Confirmative factor analysis of the dimensions of the Child Oral Health Impact Profile (Dutch version).

The aim of this study was to examine the subscales of the Child Oral Health Impact Profile (COHIP). The Dutch version of the COHIP consists of the subscales 'Oral Symptoms', 'Functional Well-being', 'Emotional Well-being', 'School', and 'Peer Interaction'. The questionnaire was administered to a sample of 510 children in Amsterdam. Missing data were replaced with the personal mean. Subsequently, questionnaire reliability was investigated by means of corrected item-total correlations and Cronbach's alpha if-item-deleted. Based on these results, six items were excluded from further analysis. Then, the questionnaire and its subscales were examined using confirmative factor analyses. One-factor models were fitted on each subscale and a five-factor model was applied to the entire questionnaire. In several cases, model fit was below an acceptable level. In conclusion, the structure of five subscales does not seem to be entirely supported in this study population.