Use of the term “population health” has surged in recent years. Much of this enthusiasm has been driven by the idea that health reform would restructure incentives and unite the priorities of health care and public health. In practice, however, population health is often invoked by various stakeholders using different definitions: in health care, it usually refers to managing the health and cost outcomes of a defined patient population attributed to a health care system. Meanwhile, in public health, population health encompasses the aggregate health status of all people in a given geographic area.

Divergent definitions of population health reflect a broader reality that, despite the potentially complementary approaches of public health and health care, health reform has not resulted in as much productive collaboration as had been hoped. There are many reasons for this, including still-conflicting financial incentives, the slow pace of institutional culture change, and an inconsistent measurement and outcome framework. These challenges are obstacles to uniting such a broad and diverse set of stakeholders.

We propose two objectives to better shape the path toward population health: (1) connecting the patients who would most benefit from health care and supportive services—such as those with undiagnosed chronic conditions—to appropriate clinical and community resources; and (2) developing strategies to address diseases with high quality-of-life (morbidity) burdens, not just those with high mortality burdens. A truly integrated population health platform, meanwhile, would move further beyond these milestones to more fundamentally address determinants such as education, poverty, and exposure to trauma.

Undiagnosed Chronic Conditions

Surveillance for undiagnosed chronic diseases is squarely in the strike zone of public health — but delivering care for those new patients relies upon high-performing health care systems, particularly primary care. Two major chronic diseases help illustrate this concept: hypertension and diabetes.

Thirteen million U.S. adults were not aware of their hypertension in 2011-12, a population prevalence of about 4.2 percent. An estimated 5.2 percent of the U.S. population has undiagnosed diabetes. Connecting individuals who suffer from these diseases with basic medical care could have outsize impact if accomplished at scale. At the same time, public health interventions (e.g., a tax on sugary beverages) could shift the underlying prevalence curves themselves.

Hypertension And Diabetes

If reducing the population prevalence of a disease like hypertension were a shared population health priority, both public health and health care sectors could take on well-defined tasks in a coordinated manner. Public health departments could take on surveillance for high blood pressure across multiple health systems and nontraditional care settings, leveraging maturing health information exchanges to identify and connect those who would benefit from primary care.

For example, the New York City Department of Health and Mental Hygiene, in partnership with the leaders of faith-based health ministries, created a Community Health Dashboard to serve as an electronic health record for congregations to record and track blood pressures of their community over time. In addition to supporting a familiar environment for screening a broader population and serving as a platform for community-led health improvement, these data can help catalyze partnership with health systems for patient engagement and care continuity.

Meanwhile, health care systems can mine their own data for effective outreach opportunities to patients with undiagnosed hypertension. Among the unaware and uncontrolled hypertensive population, 82.5 percent report a usual source of care. Current incentive structures may not encourage this type of outreach, since patients must have a threshold density of interaction with a given health care system for them to “count” in the denominator of a quality metric such as hypertension control. Yet few interventions would have greater impact than treating undiagnosed hypertension: full implementation of the 2014 hypertension guidelines for U.S. adults is estimated to prevent 56,000 cardiovascular events and 13,000 deaths annually, while reducing overall costs.

Undiagnosed diabetes presents similar opportunities for medicine-public health collaboration. A recent Swedish study of persons with Type 2 diabetes showed excess mortality risk increased with younger age and worse glycemic control, underscoring the importance of early diagnosis and treatment. The Centers for Disease Control and Prevention (CDC) and the American Medical Association (AMA) recently launched a more upstream approach to diabetes prevention, Prevent Diabetes STAT (Screen/Test/Act Today), including an effort to increase the number of physicians who screen patients for prediabetes and refer them to diabetes prevention programs. In addition to diabetes and hypertension, chronic obstructive pulmonary disease (COPD), human immunodeficiency virus (HIV), and hepatitis C are all chronic conditions with significant undiagnosed populations and effective medical interventions.

High-Need Patients

Beyond disease-specific approaches, high acute care utilization by some patients could also be thought of as an “undiagnosed” chronic condition amenable to greater public health-health care collaboration. Analyses of “super-utilizers” of emergency departments and hospitals reveal a high prevalence of multimorbidity, mental health and substance use disorders, and social barriers such as homelessness. The health care system has the unique data to identify this high-risk population with heavy acute care utilization and connect them more effectively with primary care. The behavioral and social complexity of these patients, however, means that primary care alone is unlikely to meet their needs without tight operational partnership with community-based social services.

Following this framework, the New York City Health + Hospitals accountable care organization (ACO) has begun leveraging claims and clinical data to identify high-risk patients with heavy acute care utilization and generate a checklist of workflows for primary care, care management, and social service engagement. In one example, patient addresses are cross-matched with data from the New York City Housing Authority to enable connection to resources such as senior centers, chronic disease self-management classes, and case management embedded in public housing. More broadly, the learning from such cross-disciplinary partnerships for high-risk patients could shape upstream interventions in relevant policy areas such as housing and education.

Elevating Quality Of Life

The second milestone on a path to population health, we believe, is for collaboration between public health and health care to prioritize improving quality of life, not just mortality and life expectancy. Public health has most visibly prioritized around the major causes of mortality, with remarkable results. But part of the reason that public health is “out of sight, out of mind” is because major causes of mortality, such as hypertension and diabetes, are often “silent killers” without many symptoms until advanced disease stages. On the other hand, the greatest reductions in quality of life are caused by low back and neck pain, depression, and other musculoskeletal complaints (see Figure 1). While some important public health work has focused on these areas, greater collaboration with health care presents a robust opportunity to elevate comprehensive population health strategies for these major causes of morbidity.

Public health needs the health care system to help figure out such a morbidity-focused strategy, not least because clinicians bear witness to and treat quality-of-life concerns every day: pain, disability, and psychological distress. But in turn, the health care sector can build upon decades of public health science focusing on disease burden and on measuring health-related quality of life. A joint health care-population health approach would take advantage of these reciprocal strengths.

One example in this vein is the Dartmouth Spine Center’s approach to back and neck pain, which utilizes a broad multidisciplinary team with particular focus on patient preferences and behavioral strategies. The model combines clinical expertise with a unique focus on functional status and other patient-reported outcomes, as gauged by tools traditionally used in public health to measure quality of life. This approach could serve as a natural touchpoint to better address the social, environmental, and occupational drivers of morbidity from low back pain in the general population in collaboration with public health partners. Similarly, New York City’s Thrive NYC initiative highlights the disabling burden of depression and other mental illness as a platform for bringing together public health and health care stakeholders to lay out a ‘mental health roadmap’ for the city.

Moving Toward Population Health Collaboration

Other bright spots in true population health collaboration, including geographic performance measurement and accountable health communities, have been described. We believe the two objectives outlined here, if realized, would also represent tangible steps toward blurring the boundaries between health care and public health. The major challenge with the first objective—connecting high-need, sometimes undiagnosed patients with care and needed services—is misalignment of incentives or dilution of resources to engage those patients. Elevating quality-of-life considerations, the second objective, must surmount methodologic difficulties with measuring subjective perceptions, as well as biases against psychological symptoms.

Ultimately, the shared objective of health care and public health should be to improve both length and quality of life for individuals and populations. Each discipline brings essential and complementary tools for achieving that shared objective, and will move together in concert more naturally if they have business models guiding them toward a common definition of success. Achieving the full promise of population health will require metrics that combine mortality and morbidity domains, such as health-adjusted life expectancy, and financial incentives that align the interests of a broad range of stakeholders.

Figure 1. Top Causes of Mortality vs. Top Causes of Morbidity


Rankings are based on Years of Life Lost (YLLs) and Years Lived with Disability (YLDs) per 100,000 for all ages in 2013. YLLs are years of life lost due to premature mortality. YLDs are years lived in less than ideal health. This information is based on findings from the Global Burden of Diseases, Injuries, and Risk Factors Studies, which collects and analyzes data on premature death and disability from more than 300 diseases and injuries in 188 countries.

Source: Institute for Health Metrics and Evaluation.

Authors’ Note

The views expressed in this article are those of the authors and do not necessarily reflect the policy or views of New York City Health + Hospitals.