Mark Alan Fontana, PhD

My PhD (in economics) is from the University of Michigan; I did my post doc at USC; and I’m currently a data scientist and researcher at the Hospital for Special Surgery (HSS) in NYC. I’m also an Assistant Professor in the Department of Healthcare Policy and Research at Weill Cornell Medical College.

A digital native, my interests are interdisciplinary and empirical, focused on the intersection of health, policy, economics, survey and experimental design, machine learning and prediction, and musculoskeletal disorders. I have an overarching interest in applying my technical and analytical skills to problems that involve large-scale computational tasks. I’m fascinated by data collection methods not traditionally available to social scientists.

My work at HSS focuses on (1) collecting novel patient information through digital and mobile platforms (e.g., goals, presenteeism), (2) leveraging machine learning and large datasets to predict patient outcomes for creating pre-treatment decision-support tools and augmenting post-treatment patient monitoring (e.g., predicting patient-reported outcome measures, complications, readmissions, running injury), (3) evaluating the effectiveness of those tools beyond predictive power (i.e., do they actually help patients and doctors), (4) miscellaneous policy-related research (e.g., trends utilization of services following initial diagnosis of low back pain or osteoarthritis, or evaluating consumer-facing hospital quality rating systems), and (5) helping oversee and organize our office’s portfolio of analytics projects (e.g., the creation of various dashboards for tracking patients).

My postdoctoral work at USC focused on identifying genetic variants associated with behavioral phenotypes (e.g. risk-taking, depression), as well as supervising the design and development of a mobile platform for collecting behavioral and game data. I’m broadly interested in the genetic underpinnings of risk taking and mental health disorders, specifically how genetic predispositions can (1) mediate the impact of environmental changes (so-called “gene-by-environment” interactions) and (2) be used to develop personalized healthcare regimes.

My dissertation at Michigan investigated three (pretty much) unrelated topics: the pro-health impact of mobile push notifications (using data from a large consumers wearables company), adaptive survey methods for measuring financial risk tolerance, and how genetic predispositions moderate the impact of cigarette tax policy.

Related to the first, I’m interested in how health platforms implemented over mobile devices (“mobile health”) can improve the accessibility and quality of individual healthcare information, as well as how interventions delivered through these platforms can alter personal goals and spur pro-health behavior (e.g. increased physical activity, healthier sleep schedules).

Related to the second, I’ve spent years developing adaptive surveying software that explores how framing biases affect measurements of risk aversion and optimal asset allocation. Our goal is to inform selection of default investment plans (e.g. 401(k)’s, public pensions, etc) and develop a procedure that investment advisors can use to understand their clients’ risk attitudes and better tailor portfolio advice.

See my publications: