A risk prediction model for xerostomia: a retrospective cohort study

Area of Science:

• Oral Medicine
• Dental Public Health
• Clinical Informatics

Background:

• Xerostomia (dry mouth) affects a significant portion of the dental patient population.
• Identifying risk factors and predicting xerostomia are crucial for timely intervention.
• Existing models may not comprehensively incorporate diverse risk factors.

Purpose of the Study:

• To determine the prevalence of xerostomia in a large cohort of dental patients.
• To develop and validate a predictive model for xerostomia risk.
• To identify key risk factors associated with xerostomia.

Main Methods:

• Retrospective data collection from 12,682 new dental patients (2010-2013).
• Inclusion of socio-demographic data, medical history, and self-reported dry mouth symptoms.
• Logistic regression framework for model building and external validation using ROC-AUC.

Main Results:

• Xerostomia prevalence was 12.2% among the study population.
• Increased medication use (OR=1.11) and recreational drug use (OR=1.4) were associated with higher xerostomia risk.
• Rheumatic (OR=2.17), psychiatric (OR=2.34), eating disorders (OR=2.28), and radiotherapy (OR=2.00) were significant predictors.
• The prediction model achieved high accuracy (ROC-AUC=0.816, external validation=0.799).

Conclusions:

• The developed xerostomia risk prediction model demonstrates high accuracy.
• The model effectively differentiates between high- and low-risk individuals for dry mouth.
• Clinicians can utilize this model to identify at-risk patients based on medications and systemic diseases.