A critical examination of the use of signal detection theory in evaluating a putative analgesic--transcutaneous electrical nerve stimulation.

Experiments using normal human subjects were performed to determine the effect of a putative analgesic, transcutaneous electrical nerve stimulation (TENS), on the perception of noxious thermal stimuli and also to examine critically the general applicability of signal detection theory (SDT) to the evaluation of analgesic agents. In order to provide a control for factors related to suggestion, subjects were assigned to groups in which real or sham TENS was administered. It was shown that previously suggested criteria for demonstrating actual analgesia using SDT should be reevaluated. First, discriminability changed in one of the sham groups, suggesting that discriminability can be affected by changes in mental state which may occur during the evaluation of an analgesic. Secondly, criteria changed without accompanying changes in discriminability; this was interpreted as reflecting actual analgesia, rather than only changes in response bias, produced by TENS.