Rutgers Cancer Institute of New Jersey
195 Little Albany Street
New Brunswick, NJ 08903-2681
Researchers attempting to evaluate the effectiveness of a treatment frequently suffer from difficulty finding enough people who are willing to participate in a study. One method used to address this common problem is called the crossover study.
Crossover studies are usually also double-blind studies. In the ordinary form of a double-blind study, participants receive either a real treatment or a placebo for the duration of the trial. A study of this type is said to have "parallel-group design."
In a double-blind, crossover study, however, participants receive either real treatment or placebo for a time, and then are switched ("crossed over") to the opposite treatment. Thus, researchers can get double mileage out of their participants—each person gets both placebo and treatment.
The advantage of a crossover trial is that it effectively doubles the number of people in a trial. This makes it easier for researchers to obtain statistically significant results.
However, there are some major disadvantages with crossover trials too. For example, people who take real treatment prior to placebo may carry over their benefits into the placebo period, either through the actual lingering action of the treatment, or simply by an enhanced power of suggestion. In addition, for various reasons, the longer a person participates in a study, the better the results tend to be, regardless of the treatment used. This tends to make the second treatment used in a crossover trial seem more effective than the first. These effects, as well as others, act as confounding factors, and make the crossover design less reliable (and less commonly used) than the standard parallel group design.