Objectives: To compare the cost-effectiveness of four third molar guideline implementation strategies.
Methods: Fifty-one dental practices in Scotland were randomized to one of four implementation strategies. The effectiveness of the strategies was measured by general dental practitioners' compliance with the guideline.
Results: The effectiveness of the guideline depended crucially upon the type of patient treated. In particular, for a minority of patients (14%) with no clinical signals of their 'type', the implementation strategies generate potentially large gains in evidence-based practice. However, the cost per patient of achieving these gains is large given that the costs are incurred for all patients, but benefits accrue only to a minority.
Discussion: The results show that the type of patient presenting for treatment can influence the effectiveness, cost-effectiveness and therefore policy conclusions. Consequently, the design and analysis of studies need to be sufficiently sensitive to detect subtle interaction effects. This may explain the dearth of guideline implementation trials with significant findings. The results also suggest that a more cost-effective implementation method in primary care dentistry may be to subsidize treatment conditional upon patient type.