Should the frequency, severity, or both response scales be used for multi-item dental patient-reported outcome measures?

Abstract

Background The Oral Impacts on Daily Performances (OIDP) index asks the respondents to indicate both, the frequency and severity of the impact. However, it is not clear if the two scaling methods are correlated, and if using one scale is sufficient. The purpose of the study was to investigate the correlation between frequency and severity rating scales of the OIDP instrument, and whether only one of the rating scales can be used instead of both. Methods A battery of patient-reported outcome questionnaires were administered to a consecutive sample of adult dental patients from HealthPartners dental clinics in Minnesota (N = 2,115). Only those who responded to any of the OIDP items were included in the analysis for this study (N = 873). We assessed correlations between the frequency and severity scales for all OIDP items, and for the summary scores of the two OIDP response scales. We additionally fit a categorical structural equation model (SEM) (or an item factor analysis model) and examined the correlation between two latent variables (Frequency and Severity). Results The correlation estimates for all OIDP items were greater than 0.50, indicating large correlations between the frequency and severity scores for each OIDP item. The correlation estimate between the two summary scores was 0.85 (95% CI [0.82–0.86]). When we calculated the correlation coefficient using a latent variable model, the value increased to 0.96 (95% CI [0.93–1.00]). Conclusion Our study findings show that OIDP frequency and severity scores are highly correlated, which indicates the use of one scale only. Based on previous evidence, we recommend applying the frequency rating scale only in research and clinical settings.