Autonomous virtual patients in dentistry: system accuracy and expert versus novice comparison

This article examines an autonomous virtual patient (AVP) system for identifying differences between novices and experts in dentistry. The two groups in the study were ten boarded or board-eligible experts (seven males, three females; mean±sd age 40±11) and twenty-six fourth-year dental students (fifteen males, eleven females; mean±sd age 27±3), who were defined as novices. All participants interviewed and mock-examined four randomly selected AVPs who had either orofacial pain or an oral medicine problem; they then selected needed diagnostic tests, diagnoses, treatments, and medications. The mean misrecognition rate of the software was between 13 and 19 percent. Data collected were examined for a difference between the two groups (novices versus experts) on multiple variables. Significant group differences existed in the final total score, the number of diagnostic tests ordered, and the number of medications selected. Novices reported that they found virtual patients to be a valuable educational experience. These data demonstrated that experts and novices asked essentially the same questions and spent similar amounts of time with the patients, yet the experts consistently scored higher and ordered fewer diagnostic tests and medications than the novices.