Crossover Study

Crossover studies, sometimes referred to as case-crossover studies, are a type of clinical trial design [Page 459]where a participant experiences both the intervention and control aspects of the study. Participants are first randomized into either the intervention or control arms of a study for a previously determined amount of time and observed. After that time has concluded, these participants essentially “cross over” into another arm of the study, thus serving as their own controls in the study. This type of study is appealing because of its efficiency and precision.

Crossover studies are efficient because participants are used more than once. The benefit with this fact is that the required number of participants that a study would need to recruit can be much less than that of a simple parallel trial. The number of participants is commonly referred to as a study's sample size. The sample size, along with a few other factors, directly influences the statistical significance of any measurable outcome difference. The lower the sample size is in a study, the greater the likelihood that the measured outcome will lack statistical significance.

Crossover studies may also provide an increase in the precision of the study because treatment comparisons occur within the same individual, which may reduce the normal variability that exists within two human beings.

Although the theory behind crossover studies has been established for many years, the main assumption related to this type of design, which makes them unpopular in many practical circumstances, is the fact that there must not be any therapeutic carryover from the previous treatment. If there is any residual effect present, then this effect must be eliminated before the next treatment can begin. This becomes an even more difficult scenario when there are multiple treatments. In order for a crossover study to provide valid results, without any residual effects, there must also not be any exogenous changes over time that may alter the state of the patient. Any changes, whether they are seasonal changes, changes in a participant's comorbidities, or changes that may influence the participant's adherence to the trial, could greatly alter the results of the trial.

Therefore, when designing or reviewing a crossover study design, the following criteria should be kept in mind when determining the appropriate use of a crossover clinical trial design: 1.) The participant's first allocation has not had any long-lasting effects that could alter any of the subsequent interventions; 2.) if therapeutic effects do carry over, then the washout period has been sufficient enough to eliminate these effects before the second intervention can begin; 3.) exogenous or secular changes over time have not influenced the possible outcome of the trial.