Development and Psychometric Validation of a Brief Oral Health Literacy Instrument for Use in Population-Based...

Oral health literacy (OHL) plays a critical role in shaping individuals' engagement with oral health systems and their ability to translate information into effective self-care and service use. While the Health Literacy in Dentistry (HeLD-14) scale has demonstrated strong psychometric performance, brief and psychometrically robust instruments explicitly optimized for large-scale surveillance remain limited. This study sought to develop a concise HeLD version that preserves sensitivity to low OHL while remaining operationally feasible for large-scale monitoring.

This cross-sectional psychometric study was conducted among 600 adults aged 18-65 years from diverse sociodemographic backgrounds. A sequential, theory-informed analytic approach was applied, encompassing Item Response Theory (graded response model), principal axis factoring (PAF) using polychoric correlations, and confirmatory factor analysis (CFA). Items were prioritized based on discrimination and location parameters to enhance sensitivity at the lower end of the latent OHL continuum, while structural coherence and shared variance guided further refinement. The shortened versions were recalibrated using θ estimation and their classification concordance compared with that of the original HeLD-14.

IRT-based evaluation reduced the item pool from 14 to 8 (HeLD-8), retaining items with strong discrimination and threshold locations informative at the lower end of the oral health literacy continuum. Principal axis factoring using polychoric correlations confirmed that all eight retained items exceeded the pre-specified communality threshold, supporting a coherent unidimensional structure; HeLD-8 was therefore retained in full as the final candidate instrument. Confirmatory factor analysis in two independent split-half samples supported this unidimensional structure, with acceptable-to-good fit indices. HeLD-8 showed strong concordance with the full HeLD-14 in identifying individuals with low OHL (sensitivity 92.1%, specificity 93.5%), while reducing respondent burden by more than 40% relative to the parent instrument.

A brief oral health literacy screener, HeLD-8, was developed and psychometrically validated to support identification of low OHL within population-based surveillance contexts. By preserving representation across most of the original instrument's conceptual domains while substantially reducing respondent burden, HeLD-8 may facilitate integration of OHL assessment into large-scale oral health surveillance systems; further validation in independent and more diverse samples is recommended.