Gourevitch, Marc N. MD, MPH
A central tenet of health reform is that health care delivery systems share responsibility for the health of defined populations. Framed in the context of the “triple aim” of improved health care, better health, and reduced costs, policy makers aim to increase delivery systems’ efforts to improve and sustain the overall health of persons for whose health they are at least partially accountable. In this paradigm, performance on key outcomes will be best when not only traditional quality measures but also preventive care and other “nonmedical” determinants of patients’ health are optimized. For large, multimission entities such as academic medical centers (AMCs), adapting to such shifting frameworks of accountability in clinical care poses real challenges. In fact, however, the arrival of population-oriented paradigms in health care represents a tremendous opportunity for AMCs to advance vital components of their patient care, education, research, and service missions that have too long been stymied, at least in part, for lack of alignment with external trends and incentives.
Converging Notions of Populations
Health care delivery has traditionally focused on caring for individual persons. For AMCs this has been the essential framework for medical student and postgraduate education, clinical care, and the research enterprise. Examples of exceptions include the introductory epidemiology class or, in the clinical realm, the aggregation of outcome data from individual patients cared for during a specified time period. By contrast, populations, not individuals, are the stock in trade for public health officials and health policy experts, who typically define populations on the basis of residential location or a common characteristic such as race, ethnicity, gender, age, primary language spoken, or disease status.
This difference in conceptualizing populations, on the one hand as aggregates of individuals receiving care from a delivery system and on the other as the total population in a geographic area or with a particular characteristic regardless of whether engaged in care, mirrors the divergence of the fields of medicine and public health that dates back to the early 1900s.1 Schools of medicine focus on caring for individual patients, training physicians, and defining mechanisms of pathophysiology, whereas schools of public health take on system-level approaches to improving health, training public health practitioners, and defining social and environmental determinants of health.
In recent years, a growing chorus of investigators, educators, and policy experts have argued for greater integration of the individual and public health approaches to advancing health.2 The term population health, as set forth by Kindig and Stoddart3 as well as others, encompasses “the health outcomes of a group of individuals.” Concern for the health of individuals is implicit in this framing, but illuminating and setting in motion interventions to reduce disparities in the health of subgroups, whether defined by race, socioeconomic status or geography, is its central focus.4
How does population health differ from public health? Fundamentally, population health is concerned with measuring and optimizing the health of groups, and in so doing embraces the full range of determinants of health, including health care delivery, whereas public health is concerned more generally with influences on health, and the levers to improve them, that exist largely outside of the personal health care system. Population health has grown rapidly as a framework that includes and integrates the activities of the health care sector as among the significant determinants of health outcomes among groups of persons.
Does the rapid adoption of the phrase “population health” into the health care delivery lexicon reflect sudden eagerness by health systems to take responsibility for addressing nonmedical determinants of health? The phrase cut its teeth most prominently as a core element of the “triple aim” paradigm, framed in 2008 by Berwick and colleagues5 and incorporated into the Affordable Care Act (ACA) and the concept of accountable care. The triple aim defines three linked, concurrent goals for health care delivery: improving care for individuals, improving health of populations, and reducing per capita costs. In this framework, the responsible system, whether a delivery network, large insurer, or other entity, must balance attention to these three goals such that outcomes are optimized within and across domains. Attention to preventing leading causes of morbidity and mortality, such as obesity, smoking, physical inactivity, and violence, is an expectation of the parent system. Because tackling these “upstream” causes of disease is best accomplished in concert with efforts by other sectors, such as public health, education, housing, and transportation, the population focus of the triple aim is very much in step with the broader concept of population health. In practice, however, population health under the ACA has often taken on a more clinical connotation in which the population is defined as a group of persons receiving health care from a given system or practice, rather than more broadly as, for example, the health of persons in a geographic area.6
To do justice to their missions and position themselves for the increased alignment of health care delivery and public health goals that lies ahead, AMCs must become leaders in embracing and realizing the fullest vision of population health. In practical terms, doing so will mean finding an ever-shifting “sweet spot” that best balances daily delivery system management with total population health goals, reflecting continuously evolving financial and performance targets and incentives. Specific opportunities for adopting a population health framework across clinical, education, and research missions are outlined below.
Clinical Care and Population Health
Many points of alignment exist between the AMC’s clinical mission and the goal of advancing population health. Indeed, there is wide variation in the extent to which AMCs are adopting a population focus, reflecting regional scope, urban or rural setting, referral center or community orientation, and payer mix, among other factors. As payment structures, electronic health record networks, and community benefit expectations evolve, however, alignment between delivery system and population health goals will become increasingly widespread.
Improving capacity for assuming population risk
As payment mechanisms evolve towards greater assumption of risk by health care systems for episodes of care and defined populations, the value of strategies that reduce use of costly acute care services increases. For example, in a fee-for-service paradigm, the business case is challenging for investing scarce institutional resources to mitigate asthma triggers even in a community from which a delivery system draws a substantial number of patients. But priorities shift when an increasing mix of financial risk is involved. When a single accountable care organization has broad regional penetration, delivery system investment in area-wide prevention will come more readily than for a hospital system in a large city that cares for only a modest proportion of the overall population. But even in dense urban centers, strategies are emerging that bridge traditional boundaries.7 Could there be a clinically driven business case for an urban AMC to share with the public sector the cost of combating smoking beyond the institution’s premises, or promoting access to childhood immunizations among local populations? AMCs, many with depth in prevention and in health services research, are uniquely positioned to model the potential financial impact of specific interventions that extend beyond the persons for whom they are already providing care, and to lead in implementing and evaluating interventions that bridge this divide.
Optimizing community benefit
To maintain their nonprofit status, AMCs must delineate to the Internal Revenue Service (IRS) annually their expenditures of benefit to the broader community. As access to health insurance expands under the ACA, the proportion of the community benefit contribution represented by charity care and unreimbursed Medicaid is likely to decline, and hospitals may need to increase support for other initiatives meeting community benefit criteria.8 The ACA stipulates that nonprofit hospitals must complete a community needs assessment every three years that engages stakeholders including the public health sector, and implement a plan to address the findings.9 For many AMCs, this expectation offers one of the few points of direct alignment between corporate and financial goals (IRS compliance and maintenance of tax-exempt status) and advancing population health. For AMCs that bear risk for populations in surrounding communities, neighborhood- or home-based interventions hold promise for reducing overuse of health care services while improving outcomes.10 Community benefit efforts can also align with institutional research agendas and educational goals.
Emerging models of care
Momentum continues to grow for extending health care beyond discrete office visits and into patients’ daily lives, reflecting the move towards self-management and patient-centered care, with promising early indications of impact on health outcomes and costs.11 Integrating health coaches, patient navigators, and community health workers into the patient-centered medical home can yield more continuously distributed care tailored to patients’ level of need, while fostering medication adherence, behavior change, and engagement in needed mental health or substance abuse treatment. For interventions that are strong, or even weak yet inexpensive, investments in community capacity for extending the spectrum of care are certain to deliver important benefits at the population level.
Harnessing health information technology
Tremendous potential exists for AMCs to improve the efficiency, quality, and value of the care they deliver by integrating analytics and engineering processes more deeply into their operations.12 Examples range from applying queuing theory to improve emergency department throughput, to reducing the potential for severe medication interactions by verifying electronic medication lists against those populated by patients through a Web-based portal.13,14 Indeed, leveraging the vast array of data generated in the process of care with the goal of refining and continuously improving performance—transforming into a “learning health system”—is emerging as a vital goal for large delivery systems.15 Regional linkages between delivery system health information technology (HIT) networks, or health information exchanges, offer new tools for identifying emerging trends and targeting subgroups that may benefit from tailored interventions. Patient-reported outcomes, gathered either actively (e.g., FluNearYou)16 or passively (e.g., Propeller Health)17 and aggregated creatively, can offer additional, actionable tools to delivery systems seeking to optimize population-level outcomes.18
Employees and their families
Many AMCs are major regional employers and self-insured. In providing care for lower-wage employees and their families, delivery systems reach many persons who reside in area neighborhoods from which other patients are also drawn. Forming partnerships with employees as “health ambassadors” to their communities can promote the health of both workforce and community while offering the potential for improved service utilization and heath outcomes among “covered lives.”
Relating performance measures to population health
The value of patient-reported outcomes, including metrics of functional status and social role function, is gaining acceptance alongside traditional measures of quality and safety. By extending beyond disease-specific measures to global measures of functional status, delivery systems’ attention will be drawn to dimensions of health status beyond their traditional scope. For example, if metrics from patients residing in a particular neighborhood consistently indicate low rates of physical activity, a delivery system could respond by partnering with a community organization to develop local resources that promote exercise.19
Medical Education and Population Health
Although health and effective health care delivery are heavily influenced by social, behavioral, and environmental factors, attention to these core determinants is broadly lacking in medical education.
Broadening the scope of medical student education
There is growing recognition of the importance of educating students in “nonmedical” influences on health and health care delivery. The Institute of Medicine has emphasized the need for deeper education in the social and behavioral sciences. The Association of American Medical Colleges (AAMC) announced a significant overhaul of the Medical College Admission Test to emphasize knowledge in these domains. A cooperative agreement between the AAMC and the Centers for Disease Control and Prevention (CDC) aims to integrate education in core public and population health content into medical school curricula. And new programming at some medical schools seeks to educate students in the science of health care delivery, including quality and safety.20 Yet although some exposure to epidemiology and biostatistics is widespread, education remains fragmented and often sparse in such areas as health economics and policy, health care delivery, health promotion and disease prevention, community and global health, and medical ethics.21 Efforts to expand training in these areas must align with trends to condense the duration of preclinical training and, in some cases, medical school itself.22 Medical schools must define an evolving “canon” in population health that will ensure their graduates’ and future leaders’ ability to understand as well as leverage the context in which their clinical skills are applied.23–25
Primary care and population health
A decades-long push to foster interest in primary care among medical students has recently gained momentum, spurred in part by the demand anticipated as greater access to care follows from the ACA.26 Valuable opportunities exist to align medical schools’ efforts to increase interest in primary care with initiatives to broaden education in population health.27 If new delivery system paradigms such as medical homes and accountable care are to succeed, future leaders must bridge clinic and community in managing care and optimizing health promotion and prevention.28 Engaging today’s students at the interface of primary care and population health, a stated focus of some newer medical schools and of new tracks at others,29 advances this agenda while providing rich opportunities for experiential learning. Postgraduate training can advance value-based care by supporting resource stewardship, for example by training clinicians in the Choosing Wisely initiative launched by the American Board of Internal Medicine.30
Research and Population Health
Unprecedented opportunities exist for AMCs to conduct research that advances knowledge and practice in optimizing the health of populations.
Growing emphasis on T3/T4 research
At less than 5% of the overall health-related research budget,31 total federal spending on population-health-related research remains disproportionately low and often threatened by political currents.32 Yet “downstream” translational research (often referred to as “T3” and “T4” when compared with the upstream “bench to clinic” rubrics of “T1” and “T2” research) has seen important investment in recent years.33,34 Comparative effectiveness research and its twin, patient-centered outcomes research, have received major infusions of federal support. The advent of ever-larger distributed data networks is making possible increasingly rigorous analysis of observational data derived from nonexperimental settings.35 PCORnet, recently initiated by the Patient Centered Outcomes Research Institute (PCORI), promises large-scale integration of delivery system and patient-reported data to facilitate large simple trials and comparative effectiveness research. National Institutes of Health (NIH) Common Fund initiatives such as the Research Collaboratory aim to bridge established investigative teams with frontline health care delivery systems to answer pressing challenges in implementation research. Recent funding announcements from PCORI and the Centers for Medicare and Medicaid Innovation offer additional fresh avenues for investigators. In short, population health-related opportunities represent a crucial area of growth for the AMC research enterprise.
Fostering community partnership
AMCs have long recognized the value of partnering with communities in research initiatives aligned with local health priorities. Three trends are now converging to elevate the importance of community linkages for AMCs. First, optimizing the impact of evidence-based health-advancing interventions on overall public health and health disparities requires research on implementation, dissemination, and sustainability in real-world community settings. Such research, which includes attention to culture, language, and social context, is supported by the CDC’s Prevention Research Centers and also increasingly by the NIH through institute-based funding opportunities as well as the Clinical and Translational Science Institute award program of the National Center for Advancing Translational Science. 36,37 Second, care delivery models such as accountable care organizations and health homes will need community-anchored interventions and services to optimize the health of the patient populations for which they assume responsibility.38 Finally, the community benefit requirement is gathering strength, as discussed above. Taken together, these developments argue persuasively for deepening AMCs’ investment in and alliance with neighborhood and community partners.39
Building team science
Just as achieving broad-scale health improvements requires input and engagement from multiple sectors, often referred to as a “health in all policies” approach,40 effectively tackling population health research requires integrated attention from multiple disciplines. Though this kind of “team science” approach is increasingly aspired to in many academic research settings, it can be difficult to realize.41 Drawing broadly from the health, behavioral, social, and applied sciences, population health research is inherently transdisciplinary, providing an ideal “dry lab” platform for building and strengthening research and policy initiatives that bridge departments as well as affiliated institutes and schools (e.g., business, management, arts, and sciences). Important questions span traditional boundaries. What scale of investment in the health of the general population is required to promote economic vitality and growth? Could broad-scale community-level education of lay health workers yield dividends in reducing unnecessary hospital admissions? Can advances in mobile applications be harnessed to advance personal health goals independently of literacy or educational background? These and myriad population health research challenges are uniquely amenable to a team science approach.
Defining metrics of population health
Better metrics of population health are needed to allow delivery systems, including those led by AMCs, to measure progress towards optimizing health outcomes at the population level.5 For example, defining and quantifying preventable deficits in morbidity among a population in care or local community could shape interventions for persons whose health is at once the concern of a delivery system, a community, and a local health department. Attention to intermediate outcomes such as physical activity or smoking prevalence is also critical.42 Although increasingly sophisticated measures quantify population health status at the global,43 national44 and county45 levels, they rarely focus on populations of persons receiving care from specific health care delivery systems. Conversely, measures applied to populations in care are often so narrowly construed as to exclude persons not deeply engaged in the delivery system in question.46 AMCs can lead efforts to develop population health metrics that bridge this divide.47
Measuring AMC impact on population health
How might the impact of an AMC’s initiatives on the health of one or more populations be measured? Clearly, this would depend on the populations in question: whether in care in an allied delivery system, or living in surrounding neighborhoods, or perhaps nationally or globally following dissemination and application of new knowledge arising from the institution’s research. Though challenging, modeling the impact of an AMC’s clinical, educational and research missions on the health of populations could yield valuable measures against which institutions could measure themselves and the influence of their efforts.
Academic Homes for Population Health
Where in the organizational structure of an AMC should the disciplines of population health best be housed? To date, institutional history and dynamics have been the chief arbiters of such assignments. When a medical school is not affiliated with a school of public health, core population health disciplines are often assembled under one or several departmental umbrellas. At one end of the spectrum, a dedicated department of population health (or of population medicine or public health) provides a primary academic home. Alternatively, departments of preventive medicine, community health, social medicine, health policy, biostatistics, and epidemiology address population health through their several lenses. Elsewhere still, the transdisciplinary nature of the field is reflected in a school or institute structure. At NYU School of Medicine, a new Department of Population Health with ties to other departments, institutes, and schools, including NYU’s Global Institute of Public Health, serves as the medical school’s primary academic home for health delivery science, behavior change strategies, behavioral economics, value and comparative effectiveness, health policy, community and global health, biostatistics, epidemiology, and medical ethics. Because “population health” is not formally recognized as a distinct academic discipline, time will tell whether programs bearing this name become widespread. But the value of such institutional recognition cannot be overstated. Population health embodies integration of the AMC’s social mission with its more routinely embraced missions of patient care, education, and research.48 Establishing an academic home for population health elevates consideration of the context in which medical care is delivered, while giving a home to education and research in related themes and taking a step forward in bridging medicine and public health.1
Aren’t the Broader Determinants of Health Beyond the Scope of AMCs?
If medical care exerts only a modest influence on the overall health of populations, and health-related behaviors, genetics, education, socioeconomic status, and environmental exposures together account for 90% of health outcomes, is it even reasonable to expect AMCs to focus on population health?49 Yes, for a number of reasons. To begin with, 10% is hardly trivial. If AMCs succeed in playing a central role in optimizing delivery of health care, that alone would constitute a meaningful contribution to population health. Second, health-related behaviors are responsible for an estimated 40% of premature morbidity and mortality. Academic medicine has much to offer here, through primary care and community-based interventions, integrating patient preferences into clinical priority-setting regarding behavior change, and applying principles of behavioral economics to framing health-related choices. Third, though medical care may itself be only a modest determinant of health, professionals whose passion is to improve health are disproportionately concentrated in this sector. The impact and leadership of this fundamentally committed cadre is critical to moving the needle on the health of populations. Finally, addressing nonmedical determinants of health is a shared responsibility. Because AMCs are fundamentally concerned with improving health, it is only plausible that they work to maximize their impact at the population level, acknowledging clearly their areas of primary strength as well as when partnerships with communities and across sectors are most effectively leveraged.50
The time is right for AMCs to endorse and commit to the goal of improving the health of populations. In elevating the stature of and institutional commitment to population health, AMCs can leverage tremendous opportunities for advancing core clinical, educational, and research missions, while better preparing future leaders for the complex and multiply determined context in which medical care will continue to be delivered.
Acknowledgments: The author wishes to thank David A. Kindig, MD, for his valuable comments on an earlier version of this article.
1. White KL Healing the Schism: Epidemiology, Medicine, and the Public’s Health. 1991 New York, NY Springer-Verlag
2. Institute of Medicine. Primary Care and Public Health: Exploring Integration to Improve Population Health. 2012 Washington, DC National Academies Press
3. Kindig D, Stoddart G. What is population health? Am J Public Health. 2003;93:380–383
5. Berwick DM, Nolan TW, Whittington J. The triple aim: Care, health, and cost. Health Aff (Millwood). 2008;27:759–769
7. Bazinsky KR, Herrera L, Sharfstein JM. Toward innovative models of health care and financing: Matchmaking in Maryland. JAMA. 2012;307:1261–1262
8. Principe K, Adams EK, Maynard J, Becker ER. The impact of the individual mandate and Internal Revenue Service Form 990 Schedule H on community benefits from nonprofit hospitals. Am J Public Health. 2012;102:229–237
10. Carrillo JE, Shekhani NS, Deland EL, et al. A regional health collaborative formed by New York–Presbyterian aims to improve the health of a largely Hispanic community. Health Aff (Millwood). 2011;30:1955–1964
11. Handley MA, Shumway M, Schillinger D. Cost-effectiveness of automated telephone self-management support with nurse care management among patients with diabetes. Ann Fam Med. 2008;6:512–518
12. Nelson EC, Batalden PB, Godfrey MM, Lazar JS Value by Design: Developing Clinical Microsystems to Achieve Organizational Excellence. 2011 San Francisco, Calif Jossey-Bass
13. Wiler JL, Bolandifar E, Griffey RT, Poirier RF, Olsen T. An emergency department patient flow model based on queueing theory principles. Acad Emerg Med. 2013;9:939–946
14. Schnipper JL, Gandhi TK, Wald JS, et al. Effects of an online personal health record on medication accuracy and safety: A cluster-randomized trial. J Am Med Inform Assoc. 2012;19:728–734
15. Institute of Medicine. Digital Data Priorities for Continuous Learning in Health and Health Care: Workshop Summary. 2012 Washington, DC National Academies Press
19. Silva M, Cashman S, Kunte P, Candib LM. Improving population health through integration of primary care and public health: Providing access to physical activity for community health center patients. Am J Public Health. 2012;102:e56–e61
20. Ogrinc G, Nierenberg DW, Batalden PB. Building experiential learning about quality improvement into a medical school curriculum: The Dartmouth experience. Health Aff (Millwood). 2011;30:716–722
21. Maeshiro R. Responding to the challenge: Population health education for physicians. Acad Med. 2008;83:319–320
22. Abramson SB, Jacob D, Rosenfeld M, et al. A 3-year M.D.—accelerating careers, diminishing debt. N Engl J Med. 2013;369:1085–1087
23. White KL, Connelly JE. The medical school’s mission and the population’s health. Ann Intern Med. 1991;115:968–972
24. Finkelstein JA, McMahon GT, Peters A, Cadigan R, Biddinger P, Simon SR. Teaching population health as a basic science at Harvard Medical School. Acad Med. 2008;83:332–337
25. Kaprielian VS, Silberberg M, McDonald MA, et al. Teaching population health: A competency map approach to education. Acad Med. 2013;88:626–637
26. Schwartz MD. Health care reform and the primary care workforce bottleneck. J Gen Intern Med. 2012;27:469–472
27. Shomaker TS. Preparing for health care reform: Ten recommendations for academic health centers. Acad Med. 2011;86:555–558
28. Stine NW, Chokshi DA. Opportunity in austerity—a common agenda for medicine and public health. N Engl J Med. 2012;366:395–397
30. Cassel CK, Guest JA. Choosing wisely: Helping physicians and patients make smart decisions about their care. JAMA. 2012;307:1801–1802
31. Cook-Deegan R. . Boosting health services research. Science. 2011;333:1384–1385
32. Selker HP, Wood AJ. Industry influence on comparative-effectiveness research funded through health care reform. N Engl J Med. 2009;361:2595–2597
33. Westfall JM, Mold J, Fagnan L. Practice-based research—“Blue Highways” on the NIH roadmap. JAMA. 2007;297:403–406
34. Khoury MJ, Gwinn M, Ioannidis JP. The emergence of translational epidemiology: From scientific discovery to population health impact. Am J Epidemiol. 2010;172:517–524
35. Brown JS, Holmes JH, Shah K, Hall K, Lazarus R, Platt R. Distributed health data networks: A practical and preferred approach to multi-institutional evaluations of comparative effectiveness, safety, and quality of care. Med Care. 2010;48(6 suppl):S45–S51
36. Michener JL, Yaggy S, Lyn M, et al. Improving the health of the community: Duke’s experience with community engagement. Acad Med. 2008;83:408–413
38. Springgate BF, Brook RH. Accountable care organizations and community empowerment. JAMA. 2011;305:1800–1801
39. Wartman SA. Academic health centers: The compelling need for recalibration. Acad Med. 2010;85:1821–1822
40. Rigby E. How the National Prevention Council can overcome key challenges and improve Americans’ health. Health Aff (Millwood). 2011;30:2149–2156
41. Zerhouni E. The NIH Roadmap. Science. 2003;302:62–63, 72
42. Kindig DA. Understanding population health terminology. Milbank Q. 2007;85:139–161
43. Murray CJL, Salomon JA, Mathers CD, Lopez AD Summary Measures of Population Health: Concepts, Ethics, Measurement and Application. 2002 Geneva, Switzerland World Health Organization
44. Molla MT, Madans JH, Wagener DK, Crimmins EM Summary Measures of Population Health: Report of Findings on Methodologic and Data Issues. 2003 Hyattsville, Md National Center for Health Statistics
45. Peppard PE, Kindig DA, Dranger E, Jovaag A, Remington PL. Ranking community health status to stimulate discussion of local public health issues: The Wisconsin County Health Rankings. Am J Public Health. 2008;98:209–212
46. Calman NS, Hauser D, Chokshi DA. “Lost to follow-up”: The public health goals of accountable care. Arch Intern Med. 2012;172:584–586
47. Stine NW, Stevens DL, Braithwaite RS, Gourevitch MN, Wilson RM. HALE and hearty: Toward more meaningful health measurement in the clinical setting. Healthcare. 2013;1:86–90
48. Schroeder SA, Zones JS, Showstack JA. Academic medicine as a public trust. JAMA. 1989;262:803–812
49. McGinnis JM, Williams-Russo P, Knickman JR. The case for more active policy attention to health promotion. Health Aff (Millwood). 2002;21:78–93