Prevalence, Utilities and Treatment Decision Making by Dental Students

Methods: Fourth-year dental students (n=228) using visual analogue scales reported for a patient scenario on their perception of the prevalence of caries, and their utilities for FP and TP treatment outcomes between TN (optimal health, value 1.0) and FN (worst health state, value 0.0). Their sensitivity and specificity was calculated for their restorative treatment decisions from the radiographs of surfaces of teeth with either histologically dentine caries (45) or sound surfaces (60). We used multivariate regression to analyse association between their chosen point on the ROC curve (SE minus SP) and their perceived prevalence and aversion to FN (UTP-UFN) and FP (UTN-UFP) outcomes.

Results: The students on average valued treatment outcomes of FP at 0.36 (SD 0.28) and TP at 0.78 (SD 0.21) P<0.0001). Mean subjective prevalence was 0.22 (SD 0.17). The chosen point on the ROC curve significantly depended on their perceived prevalence (p=0.05). The coefficients for the utilities of decision outcomes had the expected sign, but were not significant (p=0.12 and p=0.52).

Conclusion: The variation in the students' restorative behaviour was consistent with their perceived prevalence and utilities. From the patients' perspective, FP and TP treatment outcomes might be expected to have equal value. However the students' utilities seem to reflect their retrospective view of the decision. This might be due to isolated emphasis on preventing over treatment in the curriculum, but may also reflect the difficulty of valuing outcomes from the patients' perspective