Leveraging ICD-10 Codes for Social Determinants of Health Data | How Can We up the Ante?
There is growing consensus that social determinants of health (SDOH) are powerful influencers of healthcare outcomes, utilization and cost. According to a report issued in December 2019 by the nonprofit Center for Open Data Enterprise, healthcare organizations are increasingly realizing that SDOH data holds great potential to better understand patient needs – noting that “a person’s ZIP code can be at least as important as his or her genetic code in predicting health risks”. However, there is less agreement as to how to collect and effectively use this information in patient care, performance measurement and payment systems.
In November 2019, the University of Maryland held its annual academic Conference on Health IT and Analytics Conference, a premier event focused on some of the newest approaches in healthcare algorithms aimed at optimizing patient outcomes. One of the themes was the presence and importance of assessment of social and economic information that influence a person’s risk for adverse health outcomes and accessibility to needed healthcare services and treatments.
To understand the impact SDOH can have on overall health outcomes, it’s important to recognize the difference between SDOH, social needs and social risk factors. The Center for Open Data Enterprise report defines them as follows:
Social Determinants of Health – The conditions in which an individual is born, lives, works and ages, as well as the other aspects that shape how they live daily life, are broadly considered SDOH. SDOH affect everyone and can be either positive or negative.
Social Needs – Non-medical needs, including food, housing and transportation, are an individual’s social needs.
Social Risk Factors – Food insecurity, housing instability, residence in a high-crime area or one with a lack of recreational facilities are specific adverse social conditions that are counted as social risk factors and can result in poor outcomes and impact an individual’s health outcomes, as well as the health of an entire community.
Social Determinants of Health: The Other 80 Percent
There is increasing evidence and recognition that clinical care contributes only 10% to an individual’s overall health compared to social and economic factors, which contribute 40% to one’s overall health. Government stakeholders are demonstrating recognition and support for innovations that will positively impact SDOH, as well; starting in 2019 the Centers for Medicare and Medicaid Services (CMS) authorized Medicare Advantage (MA) plans – which now cover over one-third of all Medicare beneficiaries – to cover non-medical, supplemental benefits to address SDOH, such as food delivery, transportation, and in-home services.
Figure 1: What Contributes to Overall Health
Source: Susan Denzer, Forging a National Agenda to Advance Health Care Without Walls, AHIP’s National Health Policy Conference, March 13, 2019
As the number of Medicare Advantage members grows, so too does the number of members who are underserved, and health plans are increasingly aware of the need to implement strategies to help them understand the conditions – both medical and non-medical – that are affecting how patients are using the system. Analysis of MA plan benefits data released by CMS indicates these plans “will continue to increase their supplemental benefits offerings in 2020, including meals, transportation, acupuncture, and over-the-counter (OTC) benefits, among others.”
“Truly, in America, your ZIP code is still more important than your genetic code. And as a society, we must ‘change the map’ and finally address health disparities stagnant for generations. … We will never solve our healthcare spending addiction, or our progressive degradation of health, unless we focus on what causes 80 percent of the poor health outcomes in this country.” – Admiral Brett Giroir, Assistant Secretary for Health, U.S. Department of Health and Human Services
However, capturing information about SDOH from one data source is an often daunting – and sometimes impossible task. Today, the most commonly referenced source of data is derived from Census data aggregated at the 5-digit ZIP code level or from the American Community Survey (ACS), which is aggregated at the Census Block level. This information is most often used for marketing or financial credit score estimates. While a person’s SDOH can be gauged from information derived from the general population using public databases, this assessment comes with a fair amount of error due to the large and diverse populations covered in these aggregated databases.
Making the best decision about how to treat and engage a patient should be based on the clinical and social characteristics and needs of the individual patient. How do they want to receive care? What are his or her individual health factors or barriers that may impede access to care? For providers, this means interacting with patients on a level that is much more personal than usual, going beyond the typical conversation about symptoms and medical history, and much of this information is often documented in text within medical records at the point of care. However, large-scale analysis of medical records for social determinants is very challenging.
Is Using ICD-10 Claims Codes to Identify Social Determinants of Health the Next Frontier for Healthcare Data?
Access to patient data on SDOH has been the biggest challenge to integrating that information into patient care. However, the categorical codes derived from medical records for billing purposes are increasingly storing information about a patient’s behaviors and status and are uniquely well suited for medical research studies and care optimization. In the ICD-10 claims code library compiled by CMS, ICD codes Z55 through Z65 are specifically focused on identifying patient’s social factors, and use of these codes is growing — as demonstrated by the increasing frequency of these codes flowing into Inovalon’s proprietary dataset, the MORE2 Registry® (Figure 2). There are approximately 131 codes that fall into this range of ICD codes. Between 2015 and 2019, the number of these codes reported per year has increased nearly 10-fold.
Figure 2: Quantity of SDOH ICD Codes Reported by Year
Source: Inovalon’s MORE2 Registry®
The bulk of these recorded codes focus on housing and financial information, and secondarily on upbringing and family, but the codes can also capture exposure to noise, income level, social rejection, exposure to a dramatic event and even sibling rivalry. While the increase in the use of such codes can be attributed to greater understanding of the importance of social factors, it has also become easier to record them at the provider site due to wider use of electronic medical records at the site of care – helping to drive inclusion of SDOH data in patients’ charts.
Figure 3: 2018 Distribution of SDOH Diagnosis Codes
Source: Inovalon’s MORE2 Registry®
These codes can also be stratified by who reported them. While there is a high variation in the source of these codes, they are predominantly associated with in-home services covered by Medicaid that require a code for the care, such as meal service. These claims are reported by vendors directly to payers and in many cases these services are reimbursable. With the newly expanded ability of MA plans to offer and pay for non-medical services, the importance of using these codes is greater than ever.
Figure 4: Sources of 2018 Z55-Z65 Claims
Source: Inovalon’s MORE2 Registry®
Access to information on SDOH can provide a more holistic view of each individual patient and a better understanding of who is likely in need of specific non-medical services that would improve their health such as transportation, food delivery, and in-home assistance. As healthcare providers, health plans and health systems expand use of SDOH data, we will be better positioned to address both individual patient and overall community needs – an important step toward population health management. Effective use of this data requires appropriate infrastructure, data standardization, and resources to ensure you’re capturing the right data on the right patient and then using that data to make decisions about the healthcare delivered.
If you’d like to learn more about how Inovalon is helping healthcare organizations use claims codes to identify social determinants, social needs and social risk factors to improve quality outcomes, risk score accuracy, and economic performance, click here to contact us.