Research Methods 101: Randomized Controlled Trials
Later this month, the Patient-Centered Outcomes Research Institute (PCORI) will be releasing a draft of its methodology report, which will set the groundwork for the standards and types of research methods that can be used to develop comparative effectiveness research. This post is a part of NPC's series on research methods.
Later this month, the Patient-Centered Outcomes Research Institute (PCORI) will be releasing a draft of its methodology report, which will set the groundwork for the standards and types of research methods that can be used to develop comparative effectiveness research. This post is a part of NPC's series on research methods.
Randomized controlled trials (RCTs) answer the question: "Can an intervention work under certain controlled conditions?" To do so, an RCT employs a carefully planned experimental framework to compare an intervention with a control, investigating the effect of each treatment option on a defined outcome. After meeting strict inclusion qualifications, each participant is randomly assigned to one of the treatment groups. Ideally, the study is "double-blinded," meaning that neither the participants nor the researchers know who is receiving the active intervention and who is receiving the control treatment.
RCTs are "single" blinded when only the participants are unaware of their allocation. Appropriate randomization ensures that the groups are similar, so that observed effects are unlikely to be due to differences between those groups. In this study type, the investigator narrowly defines the patients, the interventions, and the outcomes to determine the benefits and harms of the intervention in an ideal setting. RCTs may compare an intervention to placebo or to another active therapy. Although CER focuses on comparisons of alternative therapies, the placebo-controlled RCT provides important information about the magnitude of a drug’s benefits and frequency of adverse reactions. These positive and negative outcomes may be over- or under-estimated in head-to-head randomized, controlled, active comparator studies. Moreover, placebo-controlled trials are also a key step in the route to gaining FDA approval for new therapies, or new indications for existing therapies.
Of all clinical research study designs, RCTs provide the strongest internal validity: (i.e., this drug, taken at this dose, produces this outcome in this group of people). The randomization and blinding techniques imposed at the start of a study greatly reduce the likelihood that bias or confounding influences the results. Confounding occurs when external factors are more likely in one study arm and have an independent impact on the measured outcomes (e.g., the intervention group has sicker patients than the control group). Because they reduce the impact of confounding, RCTs are considered the gold standard of clinical research, and many policymakers hope that there will be an increasing number of RCTs as CER continues to grow.
Many researchers are staunch advocates of RCTs. Dr. Michael Lauer, Director, Division of Cardiovascular Sciences, National Heart, Lung and Blood Institute, National Institutes of Health and Member, Patient-Centered Outcomes Research Institute Methodology Committee, believes RCTs are the most effective type of research and that it is critical to get the medcal community more engaged in these efforts.
[[{"fid":"380","view_mode":"teaser","type":"media","attributes":{"height":"390","width":"640","border":"0","alt":"Randomized Controlled Trials vs. Observational Studies","class":"panopoly-image-video media-element file-teaser"}}]]
However, RCTs have various potential limitations. First, the very characteristics that support an RCT’s strong internal validity can limit the study design’s external validity, or generalizability to the broader world of care beyond the study’s controlled environment. RCTs generally enroll a narrow set of patients without relevant comorbidities and rarely include children or the elderly in their study populations. second, most RCTs are modest in size due to their expense and challenging logistics. this modest size limits RCTs’ ability to assess uncommon events. third, many RCTs compare active treatments with placebo, rather than head-to-head with another active treatment, or choose to use comparators that may not be "right" or "best" for the analysis. Finally, the strict pre-set design of RCTs is occasionally abandoned by participants who may find themselves unhappy with their randomization or the constraints of the study; participants leaving a study (known as loss to follow-up) or requiring a treatment that might necessitate their transfer from the placebo to intervention arm may diminish the statistical power of the study.
Even with its disadvantages, the RCT (along with meta-analyses of multiple RCTs) provides the highest level of evidence.