Validation Of The OHIP-EDENT Using Factor Analysis Methods
Sanghavi A.1,3, de Souza R1, Feine J1, Nguyen T1, Oates T2, Nicolau B1.
1Faculty of Dental Medicine and Oral Health Sciences, McGill University, Montreal, Canada; 2School of Dentistry, University of Maryland, Batimore, USA.
Supervisors: Dr. Belinda Nicolau, Dr. Jocelyne Feine.
Background: Oral Health Related Quality of Life (OHRQoL) plays a fundamental role in the overall well-being of an individual. The OHIP-EDENT questionnaire evaluates OHRQoL in edentulous populations with 20 items on functional, psychological, and social implications of being edentulous. While several authors have argued that OHIP-EDENT scores accurately capture the latent construct of OHRQoL, conflicting findings in multiple studies challenge this idea. This results from the emergence of alternative 3- and 4-factor solutions documented in scientific literature, diverging from the originally postulated 7-factor solution of the OHIP-EDENT.
Objectives: To estimate:
1. the extent to which the OHIP-EDENT reliably reflects OHRQoL using data from two studies and
2. the optimal factor structure, potentially improving OHIP-EDENT's validity.
Methodology: Using baseline data from participants in a US quasi-experimental study (n=155) and a Canadian randomized clinical trial (n=255), the Kaiser Meyer Olkin and Bartlett tests assessed sampling adequacy. The Exploratory Factor Analysis (EFA) on data from the US population identified the latent constructs of the OHIP-EDENT. After identifying the 3-factor structure from the EFA, a
Confirmatory Factor Analysis (CFA) of the Canadian cohort data validated this hypothesized model. Factor analysis was performed on two separate datasets to ensure stability and generalizability of the 3-factor model. Subsequently, an additional step was carried out to evaluate the cross-cultural applicabilitiy of the OHIP-EDENT instrument. A 4-factor model (based on published literature) was tested on data of the Canadian population using CFA, and the 3-and 4-factor CFA models were compared.
Results: The EFA revealed a 3-factor configuration of the OHIP-EDENT (Cronbach’s alpha = 0.946); that EFA model was then validated using the Canadian cohort. 3-factor (RMSEA=0.043; CFI=0.995) and 4-factor (RMSEA=0.037; CFI= 0.996) CFA model was established.
Conclusion: The postulated 3-factor model identified constructs "Functional and Psychological Well-Being," "Social Impact," and "Physical Discomfort”, while the 4-factor model included an additional
construct, “Masticatory-related complaints”, offering a more comprehensive framework to investigate the dimensions of OHRQoL in edentulous individuals.