Use Social Determinants of Health Analytics to Inform Health Equity Strategy

Overview

Introduction

Health equity initiatives aim to enable all people to achieve their optimal health. These initiatives typically focus on reducing or eliminating health disparities by addressing common barriers to equitable health attainment, like food deserts or gaps in access to primary care.

Researchers and healthcare organizations have long recognized the relationship between SDOH and health disparities, such as the statistically significant differences in infant mortality rates across ethnicities in the U.S. Globally in aggregate, income is a primary determinant of infant mortality disparities, with high-income countries averaging four infant deaths per 1,000 live births while low-income countries average 47 deaths.

The COVID-19 pandemic has drawn mainstream media and widespread public attention to the severity of health disparities, as case and death rates along with immunization vary widely by demographics. In part due to this new visibility outside of healthcare organizations, health equity-related investments are increasing worldwide and the vendor solutions are maturing.

The success of these efforts and solutions depends on whether the organization or vendor can identify causes of disparities and recommend effective interventions to close gaps. SDOH analytics are foundational capabilities for successful health equity strategy development and execution.

CIOs, data and analytics and informatics leaders are accelerating SDOH analytics efforts to deliver insights that identify and shape specific initiatives to improve health equity (see Figure 1).

Analysis

Every organization will have a different approach to setting and measuring its health equity goals based on the various health disparities within the unique cohort of individuals and communities it serves. The first two steps in health equity strategy development are identifying which health disparities the organization will address and setting achievable outcomes the initiatives will target. Examples include reducing disparities across populations in:

• Under-five child mortality rates by 50%

• Diabetes diagnosis by 10%

• Under-30 day hospital readmissions by 15%

• COVID-19 vaccination rates by 30%

• Behavioral healthcare access by 20%

After identifying these target disparities and goals, assess the business practices that contribute to disparities and determine which of these your organization can credibly influence. These practices typically include:

• Providing minimal or no appropriate in-network care delivery services within close proximity to the adversely-affected cohort (where the definition of “close” is relative to the most prevalent means of transportation for the cohort).

• Underestimating the digital divide that limits access to virtual care, even when it is available.

• Failing to provide educational materials, customer service capabilities and healthcare-related communications at the reading level and in the language of the adversely-affected cohort.

• Limiting access to care navigators who can help close gaps in health literacy and coordinate service access and utilization across settings with formal and informal (family and community) care teams.

• Underinvesting in partnerships with community service organizations to facilitate referrals for nonmedical needs.

To be effective, health equity strategies must incorporate and address these contributing factors to disparities that are under an organization’s control. Collaborate with executive and clinical leaders to establish this strategy framework.

Most healthcare organizations recognize the socioeconomic SDOH domains such as race and ethnicity, education attainment, and food and housing insecurity. However, SDOH includes a wide and complex set of characteristics and circumstances that affect health equity, which is inclusive of behaviors. The World Health Organization (WHO) defines the following as SDOH domains:

• Income and social protection

• Education

• Unemployment and job insecurity

• Working-life conditions

• Food insecurity

• Housing, basic amenities and the environment

• Early childhood development

• Social inclusion and nondiscrimination

• Structural conflict

• Access to affordable health services of decent quality

WHO’s high-level SDOH domains include geopolitical factors that other organizations do not describe, but which researchers are increasingly demonstrating have causal effects on health outcomes. For example, structural power differences that harm or benefit groups relative to others perpetuate disadvantaged people’s mistrust in governmental and other authoritative actors. This mistrust contributes to health disparities such as vaccine uptake in Black and socioeconomically disadvantaged white individuals.

The National Academy of Medicine (NAM) has taken most of WHO’s SDOH domains — as well as a number of others from the Institute of Medicine (IoM), the NAM’s former name — and sourced or developed measures to assess them. Table 1 describes SDOH domains across five categories that the NAM defined in its consensus study report.

Data and accurate measures across these domains vary in availability, sensitivity and usefulness to health equity initiatives. Health equity teams have to weigh the effort and ethics involved in curating and analyzing the domains of data with their ability to address or influence the measure. Some measures like smoking status are routinely identified in the course of typical patient intake or member onboarding practices. These types of measures also tend to have widely-accepted evidence-based programs to improve measure performance such as smoking cessation initiatives.

However, many data points are challenging to capture — let alone address — such as exposure to violence, sexual practices, exposure to firearms or work conditions. Capturing and affecting these more elusive domains requires a long-term commitment to addressing systemic issues through partnerships with entities such as educational institutions, government organizations, faith-based organizations and competitors.

Your organization’s health equity strategy must acknowledge the complexity of identifying social and behavioral determinants and prioritize those domains that you can feasibly access and assess in the short term. At the same time, you must identify the necessary partners to tackle systemic challenges over the long term.

Whole-person needs identification is the basis for any SDOH assessment, and both individual and community-level insights are valuable. Healthcare organizations can employ three methods to develop SDOH insights:

• Directly assess needs through individual interactions, such as whether a person has access to healthy food. This method uses screening tools, such as the U.S. Centers for Medicare and Medicaid Services (CMS) Health-Related Social Needs Screening Tool,Health Leads’ Social Needs Screening Toolkitand the PRAPARE Implementation and Action Toolkit.

• Derive needs at the person level by matching and blending the data captured by screening tools, public records and consumer data sources. This approach uses the person-level blended dataset to assign risk by comparing an individual’s data to thresholds for SDOH domains, such as education level attainment. This is often packaged as a social vulnerability index or risk score. These capabilities may also append specific attributes to an individual’s longitudinal record, such as home ownership. Representative vendors for these solutions includeAcxiom, CentraForce Health, Experian, LexisNexis Risk Solutions, Socially Determined and Unite Us (via its Carrot Health acquisition).

• Infer and predict needssuch as what culturally and linguistically appropriate services are necessary to provide equitable care access to underserved communities. This approach applies analytic processes to population-level data sources including the U.S. Census Bureau’s American Community Survey Data, the United Nations Statistics Division and WHO’s Global Health Observatory data.

Using these methods, analytics teams can identify and predict nonmedical needs in four distinct ways.

Analyze community-level SDOH data sources such as census details or weather data to identify hot spots where specific nonmedical determinants (like heat waves or high pollen counts) can have a significant effect on healthcare access or health outcomes.

Apply advanced analytics to community-level SDOH data sources to derive risk and propensity scores for emerging hot spots of determinants (like internet access or food deserts) that are likely to have a significant effect on healthcare access and outcomes.

Survey or interview individuals to directly assess SDOH factors like transportation, nutrition and education. Or, infer and derive SDOH factors from person-level consumer data sources.

Apply advanced analytics to person-level data sources — including any survey or interview insights, as well as consumer data — to derive risk and propensity scores for SDOH factors.

Each of these analytic approaches support a multitude of use cases that contribute to reducing disparitiesand improving health equity.

There is an art to applying SDOH insights, as the link between the need and the outcome target is often indirect. For example, prenatal care has a statistically significant effect on maternal mortality, infant mortality, birth weight and preterm births. And there is a significant disparity in the timely access and consistent use of prenatal care across populations. Developing an approach to closing the gap should include the following elements, as seen in these theoretical examples:

• A target outcome — Reduce a health disparity gap for first-trimester prenatal care initiation and subsequent recommended utilization between non-Hispanic white pregnant women and pregnant women from historically underrepresented groups in the same county by 50%.

Evidence

About Infant Mortality, CDC.

Mortality Rate, Infant (per 1,000 Live Births), World Bank Group.

Race Gaps in COVID-19 Deaths Are Even Bigger Than They Appear, Brookings Institution.

Social Determinants of Health, WHO.

Examining Power Relations to Understand and Address Social Determinants of Vaccine Uptake, JAMA Network.

Capturing Social and Behavioral Domains and Measures in Electronic Health Records, The National Academies Press.

Validity of a Single-Item Measure of Stress Symptoms, PubMed.

AUDIT-C for Alcohol Use, MDCalc.

BMC Fam Pract., National Center for Biotechnology Information.

Social Isolation: A Predictor of Mortality Comparable to Traditional Clinical Risk Factors, National Center for Biotechnology Information.

The Accountable Health Communities Health-Related Social Needs Screening Tool, Centers for Medicare & Medicaid.

The Health Leads Social Needs Screening Toolkit, Health Leads.

PRAPARE Implementation and Action Toolkit, PRAPARE.

Mobile Reproductive Health Clinics, County Health Rankings & Roadmaps.

A Prenatal Checkup on Wheels: Are Mobile Units a Solution to Maternity Care Deserts?, Today.

A Mobile Prenatal Clinic Is Breaking Barriers for Pregnant Refugee Patients, HealthCity.

Gartner Recommended Reading