RESUMEN
BACKGROUND: In Juvenile Idiopathic Arthritis (JIA) the temporomandibular joint (TMJ) can be involved leading to pain, dysfunction and growth disturbances of the mandible and associated structures. There may be value to a three minute screening protocol allowing the rheumatologist to detect TMJ involvement systematically. Reliability and validity of the TMJ protocol for detecting TMJ co-morbidity were determined in 74 consecutive JIA patients. METHODS: The assessments of the rheumatologist and of a reference examiner (RE) were compared and validity of the TMJ protocol was established using the disease activity (JADAS-27) as an external reference. RESULTS: The internal consistency of the protocol was 0.73 (Cronbach's alpha). The inter-examiner agreement between the rheumatologist and the RE varied between 0.25 and 0.87 (Cohen's Kappa). Sensitivity and specificity, with the JADAS "3.8" indicating minimal disease activity, were 0.57 and 0.77 respectively. The area under the curve (AUC) was 0.70. A cut-off value of two positive items was found to be an optimal threshold to select the patients with likely TMJ involvement. CONCLUSIONS: The use of the protocol is feasible in everyday clinical practice. Reliability and validity aspects were satisfactory. The screening protocol for TMJ involvement provides the rheumatologist with systematic and focused TMJ information which relates to the JIA disease activity (JADAS-27).