The fact or process of investigation can result in measurements being different from the usual for a study participant.
This can partly be due to behavioural change, as described in the Hawthorne effect, which can be to some extent a conscious response to being studied. In addition, this reaction to being studied can result in altered physiological measurements, presumably largely unconsciously, and this can be termed apprehension bias.
A well-documented example of apprehension bias could be found in the measurement of hypertension. According to Cobos 2015, this was first described by Riva-Rocci in 1896.
Patients could become anxious as a result of visiting their health professional. They may also become anxious at the thought of having their blood pressure taken before it has actually been taken. As a result, this could raise their blood pressure, giving a biased record of what their physiological blood pressure actually is (Grassi 2016).
Probably the most common example of the potential impact from apprehension bias can be seen in the measurement of blood pressure – where the term “white coat hypertension” or “white coat effect” or “white coat syndrome” is sometimes used. In this context, a persons blood pressure is artificially elevated in a medical setting than when taken at home.
A review of sources of inaccuracy within blood pressure reporting among 41 studies found that 37 showed an average increase in blood pressure measured in a clinic, as compared with other settings, typically, an ambulatory measurement (Kallioinen et al 2017).
The size of the average difference between measurement methods varied widely. This discrepancy between clinic-based and ambulatory measurements may represent apprehension bias, and this example shows it can have an unpredictable effect on observations.
Apprehension bias, in the form of white coat hypertension, can have negative impacts on health. Another systematic review (Cohen 2016) examined the impact of white coat hypertension on cardiovascular events and mortality. The review found that compared with people with normal blood pressure, untreated WCH was associated with an increased risk for cardiovascular events, hazard ratio, 1.36 (95% CI, 1.03 to 2.00), all-cause mortality 1.33 (1.07 to 1.67), and cardiovascular mortality, HR, 2.09 (1.23 to 4.48)
Strategies to avoid apprehension bias include methods to reduce the anxiety of study participants. When considering blood pressure readings, the evidence for self-monitoring of blood pressures (where patients take their own blood pressure, usually in their own homes) and ambulatory automated equipment has grown. In these circumstances, measurements or observations are recorded with a low impact on the study participant (Tucker et al 2017).