More Than a ZIP Code: Socioeconomic Data for Improving Patient Outcomes

Socioeconomic data can help healthcare organizations determine how best to tailor care for individual patients.

About one-third of hospitals and health systems responding to a recent industry survey said they participate in voluntary value-based payment models that offer bonus payments to organizations that meet cost and quality targets. Research suggests that these organizations can be substantially helped toward meeting these targets if they incorporate socioeconomic data into existing predictive models and care management. The value of such data is evident when one considers report findings indicating that medical determinants account for only 20 percent of healthcare outcomes, whereas social determinants of health (SDOH) account for as much as 50 percent. Using these data, hospitals and health systems can score patients’ health risk to enhance outreach and care coordination that could increase financial returns under the Medicare Access & CHIP Reauthorization Act of 2015 (MACRA) and other payer-driven performance bonus programs.

The Shifting Needs of the Healthcare Market

Numerous changes are taking root under value-based payment, causing disruption to current business models. Care payment models are evolving to reward quality rather than paying for services that are not held accountable for outcomes. Providers are beginning to consider bonuses or penalties in payment determined by a prior period’s care performance.

The Centers for Medicare & Medicaid Services (CMS) is implementing the value-base payment approaches via two Quality Payment Program tracks under MACRA: Merit-based Incentive Payment Systems and advanced alternative payment models (APMs). In other words, MACRA has intensified the impact that the quality and value focus has on physicians’ revenue.

Care delivery models also are changing. New organizational structures such as accountable care organizations (ACOs), patient-centered medical homes (PCMHs), retail point-of-care clinics, and others are being created to help improve care quality, coordination, and results as well as business risk-sharing.

Although revenue recognition standards are clear, many healthcare organizations are not sure of how they will be affected by the shift from the fee-for-service to value-based payment. As pressure mounts to improve patient outcomes, two things have become obvious:

• Care delivery practices need to be augmented to address the need to improve outcomes.
• Better clinical data are needed to improve patient care and help practices support the new revenue basis.

Prevention: More Cost-Effective Than Cure

The value-focused care model is based on the recognition that many common health conditions or procedures could be avoided if more efficient or appropriate care protocols surrounding coordination, engagement, and timing of intervention were followed. At least 25 cents of every healthcare dollar is spent on treatment of diseases or disabilities that result from potentially changeable behaviors. Moreover, costs for patients less actively involved in their own health care are 8 to 21 percent higher than those for more actively involved patients, and the former are nearly twice as likely to be readmitted within 30 days.

Traditional “diagnose and treat” models also can be financially detrimental. Value-based payment seeks to address this problem by requiring healthcare providers to be proactive in reaching out to patients who are most likely to present with more medical complications before the cost of care required strangles the revenue associated with it. This model also provides a financial incentive for organizations to encourage patients to seek proper follow-up care.

Efforts to reengineer the provider business model should focus on proactive patient engagement because it can potentially reduce complications, costs, and even diagnostic severity. Providers should augment traditional, reactive methodologies with data and processes that enable them both to identify which patients need help first and to know how to help those patients achieve optimal results. A strategy using SDOH can enable providers to shift to a proactive focus and obtain corresponding financial benefits.

Social Factors and Health

Providers have long recognized a connection between a patient’s health and social factors— defined by the World Health Organization as “conditions in which people are born, grow, live, work, and age”—that have an impact on their health. The U.S. Department of Health & Human Services recognizes four areas of SDOH in addition to health and health care: social and community context, economic stability, education, and neighborhood and built environment. The environments in which people live have an impact on the likelihood of their developing health conditions and whether they will be able to effectively manage those conditions.

Consider the following:

• One out of every three deaths in the United States is related to social factors.
• Social isolation can increase risk of heart disease by 29 percent and stroke by 32 percent.
• Changes in income, work, and family dynamics are the top three causes of stress, and stress drives 75 to 90 percent of primary care visits.

Providers who factor the SDOH considerations into care recommendations and clinical workflows may change patient outcomes. If care recommendations do not work within a patient’s physical environment or are not affordable or conveniently located, or if the patient cannot understand them, they will not be effective at improving health.

With only medical data on individuals, providers also may be missing opportunities to intervene before a condition develops or worsens. SDOH can expose otherwise hidden risks. Of course, no one piece of data provides enough basis for understanding an individual, but by looking at socioeconomic data on various SDOH, a picture of the individual emerges that can help personalize care.

Historically, socioeconomic data were not readily available, but recent advances in data technology have made it possible for healthcare organizations to collect and analyze data and turn it into insights that can guide the actions of medical personnel.

Clinically Validated SDOH Attributes

There are hundreds of socioeconomic attributes; addresses, demographics, interests, and literacy are just a few examples. However, not all SDOH data correlate to health outcomes or are of high enough quality to be relied upon. For example, ZIP codes and other basic demographic information obtained from electronic health records (EHRs) can be useful, but that type of information provides a narrow view into patients’ complex lives. Some organizations rely on risk-based appraisals filled out by patients. These surveys are helpful but difficult to refresh and not always reliable. Furthermore, they provide point-in-time data only, which entirely misses insights available when information is collected over time.

In evaluating SDOH data for use in predictive models and care management, healthcare organizations should look for current, comprehensive, and longitudinal data that can be consistently linked to their patient populations and provided in a standardized format.

One of the most comprehensive and reliable sources of such data that can be used for assessing and predicting potential health risks is public records data. Examples include the following.

Personal finances. Financial instability causes a great deal of stress. It is also often the reason people put off physician visits and stop buying prescribed medication, potentially compromising their health.

Education. Lower levels of education are linked to lower health literacy, which can increase risk levels.

Voter registration. Individuals showing a sense of responsibility for their community may be more likely to take responsibility for maintaining their own health.

Law enforcement. Records pertaining to accident investigations could indicate future medical issues.

Derogatory records. Liens, evictions, and felonies may indicate that a person does not prioritize health.

Another key criterion for selecting SDOH attributes is ensuring they have been clinically validated against actual healthcare outcomes to confirm their predictive power. Organizations should confirm that the SDOH information has been validated or be prepared to validate it through their own internal analytics groups. For example, healthcare organizations can choose target outcomes, such as likelihood of readmission, and then compare the accuracy of predictive scores incorporating SDOH attributes with actual outcomes they see in claims data or medical records to determine if there is an improvement in predictive accuracy.

The key to using SDOH is ensuring multiple attributes are considered. Providers should look holistically across social determinants to make predictions and care plan recommendations.

Opportunities for Implementation

There are two main ways of implementing SDOH: as attributes and as predictive health scores.

Attributes. SDOH attributes can be used in clinical and analytic models to better assess and predict risk for patients. Organizations with analytics teams can add the attributes to data they already incorporate into predictive models to determine if it improves accuracy. Attributes provide a holistic view of the patient by providing insights into the patient’s educational background, social environment, economic situation, and physical surroundings. These attributes do not replace the value of medical data but can be powerful in the absence of claims data or as a supplemental data source to more comprehensively understand the individual.

Scores. Predictive health scores leverage hundreds of socioeconomic attributes to provide healthcare entities with a picture of future risk. These attributes are linked to individuals and are then used to create and deliver a composite score at the individual patient level. They can be useful for those who want a quick way to stratify their patient population for risk. For example, social determinants can be used to predict who is likely to be readmitted to the hospital within 30 days. The higher the risk score, the more likely they are to be readmitted.

SDOH attributes and scores can help with prevention initiatives as well as discharge planning and follow-up care. Both socioeconomic attributes and scores can be used in risk stratification and care management initiatives to help healthcare organizations better understand and manage risk among patient populations.

Let’s look at two hypothetical examples to illustrate how an SDOH score can be put to work.

Existing patients Ruth and Layla are both 55 years old and have the same medical conditions: diabetes and peripheral arterial disease. An outreach program would logically treat these patients the same. However, Ruth has had two bankruptcies, has no degree to assist her in the job market, lives in a high-crime neighborhood, and has no family living within a 500-mile radius. On the other hand, Layla has a master’s degree, lives in a low-crime neighborhood, and has family within 25 miles. Armed with that knowledge, a provider can now adequately develop a predictive model that it can use to allocate resources, where it will intervene immediately with Ruth and rely on more routine outreach to manage the disease condition for Layla.

The situation is similar with two hypothetical new patients, Vivian and Alexis. Without any kind of medical history (e.g., no knowledge of any existing chronic conditions), it would be tempting to provide them both with the same care recommendations or wait until a condition developed, but social determinants show a different story.

After performing an SDOH evaluation of each patient, the provider ascertains from Vivian’s SDOH score that she is at high risk for having high healthcare costs. She would be a good candidate to proactively approach for care management. On the other hand, because Alexis has a low risk score, the provider can reasonably conclude that general wellness guidance will suffice in her case. This ability to risk-stratify for outreach is particularly important as organizations across the healthcare ecosystem struggle with limited resources.

A Healthy Dose of Skepticism

The SDOH approach is new and comes at a time when healthcare organizations are overwhelmed with patients, claims, evolving payment models, and regulatory scrutiny. Innovations are regarded with careful and appropriate skepticism. At the same time, no one is likely to argue that the rapid pace of technological innovation—in health care and beyond—challenges the industry to be more receptive when considering disruptive innovation such as SDOH. Ignoring this and similar opportunities to drive change will leave an organization behind while the industry evolves in the value-based payment context.

For example, ACOs, PCHMs, and other organizations operating under risk-sharing agreements are moving beyond the medical record to improve care outcomes from the administrative level. Findings of a recent survey of clinical and administrative leaders at 17 ACOs suggest that the healthcare workforce structure is evolving as new staff roles, such as health coaches, are being introduced to focus solely on the success of patient engagement. The value-based-payment models have led ACOs to use more interdisciplinary teams, with roles stratified based on patient risk categories, and focus resources on improving care for the highest-risk patients. These teams will require tools such as SDOH to help them perform their tasks effectively.

The Value of SDOH

With an increasing emphasis on value in health care, providers must closely track value measures such as complications, hospital-acquired infections, and readmissions, and they have financial incentives to ensure their patients remain healthy. SDOH can help healthcare practitioners accurately identify high-risk patients that require action and focus on prevention. Providers can reach out to patients and encourage them to make lifestyle changes that promote good health. Providers can offer guidance to patients managing chronic illness that works within their lives, potentially reducing severity. For patients recently released from the hospital, aftercare counseling could prevent complications and readmissions.

The resultant savings in healthcare costs could be significant under a value-focused care delivery and payment model. Organizations may minimize financial risk exposure while maximizing performance incentive payments. Although hard to quantify, the care cost avoided through prevention efforts informed by SDOH also can be substantial. Perhaps most important, these new tactics can help providers engage and retain patients while keeping those patients as healthy as possible.

---

Erin Benson is director, market planning at LexisNexis Health Care, Alpharetta, Ga.

Footnotes

a. Philips, “Transforming Healthcare to a Value-Based Payment System,” The Washington Post, Accessed online on Nov. 13, 2017.

b. Booske, B.C., Athens, J.K., Kindig, D.A., Park, H., Remington, P.L., Different Perspectives
for Assigning Weights to Determinants of Health, University of Wisconsin Population Health Institute, County Health Rankings Working Paper, February 2010.

c. Weinberg, J.L.,“Medicine’s Response to Lifestyle-Related Preventable Illness,” From the Editor, AMA Journal of Ethics, Virtual Mentor, April 2013.

d. Ballou-Nelson, P., “Activate Patients in Your Practice to Improve Outcomes,” MGMA Connection Plus, Jan. 22, 2016.

e. World Health Organization, “About Social Determinants of Health.”

f. Office of Disease Prevention and Health Promotion, Healthy People 2020. “Social Determinants of Health,” Accessed June 27, 2017.

g. Heiman, H.J., and Artiga, S., “Beyond Health Care: The Role of Social Determinants in Promoting Health and Health Equity,” Issue Brief. Nov.4, 2015.

h. Cleveland Heart Lab. “Loneliness as a Risk Factor for Heart Disease and Stroke,” May 9, 2016.

i. Robinson, J., “Three Quarters of Your Doctor Bills Are Because of This,” The Blog, Huffpost, May 22, 2013, updated, July 22, 2013.

j. Sandberg, S.F., Erikson, C., and Yunker, E.D., “Evolving Health Workforce Roles in Accountable Care Organizations,” The American Journal of Accountable Care, June 12, 2017.