Predictive model for temporomandibular disorder in adolescents: Decision tree

Area of Science:

• Oral health research
• Adolescent health
• Predictive modeling in medicine

Background:

• Temporomandibular disorders (TMD) affect adolescents, potentially impacting their development.
• Understanding TMD risk factors in adolescents is crucial for early intervention.

Purpose of the Study:

• To develop a predictive decision tree (DT) model for TMD in adolescents.
• To identify high-risk and low-risk groups for TMD development in Recife, Brazil.

Main Methods:

• Cross-sectional study of 1342 schoolchildren (aged 10-17) in Recife.
• Analysis using Pearson's chi-squared test, Fisher's exact test, and CHAID algorithm for DT construction.
• Utilized SPSS statistical software for data analysis.

Main Results:

• Prevalence of TMD was 33.2% among the study population.
• Significant associations found between TMD and sex, depression, self-reported orofacial pain, and clinically examined orofacial pain.
• The DT model, incorporating self-reported orofacial pain, clinical orofacial pain, and depression, achieved 73.0% predictive power.

Conclusions:

• The decision tree model demonstrates strong predictive capacity for TMD in adolescents.
• The model effectively identifies adolescents at high risk for TMD, particularly those experiencing orofacial pain (self-reported or clinical) alongside depression.