This article was produced in collaboration with Penn LDI

Health in the U.S. is a tale of two starkly different realities. The better-off and well-connected are not only in a stronger position to receive care when they need it, but they also start off with advantages that have a tremendous effect on health — in housing, employment, stress levels, food security, social capital and more.

A 2008 report on social determinants of health from the World Health Organization said it plainly. It might not be surprising that the poor have higher levels of illness and mortality. However, the report said, “in countries at all levels of income, health and illness follow a social gradient: the lower the socioeconomic position, the worse the health.”

“Every single session I have in my clinic I see the downstream effects of social factors,” says Dave A. Chokshi, chief population health officer for NYC Health + Hospitals — New York’s public health care system — and an attending physician at Bellevue Hospital. “I think about my patient with diabetes whose blood glucose levels I haven’t been able to get under control because he can’t take the insulin I prescribe because he lives in a homeless shelter and has no place to refrigerate it. Or the person with advanced liver disease related to alcohol use exacerbated by his sporadic employment. When you trace back to the causes of the causes of illness, in so many cases you see how our social fabric itself is in need of mending.”

But broad changes are taking shape. Payers of health care are increasingly incorporating the concepts of social determinants of health into the way they think about reimbursing for health care services and providing incentives for health care delivery organizations, says Risa Lavizzo-Mourey, senior fellow at Penn’s Leonard Davis Institute of Health Economics, population health and health equity professor at Penn Nursing and the Perelman School of Medicine and Wharton professor of health care management.

That, she says, “is a new development and one that obviously creates a lot of questions and opportunities for good research to really figure out how to best incorporate social determinants of health into reimbursement plans, how to make them focused on populations and not solely on the individual, and then to look at what kinds of bundles of social determinants are going to lead to the best outcomes. It’s a very interesting and exciting time.”

For the most part in the U.S., there has been a separation between public health, which focuses more on prevention and the environmental factors that influence health, and health care, which focuses more on diagnoses and illness, says Kathleen Noonan, adjunct senior fellow at Penn’s Leonard Davis Institute of Health Economics and CEO of the New Jersey-based Camden Coalition of Healthcare Providers. “It’s not that the health care side hasn’t known for a long time that social determinants influence what they see, but health systems are not structured to respond to them,” she says. “We have a public health system that is run largely by government and smaller nonprofits that operate in parallel to larger private non-profit and for-profit health systems.”

Historically, health systems were paid on a fee-for-service model, “which did not incentivize them to think holistically at the individual or population-health level,” says Noonan. “The managed care organization (MCO) concept of a bundled payment ‘per-member, per-month’ does, at least conceptually, introduce the idea of payment for total health.”

“If the goal is to improve health in America at a lower cost, there is only so much we can do by waiting for people to get sick and then treating them,” says Shreya Kangovi, a senior fellow at Penn’s Leonard Davis Institute of Health Economics and associate professor of medicine at the Perelman School of Medicine. “It’s far more efficient both in terms of cost to human life and dollars to go upstream.”

Many are calling for the stitching together of partnerships from the health care and public health realms — a prospect Kangovi calls “the greatest opportunity to advance health in our country in a generation.”

“When you trace back to the causes of the causes of illness, in so many cases you see how our social fabric itself is in need of mending.”–Dave A. Chokshi

Seeing Pathologies Through a Different Lens

Initiatives that aim to address social determinants of health are proliferating. In Coal Township, in central Pennsylvania, Geisinger Shamokin Area Community Hospital started a fresh-food “farmacy” in 2016 to bring not just nutrition counseling to diabetic patients, but also recipes and the fresh food itself. CareMore Health in Cerritos, California, launched its Togetherness Program in 2017 to “address senior loneliness as a treatable condition by focusing on patients’ psychological, social and physical health,” as the company describes it.

UnitedHealthcare in April announced that since 2011, it has put more than $400 million into affordable housing across the U.S. to help reduce social barriers to better health in underserved communities.

The concept of social determinants of health has come to the attention of Congress with the introduction of a bipartisan bill in July. The Social Determinants Accelerator Act would provide “planning grants and technical assistance to state, local and Tribal governments to help them devise innovative, evidence-based approaches to coordinate services and improve outcomes and cost effectiveness” for Medicaid participants, according to a fact sheet from Rep. Cheri Bustos (Dem.) of Illinois.

Increasingly, looking at health in the U.S. through the lens of role of social determinants of health is considered just as useful as viewing it through traditional measures, such as how many people live and die with heart disease or cancer.

One study found that in 2000, about 245,000 deaths in the U.S could be traced to low education, 176, 000 to racial segregation, 162, 000 to low social support, 133,000 to individual-level poverty, 119,000 to income inequality, and 39, 000 to area-level poverty. “The estimated number of deaths attributable to social factors in the United States is comparable to the number attributed to pathophysiological and behavioral causes,” wrote the authors of “Estimated Deaths Attributable to Social Factors in the United States,” published in 2011 in the American Journal of Public Health.

In fact, the number of deaths in 2000 stemming from low education was comparable to the number from heart attacks (192,898) — the leading cause of death in the U.S. that year.

Programmatic innovations that address social determinants of health are increasingly popular, but these innovations should not distract from major policy decisions that may threaten gains, says Chokshi, also a clinical associate professor at New York University’s School of Medicine. A rule change proposed by the federal government, for instance, could eliminate food stamps for millions by instituting work requirements and putting up other barriers to the SNAP program.

“We can’t in the same breath talk about programs that might help dozens with food insecurity and not heed these larger dynamics on the policy side that are setting back the conversation around social determinants of health. We have to have a sense of scale in these conversations,” says Chokshi.

The reasons to care about social determinants of health go well beyond altruism or empathy.

“The circumstances that our most vulnerable citizens are in have implications for the cost of health care systems and implications for the fabric of society,” says Atheendar Venkataramani, a Leonard Davis Institute senior fellow and assistant professor of medical ethics and health policy at the Perelman School of Medicine. “Where people are not flourishing and there is inequality, those environments are a challenge for people across the income distribution. If you are in close proximity to illness that affects other people, that does affect you through how it challenges the health care system that you have and how it might redirect resources within your community. Those spillovers come to your front door.”

One animating factor behind the current discussion over social determinants of health is the rise of U.S. health care costs — projected to increase an average of 5.5% per year from 2018 to 2027, or 0.8 percentage points faster than the gross domestic product, according to the Centers for Medicare & Medicaid Services. Another was the introduction of millions more insured under the Affordable Care Act.

“For decades the big policy push was to increase the number of people who had insurance and access to care, because it was hard to imagine the impact of being able to improve health care outcomes for the population if there were so many people uninsured,” says Lavizzo-Mourey, who was president and CEO of the Robert Wood Johnson Foundation from 2003 to 2017. “And while we still have millions uninsured, it’s less than it once was.”

“It’s not that the health care side hasn’t known for a long time that social determinants influence what they see, but health systems are not structured to respond to them.”–Kathleen Noonan

Increasing attention to social determinants of health now, she says, also has to do with the state of overall health of our population. “We’ve long had the dubious distinction of paying more for care than anyone else and fair to middling health status when you compare the U.S. to similar countries in terms of wealth. And then more recently we see health care status is going down in some populations. When you put those trends together, the logical mind says: Where else do we need to look to make progress in these trends that are not going in the right direction?”

Getting to the Heart of Intervention

Progress has come in the much-praised program for which Kangovi is founding executive director. The IMPaCT program at the Penn Center for Community Health Workers hires from within local communities to provide social support, advocacy and navigation of services to high-risk patients. In operation since 2011, IMPaCT (Individualized Management for Patient-Centered Targets) has been shown through several randomized clinical trials to provide consistent improvements in quality while reducing hospital days by 65%.

In serving more than 10,000 Philadelphia-area patients to date, it boasts two dollars in return for every dollar invested annually, and has become a model for other programs around the country. Woven into its design was an initial interview process that asked 1,500 low-income patients to describe the hurdles to staying healthy and what could help.

Kangovi says she often sees a fragmented approach to assessing patients’ needs — looking in a compartmentalized way at depression, food insecurity or smoking — rather than a holistic, patient-dictated approach. As part of their ongoing contact with patients, community health workers use an interview guide developed by the IMPaCT program that allows them to “have a real conversation to learn life stories and ask the patient, ‘Mrs. Jones, what do you need to improve you health?’ It is a very logical question that is at heart of intervention,” says Kangovi. “She can say, ‘I need a reason to get out of bed in the morning,’ because her son was murdered,” rather than imposing on her a domain-centered intervention.

“It is always about what people need in their lives,” says Kangovi. “It will straddle disease types and come up with tailored action plans, and then provide hands-on support to help them achieve that. It may be that a community health worker will go with her to a crochet class to get her out of bed in the morning, and others may accompany Mrs. Smith to talk to her landlord to help get her housing. People are actually the experts of their own lives. You are a lot more cost-effective because you are not doing a cookie-cutter approach.”

It is also important who is doing the asking. Says Kangovi: “One basic premise is that the program is delivered by community health workers who come from a similar cultural and social background, so they can relate to patients. In the IMPaCT model, they are specifically engaged for their empathy and traits like being good listeners and being non-judgmental, so starting with the right workforce has been critical.”

“The circumstances that our most vulnerable citizens are in have implications for the cost of health care systems and implications for the fabric of society.”–Atheendar Venkataramani

Often, programs use a screen-and-refer protocol for identifying social determinants of health in patients. But that tool comes with significant pitfalls, says Kangovi.

“We’ve all been asked while a nurse is typing, ‘Are you safe at home?’ And what does that even really mean? People are being asked that over and over again, and that inures them — patients don’t trust them as much as they did before, and patients have legitimate fears of consequences like losing child custody, deportation, or stigma. And what are we doing when someone tells us they have struggles? The most common thing is nothing, or we are just going to refer them to some resource, like a food pantry or housing authority. The reality is there have been studies recently suggesting that screen-and-refer approaches do not work. They have low rates of uptake and do nothing for patient outcomes.”

Good Intentions, Good Outcomes?

The number of social determinants of health programs rolled out by health systems grew ten-fold in a decade, according to a 2017 study published in the American Journal of Preventive Medicine. But do they work? Which social determinants have the greatest impact on health, and which needs, when addressed, can lower costs?

Certain programs might make sense intuitively, but that doesn’t mean they will lead to breakthroughs on lower costs or better patient care. One could suspect, for instance, that providing transportation to Medicaid patients for primary care appointments would help lower missed-appointment rates. But in a 2016-2017 clinical trial conducted at two West Philadelphia clinics, complimentary ride-share services were provided for clients.

“Surprisingly, it made no difference,” said Krisda Chaiyachati, a Leonard Davis Institute senior fellow, assistant professor of medicine at the Perelman School of Medicine and clinical innovation manager for Penn’s Center for Health Care Innovation. The uptake of free rides was low, and rates of missed appointments remained unchanged at 37%, according to the research study.

“We simultaneously interviewed 45 patients, and really what we learned is that a lot of patients have pretty chaotic lives,” said Chaiyachati, the trial’s lead author. “It’s not like they missed an appointment because they didn’t know. It was because the person taking them to the appointment was sick, or they didn’t have someone to take care of an ill grandparent or didn’t have childcare. For some, missing three or four hours of work meant a meaningful dollar amount, or they were under a lot of pressure to be at work, or some patients were so sick they couldn’t leave home.”

“We’ve long had the dubious distinction of paying more for care than anyone else and fair to middling health status when you compare the U.S. to similar countries in terms of wealth.”–Risa Lavizzo-Mourey

Chaiyachati — who hasn’t given up on the idea of exploring what transportation could do to bring better health to patients and improve efficiency — notes that patients in the Penn health system on Medicaid have a nearly 50% no-show rate to primary care dates, and every missed appointment represents costs to Penn.

It also represents potentially lost lives. He is working with Penn’s Abramson Cancer Center to create screening tools to catch patients now falling between the cracks. Preliminary results of a study in progress show that one in seven uninsured cancer patients misses at least one chemotherapy or radiation treatment, while the insured miss virtually zero.

“From an equity perspective, the challenge here is that we can have two individuals who walk in our clinic doors who have the same exact diagnosis and get the same treatment, but because they have a different skin color or a different income level, one person will do much worse,” says Chaiyachati.

The disparities are just as startling for behavioral health issues as physical ones. Persons with the lowest socioeconomic status are two to three times more likely to have mental health disorders than those at the highest levels, according to the U.S. Department of Health and Human Services. At the same time, minority and rural populations in the U.S. have less access to mental heath services than others.

But behavioral health attention remains inadequate generally.

“At this point, our medical model is more adept at identifying risks for diabetes than mental health risks,” says Noonan. “The American model assumes children and young adults are healthy. After age three, you might see a doctor once a year. But for the most part, our health systems largely ignore the developmental periods when we know early signs of mental health and addiction issues begin to manifest. Figuring out how to address this gap in our health system is really important, instead of waiting until we have young adults who show up in our ERs with significant mental health and/or addiction issues.”

“People are actually the experts of their own lives. You are a lot more cost-effective because you are not doing a cookie-cutter approach.”–Shreya Kangovi

Indeed, comprehensive thinking about the entire system is critical. “We need cross-sector thinking that considers regional partnerships, and new ways for local public health departments, social service agencies and health care systems to work together to create more prevention-oriented systems,” says Noonan. “We don’t have enough affordable housing in this country, and I’m glad hospitals are thinking about the connection between health and housing, but hospitals alone can’t solve the housing crisis in our country. We also need state governors, public agencies and legislatures to use their authority to incentivize or require different types of partnerships with flexibility from the federal government around how and what is paid for in the pursuit of total health.”

It will also take some hard evidence to advance the cause of social determinants of health.

“My great worry is that there is this huge hype-to-evidence ratio in the field,” says Kangovi. A lot of data now being cited, for instance, comes from simple pre-post studies, “and if you’re measuring something, it’s not a straight line — blood sugar, rates of hospitalization, they are going to vary. And if you do a study where you take a group of people at their highest point, people who have been to the hospital a lot in the past year, and you even do nothing, their rates will go down.

“There are very good and ethical ways of using study design and to answer ‘does this work?’ You are trying to disprove your intervention. If something can really stand up in that light, then it is worth being disseminated.”

Says Kangovi: “This is a big deal, and if we don’t get it right we’re going to get the money pulled away. There are not many chances to do good for people living in poverty or address these wide-ranging issues, and I think patience is going to run out in a few years if a lot of the investments we’re making now don’t result in true outcomes.”