Relationships between nurses' observations and patients' self-reports of pain.

Pain researchers and clinicians alike are often troubled by a lack of correspondence between non-verbal behavior and patients' self-reports of level of pain. This paper discusses some of the variables which can effect the relationship between these measures. In addition, the paper reports on the reliability of nurses' observations of pain behavior and of their inferences about the intensity of a patient's pain. In general, though these observations and inferences have adequate reliability, the correspondence between such inferences and patients' reports of pain intensity are modest, though significant. Discrepancies between observers' and patients' ratings of pain are greater in a chronic pain sample (N = 37) than in an acute pain sample (N = 34). Theoretical implications of these results are discussed.