blinding concept (Schulz & Grimes, 2002). However, confusion concerning what it really means seems to lurk beyond the general understanding. Concepts, such as single blinded trial, double-blinded trials, and triple blinded trials imply very different things to various people. In addition, not only many medical researchers, but also psychologists confuse the concept of blinding with the allocation concealment (Schulz & Grimes, 2002). This confusion apparently shows the misunderstanding of both. A clear-cut practical and theoretical distinction differentiates the double-blinded randomized trials from other forms of trials, such as single-blinded and triple-blinded. Those who strongly prefer double-blinded trials have often maintained that it prevents bias of ascertainment and protects the sequence after allocation. Quite contrary, researchers in the field of psychology utilize the methods of allocation concealment majorly to prevent selection bias and protect an assignment sequence prior to and until allocation (Schulz & Grimes, 2002). According to Schulz & Grimes (2002), blinding represents a significant and distinct feature of randomized controlled trials (RCTs). The concept of blinding refers to an act of keeping the research participants, investigators or assessors unaware of the assigned intervention. This is to make them unable to be influenced by that knowledge.
Blinding, Placebo and Randomization Controlled Trials
This is a specific form of scientific experiment, and frequently a benchmark for clinical trials. The term randomization implies that all the selection of participants or research actors have an equal probability (Schulz & Grimes, 2002). These trials are frequently used in testing the effectiveness of different forms of medical intervention in patient populations. They can also offer an opportunity to collect useful information concerning adverse impacts, including drug reactions. The major distinguishing aspect of the usual randomized controlled trial is that its subjects are randomly selected to receive psychological interventions after the assessment of eligibility (Carlbring, et al., 2012). In real trials, randomization can be sophisticated, though it is conceptually easy. In fact, some theories compare it to tossing a coin (Schulz & Grimes, 2002).
The main reason for randomization is to ensure that the research outcomes are accurate and not biased. Researchers, like other people, are likely to be biased deliberately (Schulz & Grimes, 2002). However, it is also possible to be biased unwittingly. Through randomized controlled tests, researchers can avoid being deliberately biased and being biased unwittingly. This ensures that the subjects are reasonably equal.
In psychology, a blind trial refers to a trial in which the participants and the assessor take part and do not know the psychological intervention they are getting. According to Schulz & Grimes (2002), the participants might be getting a new intervention, standard intervention, or a placebo based on the trial design. All the patients might receive identical psychotherapy or counseling so that they are not capable of identifying the intervention they are getting (Carlbring, et al., 2012).
Randomized controlled testing might be blinded by procedures, which prevent the study participants or caregivers from knowing the psychological intervention they are receiving or giving. Blinding might sometimes be impossible or inappropriate to perform in RCT, unlike in allocation concealment (Schulz & Grimes, 2002). For instance, if the RCT integrates intervention in which active participation of the patient is essential, such as physical therapy, the participants cannot be blinded. According to Schulz & Grimes (2002), blinding or masking might be more appropriate in experiments involving participants who have impaired vision, and it might be less confusing in experiments in which blindness is an outcome. However, masking or blinding sends a strong message of bias prevention. According to Schulz & Grimes (2002), masking also suggests a more secure way to some researchers.
Blinded RCTs have been traditionally classified as single-, double-, or triple-blinded. RCTs that are not blinded are referred to as unblinded-RCTs (Schulz & Grimes, 2002). This paper essentially focuses on double-blinded RCTs. In this regard, the formal definition adopted by this paper is: Double-blinded RCT is an experimental study in which neither the assessor nor the subjects know the crucial aspects of the trial.
Placebo effect is also closely linked to blinding. Interventions might have no impact on the outcomes being investigated. Nevertheless, when an inefficient psychological intervention is administered to participants in the context of a well-designed RCT, the beneficial effects on the attitudes on the participants might occur, which in turn influence the outcome. Researchers call this phenomenon place effect. According to Schulz & Grimes (2002), a placebo is a pharmacological lethargic agent administered to participants in the control group. The utilization of placebo control group balances the placebo effect, which allows for independent assessment of the treatment effect. Schulz & Grimes (2002) pointed out that despite placebos having psychological impact; they are administered because they are not active. An active placebo mimics the side effects or symptoms, which might reveal the identity of the active psychological intervention (Carlbring, et al., 2012).
The Possible Effects of Blinding
If both the participants and the assessors are not blinded in psychological investigations, knowledge of the group assignment can influence the responses to the interventions received. Schulz & Grimes (2002) pointed out that participants who know they have been assigned to a group that will receive new intervention might harbor increased apprehension or favorable expectations. Those who know they have been assigned to standard intervention might feel relieved or deprived. Regardless of evidence suggesting that new interventions are as likely to be worse as they are better than standard treatments, participants probably presume that new intervention will be better than standard interventions (Carlbring, et al., 2012). In such cases, having knowledge of the interventions received, and the perceptions of that intervention, can have an effect on both the physical and psychological responses of the participants. According to Schulz & Grimes (2002), being aware of the treatment allocation can also have an impact on retention and compliance of trial participants.
Schulz & Grimes (2002) cited that blinding researchers and assessors, especially in double-blinded RCTs, who contribute to a widely defined trial team that comprises of, but not limited to, participant enrollers, health care practitioners, randomization implementers, routine data collectors, and trial designer, is also important. Researchers, specifically relevant to blinding, include healthcare practitioners, such as an attending nurse or physician; and intervention counselors. Intervention counselor can be anyone who delivers behavioral prevention message, and interacts with the study sample throughout the investigation. In this regard, Schulz & Grimes (2002) pointed out that the attitudes for or against the psychological intervention of researchers can directly be transferred to the participants if they are not blinded. Besides, the inclination of researchers can also be manifested in differential deployment of ancillary intervention the supplemental mental care or intervention, or differential decisions to withdraw participants from the trial. Schulz & Grimes (2002) added that researchers’ biases could also be manifested in their differential adjustments to the medication dose. Lastly, according to Schulz & Grimes (2002), researchers can also discourage or encourage the continuation in trial based on the knowledge of intervention group assignment.
More importantly, blinding assists in reducing the differential evaluation of outcomes that are referred to as ascertainment or information bias (Carlbring, et al., 2012). For instance, if the assessors of the trial outcome have knowledge of the intervention allocation, and believe that a new intervention is superior to the old, they are likely to register more generous responses to that intervention. In fact, in a certain placebo-regulated experiment in patients suffering from multiple sclerosis, the unblinded neurologists’ evaluation, but not the blinded ones, indicated a clear benefit of the intervention (Schulz & Grimes, 2002).
Subjective results, such as pain score, pose great opportunities for researchers to be biased. In addition, certain trial outcomes that are judged objectively can be fraught with subjectivity, for instance, salpingitis (Carlbring, et al., 2012). However, blinding becomes generally less essential in reducing observer bias. This is because the outcome becomes less subjective since the subjective results have no or less opportunity for bias. Knowledge of the treatment intervention cannot have a great affect measurement of the hard or objective results, such as death (Schulz & Grimes, 2002).
Published Psychological Experiment
Carlbring, et al. (2012) sets out to investigate internet-delivered attention bias modification in individuals suffering from social anxiety disorder using a double-blinded randomized controlled trial. The general method used involved recruiting participants by media advertisements. After a diagnostic interview, 79 participants were randomized to one of two attention training programs using the probe detection program. Participants were randomly assigned to treatment or placebo groups (Carlbring, et al., 2012). In RCTs, the assessment for eligibility is the first step. This implies that this conformed to the first general guideline by interviewing the participants to determine their illegibility. The second step is the random allocation to intervention. In this study, the eligible participants were randomized to two attention-training programs (interventions) (Carlbring, et al., 2012). The next step is receiving intervention, and during this step, the participants were trained to direct attention away from the threat. The next step is follow up, which the researchers have undertaken by calling their participants. From the flow chart below, one can see that the study followed the steps (Carlbring, et al., 2012).
The procedures and design of the study blinded both the assessors and the participants. Firstly, the participants were divided into two groups by an online random-number service independent of the researcher (Carlbring, et al., 2012). The use of a random-number service ensured that the investigator and the participants have no influence over the members of each of the groups. They were all unaware of the way they were assigned to their groups. Another element of blinding is the administration of self-report before the start of the intervention. As a result, there was no way the participants could have the knowledge concerning the treatment (Carlbring, et al., 2012). The third element is assessment that was done over the telephone (Carlbring, et al., 2012). The blind assessor had no earlier contact with the participants and no information concerning the group the participants have been allocated.
This was definitely a double-blinded randomized controlled trial with objectives of eliminating biases from both the participants and the researchers. (Carlbring, et al., 2012) Despite this, the study did not confirm the hypothesis that internet-delivered attention bias modification would be superior to a placebo condition (Carlbring, et al., 2012).
When the participants who had received the real treatment were asked to predict condition, about 81% of them thought they had been randomized to the placebo group (Carlbring, et al., 2012). One can think that such a low-level positive anticipation would influence the trial. Nevertheless, this cannot explain the lack of effect. As a result, there is a problem with believability and rationale (Carlbring, et al., 2012).
The measure of bias utilized by this research can be criticized (Carlbring, et al., 2012). It can be argued that a task in which the probe appears in the position of the neutral face would be more precise. Nevertheless, the probe detection task might be modified such that the vigilance foe disengagement and threat might be evaluated by including baseline trials (Carlbring, et al., 2012). By utilizing this new measure of bias, Carlbring, et al. (2012) pointed out that individuals with anxiety have a problem in disengaging their attention from highly threatening pictures.
The last and largest study limitation is that the participants were included without checking if they had initial attentional bias. This seems to be a serious mistake because not all the patients with anxiety show bias (Carlbring, et al., 2012).
This paper has defined double-blinded randomized controlled trials, and explained its importance in psychological trials. Double-blinded RCT is an experimental study in which neither the assessor nor the subjects know the crucial aspects of the trial. The process of conducting double-blinded RCT includes four stages that are enrollment, intervention allocation, follow up and data analysis. The main reason for randomization is to ensure that the research outcomes are accurate and not biased. More importantly, double blinding assists in reducing the differential evaluation of outcomes that is referred to as ascertainment or information bias. The trial analyzed by this was definitely a double-blinded randomized controlled trial with objectives of eliminating biases from both the participants and the researchers. This is because both the assessors and study sample did not have the knowledge on intervention, and because they never interacted.