Randomized controlled trial of a computer-assisted learning program to improve caries detection from bitewing...

Area of Science:

• Dental Education
• Radiology
• Diagnostic Accuracy

Background:

• Accurate caries detection from bitewing radiographs is crucial for effective dental treatment planning.
• Standardized calibration methods are essential to ensure consistent diagnostic performance among dental students.
• Computer-assisted learning (CAL) offers a potential tool for enhancing diagnostic skills in dentistry.

Purpose of the Study:

• To evaluate the effectiveness of a computer-assisted learning (CAL) program in improving dental students' accuracy for detecting caries on bitewing radiographs.
• To assess whether CAL-based calibration enhances diagnostic performance compared to traditional methods.

Main Methods:

• A randomized block design assigned dental students to an experimental (CAL program with feedback) or control group.
• The experimental group calibrated against expert diagnoses and histological standards.
• Diagnostic accuracy was assessed using a digital test set, analyzed with summary receiver operating characteristic (SROC) curves.

Main Results:

• The experimental group showed a significantly higher true positive ratio (TPR) for dentine caries detection (76.3% vs. 66.9%, P=0.005).
• The area under the SROC curve (A(z)) was significantly greater for the experimental group (0.832 vs. 0.773, P=0.002).
• The diagnostic odds ratio (DOR) was significantly improved in the experimental group (12.4 vs. 8.8, P=0.003).

Conclusions:

• The computer-assisted learning (CAL) program significantly improves diagnostic performance in caries detection.
• Further refinement of cognitive feedback within the CAL program is recommended prior to widespread implementation.