Simplifying Rigorous Pain Classifications for Easier Clinical Implementation

Valid diagnostic criteria exist for pain-related temporomandibular disorders (TMDs; Schiffman, 2014) and can be implemented with high inter-examiner reliability (Bair, 2013). Because multiple barriers exist for their adoption by typical clinical settings, a simplified version has been created (Durham, in review). Utilizing two datasets, we tested this simplified set of examination procedures and criteria against reference standards based on Diagnostic Criteria for TMD (DC/TMD). The DC/TMD uses multiple provocation tests: five range of motion procedures and 40 palpation procedures. We used data from the Impact Project (n=401) and OPPERA (n=812); multiple study sites and examiners were involved in both projects, providing support for generalizability, and all examiners were annually assessed for reliability. Validity was determined using AUC, sensitivity, and specificity. Comparison of demographics indicated expected differences between cases and controls, and between case groups from the two studies. Based on 2-second palpation loads to full muscle (bilateral temporalis, masseter) and the lateral temporomandibular joint pole, AUC=0.93 and sens/spec=88/98 for identification of a painful TMD. Single band palpation within each of temporalis and masseter yielded comparable estimates to full muscle examination. Different permutations of other joint pain provocation procedures did not affect the results reported here. In terms of inferential decision making from using the shortened examination protocol, LR+ was 35 and LR- was 0.15. In conclusion, an acceptable simplification of the DC/TMD can be implemented in clinical settings and thereby improve the quality of patient care in more settings. Funding: NIH/NIDCR U01DE013331, U01DE019784.