Randomized, controlled, clinical, interventional, observational, longitudinal, cross-sectional, case-control, cohort?
Understanding what goes into a clinical trial can help you know how to interpret the results. The definition of a clinical trial from the World Health Organization (WHO) is:
Any research study that prospectively assigns human participants or groups of humans to one or more health-related interventions to evaluate the effects on health outcomes.
Similarly, the definition from the United States National Institutes of Health (NIH) is:
A research study in which one or more human subjects are prospectively assigned to one or more interventions (which may include placebo or other control) to evaluate the effects of those interventions on health-related biomedical or behavioral outcomes.
Clinical trials can have a direct effect on human health, so they are subject to extensive regulation and supervision. Advance planning and coordination with institutional ethics committees are crucial for conducting a safe and ethical clinical study, and reputable journals reinforce these norms by requiring adherence to the international, country-specific, and institutional guidelines governing human research.
Some researchers plan human research studies without realizing that they are actually designing a clinical trial. These researchers may have inaccurate preconceptions about clinical trials; for example, they may believe that all clinical trials test drugs or devices, include a placebo group, or are randomized and double-blinded. By answering the four questions posed by the NIH before planning a human research study, a researcher can determine whether they are designing a clinical trial and seek additional expertise.
As an example, consider a study design in which healthy participants complete a mental health assessment, listen to a daily motivational lecture for one month, and then complete the same mental health assessment. The results of the assessments are compared to determine if any changes in mental health occurred.
The researchers should ask themselves the following questions:
First, does our study involve human participants? This question may seem easy to answer, but work with biological samples or clinical data can be considered clinical research, even if the researcher has never met the participants. Conversely, secondary analysis of existing biological specimens, data collected anonymously, or data that are publicly available are not considered to be clinical research. In our example study, healthy human participants are involved, so our answer to the first question is “Yes.”
Second, are the participants in our study prospectively assigned to an intervention? Sometimes it is straightforward to identify an intervention (for example, a drug or device), but interventions can include exercise, behavioral modifications, or diet. Although our study only has one group of participants, they were definitely assigned to an intervention (the daily motivational lecture) in our study design before the study began (that is, prospectively). Therefore, our answer to the second question is also “Yes.”
Third, is our study designed to evaluate the effect of the intervention on the participants? It is clear that our study is designed to evaluate the effect of the intervention on the participants. Therefore, our answer to this question is “Yes.” Other studies may require more consideration. For example, if a university class listened to a motivational lecture at the beginning of each class and then completed a survey about the quality of the course, the researchers appear to be evaluating the effect of the intervention on the students’ perception of the course or the instructor. If it is unclear whether a study is evaluating the effect of the intervention on the participants, the institutional ethics committee can be consulted for additional insight.
Fourth, is the effect being evaluated a health-related biomedical or behavioral outcome? In our study, a mental health outcome is being evaluated (via the mental health assessment performed at the beginning and end of the study), and our answer would be “Yes.” However, not all human research studies evaluate a health-related or behavioral outcome. If our study evaluated the participants' opinions on their job prospects instead of mental health, it probably would not be considered a clinical trial. Again, researchers should consult their institutional ethics committee for assistance if it is unclear whether they are evaluating a health-related biomedical or behavioral outcome.
After a clinical trial is properly identified, designed, and approved by the institutional ethics committee, there is another important step that is strongly recommended or required for eventual publication of a clinical trial in a scientific journal: registration. Many journals and funders prefer that a clinical trial is prospectively registered (that is, registered before patient recruitment begins) with a registry accepted by the International Committee of Medical Journal Editors (ICMJE) (including journals listed on the WHO Clinical Trial Portal). However, retrospective registration is also accepted by many journals because clinical trial registration is so valuable to the scientific community. The information collected by these registries allows researchers to identify gaps in clinical research or approaches that were previously unsuccessful, helps participants to find clinical trials that they may qualify for, increases opportunities for collaboration, and improves transparency and open communication among the global scientific community. Most registries accept both interventional and observational studies; therefore, if a researcher is unsure whether a study should be registered, the ICMJE recommends registration.
Double-blind, randomized, placebo-controlled trials: The Gold Standard
Clinical trials of treatments and vaccines for COVID-19 have been in the news constantly this year. Differences in design, size, and analysis can make it difficult to compare these studies and apply their findings to an individual’s medical care. Ultimately, the gold-standard for determining the safety and efficacy of a treatment or vaccine is a clinical trial that is designed to be double-blind, randomized, and placebo-controlled.
Double-blind indicates that both the participants and the investigators who analyze the data are “blinded” - that is, neither group knows whether the participant has received the active treatment or the vaccine. Blinding helps minimize unconscious bias in the response to the intervention; for instance, a participant may report more soreness at the injection site if they know they received the vaccine instead of the “placebo”. A placebo is a substance that is not expected to produce the same effect as the active treatment. Often a saline injection or a pill with no active ingredient is used as a placebo, but there are some trials that use “active placebos,” such as a vaccine against a different disease. Investigators can compare the overall immune response to the active placebo against the response to the experimental vaccine. This comparison is known as a control, and it is essential to accurately determine the effects of the treatment.
A controlled study has both a “treatment group” and a “control group” (which may receive a placebo or the standard-of-care treatment). Finally, many clinical trials are randomized, which means that participants are randomly assigned to the treatment or control groups of the study. Randomization helps ensure that the characteristics of the study groups are similar (for example, age, sex, weight, height, health status) and minimizes the risk of allocation bias. When you read about clinical trials of COVID-19 treatments and vaccines, examine the designs of the trials, the numbers of participants, and the statistical analyses. These characteristics will help you to compare the results of clinical trials and understand why certain treatments (and eventually vaccines) are ultimately used to fight COVID-19.
Dr. Jen Rutan is a Developmental Editor at Research Square Company.