Randomized controlled trial

A randomized controlled trial (RCT) is a type of scientific experiment most commonly used in testing healthcare services (such as medicine or nursing) or health technologies (such as pharmaceuticals or surgery). According to Lachin (1998), 'RCTs are considered the most reliable form of scientific evidence in healthcare because they eliminate spurious causality and bias'. RCTs are mainly used in clinical studies, but are also employed in other sectors such as judicial, educational, and social research. As their name suggests, RCTs involve the random allocation of different interventions (or treatments) to subjects. This ensures that known and unknown confounding factors are evenly distributed between treatment groups.

Sellers of medicines throughout the ages have had to convince their consumers that the medicine works. As science has progressed, public expectations have risen, and government health budgets have become ever tighter, pressure has grown for a reliable system to do this. Moreover, the public's concern for the dangers of medical interventions has spurred both legislators and administrators to provide an evidential basis for licensing or paying for new procedures and medications. In most modern health-care systems all new medicines and surgical procedures therefore have to undergo trials before being approved.

Trials are used to establish average efficacy of a treatment as well as learn about its most frequently occurring side-effects. This is meant to address the following concerns. First, effects of a treatment may be small and therefore undetectable except when studied systematically on a large population. Second, biological organisms (including humans) are complex, and do not react to the same stimulus in the same way, which makes inference from single clinical reports very unreliable and generally unacceptable as scientific evidence. Third, some conditions will spontaneously go into remission, with many extant reports of miraculous cures for no discernible reason. Finally, it is well-known and has been proven that the simple process of administering the treatment may have direct, sometimes very powerful, psychological effects on the patient, which is known as the placebo effect.

Additional recommended knowledge

How to quickly check pipettes?

How to ensure accurate weighing results every day?

What is the Correct Way to Check Repeatability in Balances?

Types of trials

Randomized trials are employed to test efficacy while avoiding these factors. Trials may be open, blind or double-blind.

Open trial

In an open trial, the researcher knows the full details of the treatment, and so does the patient. These trials are open to challenge for bias, and they do nothing to reduce the placebo effect. However, sometimes they are unavoidable, particularly in relation to surgical techniques, where it may not be possible or ethical to hide from the patient which treatment he or she received. Usually this kind of study design is used in bioequivalence studies.

Blind trials

Single-blind trial

In a single-blind trial, the researcher knows the details of the treatment but the patient does not. Because the patient does not know which treatment is being administered (the new treatment or another treatment) there might be no placebo effect. In practice, since the researcher knows, it is possible for them to treat the patient differently or to subconsciously hint to the patient important treatment-related details, thus influencing the outcome of the study.

Double-blind trial

In a double-blind trial, one researcher allocates a series of numbers to 'new treatment' or 'old treatment'. The second researcher is told the numbers, but not what they have been allocated to. Since the second researcher does not know, they cannot possibly tell the patient, directly or otherwise, and cannot give in to patient pressure to give them the new treatment. In this system, there is also often a more realistic distribution of sexes and ages of patients. Therefore double-blind (or randomized) trials are preferred, as they tend to give the most accurate results.

Triple-blind trial

Some randomized controlled trials are considered triple-blinded, although the meaning of this may vary according to the exact study design. The most common meaning is that the subject, researcher and person administering the treatment (often a pharmacist) are blinded to what is being given. Alternately, it may mean that the patient, researcher and statistician are blinded. These additional precautions are often in place with the more commonly accepted term "double blind trials", and thus the term "triple-blinded" is infrequently used. However, it connotes an additional layer of security to prevent undue influence of study results by anyone directly involved with the study.

Aspects of control in clinical trials

Traditionally the control in randomized controlled trials refers to studying a group of treated patients not in isolation but in comparison to other groups of patients, the control groups, who by not receiving the treatment under study give investigators important clues to the effectiveness of the treatment, its side effects, and the parameters that modify these effects.

Other aspects of control include having other members of the research team, who will typically review the test to try to remove any factors which might skew the results. For example, it is important to have a test group which is reasonably balanced for ages and sexes of the subjects (unless this is a treatment which will never be used on a particular sex or age group). Additionally, peer review and/or review by government regulators can be seen as another source of control. These bodies examine the trial results when they are presented for publication or when the drug manufacturer applies for a licence for the drug.

The importance of having a control group cannot be overstated. Merely being told that one is receiving a miraculous cure can be enough to cure a patient—even if the pill contains nothing more than sugar. Additionally, the procedure itself can produce ill effects. For example, in one study on rabbits where these subjects were receiving daily injections of a drug, it was found that they were developing cancer. If this was a result of the treatment, it would obviously be unsuitable for testing in humans. Because this result was reflected equally between the control and test groups, the source of the problem was investigated and it was shown in this case that the administration of daily injections was the cancer risk—not the drug itself.

The analysis of the trial results requires knowledge of medicine, epidemiology, and in particular statistics. The branch of statistics that deals specifically with biomedical research is biostatistics. Pharmaceutical firms employ groups of biostatisticians to try to make sense of the data. Likewise, regulators pay keen attention to the appropriateness of statistical methods used to analyze trial results.

Types of control groups

• Placebo concurrent control group
• Dose-response concurrent control group
• Active concurrent control group
• No treatment concurrent control group
• Historical control

Randomization in clinical trials

There are two processes involved in randomizing patients to different interventions. First is choosing a randomization procedure to generate a random and unpredictable sequence of allocations. This may be a simple random assignment of patients to any of the groups at equal probabilities, or may be complex and adaptive. A second and more practical issue is allocation concealment, which refers to the stringent precautions taken to ensure that the group assignment of patients are not revealed to the study investigators prior to definitively allocating them to their respective groups.

Randomization procedures

There are a couple of statistical issues to consider in generating the randomization sequences^[1].:

• Balance: since most statistical tests are most powerful when the groups being compared have equal sizes, it is desirable for the randomization procedure to generate similarly-sized groups.
• Selection bias: depending on the amount of structure in the randomization procedure, investigators may be able to infer the next group assignment by guessing which of the groups has been assigned the least up to that point. This breaks allocation concealment (see below) and can lead to bias in the selection of patients for enrollement in the study.
• Accidental bias: if important covariates that are related to the outcome are ignored in the statistical analysis, estimates arising from that analysis may be biased. The potential magnitude of that bias, if any, will depend on the randomization procedure.

Complete randomization

In this commonly used and intuitive procedure, each patient is effectively randomly assigned to any one of the groups. It is simple and optimal in the sense of robustness against both selection and accidental biases. However, its main drawback is the possibility of imbalances between the groups. In practice, imbalance is only a concern for small sample sizes (n < 200).

Permuted block randomization

In this form of restricted randomization, blocks of k patients are created such that balance is enforced within each block. For instance, let E stand for experimental group and C for control group, then a block of k = 4 patients may be assigned to one of EECC, ECEC, ECCE, CEEC, CECE, and CCEE, with equal probabilities of 1/6 each. Note that there are equal numbers of patients assigned to the experiment and the control group in each block.

Permuted block randomization has several advantages. In addition to promoting group balance at the end of the trial, it also promotes periodic balance in the sense that sequential patients are distributed equally between groups. This is particularly important because clinical trials enroll patients sequentially, such that there may be systematic differences between patients entering at different times during the study.

Unfortunately, by enforcing within-block balance, permuted block randomization is particularly susceptible to selection bias. That is, since toward the end of each block the investigators know the group with the least assignment up to that point must be assigned proportionally more of the remainder, predicting future group assignment becomes progressively easier. The remedy for this bias is to blind investigator from group assignments and from the randomization procedure itself.

Strictly speaking, permuted block randomization should be followed by statistical analysis that takes the blocking into account. However, for small block sizes this may become infeasible. In practice it is recommended that intra-block correlation be examined as a part of the statistical analysis.

A special case of permuted block randomization is random allocation, in which the entire sample is treated as one block.

Urn randomization

Covariate-adaptive randomization

When there are a number of variables that may influence the outcome of a trial (for example, patient age, gender or previous treatments) it is desirable to ensure a balance across each of these variables. This can be done with a separate list of randomization blocks for each combination of values - although this is only feasible when the number of lists is small compared to the total number of patients. When the number of variables or possible values are large a statistical method known as Minimisation can be used to minimise the imbalance within each of the factors.

Outcome-adaptive randomization

For a randomized trial in human subjects to be ethical, the investigator must believe before the trial begins that all treatments under consideration are equally desirable. At the end of the trial, one treatment may be selected as superior if a statistically significant difference was discovered. Between the beginning and end of the trial is an ethical grey zone. As patients are treated, evidence may accumulate that one treatment is superior, and yet patients are still randomized equally between all treatments until the trial ends.

Outcome-adaptive randomization is a variation on traditional randomization designed to address the ethical issue raised above. Randomization probabilities are adjusted continuously throughout the trial in response to the data. The probability of a treatment being assigned increases as the probability of that treatment being superior increases. The statistical advantages of randomization are retained, while on average more patients are assigned to superior treatments.

Allocation concealment

In practice, in taking care of individual patients, clinical investigators often find it difficult to maintain impartiality. Stories abound of investigators holding up sealed envelopes to lights or rzvzvbdvdvansacking offices to determine group assignments in order to dictate the assignment of their next patient^[2]. This introduces selection bias and confounders and distorts the results of the study. Breaking allocation concealment in randomized controlled trials is that much more problematic because in principle the randomization should have minimized such biases.

Some standard methods of ensuring allocation concealment include:

• Sequentially-Numbered, Opaque, Sealed Envelopes (SNOSE)
• Sequentially-numbered containers
• Pharmacy controlled
• Central randomization

Great care for allocation concealment must go into the clinical trial protocol and reported in detail in the publication. Recent studies have found that not only do most publications not report their concealment procedure, most of the publications that do not report also have unclear concealment procedures in the protocols^[3][4].

Difficulties

A major difficulty in dealing with trial results comes from commercial, political and/or academic pressure. Most trials are expensive to run, and will be the result of significant previous research, which is itself not cheap. There may be a political issue at stake (compare MMR vaccine) or vested interests (compare homeopathy). In such cases there is great pressure to interpret results in a way which suits the viewer, and great care must be taken by researchers to maintain emphasis on clinical facts.

Most studies start with a 'null hypothesis' which is being tested (usually along the lines of 'Our new treatment x cures as many patients as existing treatment y') and an alternative hypothesis ('x cures more patients than y'). The analysis at the end will give a statistical likelihood, based on the facts, of whether the null hypothesis can be safely rejected (saying that the new treatment does, in fact, result in more cures). Nevertheless this is only a statistical likelihood, so false negatives and false positives are possible. These are generally set an acceptable level (e.g., 1% chance that it was a false result). However, this risk is cumulative, so if 200 trials are done (often the case for contentious matters) about 2 will show contrary results. There is a tendency for these two to be seized on by those who need that proof for their point of view.

See also

• Drug development
• Double-blind
• Evidence-based medicine
• Hypothesis testing
• Intention to treat analysis
• Medicine
• Meta-analysis
• Randomization
• Statistical inference
• Systematic review

References