Search

Background: There is great potential for routinely collected data to be used to examine oral-systemic conditions that share common determinants such as childhood bodyweight and dental caries. However, knowledge of the current use, quality and collection of such routine data is limited. The concept of this thesis was that of the student and the purpose was to conduct a quality assessment of routinely collected data as it is currently being used within research that investigates the childhood dental caries-bodyweight relationship. National Dental Inspection Programme (NDIP) dental caries and Child Health Surveillance Programme (CHSP) bodyweight data were used as examples of routine data throughout. Aims: The research had two aims:i. To appraise and summarise the available evidence on the childhood dental caries-body weight relationship and to determine the use of routine data variables within studies which look at this relationship.ii. To explore, quantitatively and qualitatively, the quality of the routinely collected data used within a data linkage study that investigates childhood dental caries-body weight relationship. Methods: Three sequential studies were conducted:i. An umbrella review: Five electronic databases were searched together with systematic review databases, reference lists and grey literature to identify systematic reviews that investigated any relationship between body weight and childhood caries experience. The records were retrieved and reviewed independently by two reviewers using a pre-piloted data extraction tool, the AMSTAR 2 checklist and additional reporting items from the PRISMA statement. Routine data sources were identified by reviewing methodology section of individual studies included in each identified systematic review.ii. A data quality assessment, linkage and exploratory analyses of a dental caries-bodyweight dataset: An assessment of data quality (completeness, consistency, timeliness, integrity and validity) of the linked routinely collected data was undertaken using the Data Quality Assessment Framework (DQAF). NDIP dental caries, CHSP bodyweight and local demographic data were linked. Finally, the linked data were analysed to explore whether a relationship exists between dental caries experience and bodyweight when routine data is used. iii. A qualitative exploration of oral health data collection by NDIP dental teams: A purposive sample of 12 participants were recruited to discuss behaviours, personal experiences and perspectives of collecting local routine data within the school environment to ascertain how these could affect the quality of the data collected. Data were collected through focus groups/one-to-one interviews and analysed using the Framework Method. Results: The umbrella review found the reporting and methodological quality of the included systemic reviews to vary. An inconsistent pattern in the direction and magnitude of the relationship between childhood dental caries experience and bodyweight was found. The review of routine data use within individual studies found that 21/75 index publication analyses made use of a routine data sources to gain information on either bodyweight or dental caries experience at the level of the individual child. The second study found that NDIP data can be linked to other routinely collected datasets such as CHSP and demographic information via a unique identifier. Using the DQAF criteria for data quality, the quality of the dataset used within the analysis was found to be poor with regard to completeness, integrity and consistency. The dataset was found to be satisfactory with regard to timeliness and validity. A statistically significant relationship was not found to exist between dental caries experience in children in Tayside, Scotland when looked at through linked routinely collected NDIP and CHSP data. The final study found that aspects of the NDIP school inspection process before, during and after the examination visit could all impact on routine data quality. Features such as the environment where data were collected, team working, data protection and communication between the dental team and school could all impact on collection and therefore data quality. A significant finding was that data completeness could be compromised where data were received or transferred between services, systems or individual people. Conclusion: The available evidence found that routine data variables are frequently used to examine the childhood dental caries relationship. Quality assurance and knowledge of the exact origin of the data influenced by the circumstances in which it is collected are important factors to consider. These findings raise a concern around the accuracy of routinely collected data used a large observational study and therefore may limit its use within epidemiological research that is looking to provide accurate and reliable results.

Exploring the association between neighborhoods and dental caries could provide a way to address oral health inequalities in a population by developing multilevel oral health community interventions. This study examines neighborhood effects on the severity of caries among African Americans, selected in a multistage probability sample of African-American families living in Detroit. This study used cross sectional data on 1,021 adults obtained in the first wave of interviews and examinations conducted in 2002--2003 by the Detroit Center for Research on Oral Health Disparities. Multilevel analyses focused on 27 neighborhood clusters derived from the original sample and involved a combination of individual (level-1) and neighborhood data including Census and geocoded information (level-2). There is statistically significant variation in the severity of caries (DS and DFS scores) among low-income neighborhood clusters; however, the magnitude of this variation is less than 3%. On average, 36% of the neighborhood level variance in caries severity was explained by level-2 predictors. Direct effects on caries severity, after accounting for individual risk factors, were observed with the number of churches and grocery stores in the clusters. In particular, lower mean DS and DFS are associated with an increasing number of churches in neighborhood clusters, while there is an increase in untreated dental caries with a higher number of grocery stores in the neighborhoods. Individual predictors such as age, socioeconomic status, total sugar intake, oral hygiene and oral health self-perception explained approximately 15% of the caries variance within the neighborhood clusters. The great inter-individual variability in caries is still mostly unexplained by classical risk factors for this condition. This study suggests that neighborhoods contribute something unique to the oral health status of caregivers, beyond their own socioeconomic position and individual risk factors, and adds to growing evidence that community and individual-based interventions are necessary to reduce disparities among African-Americans. Churches may offer a promising venue in which to conduct these strategies.