Validity of the grading of severity of chronic pain in Japanese patients with temporomandibular disorders

Abstract

Objectives: Recently, many authors have tested the performance of the Research Diagnostic Criteria for Temporomandibular Disorders (RDC/TMD). The RDC/TMD have been translated into 18 languages and are being used by a consortium of RDC/TMD-based international researchers. We set out to validate the Grading of Severity of Chronic Pain (GSCP) in RDC/TMD for Japanese patients with TMD.
Methods: We used the GSCP in the Japanese version of RDC/TMD. Four hundred and forty-eight patients out of a total of 542 patients with TMD were analyzed. The GSCP included 7 questions on 3 items: Characteristic Pain Intensity (item 1, three questions), Disability score (item 2, three questions), and Disability days (item 3, one question). To assess cross-validity of the GSCP, patients were split randomly into two groups of roughly the same size (groups 1 and 2). We used three statistical methods to check the validation: the Mokken analysis method for cross-validity, structural equation modeling (SEM) for factor-validity, and the Spearman correlation coefficient for criterion based validity.
Results: Using the Mokken scale analysis, the item coefficient H (item H) level for group 1 showed a range of 0.64≤item H≤0.72 (0.43≤item H≤0.62 for group 2), and the scale coefficient H (H) was 0.70 and 0.57 for each group, respectively. The results indicated monotone homogeneity. Furthermore, reliability of the scales of groups 1 and 2 were 0.83 and 0.77, respectively, indicating double monotonicity. SEM yielded a substantially excellent model fit. GSCP showed a criterion based validity with a visual analog scale and a questionnaire to assess pain-related limitations of daily functions.
Conclusion: In this series of Japanese TMD patients, cross-validity, factor validity, and criterion based validity of GSCP were demonstrated, and the GSCP subscales were used to compare Japanese and English speaking patients with TMD.