Objectives: The early childhood oral health impact scale (ECOHIS), a proxy measure to assess the impact of oral health problems and on the quality of life of preschool children and their families, has been translated into Chinese and validated employing the classical test theory. However, some invariable problems are related to this theory. The objective of this study was to assess the metric properties of the ECOHIS by applying the item response theory.
Methods: A random sample of 1296 Chinese preschool children in Hong Kong participated in the study and were subjected to an oral examination for their caries status. Their parents were asked to respond to the ECOHIS and an extra set of questions concerning their socio-demographic backgrounds. The collected data were analysed using Rasch model.
Results: The obtained results support the interpretation of adequate total fit of persons and items, because the mean infit MNSQ (1.08) and outfit MNSQ (0.94) for persons, and the mean infit MNSQ (0.99) and outfit MNSQ (0.94) for items, are within the acceptable ranges (0.6-1.4). The difficulty spread of the items was between -0.81 and 1.96 logits, and the levels of respondents spanned from -5.56 to 6.72 logits on the person-item map. The spread of respondents was skewed and mistargeting, between the items and respondents, was confirmed. Combination of responses of 'often' and 'very often' was suggested because of their flat probability showing on the characteristic curves.
Conclusions: Rasch model has been shown to be a feasible and practicable way of assessing the metric properties of the ECOHIS. The results allow the use of ECOHIS to be recommended for clinical research in oral health-related quality of life in young children with minor modifications.
© 2011 John Wiley & Sons A/S.