EBM Consult

Cross-Over Study Design

    Description of a Cross-Over Study

    • A study design where all patients from a population of interest are initially assigned to one of two groups, one group who gets exposed to the intervention or a second group that does not get exposed to the intervention. 
    • After a period of time, an evaluation of the outcome is done and the patients from both groups undergo a period of washout so that the effect from the initial group intervention has been removed. Once this occurs, the subjects will then cross-over into the other group where the process starts over.
    • This study design provides the advantage of reducing the variability in outcome measures from outside confounders because each patient will serve as his or her own control.
    Study Diagram

    Cross-Over Study Design Image


    • Reduced influence by confounders since patients serve as their own controls
    • Reduced variability in the outcome(s) being measured, thus increasing the precision of estimation
    • Smaller sample sizes required
    • Having the opportunity to receive both treatments can sometimes be attractive for subjects

    • Cannot be done when the subjects can only receive one treatment
    • Assumption of no carryover effects (from washout period) is difficult to sometimes accurately test
    • May take longer than a randomized clinical trial since patients have to cross over into each arm after an appropriate washout period
    • Can be subject to period effects where differences in the effectiveness of an intervention can occur due to the passage of time.  For example:
      • Development of tolerance
      • Resistance
      • Dropouts
      • Changes in the disease process being evaluated or treated
    Related Content
    • Noordij M et al. Study designs in clinical research. Nephron Clin Pract 2009;113(3):c218-21.
    • Tinmouth A et al. Interventional trials: an overview of design alternatives. Transfusion 2007;47:565-67.
    Editors & Reviewers


    • Anthony J. Busti, MD, PharmD, FNLA, FAHA

    Last Reviewed:  July 2015