Related Experiment Videos
Can low accuracy disease risk predictor models improve health care using decision support systems?
D K Benn1, D D Dankel, S H Kostewicz
1University of Florida College of Dentistry 32610-0414, USA.
Proceedings. AMIA Symposium
|February 3, 1999
Summary
A new decision support system aids dental caries management by accurately identifying low-risk individuals. Monitoring early decay in this group and selectively filling deep lesions could halve annual fillings.
Area of Science:
- Dentistry
- Public Health
- Biostatistics
Background:
- Dental caries is a multifactorial disease.
- Current risk assessment models for caries have limitations in sensitivity and specificity.
- Existing models are inadequate for precise resource allocation to high-risk populations.
Purpose of the Study:
- To design a prototype decision support system for dental caries management.
- To improve the identification of low-risk individuals for caries.
- To explore a strategy for reducing annual dental fillings through selective monitoring and treatment.
Main Methods:
- Development of a decision support system integrating a risk assessment model.
- Evaluation of risk prediction model performance (sensitivity: 65%, specificity: 80% for predicting 2+ new lesions).
- Proposed management strategy: monitoring early caries in low-risk individuals, filling lesions only beyond 1/3 dentin depth.
Main Results:
- Risk models accurately identify low-risk individuals.
- Selective treatment of early caries in low-risk patients could reduce annual fillings by 50%.
- Current dental education often lacks caries risk assessment training.
Conclusions:
- A decision support system combined with a specific risk model can enhance caries management.
- Improved identification of low-risk individuals is key to optimizing dental resource allocation.
- Shifting from early intervention to risk-based monitoring may significantly reduce restorative procedures.