Figure 1-4 demonstrates the major difference between clinical trials and cohort studies: clinical trials randomly assign patients to different treatment groups (exposure groups). Random assignment yields treatment groups with similar characteristics in regard to variables that may alter outcomes or the risk of complications from the treatment. This control of confounding factors is a major advantage of clinical trials over other study designs.
Figure 1-4 Simplified schematic of a randomized controlled trial.
All the previously mentioned features of high-quality observational studies, such as the following, should also be applied to randomized controlled trials:
a well-defined research question and objectives
explicit inclusion and exclusion criteria
an adequate sample size
predefined, objective primary and secondary outcomes
masking of patients, treating clinicians, and evaluators to the assigned treatment
complete follow-up of all patients
The CONSORT (Consolidated Standards of Reporting Trials) Statement, an evidence-based set of recommendations, includes a checklist of features that should be included in the design and reporting of clinical trials.
When evaluating a clinical trial, the clinician should consider 2 issues in addition to the other features of high-quality studies. The first is whether the study excluded patients from data analysis because they did not meet all of the eligibility criteria, experienced adverse effects and stopped treatment, or did not adhere to the treatment regimen. Exclusion of these types of patients creates biased results because the excluded patients’ results may differ from those of the patients included in the analysis. For this reason, clinical trials should include an intention-to-treat analysis, which includes the data from all enrolled participants, and separate analyses of those who completed the trial and those who did not.
Results from subgroups of patients (eg, young vs old, hypertensive vs nonhypertensive) should be regarded with suspicion. By statistical chance alone, a study can identify a subgroup of patients for whom the benefit of treatment is statistically significant. A subgroup evaluation may be considered valid if the investigators identified the subgroup a priori in the study design, treatment results vary similarly across subgroups (eg, success steadily decreases in each age stratum as the participants become younger), and a biologically plausible explanation exists for the finding.
Excerpted from BCSC 2020-2021 series: Section 1 - Update on General Medicine. For more information and to purchase the entire series, please visit https://www.aao.org/bcsc.