Interobserver variability in assessment of signs of TMD

PURPOSE: The aim of this report was to study the ability of examiners to measure reliably the clinical signs of temporomandibular disorders (TMD). Four examiners participated in this study of 11 TMD patients and 25 nonpatients. MATERIALS AND METHODS: Vertical and lateral excursions of the jaw were measured using a millimeter ruler. Joint sounds during vertical jaw movements were assessed using digital palpation. The reliability of delivering appropriate degrees of digital pressure to assess masticatory muscle pain was assessed using a manometer after training examiners to exert specified pressures. RESULTS: Intraclass correlation coefficients for the measurement of vertical and protrusive jaw movements were > or = 0.87, which was considered excellent. The intraclass correlation coefficient for measurements of left and right lateral jaw excursions varied between 0.73 and 0.85, which was considered acceptable. The interobserver agreement for detecting the joint sounds showed overall agreement across examiners of 78%. Kappa for every possible pair of examiners varied between .52 and .86 (median .75, interquartile range .18). Reliability for diagnostic categories from the Helkimo index and Research Diagnostic Criteria for Temporomandibular Disorders involving joint noises showed modest reliability. CONCLUSION: Point estimates and measures of spread for reliability measures of single clinical TMD signs as well as combinations of signs into diagnostic categories from the Helkimo index and Research Diagnostic Criteria for Temporomandibular Disorders involving joint noises were sufficient in a group of four examiners.