TITLE:
Predicting Non-Attendance: A Model of the Complex Relationships in Dental Care Non-Attendance among Adolescents in Örebro County, Sweden
AUTHORS:
Annika Gustafsson, Carina Persson, Carina Källestål
KEYWORDS:
Adolescents, Dental Attendance, Broken Dental Appointments, ROC, Partial Area under the Receiver Operating Characteristics, Predicting Non-Attending Behaviour
JOURNAL NAME:
Psychology,
Vol.11 No.9,
September
17,
2020
ABSTRACT: Background: Understanding why adolescents miss dental appointments is central to promoting people’s oral health into adulthood. Studies on non-attendant adolescents are rare. Method: In our previous study using data from the 2014 population-based cross-sectional survey of schoolchildren in ?rebro County, Sweden we suggested a risk-factor model for non-attendance with five components (sociodemographic factors, self-rated health, perceived life events, health behaviour, social cohesion and social capital). The present study tested the external validity of the multivariate model for all components simultaneously developed from the 2014 study using a new population-based cross- sectional survey (n = 7576) of schoolchildren in ?rebro County, Sweden 2017. Using the predicted values from the 2014-logistic regression, we computed ROC-curves for the 2014 and 2017 study populations, respectively. Results: Of the 6 304 adolescents who answered the question of attendance to dental health care and all eleven independent items in 2017, 324 (5.1%) reported that they sought dental care only for acute pain or not at all, thus regarded as non-attendant. When using the risk model based on the 2014 data for the 2017 data, the area under the receiver operating curve (AUC) for non-attendance adolescents was 0.74 (95% CI 0.71 - 0.76). Conclusion: The present study shows that the model created from the 2014 population-based survey has excellent discriminative performance (AUC) in the 2017 population-based survey predicting non-attending adolescents. The methodology applied in this study may be useful to other health care services to develop predictive non-attending models based on their specific population.