Study setting, sample size and study design
We report on a cross sectional study conducted in the Gampaha district, Western Province of Sri Lanka. The study population was 15-19 year old adolescents who were attending secondary government schools in the district, encompassing a total of 377 functioning schools. The sample size for the study was calculated by using the formula n = z2 p (1-p)/ d2 [18] and minimum sample was found to be 384. Since the study was designed to have cluster sampling technique, sample was adjusted by including the designed effect of 2.9[19]. Hence the final sample size was calculated as 1337 with the inclusion of the 20% non-response rate. As this study involved secondary school children (children above Grade 6) the minimum number of children was identified as 20, thus the number of children in a cluster or cluster size was decided as 20. Due to feasibility issues even though cluster size was determined as 20, two clusters were taken from one selected school. Sixty-seven (67) clusters were selected using the multi stage cluster sampling technique with probability proportionate to size of grade, across grades 10, 11, 12 and 13. Clusters were selected according to the school sampling frame.
This study received the approval from the Ethics Review Committee of the Faculty of Medicine, University of Colombo, Sri Lanka (Ref No EC 15-171). Administrative clearance was obtained from the Educational and Health ministerial personnel. The objectives of the study and the data collection methods were explained to participants, along with their parents, with written consent obtained prior to commencement. The field team was formed by a dental surgeon who carried out an oral examination on all participants; a retired school dental therapist for recording clinical data; and an assistant. Data were collected from the adolescents’ clinical oral examination as well as from two self-administered questionnaires: the OIDP to measure oral health related quality of life; and a second questionnaire developed for the study to collect information on socio economic characteristics and oral health care behaviours. Adolescents with dental problems were referred to the nearest government dental clinic where treatment was guaranteed.
Oral Impact on Daily Performance scale (OIDP scale)
The OIDP index was initially developed by Adulyanon [6] and later modified and validated in a 220 sample of 15 to 19 year-aged cohort of Sri Lankan adolescents [20]. During cross-cultural adaptation few items were modified with the consensus of the members of the expert committee. In addition, scoring system was revised to report only the severity of the impact and period was shorten to three months. This was based on the results of the pre-testing of the modified OIDP scale among a sample of 15-19 year-aged adolescents and in agreement with the expert committee members.
An assessment of factorability found that all correlation coefficients were > 0.30 with no item found to increase Cronbach’s alpha when deleted. The KMO measure of sampling adequacy was 0.87 and Bartlett’s test of Sphericity was significant (p<0.001), indicating that the data is suitable for factor analysis. Calculated Cronbach’s alpha for the study was 0.88, indicating good internal consistency reliability of the scale. No correlation was negative indicating homogeneity among the items. Concurrent validity was assessed by testing modified OIDP scale against self-reported perceived oral treatment need and perceived oral health problems. The relationships were significant (p<0.05) indicating that the instrument could adequately discriminate between adolescents who had did not have perceived dental treatment needs and adolescents who had different perceptions of overall health problems.
It contains eight items distributed between two domains: functional; and social and psychological. The functional domain includes items assessing the impact of oral health on: chewing and enjoying foods; talking and pronouncing clearly; and cleaning teeth. The social and psychological domain includes items assessing the impact of oral health on: good sleep without disturbances; being able to smile without embarrassment; maintaining usual emotional state without being irritable; school and household activities; and enjoying time with friends. See the additional file 1 to 5 for more details.
Revised OIDP scores were recorded on a six-point likert scale to reflect how severe the impact of each event was over the past three months, ranging from 0 indicating no impact, to 5 indicating a very severe impact. The total OIDP scores for individual domains were calculated as a simple sum of the response code. The potential functional domain and social/psychological domain scores ranged from 0 to 15 and 0 to 25 respectively. Total OIDP scores could range from 0 to 40. Higher OIDP scores indicated poorer OHRQoL. The primary outcome in this study was the total OIDP score. Domain specific scores were analysed as secondary outcomes.
Socioeconomic characteristics, oral health care seeking and oral health behaviour questionnaire
An additional questionnaire was developed to collect information about adolescents’ age, gender, family income (measured in Sri Lankan rupees and categorized in tertiles), oral hygiene habits (brushing frequency), consumption patterns of soft drinks, sugary items and fruits (categorized based on the frequency of consumption: occasional or regular); oral care seeking pattern (frequency of seeking care categorized according to the number of visits per year).
Clinical examination
The children were examined in a classroom at the school lying on ordinary desk under natural light. The clinical examination was exclusively visual, with the help of a dental mirror, CPI probe and a millimetre ruler. Biosafety measures established by the World Health Organization (WHO) were strictly followed [21]. The WHO criteria for the diagnosis of Decayed, Missing and Filled teeth (DMFT) were applied. The DMFT was categorized into caries free (DMFT=0), low severity (DMFT=1-4) and high severity (DMFT>4). Oral hygiene was assessed by Oral Hygiene Index-Simplified (OHI-S). Debris and calculus indices were calculated and there by OHI-S were calculated using the standard formula. OHI-S was categorized into good oral hygiene (OHI-S=0) and poor oral hygiene (OHI-S>0). Periodontal status was measured by assessing the bleeding status and pocket depth. The dental trauma data were analysed according to the presence of at least one kind of trauma or the absence of trauma. Dento-facial anomaly data were classified according to the need for professional intervention and the criteria were, 0= none, 1 = slight but no treatment needed, 2 = severe anomalies needing treatment.
Quality control
Quality control measures included a discussion of all possible classifications and criteria used in the study for the diagnosis of each oral health condition through an analysis of pictures of clinical cases for the disorders and diseases. An instruction manual for the field team was prepared and used during the training and throughout the data collection. Preceding the study, the inter examiner agreements were established. A specialist in community dentistry at Dental Institute, Colombo was considered as the gold standard and training involved 20 children of the same age and not belonging to the sample was performed to the test methodology. The inter examiner reliability was assessed using Kappa statistics. It showed a perfect agreement for both dental caries and bleeding. There was an 85.7% agreement for dental caries and 88.3% agreement for bleeding.
Data analysis
Statistical analyses were carried out using the Statistical Package for Social Sciences (version 23). Total OIDP scores, the overall mean OIDP score and scores for the individual domains were analysed for difference between specific oral diseases and disorders, and adjusted for socio economic characteristics and oral health behaviours. After applying statistical and graphical tests for normality, it was observed that distribution was positively skewed; hence non-parametric tests were used predominantly. Mann Whitney tests were used to compare the OIDP scores between different levels of dental caries, oral hygiene, trauma and malocclusion, with the level of significance was set to 5% (p <0.05). Spearman correlation was used to assess the correlation between the OIDP scores with the DMFT, OHI-S, debris and calculus indices.
Univariable and multi-variable logistic regression models were used to determine factors associated with the OIDP score. For analytical purpose the OIDP score was divided in to two categories, ‘good OHRQoL’ (OIDP score = 0; n=551) and ‘poor OHRQoL’ (OIDP score > 0; n=781). ‘good OHRQoL’ was considered as the reference category, thus, odds ratios are presented as risk factors/ protective factors of ‘poor OHRQoL’. The independent variables in these regressions included socio-economic characteristics (adolescents’ age, gender, family income and mothers’ education level), oral hygiene habits (daily brushing frequency), consumption patterns of soft drinks, sugary items and fruits and oral disease conditions (presence of dental trauma, anomaly, bleeding and pocketing and as well as number of decayed, filled and missing teeth due to caries and OHI-S index). A forward conditional process was used in the multivariable logistic regression analysis to select a parsimonious set of variables that predict the dependant variable. The study conformed to the STROBE guidelines.