#037: The unlikely evolution of Population Health
About health EQ
Every month, you'll get actionable ideas and resources to drive change and expand your knowledge base.
Greetings, my Gurus!
ICYMI: I just announced that Health Data Guru is evolving into health EQ, a name that better reflects my focus on exploring the data, decisions, and dynamics shaping U.S. healthcare.
Naturally, I want to immediately start exploring all the meaty topics I’ve been contemplating this year, like unpacking social and cultural contexts of health, interrogating determinants of health, thought-provoking article reviews, and community-based participatory methods (to name just a few).
But instead, I’ve decided to start with something more fundamental to set the context for all future conversations you’ll find here at health EQ.
Subscribe so you never miss a new post!
Today we’re diving into Population Health.
It’s a topic we think we know well, but do we really?
Its roots—along with the limitations and distortions of modern interpretations—might make you see it in a whole new way.
Population health has gained significant traction over the last 10 years. Just check out this Google Trend showing the relative popularity of the term (compared to Health Equity and Value-based Care for giggles).
SO WHAT DOES IT MEAN?
Population Health: a Definition
Population health emerged prominently in the 1990s, influenced by Canadian frameworks prioritizing social determinants of health (Kindig & Stoddart, 2003). Its adoption in the U.S. gained momentum with a growing focus on health inequities and systemic drivers of poor health.
David Kindig and Greg Stoddart’s succinct and widely-cited definition of population health is:
“the health outcomes of a group of individuals, including the distribution of such outcomes within the group.”
This definition incorporates four important components:
Health Outcomes: Including both quantity (e.g., life expectancy) and quality (e.g., health-related quality of life).
Determinants of Health: Considers medical care, individual behaviors, genetics, social and economic factors, and the physical environment.
Policies and Interventions: Focuses on targeting more structural determinants to improve health outcomes.
Distribution and Variability: Focuses not just on the outcomes of the population as a whole, but specifically exploring the variation in outcomes within groups and between groups, and understanding why those variations exist.
Despite this definition’s broad applicability and attribution in academic literature—scope and applications still vary widely in practice, partly due to competing frameworks and priorities among stakeholders.
At first blush, this may not seem like a big deal because different perspectives, different focuses, right?
Until you realize that you could ask 10 people about “pop health” and probably get 10 different answers—largely depending on where that person sits in the healthcare ecosystem and what their individual background or experience is.
Population Health Management: Missing the Mark?
You may also notice that the simple definition overlooks critical aspects of healthcare operations. As such, "population health" evolved over time from academic discourse to practical applications and implementations, including the rise of the term "population health management," which seeks to align healthcare delivery with population health goals (Itchhaporia, 2021).
Population health management is a derivative but distinct concept that refers to the operationalization of population health principles specifically within healthcare systems, with typical healthcare concerns of cost, quality, and utilization—such as:
Total Cost of Care: Understanding how interventions affect overall expenditures.
Medical Economics: Balancing quality improvements with cost containment.
Quality Measures: Utilizing frameworks like HEDIS to track gaps in care.
Alternative Payment Models and Contracting: Aligning financial incentives and networks with population health goals.
Disease and Case Management: Addressing chronic conditions through proactive care coordination.
Risk Stratification: Categorizing populations based on their health risks to tailor interventions effectively.
Population health management is a more narrow, biomedical perspective that focuses on using data-driven approaches to segment populations, identify risks, and implement targeted interventions for specific disease conditions (like diabetes) or adverse event outcomes (like emergency department visits), commonly tied back to a payment incentive or quality measure (which drive many goals and operational programs in healthcare).
Though more narrow and geared toward the healthcare system, “population health management” also suffers from a lack of clarity around its exact scope and boundaries. For some, it represents a pathway to value-based care through initiatives like care coordination, disease management, and addressing gaps in care. For others, it remains narrowly tied to cost containment and quality metrics, such as HEDIS measures.
So here’s the rub:
All of this ambiguity can muddy (even distort) the original “intent” of population health. Some of this is to be expected, as different perspectives reflect the varied goals and unique interpretations people bring to the conversation, shaped by their objectives, experiences, and contexts. But it is still important to note because it creates barriers to mutual understanding, incentives, and alignment on scope, responsibility, and measuring success effectively across the board—not only between public health and healthcare, but even between like-minded organizations working together on a “population health” initiative. And in some cases, even between different roles within those organizations (i.e. clinicians vs. administrators).
More specifically, consider how:
Public Health Practitioners might see it aligned with their mission to prevent disease, promote health, and protect communities through organized interventions and policies at the population level. They may focus on quality of life indicators, ensuring community access to essential resources, and tackling social drivers of health with a strong emphasis on prevention, health promotion, and harm reduction.
Managed Care Professionals or those on the payer side might see population health as systematically improving outcomes for a defined group of members through data analytics, risk stratification, and targeted disease or care management interventions. Their lens often focuses on high-risk patients, Quality programs (such as HEDIS), and managing total cost of care through utilization management and network contracting or alternative payment models.
Healthcare Delivery Professionals on the administration side may see it as taking responsibility (and possibly even financial risk) for the health outcomes of patients attributed to their system or certain patients attributed to their system, such as members of a certain health plan—not just those actively seeking care at the moment within their four walls. They may emphasize proactive outreach for preventive services and other "gaps in care," capitation management, managing to certain types of episodes or reduction in certain outcomes, like readmissions.
Clinicians and Practitioners serve as the point-of-care implementers for many of the programs mentioned above, connecting individuals to broader initiatives. They are uniquely positioned to see how each individual is part of the larger population (and likely, community), which means they understand community needs more deeply and have a desire to connect patients with appropriate resources. Their role in linking individual care to population-level strategies is integral to the success of population health efforts, but they may not resonate as much with the administrative or payment-based incentives.
Any combination of the above, plus others.
The definition of a population is just as broad and ambiguous as the definition of population health.
I've seen populations defined as:
All members of a health plan
A subset of members of a health plan, like an employer group
A cohort or subgroup within the subset, like members in an employer group with diabetes
A line of business, like Medicare Advantage
A geographic region, like a state, city, or region
A specific patient registry, like people with diabetes who also have other comorbidities and are due for their eye exam
A specific group of patients you’re “responsible for,” such as those with a certain PCP or those whom you are paid for in a capitated arrangement because they belong to some kind of group
Any group of something you’re studying
Any combination of the above, plus others.
None of these definitions are inherently wrong, but they do show the many ways a "population" can be interpreted within the may ways population health itself can be interpreted.
An Example to Ponder
Consider social determinants of health, like housing, education, and employment opportunities.
These factors are certainly encompassed under the original intent and focus of population health. So whose job is it to address—let alone fix—them? Depending on where you sit, you may say the healthcare organizations should step up to the plate and invest (which they’re starting to do, albeit more in a typical ‘address the symptom’ manner). Or you may say public health or government, because that is a social issue beyond the scope of healthcare. Or you may say even say no one in healthcare or public health at all, because that is a cross-sector policy issue and how could healthcare impact affordable housing, anyway?
And even if we could agree that it still does belong in our definition and scope of population health work, the challenge then becomes aligning it to incentives, ROI, how far “upstream” we really go to “address” them—do we provide some housing vouchers in our community to certain people, or do we start advocating for and putting our company’s name to policy recommendations that address affordable housing?
You can start to see how the variability in definition, alignment, and scope (and frankly, financial incentive and sociopolitical will) may end up pushing us to smaller scale, more palatable, more operationally feasible, and more financially/sociopolitically “sound” initiatives.
Of course, we all understand the various pressures of why it is this way, but it leaves me wondering how we might actually make a dent in some of these issues when we’re approaching it this way—which illustrates the issue I’m contemplating about scope.
THINKING DEEPER ABOUT WHAT WE TAKE FOR GRANTED
Historical Roots: From Astronomy to Darwin
The concept of analyzing populations traces back to early scientific disciplines like astronomy and evolutionary biology, which started to explain patterns in groups rather than zeroing in on individual components. Necessary, absolutely—but what we would come to find is it set the stage for a challenge we still wrestle with today: seeing “populations” as an aggregate group of sums and averages, instead of a collection of unique, complex individuals interacting with and influenced by one another and their surroundings in infinitely combinatorial ways that sometimes seem to be anything but the expected (average).
Despite these challenges, the evolution of the science is interesting and can help us rethink how we approach it today.
This begs the question: is it a function of the overly-simplistic modeling we use, the complexity of the population’s behavior, or both?
Adolphe Quetelet, a Belgian astronomer and mathematician, adapted astronomical methods—originally developed to predict celestial behavior—to study human populations. For example, just as astronomers used averages to predict planetary orbits, Quetelet used statistical averages to highlight societal norms, though often at the expense of individual variability. Quetelet’s concept of the "average man" (l'homme moyen) became a statistical abstraction that described central tendencies within populations, revealing societal dynamics like crime rates and mortality influenced by social and environmental factors. These methods helped identify patterns and averages in human groups—very cool.
Charles Darwin’s theory of natural selection took it further, emphasizing the interplay between organisms and the influence of their environments. For example, Darwin demonstrated how environmental pressures shaped survival and reproduction of an individual organism, which parallels how environmental, social, and genetic factors shape health outcomes for an individual in population health terms.
While Quetelet’s and Darwin’s contributions revolutionized the study of populations, they reveal limitations that persist today.
The early reliance on averages in population models often obscured the variability within groups. For example, Quetelet’s "average man" concept focused on statistical means to describe societal norms but failed to account for the dynamic and relational aspects of populations, often treating them as static aggregates of individuals (Krieger, 2012). Darwin, by shifting from "errors" to "variation" in his work on natural selection, started to introduce a nuanced understanding of diversity within populations and their environmental interactions, paving the way for more complex considerations of population dynamics that will be critical to the future.
Limitations of Sums and Averages: the False Dichotomy of the Individual and the Population
Nancy Krieger, a thought leading scholar in the field, critiques these over-simplified, or reductionist, approaches, pointing out that populations are not merely averages or the sums of their individual parts, but dynamic, relational entities—each shaped in unique ways by historical, political, and socio-economic forces. In fact, she metaphorically posits that a population is more like a sentient organism in this way, introducing the concept of "structural chance" to describe how systemic inequities create opportunities and constraints that influence health outcomes at both individual and population levels in confounding ways (Krieger, 2012).
Similarly, Arah (2009) also challenges the notion of a strict separation between individual and population health, emphasizing their inherent interconnection that functions like a ripple effect. He reminds us of "system causation," wherein individual health outcomes and population trends are dynamically linked. Individual behaviors and risks cascade upward to influence broader population patterns, while systemic and environmental factors also trickle downward to shape individual opportunities and exposures (Arah, 2009). The interconnectedness of this must be taken into account. Krieger also points out that the effects of systemic stressors on individuals at a population level may actually be eventually embodied into their physical manifestation. For instance, chronic stress from systemic inequities—such as living in racially segregated neighborhoods—may influence epigenetic expression, in essence physically incorporating what were previously external factors.
Consider hypertension in a community: individual cases might stem from genetic predispositions, poor diet, or stress. However, these factors are intricately connected to systemic determinants, such as limited access to fresh food, unsafe environments that discourage physical activity, and economic instability. These dynamic scenarios create a continual feedback loop where population-level patterns, shaped by systemic inequities, amplify individual risks, and individual health outcomes contribute to broader community trends (Krieger, 2012; Arah, 2009). If we rely simply on sums, averages, and health outcomes without incorporating the compounding impact of systemic/external factors, we are not only left with half a picture, but we’re also left with oversimplified aggregate models that obfuscate the rich complexity that influence outcomes and, if better understood, could possibly inspire better solutions—could that be one reason why we’ve not been able to drive large-scale change?
You can start to see where our current models start to fall short.
These examples highlight the inadequacy of models that rely solely on averages and simple summations, as they fail to capture complex interplay (Krieger, 2012). Current models often fail to account for the dynamic interactions between structural, societal, individual, behavioral, epigenetic, and the various combinations of those factors.
This is why we need more robust models for the future.
These perspectives and examples highlight the need for more robust models that can capture, model, and address the dynamic individual and systemic determinants simultaneously.
As a start, simply disaggregating data can reveal disparities within populations and subpopulations that aggregated sums and averages commonly obscure. But more advanced stochastic modeling techniques, like agent-based modeling or Monte Carlo simulations, could offer new pathways to address these limitations. Stochastic models capture random variability in health outcomes and risks, adding a layer of realism to predictions. Agent-based modeling, for instance, can simulate interactions between individual agents (e.g., people, organizations) and their environments, providing insights into how behaviors and systemic factors may influence health outcomes over time. Monte Carlo simulations, on the other hand, evaluate the potential impacts of different interventions or policy changes by running thousands of simulated scenarios to assess outcomes under varying conditions.
Learn more about Agent-Based Modeling and find some resources from Columbia's Mailman School of Public Health here.
Ultimately, the way we think about population health—and the methods that we use—must evolve to embrace models capable of addressing the multifaceted influences on health: social policies, community environments, and biological responses. This integration of more social and structural factors into our models (rather than seeing them as ancillary to), is crucial for advancing actionable, equity-driven strategies amidst a modern, rapidly evolving healthcare system slowly placing more money and effort into addressing such health-related social needs and other social determinants of health—though how effectively they address structural dynamics is another post for another day.
SO WHERE DO WE GO FROM HERE?
Population Health Science
For me, it’s Population Health Science: an interdisciplinary field focused on understanding patterns and improving outcomes for entire populations to identify and address the structural and systemic factors that shape health outcomes both across and within diverse groups. —what works, what doesn’t, and why or why not. Unlike population health management-oriented or health promotion approaches, population health science can help us uncover the underlying quantitative and qualitative causes of outcomes—whether that be epigenetic, social, structural, cultural, etc.
To get at modern-day issues more deeply and effectively, we must draw on a more diverse array of fields and disciplines—such as public health, epidemiology, sociology, economics, political science, history, data science, and others—to examine the interconnected drivers of disparities and outcomes and design evidence-based interventions and systemic solutions that are human-centered but measurable, scalable, and effective.
And that is what Population Health Science aims to do. This integration of disciplines and advanced modeling makes it a natural home for health equity, which has roots, influences, dynamics, and manifestations to consider that span well-beyond a public health or healthcare-centric view alone, and certainly well-beyond quantitative thought. Again, another post for another day.
Looking Ahead
If we can continue to evolve our thinking and collaborative approaches a bit, population health is quite uniquely positioned to address evolving challenges at the intersection of data, technology, community, and healthcare improvement. This is the future of healthcare, but it will take some work, first in reflective thought and then with intentional action.
Here are some areas I’m watching in the space:
Health Equity as a Core Metric: Population health frameworks are increasingly prioritizing reducing disparities as a measure of success. This shift challenges systems to focus not just on aggregate improvements, but on ensuring that vulnerable populations experience tangible, targeted support based on their unique needs. Empowering communities to actively participate in the design and decision-making process of these programs will also be crucial to ensuring that new interventions are effective, and is an oft-overlooked strategy. We will talk in future posts about how there is more to health equity, but this is a palatable and feasible foray, particularly as other initiatives involving similar semantics are having mixed reception.
Social and Structural Determinants: Social determinants are the conditions in which people are born, grow, work, live, and age, such as access to housing, education, and employment opportunities. Structural determinants, by contrast, encompass the broader systemic and policy frameworks that shape, and have shaped, these conditions, including economic systems, political governance, and societal norms. Both are critical to understanding and improving health outcomes, but they require distinct strategies for intervention. Healthcare systems are beginning to address social determinants by integrating needs assessments into clinical care, investing in community-based partnerships, and advocating for policies such as universal housing programs and equitable education funding. However, there is still much room for growth in upstream structural solutions such as policy and advocacy, cross-sector collaboration, and sustainable investments.
Digital Health Transformation: Advances in health data interoperability, artificial intelligence, and predictive analytics are revolutionizing how population health data is collected, analyzed, and acted upon. For instance, using machine learning to stratify patients based on risk factors allows for more precise and timely interventions, and increasingly disaggregated, yet incredibly rich, data will allow us to analyze things with more nuance than ever before, unearthing pockets of opportunity that are currently flying under the radar. Digital Health tools are also showing some promise to engage with individuals in new and hopefully more effective ways, though much study on success factors and long-term outcomes remains to be done here.
Advanced Data Modeling: As I mentioned before, we can start to dabble in more advanced data modeling techniques to uncover deeper insights into health determinants and possible interventions. These approaches could incorporate lesser used mathematical models, such as agent-based modeling, stochastic modeling, and Monte Carlo simulations. Additionally, with the increase in interoperability, data collection, and data exchange, disaggregated datasets could reveal disparities within populations that aggregated data currently obscure, while more robust simulations of 'what-if' scenarios could allow decision makers to better predict and refine the effects of interventions, policy changes, etc.
Call It What You Want
Regardless of what we call it or how we define it, the future of health lies in combining the strengths of all these ideas, context, and disciplines to build better systems that benefit the health, wealth, and wellbeing of both the “individual” and the “population.”
For me, this post sets the stage for more dialogue about health equity and innovation in population health, which represent a future healthcare paradigm that addresses not only the what, but also the why and the how. And that is exactly why I’m doubling-down on this as the foundational lens for all our future conversations at health EQ, and why I chose to cover this topic first and foremost under the rebrand.
Future content will dive deeper into many of the topics discussed throughout this article because I believe that the interdisciplinary quantitative, and qualitative approach will help us pave a better path to more effective, equitable systems that prioritize community needs and actionable insights—driven by data but informed by a deep exploration of context and resultant new strategies.
That is what health EQ is all about, after all.
Until next time,
-Stefany
P.S. If you have any ideas, suggestions, feedback, or requests for specific topics, I’m always open. Hit reply and let me know!!
. . .
Sources:
Arah, O. A. (2009). On the relationship between individual and population health. Medicine, Health Care and Philosophy, 12(3), 235–244. https://doi.org/10.1007/s11019-008-9173-8
Baba, Z., Belinske, S., & Post, D. (2018). Public Health, Population Health, and Planning: Delaware Journal of Public Health, 4(2), 14–18. https://doi.org/10.32481/djph.2018.03.004
Bhosale, A. S., Urquhart, O., Carrasco‐Labra, A., Mathur, M. R., Rafia, K., & Glick, M. (2024). Population health and public health: Commonalities and differences. Journal of Public Health Dentistry, jphd.12651. https://doi.org/10.1111/jphd.12651
Chokshi, D. A., & Mohta, N. S. (2021). Public Health and Population Health: Are They the Same Thing? NEJM Catalyst, 2(2), CAT.20.0653. https://doi.org/10.1056/CAT.20.0653
Cullen, M. R., Baiocchi, M., Chamberlain, L., Chu, I., Horwitz, R. I., Mello, M., O’Hara, A., & Roosz, S. (2022). Population health science as a unifying foundation for translational clinical and public health research. SSM - Population Health, 18, 101047. https://doi.org/10.1016/j.ssmph.2022.101047
Diez Roux, A. V. (2016). On the Distinction—Or Lack of Distinction—Between Population Health and Public Health. American Journal of Public Health, 106(4), 619–620. https://doi.org/10.2105/AJPH.2016.303097
Itchhaporia, D. (2021). Population Health. Journal of the American College of Cardiology, 78(15), 1569–1572. https://doi.org/10.1016/j.jacc.2021.09.00
Kindig, D., & Stoddart, G. (2003). What Is Population Health? American Journal of Public Health, 93(3), 380–383. https://doi.org/10.2105/ajph.93.3.380
Krieger, N. (2012). Who and What Is a “Population”? Historical Debates, Current Controversies, and Implications for Understanding “Population Health” and Rectifying Health Inequities. The Milbank Quarterly, 90(4), 634–681. https://doi.org/10.1111/j.1468-0009.2012.00678.x
Steenkamer, B. M., Drewes, H. W., Heijink, R., Baan, C. A., & Struijs, J. N. (2017). Defining Population Health Management: A Scoping Review of the Literature. Population Health Management, 20(1), 74–85. https://doi.org/10.1089/pop.2015.0149
Zimmerman, F. J. (2021). Population Health Science: Fulfilling the Mission of Public Health. The Milbank Quarterly, 99(1), 9–23. https://doi.org/10.1111/1468-0009.12493
World Health Organization (WHO). Promoting Health in All Policies and intersectoral action capacities. (n.d.). https://www.who.int/activities/promoting-health-in-all-policies-and-intersectoral-action-capacities
Like this content? Subscribe for new posts.
If you want to learn more about health data, equity, population health innovation, or just plain level up your healthcare game, check out my free Guides. Courses and more resources are coming soon, so check back often!