Examiner expectancy effects in the measurement of pressure pain thresholds

The ascending Method of Limits, used for the determination of pressure pain thresholds (PPT), is not a psychophysically robust method. The present study sought to determine if the examiner's expectancy, based on whether the measurement site was clinically 'painful' or 'non-painful', would bias the obtained PPT values. Twenty-eight patients with facial or temporal area pain served as subjects, and in each subject, a pain site and a control site were identified and marked. According to a randomization schedule, the pain and control sites were correctly marked in half of the subjects and were mis-labeled in the other half, thereby controlling the examiner's knowledge of a site and thus the examiner's expectancy of what the PPT should be. Two examiners, shown to be reliable with each other in both pre-clinical and post-clinical reliability studies, were blind to the true purpose of the study and to the marking procedures. Each examiner made one PPT measurement at each marked site in a counterbalanced measurement order. Manipulating the examiner's prior knowledge of the measurement site's characteristics significantly lowered the obtained PPT values for control sites but did not significantly alter the PPT at the clinically painful sites. Nevertheless, the pain sites still had significantly lower PPTs than did control sites. We conclude that: (i) PPTs at pain sites are robust to a major source of measurement bias associated with the ascending Method of Limits; (ii) measurement order and knowledge of measurement site characteristics can influence obtained PPT; and (iii) the common protocol in which the examiner monitors the amount of pressure during PPT measurement in order to control the force application rate may serve as a mechanism that can bias the obtained values.