Who attends a Children's Hospital Emergency Department for dental reasons? A two-step cluster analysis approach

Emergency departments (EDs) have been identified as key providers of dental care although few studies have examined patterns of attendance or clusters of characteristics. The aim was to identify the reasons for visits to an ED, whether these remained stable over time, and characterize clusters of patients by socio-demographic and attendance variables.

Pseudonymized data were obtained for children who attended the ED in 2003-2004, 2004-2005 and 2012-2013. Presenting complaint was categorized as attending for dental or nondental reasons. Other variables analysed included patient (age, sex, ethnicity and deprivation) and attendance characteristics (distance travelled, season, nature of complaint, time elapsed since onset of symptoms, day of week and hours of attendance), together with treatment outcome (advice, antibiotics and referral). To assess trends over time, analyses were conducted on patient, attendance and treatment outcome variables. To examine whether patients could be characterized by socio-demographic and attendance variables, a two-step cluster analysis was undertaken on 2003-2004 data set and validated on 2004-2005 and 2012-2013 data sets.

In 2003-2004, 550 children attended the ED for dental reasons rising to 687 in 2012-2013. The most important predictors of dental attendance were as follows: nature of complaint, ethnicity, time elapsed, sex and deprivation of the area in which children lived. The analysis showed two clusters: cluster 1 was comprised of children who attended the ED for dental injury, were of White ethnicity and attended within 24 h of onset of symptoms. Children in this cluster were likely to be from the least or less deprived areas (compared to Cluster 2) and were more likely to be males. Cluster 2 comprised of children attending the ED for caries, oral mucosal lesions or other complaints, were likely to be of other (non-White) ethnicities and were likely to attend more than 24 h after symptoms began. Children in this cluster were more likely to come from the most deprived areas and were both males and females. The clusters varied according to treatment outcome; those patients in Cluster 2 were more likely to be prescribed medication, whilst those children in Cluster 1 were more likely to be referred to another specialty.

A significant number of visits to the ED were for dental reasons with two clusters of children. The results have identified groups of patients for whom appropriate dental provision is lacking and where targeted services are needed to improve outcomes for children and reduce the burden on EDs.