|
Abstract:
|
Tooth loss from periodontal disease is a major public health burden in the United States. Standard clinical practice is to recommend a dental visit every six months; however, this practice is not data-driven, and poor dental outcomes and increasing dental insurance premiums indicate room for improvement. We consider a tailored approach that recommends treatments if, when, and to whom they are needed thereby simultaneously reducing periodontal disease and resource expenditures. We formalize this tailoring as dynamic treatment regime which comprises a sequence of decision rules. The dynamics of periodontal health, visit frequency, and patient compliance are complex yet the estimated optimal regime must be interpretable to domain experts. We combine flexible non-parametric Bayesian dynamics modeling with policy-search algorithms to estimate the optimal dynamic treatment regime with an interpretable class of regimes. Both simulation experiments and application to a rich database of electronic dental records from the HealthPartners confirms the effectiveness of our proposed methodology relative to existing methods.
|
