The primary aim was to determine new estimates for the measurement reliability of the Research Diagnostic Criteria for Temporomandibular Disorders (RDC/TMD) Axis I diagnostic algorithms. A second aim was to present data on the reliability of key clinical measures of the diagnostic algorithms.
Kappa (k), computed by generalized estimate equation procedures, was selected as the primary estimate of interexaminer reliability. Intersite reliability of six examiners from three study sites was assessed annually over the 5-year period of the RDC/TMD Validation Project. Intrasite reliability was monitored throughout the validation study by comparing RDC/TMD data collections performed on the same day by the test examiner and a criterion examiner.
Intersite calibrations included a total of 180 subjects. Intersite reliability of RDC/TMD diagnoses was excellent (k > 0.75) when myofascial pain diagnoses (Ia or Ib) were grouped. Good reliability was observed for discrete myofascial pain diagnoses Ia (k = 0.62) and Ib (k = 0.58), for disc displacement with reduction (k = 0.63), disc displacement without reduction with limited opening (k = 0.62), arthralgia (k = 0.55), and when joint pain (IIIa or IIIb) was grouped (k = 0.59). Reliability of less frequently observed diagnoses such as disc displacements without reduction without limited opening, and osteoarthrosis (IIIb, IIIc), was poor to marginally fair (k = 0.31-0.43). Intrasite monitoring results (n = 705) approximated intersite reliability estimates. The greatest difference in paired estimates was 0.18 (IIc).
Reliability of the RDC/TMD protocol was good to excellent for myofascial pain, arthralgia, disc displacement with reduction, and disc displacement without reduction with limited opening. Reliability was poor to marginally fair for disc displacement without reduction without limited opening and osteoarthrosis.