131 - Predictive Models for Pediatric Caries Risk: A Scoping Review

Abstract:

Purpose: This scoping review synthesized evidence on the diagnostic accuracy, predictive validity, and clinical utility of caries risk assessment (CRA) tools and predictive models for children and adolescents. The objective was to evaluate the performance of established instruments and assess advances in computationally informed approaches.

Methods: A structured PubMed/MEDLINE search identified English language publications through February 2025 using predefined terms related to dental caries, risk assessment, predictive modeling, and pediatric populations. Seventy-seven records were retrieved, with 29 studies meeting inclusion criteria after screening. Extracted data included tool characteristics, validation methodology, performance metrics, and implementation considerations. Reference lists of key studies were also reviewed to ensure comprehensive coverage.

Results: Established CRA tools, including the Cariogram, age modified programs, and simplified checklists, demonstrated moderate predictive validity, with AUC values generally ranging from 0.65 to 0.75. Performance varied across age groups and settings. Emerging approaches incorporating salivary biomarkers, microbiomic profiling, and machine learning algorithms showed higher predictive potential, often achieving AUC values above 0.80 in development cohorts. However, these innovations remain limited by insufficient validation, inconsistent methodology, and uncertain feasibility for broad clinical use. Implementation analyses identified barriers involving workflow integration, clinician training, cost, and acceptance.

Conclusions: CRA methodologies in pediatric dentistry range from clinical instruments to computational models. While traditional tools offer imperfect but useful stratification, emerging approaches may improve predictive precision. Continued progress requires external validation across diverse populations, cost effectiveness analyses, and implementation frameworks that support translation into clinical practice.

Identify Supporting Agency and Grant Number: