In this month’s Q&A, we sit down with Dr. Sussell to discuss his approach to research, causal inference, and his take on the most important drivers of change in the health policy landscape over the next 5-10 years.

Q1: What interests and experiences led you to PHE?

As a researcher, I was drawn to economics because I find the underlying premise of the discipline so appealing. At its core, economics is about using mathematics to describe and predict how different actors in society—consumers, manufacturers, insurers—will behave in different situations. Mathematics provides a great framework for thinking about public policy questions because it allows you to move beyond theory and conjecture. When you have an empirical approach, you can take this additional step of falsification—you can apply your model to some data and actually get a sense of whether your model was correct.

And this is what we do at PHE every day. Our whole model is applying multidisciplinary expertise to finding the right answer to difficult health policy questions. I would add that for me as an economist, the coolest thing about working at PHE is the ability to continually learn from colleagues with training that differs from my own—epidemiologists, statisticians, clinicians, and so on.


Q2: Why are economic methods for causal inference important in healthcare economics and outcomes research (HEOR)?

Causal inference is the branch of statistics dealing with identification of treatment effects. So a causal inference question might be something like “How does receipt of some treatment (like a drug) affect some outcomes (like overall survival)?” Randomized controlled trials (RCTs) are widely understood to be the gold standard design for causal inference, but there are frequently situations where RCTs are logistically or ethically infeasible. For example, patients who participate in clinical trials are known to be very different from patients who don’t. If a researcher is interested in understanding efficacy in a real-world population, prospective randomization just isn’t going to be possible. Economics provides a really valuable set of tools for conducting nonexperimental causal inference research in these situations. For example, some of my PHE colleagues use a strong design called “instrumental variables” to estimate how receipt of oral nutritional supplements affects healthcare spending for Medicare patients hospitalized with chronic obstructive pulmonary disease (COPD)  https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4451705/.


Q3: What do you think will be the most important drivers of change in the health policy landscape over the coming 5-10 years?

The Affordable Care Act (ACA) was the largest federal health policy intervention in several decades. It was enacted 8 years ago, but it remains controversial as evidenced by the elimination of the tax penalty associated with the individual mandate, and the near passage of “skinny repeal” in 2017.

So, to my mind, the major driver of change is going to be political. There is a high likelihood of new legislation to undercut the ACA, as well as some type of block-granting of Medicaid. On the flip side, there are increasingly vocal arguments for some form of “Medicare for All.” The outcome of upcoming elections will certainly influence the stance the United States takes on the role of government in health insurance coverage.


Q4: What do you like to do in your time off?

Weekends generally involve hiking or heading to the park. My wife and I have two small children—our daughter Nico is 3 and our son Desmond is 18 months—and a giant Labrador/Great Pyrenees mix named Charlie. We’ve found that sun exposure and exercise are strong predictors of long naps, which are in everyone’s best interests.