Validity of patient self-reports and clinical signs in the assessment of sleep bruxism based on home-recorded electromyographic/electrocardiographic data

Abstract

Background: Reliable and suitable bruxism assessment would be desirable, but available studies presented heterogeneous results. Objective: To determine the agreement of patient self-reports and clinical signs of sleep bruxism (SB) with electromyographic/electrocardiographic data. Methods: Two hundred individuals from a German dental clinic population (120 women and 80 men) participated in the study. Participants completed different SB questionnaires, had a clinical examination to evaluate bruxism signs and used the Bruxoff® device to record electromyographic/electrocardiographic data for five nights. To investigate interrater reliability for clinical diagnosis of bruxism, 126 of the 200 participants were assessed for clinical signs of bruxism by two independent uncalibrated examiners. Statistical evaluation included calculation of sensitivity, specificity and accuracy and of Cohen's kappa. Results: Based on the Bruxoff® data, 106 participants were identified as bruxers and 94 as non-bruxers. The 106 bruxers were further classified into 47 moderate and 59 intense bruxers. The highest accuracy and sensitivity values were recorded for the overall score for clinical bruxism signs (accuracy: 72.0% and sensitivity: 70.8%). The best specificity (96.8%) was seen for the question regarding tooth grinding in the last two weeks reported by others, but concurrent sensitivity was very low (3.8%). Analysis of interrater reliability revealed a substantial agreement (Cohen's kappa of 0.6). Conclusion: The study results indicate that self-report questionnaires and clinical signs have moderate sensitivity, specificity and accuracy for diagnosing bruxism comparing with an ambulatory device for current SB (Bruxoff®). Regarding interrater reliability for clinical signs of SB, substantial agreement was found between the two examiners. Clinical trial no. NCT03039985.