Confirmatory factor analysis on factor structure of OHIP

Background: The Oral Health Impact Profile (OHIP) with 49 items (full questionnaire) and short forms with 14 items were developed as measures of the oral health-related quality of life (OHRQoL) for elderly. Previous study showed that OHIP demonstrated good validity and reliability. Objectives: To confirm the hypothesized factor structure with seven subscales (functional limitations, physical pain, psychological discomfort, physical disability, psychological disability, social disability and handicap) of OHIP in measuring OHRQoL for elderly in Hong Kong. Methods: The instrument was interviewer-administered to 586 elderly persons aged 60 - 80 years. Confirmatory factor analysis (CFA) was used to evaluate the goodness-of-fit of the measurement models of factor structure with seven subscales using the full and short forms of OHIP. For an adequate model fit, the ratio χ²/df should be < 3, RMSEA (Root mean square error of approximation) should have a value < 0.08, while the GFI (Goodness of fit index), CFI (Comparative fit index) and NFI (Normed fit index) have a value > 0.90. Results: The results showed that the measurement models for the subscales of functional limitations, psychological disability, social disability and handicap fitted the data adequately while the models for the other three subscales fitted the data poorly (χ²/df > 3). Since not all subscales in the full form were adequately fitted, the evaluation of the measurement model of the full form was not proceeded. The measurement models of short forms with 14 items fitted the data well in all respects. Conclusion: The short forms of OHIP fitted the hypothesized factor well unlike the original form. The results indicated that the items used in the short forms contain sufficient information in measuring OHRQoL for elderly in Hong Kong.