Reliability of visual analog and verbal descriptor scales for "objective" measurement of temporomandibular disorder pain.

Eight dentists viewed standardized videotapes showing palpations of the temporomandibular joint and muscles of mastication and recorded their judgments concerning the amount of pain the patient was experiencing. Judgments were recorded using a four-point verbal descriptor scale (VDS) ("none", "mild", "moderate", "severe" pain) or a 100-mm visual analog scale (VAS) anchored with the terms "no pain" and "worst pain possible". Test/re-test reliability over a one-week period and interjudge reliabilities were calculated for each scale; reliabilities of the two scales were directly compared based on the statistical equivalence of weighted kappa and the Intraclass Correlation Coefficient. Neither scale showed satisfactory reliability. Median test/re-test reliabilities were k = 0.590 for the VDS and r = 0.822 for the VAS. Interjudge reliabilities averaged k = 0.394 for the VDS and r = 0.735 for the VAS. Direct comparison of reliabilities for the two scales showed no clear advantage for either scale. The marginal reliabilities of these scales, when used by dentists to quantify the patient's pain, suggest that neither scale should be regarded as an "objective" pain measure.