Early Childhood Ear Diseases and Traumatic Dental Injuries: A Machine Learning Approach
Dental Traumatology · 2026-06 · 42 (3) · 347-355
DOI: 10.1111/edt.70023 ↗Kurz gefasst
Fachartikel in der Zeitschrift Dental Traumatology, online erschienen am 21.10.2025. Verfasst von Xiangqun Ju und 5 weiteren Autor:innen. Beteiligte Einrichtung: Department of Oral Health Policy and Epidemiology, Harvard School of Dental Medicine, Boston, Massachusetts, USA. Hauptthemen laut MEDLINE: Machine Learning, Tooth Injuries, Ear Diseases. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 28 Literaturangaben. Gefördert von: The Commonwealth Department of Families, Community Services and Indigenous Affairs.
Abstract (Originalsprache)
Objective
Traumatic dental injury (TDI) is the world's fifth most common injury affecting children and adolescents. This study aimed to predict TDI based on ear problems and other factors among Australian children aged 4 years or over.
Methods
We used two longitudinal studies: the Longitudinal Study of Indigenous Children (LSIC) and the Longitudinal Study of Australian Children (LSAC). The outcome was the prevalence of TDI in children aged 14 years. We assessed 50 features, including demographic characteristics, health conditions (such as ear diseases) and health‐related behaviours of the children, their parents, and other family members in children aged 4–6 years. The performance of machine learning algorithms was evaluated using the area under the receiver operating characteristic curve (AUC) with its 95% confidence interval, along with sensitivity (recall), specificity, precision (positive predictive value), F1 score (the harmonic mean of sensitivity and precision) and accuracy (the proportion of correct predictions). These metrics were evaluated for the full sample and separately for Indigenous and non‐Indigenous subgroups, using both the full sample and stratified models. A sensitivity analysis was performed to compare model performance across the groups.
Results
This study included 1746 Indigenous and 8357 non‐Indigenous children and their parents and family members. The prevalence of TDI among Australian children at age 14 was approximately 13%. The full sample model applied to the full sample and the non‐Indigenous subpopulation demonstrated high predictive performance, with AUCs of 0.85 (95% CI: 0.82–0.88) and 0.86 (95% CI: 0.82–0.89), respectively. In contrast, the model applied to the Indigenous subpopulation demonstrated lower performance, with AUC values ranging from 0.76 (95% CI: 0.68–0.83) in the full sample model to 0.80 (95% CI: 0.70–0.83) in the stratified model.
Conclusions
Our findings indicated that early childhood ear problems were strong predictors of traumatic dental injuries in Australian children.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Department of Oral Health Policy and Epidemiology, Harvard School of Dental Medicine, Boston, Massachusetts, USA.
- Australian Research Centre for Population Oral Health, Adelaide Dental School, The University of Adelaide, Adelaide, South Australia, Australia.
- Department of Health Policy and Management, Harvard T. H. Chan School of Public Health, Boston, Massachusetts, USA.
ORCID-Profile: Xiangqun Ju
Förderung
- The Commonwealth Department of Families, Community Services and Indigenous Affairs
Zitieren
APA
Ju, X., Zhao, N., Santiago, P. H. R., Mejia, G. C., Jamieson, L., & Elani, H. W. (2026). Early Childhood Ear Diseases and Traumatic Dental Injuries: A Machine Learning Approach. Dental Traumatology, 42(3), 347-355. https://doi.org/10.1111/edt.70023
Vancouver
Ju X, Zhao N, Santiago PHR, Mejia GC, Jamieson L, Elani HW. Early Childhood Ear Diseases and Traumatic Dental Injuries: A Machine Learning Approach. Dental Traumatology. 2026;42(3):347-355. doi:10.1111/edt.70023
BibTeX
@article{Ju2026,
author = {Ju, Xiangqun and Zhao, Ningsheng and Santiago, Pedro Henrique Ribeiro and Mejia, Gloria C and Jamieson, Lisa and Elani, Hawazin W},
title = {Early Childhood Ear Diseases and Traumatic Dental Injuries: A Machine Learning Approach},
journal = {Dental Traumatology},
year = {2026},
volume = {42},
number = {3},
pages = {347--355},
doi = {10.1111/edt.70023}
}
EUVIMED