Randomized controlled trials (RCTs) or randomized clinical trials are now widely considered the gold standard by which we judge efficacy. The U.S. Food and Drug Administration (FDA) requires them for drug approval4.1; the National Institutes of Health rewards them with funding; the journals encourage them by publication; and increasingly, practitioners read them and apply their results. When feasible and ethical, RCTs are a standard part of health research. Thus, it is critically important to appreciate what these trials can tell us, what can go wrong, and what questions they cannot address.
RCTs today are usually conducted using an elaborate set of rules and procedures. These apply not only to drugs but increasingly also to a wide range of interventions from prevention to palliation. Much of what we discussed in the first three chapters applies to RCTs. In this chapter, we will emphasize the approaches and issues that are unique or especially important for RCTs.
The reporting of RCTs in high-quality journals generally follows the recommendations of what is known as the CONSORT (CONsolidated Standards Of Reporting Trials) Statement.[(1)] The details for conducting the study need to be defined in what is called the study’s protocol.
Before beginning a RCT, the investigation must be reviewed by an Investigational Review Board (IRB) to evaluate the quality of the study design, the ethics of conducting the study, and the safeguards provided for patients, including a review of the informed consent statement that potential participants will be asked to sign. The IRB is determining whether it is reasonable for a potential participant to be asked to participate. IRBs have taken on increasingly important roles in the conduct of research especially RCTs. Learn More 4.1 reviews roles that IRBs play.
Learn More 4.1: Institutional Review Boards (2,3)
Before beginning an investigation on human beings, a researcher must obtain permission from the Institutional Review Board at each institution in which the research will be conducted. Subsequently, the researcher must obtain the informed consent of all those who are asked to participate. Let us take a look at what is involved in this process.
IRBs are governed by U.S. federal regulations. These regulations derive from the 1978 Belmont Report, which resulted in large part from the Tuskegee study. The Tuskegee study followed approximately 600 poor black males with syphilis beginning in 1932. They were followed without treatment for 40 years despite the fact that effective treatment, penicillin, was available more than two decades earlier.4.2
The Belmont report outlined the following principles:
Respect for persons involves recognition of the personal dignity and autonomy of individuals, and special protection of those persons with diminished autonomy.
Beneficence entails an obligation to protect persons from harm by maximizing anticipated benefits and minimizing possible risks of harm.
Justice requires that the benefits and burdens of research be distributed fairly.
The fundamental role of the IRB is to establish that the benefits to the individual and to society are worth the risks. Risk in this context means the potential harm to the individual study participant. To accomplish this goal, the IRB needs to be assured of the quality of the study design since without a well-designed study, little or no benefit can be expected.
An investigator also is expected to establish that there is a need to conduct a RCT. This includes establishing what is called clinical equipoise. Clinical equipoise implies that there is sufficient controversy within the expert clinical community about the preferred treatment to justify a RCT. Notice that clinical equipoise does not imply that the investigator is neutral as to the best treatment. Often the investigator will hold a strong opinion as to the best treatment. It is the IRB’s responsibility to ensure that safeguards are in place to ensure that the investigator’s opinion does not influence the conduct of the investigator.
The federal guidelines outline specific roles for the IRB which include the following:
- Identify the risks associated with the research, as distinguished from the risks of therapies the subjects would receive even if not participating in research
- Determine that the risks will be minimized to the extent possible
- Identify the probable benefits to be derived from the research
- Determine that the risks are reasonable in relation to the benefits to subjects, if any, and the importance of the knowledge to be gained
- Ensure that potential subjects will be provided with an accurate and fair description of the risks or discomforts and the anticipated benefits
- Determine intervals of periodic review, and, where appropriate, determine that adequate provisions are in place for monitoring the data collected
- Determine the adequacy of the provisions to protect the privacy of subjects and to maintain the confidentiality of the data
If vulnerable populations such as pregnant women, children, or prisoners are included in the research, additional protections are required. Financial compensation is limited to modest compensation for time and effort and is not designed as an incentive to participate. Once the IRB has approved a study, the investigator may seek out individuals who fulfill the inclusion and exclusion criteria and ask for informed consent to participate in the investigation from the individual or their legal representative.
The content and process of informed consent is also governed by federal regulations and guidelines. Informed consent is a process and not just a document. The process of obtaining informed consent must allow potential participants adequate time to ask questions and deliberate on their decision. It must avoid “coercion or undue influence.” Situations in which undue influence may occur and special protections are required include those in which potential study participants are in subordinate positions.
For instance, undue influence may occur when an investigator is involved in a potential participant’s health care or when a faculty requests informed consent from a student or an employer encourages an employee to participate. Informed consent information must be provided clearly and in a language understood by the potential study participant.
Federal guidelines also outline minimum content that needs to be included. The content of informed consent must include the following:
- A statement that the study involves research, an explanation of the purposes of the research and the expected duration of the subject’s participation, a description of the procedures to be followed, and identification of any procedures that are experimental.
- A description of any reasonably foreseeable risks or discomforts to the subject.
- A description of any benefits to the subject or to others that may reasonably be expected from the research.
- A disclosure of appropriate alternative procedures or courses of treatment, if any, that might be advantageous to the subject.
- A statement describing the extent, if any, to which confidentiality of records identifying the subject will be maintained.
- For research involving more than minimal risk, an explanation as to whether any compensation is provided for adverse events.
- A statement that participation is voluntary, refusal to participate will involve no penalty or loss of benefits to which the subject is otherwise entitled, and the subject may discontinue participation at any time without penalty or loss of benefits to which the subject is otherwise entitled.
Approval by the IRB and obtaining informed consent from a study participant is a required part of the recruitment process. Once approved by the IRB, those who are asked to participate in the study must be informed and provide their informed consent. An additional review under the Health Insurance Portability and Accountability Act regulations is also now required to ensure the confidentiality of study data.
The CONSORT Statement now expects that a journal article reporting a RCT includes a flow chart that displays the passage of participants through the trial.
Figure 4.1 is the recommended template for reporting the data from RCTs. The terms used are parallel to the first four components of the M.A.A.R.I.E. framework.
FIGURE 4.1.Template of the CONSORT diagram showing the flow of participants through each stage of a randomized trial.
(Adapted from Consort Statement www.consort-statement.org. Accessed July 20, 2011.)
- Enrollment = Method
- Allocation = Assignment
- Follow-up = Assessment
- Analysis = Results
Now that we have taken a look at the planning and preparation that is required before beginning a RCT, let us use the M.A.A.R.I.E. framework to examine the unique features of RCTs.
4.1 The FDA generally requires convincing results from two independently conducted, well-designed RCTs for approval of a new drug. These investigations may be conducted in the United States or abroad.
4.2 Note that the Tuskegee study was a cohort study. IRB rules are not limited to RCTs; they cover all studies involving “human subjects research,” although exceptions and expedited reviews are available for studies that involve minimal risk to study subjects. Minimal risk is defined by the federal guidelines as “… the probability and magnitude of harm or discomfort anticipated in the proposed research are not greater, in and of themselves, than those ordinarily encountered in daily life or during the performance of routine physical or psychological examinations or tests … .” [(2)] For example, the risk of drawing a small amount of blood from a healthy individual for research purposes is no greater than the risk of doing so as part of routine physical examination.
RCTs generally are used to establish the efficacy of interventions from prevention to cure to palliation. Thus their study hypothesis usually indicates that on average those in the study group will have a better outcome than those in the control group. RCT may include more than two groups allowing comparisons of more than one dosage of a treatment or multiple interventions aimed at the same disease. In addition, a type of RCT called a factorial design may allow a RCT to address more than one study hypothesis.4.3
RCTs are capable of demonstrating all three criteria of contributory cause. When applied to a treatment or other intervention, the term efficacy is used instead of contributory cause. Efficacy means that in the study group being investigated, the intervention led to an increase in the probability of a desirable outcome. Efficacy, however, needs to be distinguished from effectiveness. Effectiveness implies that the therapy works under usual conditions of practice as opposed to the conditions of an investigation.
RCTs usually have a very specific study hypothesis because they seek to determine whether the intervention works when given according to a defined dosage schedule, by a defined route of administration, and to a defined type of patient.4.4
Thus RCTs are expected to have a detailed protocol, including specific inclusion and exclusion criteria. These inclusion and exclusion criteria define the population being investigated. All participants are expected to fulfill these criteria.
A large number of individuals may be evaluated to determine if they meet these inclusion and exclusion criteria. All those who are assessed for eligibility usually do not end up being participants in the investigation. For instance, they may not meet the inclusion or exclusion criteria or they may refuse to participate despite being eligible. Thus the CONSORT Statement’s flow chart begins by identifying the number assessed for eligibility and then indicates the number who were excluded and the reasons for their exclusion.
RCTs are not suitable for the initial investigation of a new treatment. When used as part of the drug approval process, RCTs are referred to as phase 3 trials. The FDA has traditionally required two independently conducted RCTs conducted for the same indication before reviewing a drug for approval. Ideally, a randomized clinical, or phase 3, trial should be performed before the drug is widely used for the indication being addressed. We will examine the FDA phases of review in greater depth in Chapter 6 when we take a look at harms or safety.
For new drugs that do not have market approval, identifying volunteers may be relatively easy. However, for many procedures and drugs that have been previously marketed and used for other indications, the treatment may have been widely used before RCTs could be implemented. This is a problem because once the treatment has been widely used, physicians and often patients have developed firm ideas about the value of the therapy. In that case, they may not believe it is ethical to enter into a RCT or to continue participation if they discover that the patient has been assigned to the control group.
Once the time is considered right for a RCT, the next question is whether it is feasible to perform one. To answer this, the investigator must define the question being asked in a RCT.
Most RCTs aim to determine whether the new or experimental therapy results in a better outcome, on average, than a placebo or standard therapy. To determine whether a trial is feasible, investigators need to estimate the necessary sample size. They must estimate how many patients are required to have a reasonable chance of demonstrating a statistically significant difference between the new therapy and the placebo or standard therapy. The required sample size as calculated before conducting the investigation depends on the following factors4.5:
- Size of the Type I error that the investigators will tolerate. This is the probability of demonstrating a statistically significant difference in samples when no true difference exists between treatments in the larger population. The Type I error is usually set at 5%.
- Size of the Type II error that the investigators will tolerate. This is the probability of failing to demonstrate a statistically significant difference in study samples when a true difference of a selected magnitude actually exists between interventions in the larger population. As we discussed previously, investigators should aim for a Type II error of 10% and accept no more than 20%. A Type II error of 20% indicates an 80% statistical power since the statistical power plus the Type II error add up to 100%. The 80% statistical power implies 80% probability of being able to demonstrate a statistically significant difference between the samples if a true difference of the estimated size actually exists in the larger populations.
- Percentage of individuals in the control group who are expected to experience the adverse outcome (death or other undesired outcomes) under study. Often this can be estimated from previous studies.
- Improvement in outcome within the study group that the investigators seek to demonstrate as statistically significant. Despite the desire to demonstrate statistical significance for even small real changes, the investigators need to decide the minimum size of a difference that would be considered clinically important. The smaller this difference between study group and control group therapy that one expects, the larger the sample size required.4.6
Let us take a look at the way these factors affect the required sample size. Table 4.1 provides general guidelines for minimum sample size for different levels of these factors.