Early Childhood Ear Diseases and Traumatic Dental Injuries: A Machine Learning Approach
Xiangqun Ju1,2, Ningsheng Zhao1, Pedro Henrique Ribeiro Santiago1,2
1Department of Oral Health Policy and Epidemiology, Harvard School of Dental Medicine, Boston, Massachusetts, USA.
Summary
Early childhood ear problems are significant predictors of traumatic dental injuries (TDI) in Australian children. This finding aids in predicting and preventing TDI among youth.
Area of Science:
- Pediatric Health
- Dental Public Health
- Epidemiology
Background:
- Traumatic dental injury (TDI) is a prevalent issue affecting children and adolescents globally.
- Understanding predictive factors for TDI is crucial for targeted prevention strategies.
Purpose of the Study:
- To predict the occurrence of traumatic dental injuries (TDI) in Australian children.
- To investigate the association between early childhood ear problems and later TDI development.
Main Methods:
- Utilized data from two large Australian longitudinal studies (LSIC and LSAC).
- Assessed 50 features including demographics, health conditions (ear diseases), and behaviors from early childhood (4-6 years).
- Employed machine learning algorithms to predict TDI prevalence at age 14, evaluating performance using AUC, sensitivity, specificity, precision, and F1 score.
Main Results:
- The study included 1746 Indigenous and 8357 non-Indigenous children.
- TDI prevalence at age 14 was approximately 13% in the Australian cohort.
- Machine learning models showed high predictive performance for the general and non-Indigenous populations (AUCs 0.85-0.86), with slightly lower performance for the Indigenous subpopulation (AUCs 0.76-0.80).
Conclusions:
- Early childhood ear problems were identified as strong predictors of traumatic dental injuries.
- These findings highlight the importance of addressing ear health in early childhood for TDI prevention.
