To our best knowledge, this is the first study to assess the psychometric properties of the C-OIDP in a convenience sample of primary school children. Until now, no national oral health survey on the prevalence of oral health impacts of Turkish children using a validated OHRQoL has been conducted in Turkey. Therefore, we choose to use the C-OIDP in this study as it is designed to be incorparated into oral health needs assessment [7,8,9]. Using this scale in population surveys could help professionals for planning and evaluating oral health promotion activities and oral health services for the community [7,8,9]. The validity and reliability of measure should be confirmed before quality of life is used as an outcome .
The Turkish C-OIDP showed acceptable internal consistency and test-retest reliability. The Cronbach’s alpha was higher than the values reported from previous studies conducted in France , England , Brazil , Peru , Spain , Italy , Chile , South India  and Moroccan . The ICC was similar to the validation studies in Chile , France  and South India .
The face-to-face interview format was preferred for data collection, because more than half of students aged 11–12 years had difficulty responding to the questions in the second part of the C-OIDP in the pretesting of the Turkish C-OIDP. Similar approach was used in many validation studies [10,11,12,13, 15, 16, 18, 20, 22, 29] conducted in similar age groups.
As used in previous studies [10,11,12,13,14, 16,17,18,19,20,21,22,23,24, 47], we chose to use the weighted OIDP scores in this study, because this scoring method was found to be a better predictor than the unweighted scores (frequence and severity scores) for DMFT .
Only four validation studies examined the factor structure of the C-OIDP [13, 20, 22, 28], three used EFA only [13, 20, 26], one used both EFA and CFA . Studies using only EFA [13, 29] suggested a three-factor model consisting of physical, psychological and social health components, whereas Agrawal et al.  and Amilani et al.  proposed a two factor model which represented the physical and psychosocial health components. Our study followed a similar approach that was used by Mtaya et al.  to identify the factor structure of the C-OIDP. We conducted firstly an EFA. The results of the EFA were then tested using CFA on the same sample to obtain an estimate of goodness of fit and to compare the extracted model to the previous model identified in the literature . In our study, a two-dimensional structure was identified with EFA. The first factor reflected the impact of oral conditions on functional limitations while the second factor consisted of items reflecting the psychosocial aspects of OHRQoL. The similar two-factor structure as in Agrawal et al.’s study  and Amilani et al.’s study  were found. The findings of this study agree with previous studies [13, 20, 29], reporting that the C-OIDP has a multidimensional structure which represent in the theoretical model of oral health consequences . In the study of Mtaya et al. , CFA indicated better fit for a three-factor solution than the two- and one-factor model of the C-OIDP. We compared a two-factor solution obtained from the EFA to a three-factor structure . CFA identified the new two- factor model which fit the data better than the previously proposed three-factor model . Our findings based on EFA and CFA suggest that C-OIDP is a multidimentional measure covering functional and psychosocial dimensions. These two dimensions represents the ultimate impacts of oral health consequences. This study provides insight into the underlying factor structure of the total score version of the C-OIDP .
Differences in factor structure between studies may be attributed to sampling differences, having various functions and meanings of the C-OIDP items in different cultures, and use of different scoring versions [26, 30, 49] Further study is required to generalise and confirm the two factor structure of the C-OIDP in a large, nationally representative sample of Turkish children.
The prevalence of oral impacts observed in this study is much greater than what was reported in earlier studies in Brazil , Thailand , France , Israel . The most prevalent oral problem reported was sensitive teeth followed by toothache, which were similar to previous studies conducted in Brazil , Sudan , Italy , Israel , Malaysia , North India , and Moroccan . In agreement with most studies [11, 12, 14, 15, 17, 18, 20,21,22, 29], we found that the performances with the highest frequencies impacts were eating, cleaning mouth and smiling; while the performances with the lowest impact were “studying” and relaxing. This is not surprising, because impacts on functional limitations and psychological well being were more prevalant than impact on social well-being social during children’s transition to adolescence [47, 50]. Similarly, some studies reported that social contact [11, 15, 17,18,19], emotional stability [14, 29] and speaking [19, 22] were the least frequently reported impacts. These differences in self- reported oral impacts may be related to childrens’ perception about health and ilness which are affected by their stage of development and social context in which they live [47, 51].
Knowledge in existing validation studies is limited, especially in terms of behavioral and socio-demographic predictors of impaired OHRQoL. In previous validation studies using multivariate analysis methods, socio-economic status, child’s age and gender, district of residence, area of residence and type of school were found be important predictors of impaired OHRQoL [15, 16, 20, 29]. In the bivariate analysis, children whose mothers had lower educational level and more children had worse OHRQoL. This may be explained by the fact that these mothers face a number of barriers to accessing oral health care for children and they have lower health literacy, resulting in worse oral health knowledge and their children’s oral health status .
Consistent with previous studies used bivariate analysis method, we found that use of dental floss, regular dental check-up , and consumption of sugars between meals [16, 29], were associated with worse OHRQoL. So far, only one study reported that the fruit intake frequency and mouthwash were significant behavioral predictors of OHRQoL . In contrast to this study, we found that only routine dental attendance was an important factor of improved OHRQoL.
Consistent with previous studies, we found that perceived need for dental treatment [9,10,11,12,13,14,15, 18, 24], dissatisfaction with oral health [9, 11, 12, 14, 24, 29], having oral problems [9, 12, 14, 19, 21, 29], poor perceptions of self-rated general health [12, 29] and oral health [11,12,13, 16,17,18, 24, 29] were associated with worse OHRQoL in the bivariate analysis. However, subsequent multivariate analysis showed that SROH, satisfaction with oral health and self- reported oral problems were important subjective factors of OHRQoL. Our findings are consistent with Castro et al. , who reported that self-reported oral health problems explained more of the variation in OHRQoL than clinical normative measures.
In some studies, the clinical indices were used in combination with the C-OIDP index [10, 15, 16, 19, 20, 29]. Bivariate analysis revealed that presence of malocclusion, gum disease and dental caries were significantly associated with children’s OHRQoL. Only malocclusion was found to be a significant predictor of impaired OHRQoL of Turkish children in the multivariate analysis, which is consistent with previous studies employing multivariable analysis [15, 20].
This study has some limitations and strengths that must be taken into account when interpreting its results. This study provided initial support for the reliability and validity of the Turkish C-OIDP in a convenience sample of primary school children aged 11–12 years from two public schools in Istanbul. Thus, our findings could not be generalized to the population of interest. This study was conducted on students at two primary public schools, one of that located in a deprived area but others in a semi-deprived area. This could lead to bias because children from families living in poverty and deprivation are more likely to have oral diseases, resulting in reduced children’s and parents’ quality of life . Future representative population-based studies are needed to understand the impacts of family socio-economic status and school-related factors on children’s OHRQoL. Due to cross-sectional design, this study did not verify any cause-effect relationship among the assessed variables and any changes in scores over time. Population-based cohort studies and further validation study in children of different age groups are needed. Future study using the Item Response Theory may provide additional information to the Classical Test Theory and allow performance assessment of individual items [8, 49]. Additional research aiming to compare psychometrically different administration modes of the Turkish C-OIDP index may be useful for selection the effective data collection mode in future epidemiological studies of child populations as well as in clinical settings [52, 53]. We did not use the intensity and extent of impacts as an alternative method of reporting the severity of oral impacts. Future studies using this scoring method may provide useful information on the extent and severity of oral health conditions in the target population when planning and evaluating oral health programmes and services . Additional study using the condition-specific specific C- OIDP may provide an oppurtunity to discrimate between groups with different levels of normative treatment needs in children when planning oral heath services .
The CFA and EFA conducted on the same data set in this study. This approach are accepted as less informative . Therefore, further research using CFA is necessary to investigate whether the factor structure can be replicated in the new dataset. The factorial structure of the OIDP in the total score version was examined in this study. Future study on understanding the cross-cultural differences in factor loadings and the impacts of use different scoring methods may provide additional insight into the interpretation of the OIDP factor structure [30, 49].
The main strengths of this study are that multivariate analysis was used to analyze the factors of OHRQoL and the factor structure of the C-OIDP was evaluated through both EFA and CFA methods. This study may provide a compherensive evaluation of the clinical, behavioural, socio-demographic and subjective factors that influence OHRQoL. Furthermore, the findings from this study revealed additional insights into the factor structure of the C-OIDP in the total score version.