MISTAKEN IDENTITY BIAS:
In compliance trials, strategies directed toward improving the patient’s compliance may, instead or in addition, cause the treating clinician to prescribe more vigorously: the effect upon achievement of the treatment goal may be misinterpreted1. A classic problem that may occur in unblinded studies; when scientists or participants change the frequency, type, or amount of their exposure (in this case prescription of a treatment) to suit personal preferences or study objectives. Blinding of the treatment/exposure may not completely address this bias as it may not be feasible in a study of compliance (clinicians and patients would need to be aware of treatments they are taking for ethical reasons). Blinding clinicians and participants to the study objectives, or purpose (but not treatment), or addressing this bias at the end of study using a quantitative bias analysis2, may be the solution to address this bias. Also see: Information Bias.
References:
1. Sackett DL. Bias in analytic research. J Chronic Dis. 1979;32 (1-2):51-63. (Link to Reference)
2. Lash TL, Fox MP, Fink AK. Applying Quantitative Bias Analysis to Epidemiologic Data. New York: Springer 2009. (Link to Reference)