Population Health and Individual Health Are At Odds With Each Other
What’s Good For People May Not Be Good For You
Key Takeaway[s]:
· Population health and individual health can be at odds with each other
o Consider how direct primary care may appear as a solution to primary care concerns in the U.S. and can be considered good or bad depending on whether you look at it from the individual vs. population perspective
o Consider how individual factors—such as socioeconomic status and values of what constitutes a “good life”—
can inform the kind of individual risk they want to take
§ Certain treatments that may provide marginal improvements based on population data, but a physician and patient deem it worthwhile to try out and “see what happens”.
· In the world of guidelines and evidence-based medicine, only the physician can fill the role of re-particularizing population recommendations for individual contexts.
I always had an idealistic notion that if we improve population health—the focus on health outcomes for populations—rather than individual health—the focus on health outcomes of individuals—it would improve almost everyone's individual outcomes. I don't know where I developed these intuitions—but I felt them strongly as I interviewed at various medical schools and completed secondary essays asking for my opinions on key healthcare issues.
Like most of my beliefs, this turns out to be wrong.
Here's why.
The relationship between population health and individual health is not always chummy. The reason for this is straightforward. People are not averages since individual responses vary. Hence, aggregate data cannot be assumed to extrapolate to everyone else1. Yet the nature of public health initiatives requires trade-offs that may benefit a significant number of patients but hurt the individual health of certain patients.
One example occurred when I sat down to chat with another physician who runs a direct primary care practice [DPC]. Many physicians who are sick of the fee-for-service model in private practice move to a DPC that uses a membership fee-based service. However, some population health experts don't like the growing trend of direct primary care. Why? The average physician who runs a DPC takes care of fewer patients overall than if they worked in a hospital system or standard private practice. While the patients may feel like they get more time with their doctors and the doctor may feel less burnt out due to having more autonomy in their practice, from a population health perspective, it can still be seen as a net negative because it impedes adequate healthcare coverage.
Another example of population health and individual health clashing can do with specific treatments that are newer and provide marginal improvements based on population data, such as specific alternative healthcare treatments or novel pharmaceutical or procedural interventions.
I remember a clinical interaction in a volunteer clinic where a family practice doctor explained how a recent nutrition trial found no difference in weight loss between patients who went on a ketogenic diet and other dietary patterns. The patient shook her head and responded, saying, "Well, Doc! The ketogenic diet worked for me!". There might not be a conflict here at all. The doctor was speaking from a population perspective and could be correct. It could very well be the case that on the population level, the ketogenic diet does not outperform other dietary patterns when the goal is weight loss. At the same time, the patient was speaking from an individual health perspective and could be correct. It could be the case that for her context, due to support factors such as her lifestyle constraints, cultural belief system, or prior dietary habits, the ketogenic diet was a uniquely effective intervention to help her lose weight in her local context. Hence, the conflict materializes in the doctor's office. The only way to resolve it is for healthcare providers to bridge this divide by re-particularizing population health claims for individual cases as they interact with the patient.
Now that I have recognized the apparent schism between population and individual health across time, two key takeaways emerge for me.
The step of re-particularizing evidence is a vital part of evidence-based practice that is much more than knowing p-values and regurgitating the methods of a meta-analysis.2
It's not immediately apparent that an effective population health intervention will help people in your local context. Maybe it does. Maybe people around you slip through the cracks. Some people—like Geoffrey Rose—consider this a necessary evil. Others view this as something that must be actively fought against.
I refer to philosopher of science Nancy Cartwright’s paper to refresh my memory on why this is the case.
I cringe as I think about how many seasoned clinicians may roll their eyes at how obvious a point this is. However, to a lowly second year medical student like me—this blew my mind.