Standardizing Social Determinants Of Health Assessments
Care delivery to patients with unmet social determinants of health can be more challenging, given that 80 percent of factors that impact a person’s health are due to socioeconomic, environmental, or behavioral factors—all of which are outside the walls and outside the “control” of the medical office or hospital.
by Douglas P. Olson, Benjamin J. Oldfield, Sofia Morales Navarro | Mar 18, 2019 | This article originally appeared in Health Affairs
SHARE THIS POST
Standardizing Social Determinants Of Health Assessments
Screening for social determinants of health (SDOH) is increasingly done in primary and specialty care settings. Payers and health care organizations have recognized the importance of these determinants not just to clinical outcomes but also to cost and use of services. Toward that end, both clinical and financial cases have been made for an expanded focus on SDOH for many, if not all, patients. Which patients should be screened, how to screen them, and how to incorporate these data in medical records and care plans remain unclear. Harmonizing documentation of social determinant of health data—for health care organizations, community-based organizations and payers—is critical to ensure SDOH assessment and mitigation are standardized, trackable for individuals and populations, and formalized in health reform efforts.
Health care institutions seeking to implement SDOH as part of a routine medical visit have historically faced three major systems challenges:
- Lack of a standardized SDOH screening tool in the electronic health record;
- Reliance on clinical provider staff (that is, doctors, advanced practice registered nurses, and physician assistants only) to screen and document for SDOH; and
- Lack of a standardized crosswalk between SDOH and diagnostic codes for documentation.
Here, we describe how systems might overcome these challenges, based on our experiences at Fair Haven Community Health Care (FHCHC) in New Haven, Connecticut, a community health center and patient-centered medical home providing integrated medical, behavioral health, dental, and addiction medicine services to more than 18,000 individuals and their families. We also describe the implementation of a novel crosswalk that links SDOH to International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10) diagnoses.
Lack Of A Standardized SDOH Screening Tool
A variety of validated questionnaires exist to screen for SDOH. Two of the most widely used are the accountable health communities (AHC) instrument and the protocol for responding to and assessing patients’ assets, risks, and experiences (PRAPARE) instrument. Both assess for multiple, overlapping social needs and are applicable to diverse care settings. Using a widely accepted, validated tool achieves many objectives of screening. Where possible, clinics and health systems that share patient populations and catchment areas should implement the same screening instrument to facilitate the data aggregation and the standardization of practices across the levels of the health care system (for example, outpatient, emergency, and inpatient, as well as primary care, dental care, and behavioral health care). We have implemented the AHC screening tool in partnership with other local practices in New Haven, Connecticut.
Engaging More Staff To Screen And Document For SDOH
Assessing and documenting SDOH has often been thought of as the job of the clinician, generally during the collecting of a social history. One reason for this is that, historically, collecting information that may lead to an assessment or diagnosis has been restricted to those professionals who make medical diagnoses.
Official ICD-10 coding guidelines now allow the entire clinical care team—care managers, community health workers, medical assistants, nurses, discharge planners, behavioral health clinicians, dental staff, and so forth—to document social risk in the electronic health record. The ICD-10 provides an expanded set of codes reflecting patients’ social characteristics in the form of “Z-codes” (Some examples: Z59.9 is “Problem related to housing and economic circumstances, unspecified”; Z60.4 is “Social isolation, exclusion and rejection”). These ICD-10 codes provided more specific diagnoses, and as a result, greater opportunities to document and detail both biomedical and social risk. As noted in the ICD-10-CM Official Guidelines for Coding and Reporting, FY 2019from the Centers for Medicare and Medicaid Services (CMS): “For social determinants of health … code assignment may be based on medical record documentation from clinicians involved in the care of the patient who are not the patient’s provider since this information represents social information, rather than medical diagnoses.”
As a community health center where care coordinators and behavioral health clinicians work alongside primary care providers, we have implemented both team-based care and team-based responsibility for that care. For example, patients may get screened for SDOH by a clinical assistant during an annual visit, by a care coordinator who has received a referral out of a clinician’s concern for social risk, or by front desk personnel, depending on the risk profile of the patient. Integrated delivery systems such as ours are ideal for SDOH screening. Other organizations that practice team-based care can make use of different professionals to implement SDOH screening and document Z-codes at various points along the flow of patients through services.
Lack Of A Standardized SDOH/ICD-10 Crosswalk For Documentation
A widely agreed-upon crosswalk that links SDOH to corresponding ICD-10 codes does not exist. Among a broad array of ICD-10 codes, our organization developed such a crosswalk (Exhibit 1) to link positive screens using the AHC screening questions to diagnoses that are added to the patient’s problem list in his or her medical record. Candidate diagnoses were identified by clinical care teams and then individual diagnoses were chosen from these by consensus among an interprofessional group. In early 2018, agreed-upon diagnoses were then linked to our screening tool so that those performing the screening can enter corresponding diagnoses. As a result, various types of professionals can now systematically code for SDOH using agreed-upon ICD-10 codes. While internal evaluation is ongoing, these early efforts have already allowed us to identify a SDOH need in more than 75 percent of those screened. We are exploring models to examine risk with and without SDOH needs. These screenings have also affected our organizational strategy for scaling these efforts and enhancing and developing new community collaborations based on prevalence.
Exhibit 1: Social Determinant Of Health: ICD-10 Crosswalk
Source: Fair Haven Community Health Care.
Challenges And Limitations
As specific as they may be, however, ICD-10 codes do not tell the complete story of a patient’s social risk. Specificity—and the multitude of diagnostic choices offered by ICD-10—may make population-level assessments difficult. As noted by Laura Gottleib and colleagues, “ICD-10 was not initially designed to ensure sufficient specificity to inform action, whether at the individual or population level.” A population health approach to SDOH screening requires unified, category-based diagnoses that facilitate an identification of populations based on their social risk. For example, the National Health Care for the Homeless Council has provided a pragmatic, detailed brief on screening and documenting for homelessness. This focuses on a single ICD-10 code, Z59.0. However, a total of 11 codes can be used to describe homelessness and housing inadequacy within the Z59 code set (Z59 to Z59.9), with additional codes in the Z77 series to indicate housing inadequacy. Balancing population impact with specificity is a challenge and a shortcoming for many cross-walked codes.
Opportunities For Improvement
There are social determinants of health that lack ICD-10 codes altogether, such as language barriers, immigrant or migrant status, transportation difficulty, and vulnerabilities due to insurance status. Adding these codes to future editions of ICD classifications will help to more fully capture social risk and challenges that have important, clinically appreciable impact on care delivery. Some SDOH are driven by complex social and structural phenomena, such as racism or fear of deportation, and may be difficult to describe in a diagnostic construct such as ICD-10.
Creating a “perfect” crosswalk is a Herculean task and will require approaching the process from a systems-based, clinically focused viewpoint that will engage the expertise not only of clinicians but also of those who provide other critical social services, such as public housing and legal aid. We suspect that other organizations have cross walked ICD-10 codes and SDOH like us, and others will do so more frequently. To prevent an abundance of discordant crosswalks, it would be advantageous for those already providing leadership in this space (community health centers, the Health Resources and Services Administration, the Center for Medicare and Medicaid Innovation, CMS, and payers) to offer guidance swiftly to align implementation with alternative payment methodologies. The crosswalk we have implemented is a start to this process.
Next Steps: Upstream Potential
Care delivery to patients with unmet social determinants of health can be more challenging, given that 80 percent of factors that impact a person’s health are due to socioeconomic, environmental, or behavioral factors—all of which are outside the walls and outside the “control” of the medical office or hospital. Health has deep roots in communities, non-health care institutions, policies, and social hierarchies such as those posed by race, class, and gender. Quantifying that complexity is difficult.
As we seek to recognize the importance of SDOH on health outcomes, we must create systems that support their integration into care delivery. While primary care may take the lead, specialty care, behavioral health, dentistry, pharmacy, and all other interactions that people have with our care delivery system need to collectively own these efforts. By standardizing our approach to SDOH across professional roles and health care systems, we can facilitate the aggregation of data that will ultimately inform payment redesign, help patients meet social needs, and support healthier communities.