Improving treatment decisions from radiographs: effect of a decision aid

Abstract

Purpose: To test whether dental students' accuracy of treatment decision making for dentine caries using radiographs improved after using a decision aid (DA).

Methods: Dental students (n = 227) assessed the need for treatment for proximal surfaces of a test set of radiographs for a scenario patient. These 20 simulated bitewings contained 56 proximal test surfaces of which 14 were sound, 19 had enamel lesions and 23 had dentine lesions according to the histological "Gold standard". The students next used a DA, followed by re-evaluation using the same test radiographs. In this test-retest design, paired two-sided T tests were used to test for improvement in sensitivity (Se), specificity (Sp) and log diagnostic odds ratio (DOR) of treatment decisions for dentine caries.

Results: The students' initial mean Se, Sp and DOR were 0.66, 0.82, and 12.9 (SD 0.11, 0.10 and 8.2). After using the decision aid, mean values improved (0.67, 0.83 and 14.1) (P = 0.83, 0.07 and 0.04).

Conclusions: Using the stand alone DA led to a slight improvement in student performance.