When people think of health inequity, it is often in terms of racial disparities—Black men being two times more likely to die from prostate cancer, for instance. The causes of these inequities comprise a range of factors, from the biological to the social, environmental, and economic. In short, good health is not solely determined by whether a person has access to quality healthcare. Rather, it can be affected by a range of situations that happen to people outside of the health system in their everyday life.
These social determinants of health aren’t static but, much like the growth rings on a tree, are present throughout a patient’s lifetime. According to the World Health Organization, social determinants of health are “the conditions in which people are born, grow, work, live, and age, and the wider set of forces and systems shaping the conditions of daily life. These forces and systems include economic policies and systems, development agendas, social norms, social policies, and political systems.”
While research into the social determinants of health harks back more than 60 years to include the work of S. Leonard Syme, a noted epidemiologist at the University of California, Berkeley, School of Public Health, it has only been since the mid 1990s and early 2000s that interest in the field began to take off. And in 2011, research from co-lead authors Sandro Galea of the University of Michigan and Melissa Tracy, then at Columbia University, published in the American Journal of Public Health presented a stark conclusion: “The estimated number of deaths attributable to social factors in the United States is comparable to the number attributed to pathophysiological and behavioral causes,” the
authors wrote.
These factors have been known to affect health and wellness for some time, however. The question now is: What can we actually do within the context of providing healthcare to help mitigate the negative aspects of the social determinants of health?
Enter the use of real-world evidence (RWE) and real-world data (RWD).
How RWE can improve health inequity
The fundamental understanding that social, economic, and environmental factors play a central role in a person’s health naturally leads to the consideration of how non-medical data—i.e., data not found in a patient’s medical record or from a clinical study—can be used by healthcare providers to improve a patient’s health.
According to Robert Hiatt, MD, PhD, a professor in the department of epidemiology and biostatistics, University of California, San Francisco, and associate director of population sciences at the university’s Helen Diller Family Comprehensive Cancer Center, applying RWE in the clinical setting requires a broad approach. “Real-world evidence, to me, means studying whole populations, collecting data on individuals of all sorts, not just race and ethnicity, but also other backgrounds and conditions and then looking at that data to generate evidence of some relationship, or some effect of an intervention,” he said.
While this may sound straightforward, considering the breadth of factors that contribute to health and wellness, the difficulty comes in figuring out which have the greatest effect on a patient’s health.
“One of the big questions is, What is the impact of non-clinical factors on the course of treatment a patient is on? It turns out if you don’t have access to the right food, if you don’t have access to transportation, if you live on your own, those can be risk factors or predispositions for not being able to go through a particular journey of care as well as someone who has transportation, has the right food, and has someone to help them out,” said Jean Drouin, CEO and founder of Clarify Health, an analytics and value-based payments company. “Those have rightly been called social determinants of health—factors that can impact outcomes.”
But once a picture of an individual patient is developed, how can this information be used to help address health inequality? A simple measure, Drouin noted as an example, is patients who don’t have a means of getting to appointments being provided with vouchers for transportation services to remove that barrier to care. Other interventions could include finding ways to help relieve food insecurity to provide people with adequate nutrition.
While such interventions would start to address health equity, they require policy decisions by both public and private payers—a difficult shift as these aren’t currently considered health services.
Grammati Sarri is head of real-world advanced analytics, extended research partnerships at Cytel, a statistical software and advanced analytics company. Grammati noted that while the use of RWE has been a hot area within academic research, there are a couple of reasons why incorporating RWE to address health inequity has yet to gain traction among healthcare payers as a matter of policy.
“The state of research capabilities within these decision-makers has, I believe, had an impact on the use of RWE. Some of these methodologies require advanced analytics, or epidemiology capabilities, and many organizations don’t have this critical expertise in-house.” she said.
“Additionally, there has been a resistance to change, which has extended beyond RWE to other methodologies, for example how to incorporate data on access inequalities in decision-making. Payers appear to be hesitant to try new methods outside what has been used,” Sarri continued. “I think their desire to ensure consistent decision making is a driving reason. There is concern that these methods might not be applied consistently.”
RWE requires RWD
Developing evidence that specific interventions work for cohorts of patients that share similar race, ethnicity, health history (from the electronic health record), and social determinants of health is a data-hungry exercise. While the intent is to provide the best possible healthcare pathway for an individual patient, improving health equity requires information from large populations, then finding was to analyze disparate, seemingly unrelated data to stratify patient populations. Done without bias, the goal is to develop information on how care delivery should vary from patient to patient to achieve the best possible outcomes.
While many scientists in healthcare are accustomed to using data that have been meticulously collected, such is not always the case in the realm of real-world data, which are almost exclusively collected outside the bounds of a formal scientific study.
“I think the trick with real-world data—as opposed to highly formalized, restricted data—is that you accept a little bit of messiness in the data in order to come up with a conclusion,” said Hiatt. “Part of what science does is to seek truth, but the other part is seeking information to make decisions. I think when you’re working with real-world data, you’re happy to collect enough data from enough less-well-curated sources to allow you to make a decision.”
Ultimately, the goal of collecting real-world data of patient populations is to analyze it against patient outcomes—the real-world evidence that a particular therapeutic intervention is either effective or ineffective. Past research on the effects of the social determinants of health can help inform new therapeutic pathways that will be more effective for populations known to experience poor outcomes under current approaches.
Not all the data needed to derive real-world evidence of effective health interventions are found outside the healthcare system. Data contained in the electronic health records (EHRs) of individual patients should be included in order to have a solid understanding of an individual patient’s health history.
But a big problem with EHR data, particularly in the U.S., is they are often spread among general practitioners and specialists who the patient has visited, so rarely all in one place. According to Drouin, there is a simple fix to this—instead of using EHR data to have a longitudinal view of the patient’s health history, use payer claims data instead. “In the U.S., claims data, ultimately, is the only record that contains everything that happened to you over the course of the last year related to your healthcare,” he said. “Then, after the age of 65, you are in Medicare.
“If what you’re looking to do is to gain massive insights on things that we can do, generally, to help very specific subpopulations of individuals—including those with specific social determinant characteristics—what is better than to have the largest possible data set, of 300 million plus Americans?” Drouin asked.
As he sees it, these data can provide the kinds of insights that can help guide treatment for a 75-year-old African American man with diabetes to achieve the best outcomes.
In addition to these data on health journeys in the U.S., a growing data set of patient information that is already yielding actionable insights is found in the ongoing All of Us Research Program run by the National Institutes of Health (NIH). This program seeks to collect health and demographic data from at least 1 million Americans. These data have the potential to help reduce healthcare inequity, as the program has made participant diversity a priority.
“All of Us promises to generate a large amount of information not only on white populations, but other major races, ethnic populations, and people from lower socioeconomic status, people from rural areas, people who have different gender preferences, or some disabilities. They may sustain different social conditions, food insecurity, housing insecurity, racism, all these things which we know are important to society that create patterns of disease,” said Hiatt, who is a principal investigator with the All of Us program.
While All of Us asks participants to provide blood and urine samples as well as access to their electronic medical records, it has also developed patient questionnaires developed specifically to collect better data on the social determinants of health. “We collect data in what we call modules via interviews and questionnaires,” Hiatt noted. “The basic modules have to do with medical history, or family history, or chronic conditions. But then we have a module for environmental exposures and another module that asks questions about housing insecurity, food insecurity, racism, religiosity, social isolation, social support, as well as income, education, and occupation, which are the standard for assessing socioeconomic status.”
The program now has more than 400,000 people enrolled and has made the data collected so far, including some outcomes data, available to qualified researchers, which should help shape future health interventions that can close the equity gap.
Peering into the future, Drouin sees the potential to develop patient scoring systems similar to credit scores that take into account both health information and associated social determinants, then matching the score with a patient’s health interventions and feed these data into AI-based clinical decision support systems to help guide appropriate care.
Sarri, too, is optimistic about the potential to help guide care and begin leveling the healthcare playing field.
“There are, of course, some challenges to overcome,” Sarri said. “But I think the more we talk openly about how to improve techniques, rather than focus on these potential challenges, the faster implementation of these techniques will unravel the full potential of RWE in improving health among disadvantaged groups. Collaboration is the key and I’m encouraged that methodologically recent recommendations have come from key stakeholders like payers, the public sector, industry, and research institutes, not just one sector. This will help ensure that health improvements are enjoyed by everyone in our society while ensuring the sustainability of healthcare systems.”
Chris Anderson, a Maine native, has been a B2B editor for more than 25 years. He was the founding editor for Security Systems News, and Drug Discovery News, and led the print launch and expanded coverage as editor in chief of Clinical OMICs, now name Inside Precision Medicine.
