Subjects: The sample was a multistage stratified random one consisting of children attending primary schools. A group of mothers with their children aged 6 - 12 years participated in the study. The children are healthy ASA I and their native language is Arabic. They are joining primary schools either private or public in Jeddah city, Saudi Arabia.
According to Ministry of Education files, the number of registered children in the primary schools in the academic year 2017-2018 was 293,373 children (147,525 boys and 145,848 girls) in 737 schools in Jeddah districts. Based on the gender, schools were divided into boys’ schools (344) and girls’ schools (393) following the country’s regulations. Sample size was calculated based on previous research (18). A two-sided 95% confidence interval for a single proportion using the large sample normal approximation modified for a finite population of size 300,000 will stretch 0.025 from the observed percentage for an expected proportion of 35% when the sample size is 1392. The schools were stratified on three levels:
-
Four strata, by district (west, east, north, south).
-
Then by type each strata was stratified into public or private producing 8 strata.
-
Finally, by gender each of the 8 strata was stratified into boys or girls resulting in 16 strata.
Random number generator (19) was used to select one schools from each of the 16 strata resulted into 16 primary schools with approximately 7,181 children from different regions in Jeddah. To compensate for the drop out of participated mother-child pairs due to incomplete questionnaire response or the child resistance to oral examination, 1728 questionnaires were distributed. From each grade, one class was randomly chosen to join the study by the bowl method. In cases of a small number of children in any class (less than 18), another class was randomly selected, and in cases of a small number of children in the school (less than 108), the whole school children were included in the sample.
Questionnaire
The questionnaire had a cover letter describing the aim of the study and stressing confidentiality of the responses. An Arabic questionnaire comprised of four sections was used for this study. The questionnaire was completed by mothers. Section I, to record the demographic characteristics like child gender; and mothers’ age (40 years or less, 41years or more), educational level (High school or less, Diploma/University or higher) and monthly family income; (low, medium, or high). The informed consent was attached to this section. Section II, Hiroshima University– Dental Behavioral Inventory (HU-DBI)-questionnaire, to evaluate attitudes and behaviors of mothers pertaining their own oral health based on the sum of agree/disagree responses by giving one point to each favorable response of good oral health (20). This section contains 20 items asking about the protection of her teeth, regular visits to the dentist, ideal toothbrush, and toothpaste. To clarify item 5 (I use child-sized toothbrush), it was modified; into (I use recommended-sized toothbrush). High scores indicate good oral health attitudes and behaviors (21–29). Section III of the questionnaire was to recognize mothers’ knowledge about their children oral health (30). The present study used the last 2 sections (12 questions) concerning dental health and dental care for children. Mothers selected the correct response from multiple choices. One point was given for each correct answer. Section IV for identifying the mothers’ attitudes towards their children’s oral health, (31). It was composed of 13 agree/disagree questions. One point was given for each correct answer.
Arabic Translation of the questionnaire was carried out by two Saudi linguists. Forward and back-translation was used. The first linguist, who is fluent in both the Arabic and English languages, translated the questionnaire from English to Arabic, forward translation. The other linguist translated back the Arabic version into English language, backward translation. Then, a dental public health professor compared the two English questionnaires (the original one and translated one) and made modifications until the final Arabic translated version was formulated which was considered the final Arabic questionnaire (32).
Test-retest Reliability of the Arabic questionnaire was done. The final Arabic version of the questionnaire was completed twice with two weeks’ interval by a group of mothers (30 mothers), not included in the study. The results were compared using Pearson's correlation coefficient (Pearson's r) as a reliability test. Pearson's r between the two questionnaires’ responses was referred to as the coefficient of stability and were found to equal 0.90 which is considered excellent. Moreover, internal consistency was done to reflect the inter-correlation between items in the questionnaire and was quantified using the Cronbach's alpha to be 0.86 (33, 34).
Content validity of the Arabic translated version of the questionnaire was performed to recognize if the questionnaire items were representative of the entire theoretical construct that the questionnaire was proposed to measure. A panel of experts in pediatric dentistry were asked to rate each item in the questionnaire concerning relevance, clarity, simplicity, and ambiguity on a four-point Likert scale. The Content Validity Index (CVI) was calculated according to the methodology of Lynn MR and Polit et al and was found to be 0.84 (33, 34).
Three school visits were performed. The first visit for distribution of the questionnaire for the selected children in the class to submit to their mothers with the help of the class teacher. Each questionnaire was assigned a serial number which was matched with an examination sheet for confidentiality purposes. The second visit was carried out after one week, for collection of the questionnaires and clinical oral examination for the children who had a permission from their mothers to assess dental caries experience using two indices (dmft) and (DMFT) (35), and oral hygiene status of the children by Simplified Oral Hygiene Index (OHI-S); Debris Index component (DI); (Greene and Vermillion, 1964). A brief report was sent to the mothers describing the oral health status of their children and instructions for improvement and/or maintaining the oral health condition. Two groups of examiners, one for girls’ schools (two examiners) and one for boys’ schools (two examiners) shared in this study. Inter and intra-rater reliability test was done for Calibration and training of the examiners. Kappa statistic was calculated and found excellent (0.93 for female and 0.95 for male examiners). The inter-rater reliability between the examiners was evaluated using the Intra-class Correlation (ICC) and was found to be 0.96. On the third school visit, mother-child pairs were gathered for presenting an oral health education lecture and distribution of oral health instruction pamphlets to increase their awareness of the proper oral health knowledge, attitude and behavior regarding themselves and their children.
The research protocol was approved by The Research Ethics Committee at the Faculty of Dentistry, King Abdulaziz University with proposal number (008-16). In addition, approval was obtained from the local School Health and Education Directorate Authority; Ministry of Education; to provide the primary schools’ lists in Jeddah city districts and to implement the research among the schools.
Statistical Analysis
Data was entered, coded, and analyzed using the Statistical Package for Social Science SPSS (IBM Statistics for Windows, Version 23.0 Armonk NY: IBM Corp). Univariate analysis of data was carried out and presented as frequencies, percentages or means and standard deviations. The responses of the three questionnaires from public schools were compared to private schools using chi-square test. The association between the demographic variables and questionnaire score, DMFT, and OHI-S was evaluated using independent t-test or ANOVA test which was followed by post hoc analysis. A multiple linear regression was modelled to assess the effects of all the significant independent variables on the DMFT (dependent variable). The significance level was set as p<0.05.