Development of a quality-assessment tool for experimental bruxism studies: reliability and validity.

AIMS To combine empirical evidence and expert opinion in a formal consensus method in order to develop a quality-assessment tool for experimental bruxism studies in systematic reviews. METHODS Tool development comprised five steps: (1) preliminary decisions, (2) item generation, (3) face-validity assessment, (4) reliability and discriminitive validity assessment, and (5) instrument refinement. The kappa value and phi-coefficient were calculated to assess inter-observer reliability and discriminative ability, respectively. RESULTS Following preliminary decisions and a literature review, a list of 52 items to be considered for inclusion in the tool was compiled. Eleven experts were invited to join a Delphi panel and 10 accepted. Four Delphi rounds reduced the preliminary tool-Quality-Assessment Tool for Experimental Bruxism Studies (Qu-ATEBS)- to 8 items: study aim, study sample, control condition or group, study design, experimental bruxism task, statistics, interpretation of results, and conflict of interest statement. Consensus among the Delphi panelists yielded good face validity. Inter-observer reliability was acceptable (k = 0.77). Discriminative validity was excellent (phi coefficient 1.0; P < .01). During refinement, 1 item (no. 8) was removed. CONCLUSION Qu-ATEBS, the seven-item evidence-based quality assessment tool developed here for use in systematic reviews of experimental bruxism studies, exhibits face validity, excellent discriminative validity, and acceptable inter-observer reliability. Development of quality assessment tools for many other topics in the orofacial pain literature is needed and may follow the described procedure.