Population health was defined in 2003 by David Kindig and Greg Stoddart, who described it as the health outcome of a group of individuals, including the distributions of such outcomes within the group. Today, I see the majority of companies – and that includes providers, insurers, and care management companies – equate population health with actions taken at the individual level to achieve those outcomes. For example, we find a group of people with diabetes, find out who has gaps in care, and get to each one of them through in-person, phone, video, or digital interactions. Because we focus on a cohort of people with diabetes, we say we are doing population health. As technology advances, we are developing new modes of interaction that are cheaper in unit cost and have greater reach (enabling us to access a broader group of people) or technologies that make for a better consumer experience. Much of the innovation I see though is still focused on making interactions better one person at a time. While this focus is important (and even critical), it leaves important opportunities on the table. 

True population health acts at both an individual and a structural level

Here’s an exaggerated example to make this point. Let’s say there is a four-way stop sign intersection with a high number of accidents every year. The individual (one person at a time) approach could result in putting ambulances right next to the intersection. That would be a great way to drive value and saving lives. A better idea may be to have caution signs just before the intersection to prevent the accidents from happening. Or even better may be just to replace the whole intersection with traffic lights, cameras and a delay between green lights. Or still better yet may be to not let the roads intersect at all – let one pass over the over and skirt the entire problem. All of these could work, but the degree of structural intervention increases dramatically in the latter cases. The structural intervention in this example is one that changes the external context (the intersection) in order to change the outcome (traumatic injury and death). 

Social determinants, because they are contextual in nature, are naturally suited for structural intervention

The CDC describes social determinants of health (SDOH) as conditions in the places where people live, learn, work, and play that affect a wide range of health risks and outcomes. In other words, context really matters, and we should address it if we want to improve health and outcomes. 

As an example, giving individuals a Lyft ride to their office visit like Humana recently did is a nifty structural intervention delivered at the individual-level using new technology (providing transportation via an app to enable access to care when they need it) and a great innovative idea. Creating better public transportation infrastructure is a larger-scale structural intervention (increase transportation options and thus mobility in an entire geographic area to enable access to care) that also turns out to have other benefits as well, like wage growth. Which is better? It depends (although for Humana specifically, the infrastructure option is not really a viable option given their distributed membership). These two approaches focus on lack of transportation and access to care as the barriers, but go about solving that problem in very different ways, with different timelines and different cost structures. In the past, structural changes were much harder and costly to do, but were also highly leveraged and durable once in place. Technology has brought new options to the table, and Humana’s use of Lyft is a very interesting example. We should expect to see a lot more of these as we begin to figure out how to go after social determinants in population health settings. 

Can you come up with examples where structural interventions might make sense? What if better air quality could prevent inpatient visits and asthma flares? What if free housing complexes could break the cycle of recurrent emergency department and hospital visits? What if we could build safe walking and biking paths instead of giving everyone a step tracker and asking them to exercise? These have all been done in isolated cases, but run up against the traditional ways we finance healthcare and the distorted business economics that make it challenging to justify at a larger scale. And yet, times are changing fast. Structural interventions might just make more financial sense (they have always made clinical sense) in a healthcare world moving towards value-based care and a greater appreciation of social determinants of health.