Study population and sample
We undertook a cross sectional study on adolescents aged 12–15 years, living in the National Capital Territory (NCT) of Delhi, and selected adolescents residing in areas representative of three different socioeconomic groups: urban slums, resettlement colonies, and middle-class neighbourhoods.
According to the Census of India in 2001 , urban slum areas lack security of environment, livelihood, amenities and tenure. These slums have been characterized by harsh physical and social environmental conditions like poor housing, insecurity of tenure, poor access to safe drinking water, sanitation and severe overcrowding . Resettlement colonies are located adjacent to a slum. The Government provides more infrastructure (e.g., water and electricity) in these colonies, however living conditions can still be very difficult. To summarize, a resettlement colony is much larger than a slum, with thousands of households organized into blocks.
Slums and resettlement colonies were recruited from a list of registered resettlement colonies and urban slums in Delhi. The inclusion criteria were: (a) colonies must be within a radius of 25 km from the research office, (b) both slum and resettlement colony should present together as a cluster, (c) colonies should have more than 500 households in both the components of the cluster (slums and resettlement colonies), and (d) a known non-governmental organization should be working in the community, willing to participate in the research. Fourteen colonies and slums were found to be eligible. In order to assess the variation in their population features, demographic data from two blocks in each of these slums and resettlement colonies were collected. All the slums and resettlement colonies were found to be demographically comparable (in terms of ethnicity, religion, language, number of households, population per block, and school going/non-school going children per family). For the adolescents from middle class neighbourhoods, private schools which have English language as medium of education and charge higher fees (‘English Medium Schools’) were selected. Multi-stage random sampling technique was used to select the study sample. Four slums and resettlement colonies were randomly selected from the 14 identified colonies and four English medium private schools were randomly selected from a sampling frame of 48 eligible schools.
The parents of adolescents in urban slums and resettlement colonies were approached with the help of local NGO representatives working actively in that area and were introduced to the study through a written information sheet. An informed consent form was given to the participants and interviews were conducted once the parents signed the consent form. For parents who are unable to read and write, the interviewer/field worker explained the study, including benefits and possible risks in a simple and easy to understand language. Thumb impression of the parents on the consent form was taken if they were illiterate after fully explaining the study. For obtaining the sample of children from middle and upper middle-class homes, the relevant school authorities were contacted. The selected children were then given an information sheet providing details of the study and consent form, to be signed by their parents. Once parental consent was received, the children were also asked to provide their consent to be a part of the study. Adolescents were recruited only when both the parents and the adolescent had signed consent forms.
We used relevant data from a large sample of adolescents residing in urban slums, resettlement colonies, and middle-class neighbourhoods in NCT Delhi. Overall, 1,600 adolescents were contacted and 1,386 agreed to participate (response rate: 86.6%). Of them, 840 participants reported “Yes” to the question “Do you have any of these problems (had toothache or sensitive teeth, had bleeding or swollen gums, been aware of decay in your teeth or a broken adult tooth, had ulcers or a loose baby tooth, had a problem because of tooth colour, shape, size or position) in the last 3 months” (Appendix). These 840 participants with self-reported oral health problems constitute the analytical sample for this study as the Child-OIDP questions (“Have any of these problems with your teeth and mouth led to difficulties with Eating, Speaking, Cleaning your mouth, Relaxing (including sleeping),Your feelings (for example being more impatient, irritable, easily upset), Smiling or laughing, Doing your schoolwork, Mixing with friends and other people”) were addressed only to them.
Data for this analysis were collected through an interviewer-administered questionnaire. The questionnaire measured material resources, neighbourhood social capital, social support, health-related behaviours (alcohol and tobacco use, diet, frequency of tooth brushing), and key health and sociodemographic variables.
The questions on health behaviours were derived from the WHO HBSC survey questionnaire and were aimed to assess habits like frequency of brushing, tobacco and alcohol consumption, dietary pattern and involvement in violent activity. Self-rated oral health of adolescents was measured by asking respondents as to how do they describe the overall condition of their teeth and gums .
The questionnaire included pre-existing questions and scales which were checked for reliability and validity in the study population during a pilot study to assess material deprivation in India . The response from the students was very positive and we achieved a response rate of 86.6%.
Translation of the questionnaire in Hindi and back translation to English was independently done by two people who were proficient in both languages (English and Hindi). It was also back translated in order to ensure that the true meaning of the questions was not lost in the Hindi version. Translation of questionnaire in Hindi was done keeping in mind the sensitivity to the local culture.
The Child–OIDP questionnaire was used as the measure of OHRQoL in this study. It assessed oral impacts on the following daily performances: eating, speaking, cleaning teeth, smiling, emotional stability, relaxing (including sleeping), doing schoolwork/homework, and social contact.
The Child-OIDP extent was created by counting the number of oral impacts on the aforementioned daily performances and ranged between 0 and 8, with every unit representing one more oral impact reported. The explanatory variables included in the study were ‘Age’, ‘Gender’, ‘Educational level of adolescent’, ‘area of residence’, ‘wealth index’, ‘bothered by oral health problem ‘, self-rated oral health’ and ‘self-rated general health’.
The wealth of the adolescent’s households was assessed by asking them questions about various material assets (television, car, electricity at home, bicycle, built-in kitchen sink, hot running water, washing machine, dishwasher, refrigerator, domestic help, mobile/cellular phone, bullock cart, computer, stereo system, livestock, internet access, motorbike and a second home). The responses on these households’ assets were used to create the variable “wealth index” using Principal Component Analysis (PCA). The variable ‘wealth index’ which was used to understand the socio-economic status (SES) of adolescents, was categorized as ‘poor’, middle’ and ‘rich’ . The perception on difficulty due to oral problems was assessed by the question- “To what extent have you been bothered by the problems of your mouth and teeth?”, and its related variable ‘bothered by oral health problems ‘was generated on dichotomizing the responses into ‘yes = 1’ and ‘no = 0’. Similarly, the perception on oral health and general health were assessed by the questions “How would you describe the overall condition of your teeth, denture and gums?” and “How would you describe your overall health?”. We also dichotomised the responses of these variables (self-rated oral health and self-rated general health) into ‘good = 1’ and ‘bad = 0’ for analysis purposes.
The internal consistency reliability was tested by using the standardised Cronbach’s alpha coefficient, item–total and inter–item correlations.
As mentioned earlier, of the 1,386 adolescent children 840 adolescents reported experiencing an oral health problem (toothache or sensitive teeth, bleeding or swollen gums, decayed or broken tooth, ulcers, tooth discolouration) and were subsequently asked the Child-OIDP questions. These adolescents comprise the analytical sample for this study. Descriptive analysis included percentage distribution of categorical variables. Bivariate analysis using Chi-squared test at 5% significance level (two-tailed) was used to understand the association between outcome and demographic (age, gender), socio-economic (education, place of residence, wealth index) and health variables (self-rated oral health, self-rated general health, bothered by oral health problem). The Child-OIDP extent score varied from 0 to 8. We further dichotomised the Child-OIDP extent into those with no oral impact (Child-OIDP = 0) and those with at least one oral impact (Child-OIDP = 1) to understand the association between prevalence of oral impacts and other selected variables.
Content and Face validity
Content validity of the Indian version of the Child-OIDP was established by seeking feedback from subject matter experts in oral health, non-communicable diseases and social determinants of health. A pilot study was also undertaken on an independent sample of 50 adolescents to further assess the acceptability and confirm the appropriateness of the layout, translation and sequence of questions.
During the pilot study, the respondents were asked to provide input on the way questions were asked, their perceived difficulty in understanding a particular question, the sequencing of questions and whether anything important has been missed. This helped us to establish the face validity of the Indian Child-OIDP.
The construct validity was assessed by looking at the associations between the Child-OIDP (extent) and three variables (‘self-rated oral health’, ‘self-rated general health’ and ‘bothered by oral health problems’) through a non-parametric test (median test) on the distribution of the OIDP extent.
Minimally important difference (MID)
The study utilized the distribution-based approach to calculate the MID for the Child-OIDP extent. Distribution based MID methods, i.e., Effect Size (ES) and Standard Error of Measurement (SEM) were used due to the cross-sectional nature of data . To estimate the ES and SEM, we dichotomized the variables ‘self-rated oral health’, ‘self-rated general health’ and ‘bothered by oral health problems”. ES is defined as the difference in mean scores between groups divided by the standard deviation of both groups. The magnitude of ES < 0.20 SD, 0.20–0.49SD, 0.50–0.79 and ≥ 0.80 SD is classified as negligible, small, moderate, and large, respectively. SEM was estimated by standard deviation multiplied by a square root of one minus the internal consistency of the Child-OIDP extent. All statistical analyses were performed using Stata software (version-14.0) .