We demonstrated that geo-maps based on caries risk can be used to monitor changes in caries risk over time. An important finding from the present study was that the geo-map concept was sensible enough for disclosing changes in geographical caries risk patterns among preschool children over a 4-year period, in spite of minor changes in absolute caries risk and stable contextual socioeconomic characteristics. We also underline that the Bayesian smoothing of the relative risks yielded essentially conservative estimates, meaning moderate sensitivity but high specificity [16
]. The regression to the mean problem, i.e. the phenomenon that a by-chance elevated (or lowered) relative risk changed towards the null from one occasion to another, was thereby tackled.
The geo-mapping methodology proved to be rapid and robust albeit, as always, depending on the quality of the input data. The high coverage of data (>75%) indicates that the findings are valid for the entire population. There is always a risk that non-examined children or children of families avoiding dental visits have an impaired oral health but that risk was low in the present study groups. In fact, the dominating reason for not being examined was a prolonged recall interval (up to 18months) due to an estimated low caries risk [10
The finding of an increased polarization of caries risk in spite of a generally improved dental health over the 4-year period was unwanted but not unexpected in the light of current papers on caries trends [17
]. The Public Dental Service in the region is strongly encouraged to work actively with CRA and individual recall intervals. It should however be stressed that the overall cumulative caries burden in the present study population was low from an international perspective [21
]. Nevertheless, the present time-trend geo-maps enable more detailed analyses for the possible local reasons for the certain changes in SmRRs on parish and/or clinic level. For example, the parish in the northern part of the region that fell from top-10 to the mid-range in four years has seen a rapid growth and urbanization with relocation of many young families. It is also important to realize that the ranking is relative. One parish with unchanged and low caries risk might be passed by one slightly improving its caries risk. Likewise, a decreased SmRR ranking might not necessarily mean that the caries risk de facto is significantly increased.
Our purpose was to disclose changes geographical patterns in childhood caries risk over time, without adjusting the parish-level relative risks (SmRRs) for contextual socioeconomic characteristics (which are geographically confounded factors [22
]). If a parallel change in a socioeconomic determinant of parish-level relative risk occurs, further adjustments of the SmRR can provide insights to what extent the altered contextual position had an impact. However, we did not observe any considerable parallel changes in the available contextual socioeconomic indicators. Beyond the contextual socioeconomic data, we had lack of data on possible explanatory variables. Of course, it is essential to collect such data for planned assessments of preventive care. In fact, in the Halland region, there is a plan to collect data from the dental clinics regarding preventive care and, also, a proposal to reallocate resources for preventive care. Notwithstanding those planned efforts, we have demonstrated that our proposed method based on repeated geo-maps provides a useful assessment tool.
We are fully aware that there is limited evidence for methods to bridge inequalities in early childhood caries and that compliance for both self-applied and professional measures is a paramount concern [23
]. However, for the preschool ages with adult responsibility, several controlled field trials have demonstrated that education, motivational interviews, coaching, tooth brushing and fluoride can significantly improve oral health and especially in vulnerable and deprived populations [24
]. Geo-maps addressing time trends can hopefully be used to increase the understanding and improve the quality of follow-ups in future studies with the goal to optimize the strategies to eradicate early childhood caries. Indeed, our findings strongly indicate that repeated geo-maps over time could be useful to monitor the effect of population-based interventions as well as targeted programs for individuals with increased caries risk. Interestingly, feed-back on interventions and programs might be available in a relative short time span. The effectiveness of such programs is probably not limited to dentistry but can also be valid for current and future school-based activities such as promoting a healthy lifestyle and fighting over-weight and obesity in childhood. We reinforce the assumption that the common risk factor approach (high-sugar diet, fats, etc.) should be at focus for the prevention of both dental and medical diseases [10