CARRYOVER BIAS:
Typically referred to as Carryover Effect; a bias that occurs when the effect of one exposure (e.g. earlier one) carries over into the domain of when another exposure is acting (e.g. carries over into a later period); such that measurement of the later exposure is contaminated by the effect of the earlier exposure1,2.
This bias is often relevant to experimental studies which use multiple exposures/treatments in sequential periods (e.g. cross-over clinical trials); such that the treatment in the first period may unintentionally act into the second period etc. Implementing a sufficiently long washout period between treatments may be a way to mitigate this bias in the case of a clinical trial. However, Carryover Bias may also occur in prospective (forward temporality) observational studies as well, if more than one exposure is of interest, and relevant to the disease/outcome in the study. Further, Carryover Bias may occur in studies where the exposure acts locally within an enclosed system, and the comparison groups are all contained in the enclosed system. For example, surgery on one paired-organ (e.g. one kidney) when the other paired organ is acting as the control in the same individual3.
Carryover Bias/Effect may be classified as a bias if the contamination of exposures is systematically different between comparison groups, otherwise it may be classified as an effect. Also see: Bogus Control Bias, Information Bias, and Confounding Bias.
References:
1. Porta M, ed. A Dictionary of Epidemiology. Sixth ed. New York, NY: Oxford University Press 2014. (Link to Reference)
2. Ahrens W, Pigeot I, eds. Handbook of Epidemiology. Second ed. New York: Springer 2014. (Link to Reference)
3. Haynes RB, Sackett DL, Guyatt GH, P. T. Clinical Epidemiology. How to Do Clinical Practice Research. Third ed. Philadelphia: Lippincott Williams & Wilkins 2006. (Link to Reference)