Health systems can address health disparities by improving their data collection

At both the local and federal level, health systems and agencies are looking to develop programs to promote health equity—and improved patient data collection may be a good place to start.

Healthcare organizations can look at their patient population and social determinants of health (SDOH) data to help identify and find solutions for health disparities, which cost the industry about $93 billion in excess medical expenses each year, per a 2018 report from the WK Kellogg Foundation.

“Once you have the data, the sky’s the limit,” Lyndia Hayden, senior director of data integrity and equity analytics at the New York-based Mount Sinai Health System, told Healthcare Brew. “There’s a lot that you can do toward equitable practices and policies once you have data integrity, and that’s what we’re focusing on.”

Why this data is important

Many health systems now collect patient demographic data such as race, ethnicity, and language preference through voluntary surveys. But few have been gathering this data systematically, according to a 2022 study from Hayden and her team, who are aiming to improve data collection at Mount Sinai. And health systems may have limited chances to collect this data, as patients only visit the hospitals a few times a year—if they go at all—the study found.

“The foundational step toward reducing health disparities is the establishment of the basic information needed to identify and measure those differences, i.e., the accurate capture of race and ethnicity information of all patients,” the researchers said in the study.

But patient demographic data is just one piece of the puzzle. When healthcare organizations account for SDOH, patients have been able to reduce spending by about 11%—or a mean of $2,601—in a year, according to a 2018 study from HealthConnections, a social services coordination program from WellCare Health Plans.

“If we put even compliance aside and we put all the right feelings aside, addressing SDOH simply makes good business sense,” Hillit Meidar-Alfi, founder and CEO of software company Spatially Health, told Healthcare Brew.

Spatially’s clients include health systems, health insurers, public health officials, and other healthcare organizations. When working with health systems, Spatially combines its own observational SDOH data with patient data from hospital surveys to paint a fuller picture of that specific patient population and their needs.

For example, if a patient with a chronic condition is regularly missing appointments, Spatially can help a health system use SDOH data to identify barriers, like if the patient is an hourly worker who is “most likely not going to leave work for a couple hours to go for an appointment,” Meider-Alfi said.

The health system can then see what options may help the patient, including trying telehealth, giving the patient a doctor’s note to ask for time off, or even referring the patient to a weekend mobile clinic if available, she added.

“It’s really important to connect the dots to give this comprehensive view, which [the Centers for Medicare and Medicaid Services (CMS)] is really driving here. What they’re really driving with health equity is being able to treat the whole patient,” Meidar-Alfi said.

Data-driven policy

At the federal level, agencies are finding gaps in data collections that may hinder the creation of health equity-related policies.

But CMS is raising the bar for other healthcare organizations with the 2022–2032 Framework for Health Equity. The framework sets guidelines for collecting and analyzing demographic and SDOH data, and helps the agency address health disparities across Medicare, Medicaid, and other CMS programs. The first step is to use health data to gain insight into the state of persistent disparities and track progress toward mitigating inequities.

The agency, however, doesn’t collect data on some SDOH (e.g., income), and patients may also opt out of providing the self-reported data.

The framework outlines how CMS plans to improve data collection. Organizations, according to the framework, can fill in data gaps with other sources including sexual orientation and gender identity (SOGI) data from the Medicare Current Beneficiary Survey, which tracks socioeconomic and demographic data of Medicare beneficiaries, or SDOH data gleaned via certain ICD-10 codes.

Local solutions

Hayden and her team found in their study that population health data at one of the largest health systems in New York City is often inconsistent and incomplete, due in part to patient information that’s migrated from other practices.

“We have some practices that we acquired—small doctor offices for network practices—and they don’t collect this data. So when you want to migrate into the health record, which is Epic, it comes in as ‘unknown,’” Hayden said, adding that the unknowns are often in categories like ethnicity or language preferences.

Race and ethnicity data can be sensitive subjects for some patients, which Hayden said can make it difficult to collect, but she still recommends that medical professionals looking to get this information to at least try—and said that sometimes it requires a little creativity to build a complete patient population data profile.

A 2021 report from the American Hospital Association’s (AHA) Center for Health Innovation found that some health systems don’t look beyond the clinical data in their own electronic health records to understand health disparities. But they can look at other datasets, such as community health needs assessments, clinical data from affiliated and unaffiliated providers, and medication adherence data from pharmacies, per the AHA report.

“We want to find out as much [about] our patients as we can so that we can, if we have to, customize their health plan,” Hayden said.