Hidden truths: norms in clinical practice

The idea of the norm holds promise of an elegant and easy to understand method of identifying expected and unusual behaviors. This thesis focuses on understanding the underlying community based data driven norms - the Hidden Truths, and their relationship with clinical practice. We are motivated by the significance of norms in healthcare and the impact of studying the prescription norms to solve real world problems. In this thesis, we propose the notion of a community based data driven de facto norm that can be used to understand patterns of practices in a health care setting. We propose desirable properties of de facto norms, and empirically investigate well-known statistical and machine learning methods to extract and validate these norms, and explore their real-world applications. The thesis utilizes two real world challenges, as example cases, with respect to over-use of drug prescriptions, one is the antibiotic resistance and second is the opioid epidemic.

De facto norms can be a useful tool for comparing and understanding differences in behavior and practices over time as well as across different settings or groups. Overall, community based de facto norms can provide a valuable complement to medical guidelines in shaping behavior and decision-making, as they can capture the complexity and diversity of individual situations and help to identify potential areas for improvement or intervention. This research highlights the applications of de facto norms in real-world scenarios and the importance of studying them for medical decision and policy making and stewardship programs.