The efficacy of dental treatments often varies according to patient characteristics (e.g., age, gender, race/ethnicity). These variations are called heterogeneity of treatment effects (HTEs). Understanding how patient characteristics interact with treatment is key to improving treatment decisions for individual patients and for developing targeted treatments to reduce oral health disparities. However, studies that purport to show HTEs include numerous examples of misleading claims that have been later overturned (e.g., the lack of effectiveness of antihypertensive treatment in elderly patients has been refuted). Unfortunately, we lack the design and analysis strategies that can identify and confirm patient characteristics that cause HTEs. Our long-term goal is to develop HTE analytic strategies that can be used to personalize disease management and treatment, including preventive treatments. In pursuit of that goal, our primary aim is to develop an HTE framework that can identify a panel of patient characteristics that cause HTEs. Compared to traditional exploratory HTE methods, our new framework will identify complex interactions among patient characteristics.
Our second aim i s to demonstrate how the framework can be used to identify and confirm a panel of patient characteristics that cause HTEs in two NIDCR-funded caries prevention trials. Finally, our third aim is to develop freely available software to provide researchers with the needed tools to apply this framework to all dental trials. The proposed research is significant because it will provide dental researchers with the analytic tools they need to rigorously and effectively identify and confirm a panel of patient characteristics that cause HTEs. This study is aligned with the NIDCR's 2014-2019 strategic plan (Goal II) to enable precise and personalized oral healthcare through research and the National Research Council's endorsement of the emerging field of precision health. This project will also be highly time and cost effective because it will leverage the resources of two existing RCTs, an efficient use of current NIDCR studies.
Patients and dentists often make treatment decisions based on data representing the average (or overall) study results. This decision assumes that the average study population represents an individual patient even though a 'real' patient is likely different from the average patient, in terms of demographic or health behavior characteristics, known as heterogeneity of treatment effects (HTEs). This new strategy to identify patient characteristic that cause HTEs, proposed in this study, is aligned with NIDCR's 2014-2019 goal to enable precise and personalized oral health-care through research. NIDCR is committed to being at the leading edge of this development to embrace personalized, precision healthcare. A first step toward developing effective and personalized disease-management is improving the analytic strategies that can be used to identify and confirm HTEs, as proposed in this study.
