2.2. Experimental study designs
In contrast to observational studies, an experimental study assigns subjects to different treatment or exposure levels. This type of study design can be used to investigate the change in health status due to disease screening programs, prevention plans, interventions, diagnostic techniques or treatment procedures. Ultimately, a research team decides who will be treated or exposed, which consequently results in experimental intervention, not just observation of natural events.
Randomized studies are very powerful for investigating cause and effect because of the random assignment of study subjects to two or more intervention strategies, which leads to a compelling test of causality. The most simple randomized trial design consists of participants being randomly assigned to one of two treatment arms, the experimental arm (receive treatment of interest) or the control arm (receive no, placebo or standard treatment). Data from randomized trials can be utilized to calculate incidence of outcomes per treatment arm and then compare the incidence using the relative risk or risk differences. Randomization helps protect against bias, because it is likely that potential confounders are equally distributed across the treatment and control study groups. The scope of randomized studies is limited because these studies aim to confirm or disprove a specific hypothesis. Additionally, the cost and time needed to conduct trials are two primary disadvantages of this study design. A third concern is that the results from a controlled randomized trial may not be generalizable to uncontrolled real-world settings. There are many different variations on the simple randomized study design in which randomization schemes are modified and researchers are blinded to study conditions.