Over the last five years, clinicians, health plans and policymakers have ramped up their focus on the social determinants of health (SDOH). SDOH are non-medical factors (food security, housing security, reliable access to transportation, etc.) that influence a person’s health status. Across the nation, governments, health insurers and health systems are investing in SDOH interventions in the belief that addressing SDOH will improve the lives of vulnerable people and, by doing so, lower the cost of care.
The Trump administration is encouraging creative approaches for addressing SDOH without increasing federal health care spend. For example, in Medicaid, several managed care organizations are now experimenting with providing housing for plan members who are homeless and ill. In Medicare Advantage, health plans have new opportunities to use Medicare dollars to provide SDOH-oriented benefits, including meals, companionship and even pest control for members with specific diagnoses. At least a half dozen vendors now offer SDOH platforms to help health plans and health systems coordinate patient medical and social services. And fee-for-service Medicare’s Accountable Care Organizations have the ability to use Medicare payments to provide a range of SDOH interventions and to creatively contract with the providers of those services.
On January 23, the Centers for Medicare and Medicaid Services (CMS) quietly released the government’s first report on the Z codes of Medicare beneficiaries, Codes Utilization among Medicare Fee-for-Service (FFS) Beneficiaries in 2017. Z codes, a recent addition to the taxonomy of medical conditions used for billing (called ICD-10), permit physicians and other health care providers to identify non-medical concerns in their patients. While Z codes do not align perfectly with the social determinants of health, they are the only standardized mechanism currently in place for gaining a national view of SDOH from the perspective of health care providers.
To date, a few studies examined the (under-)use of Z codes, and one study was recently published on Z codes for the Medicare population in the Journal of General Internal Medicine. While CMS’s report is comparatively modest in its level of analysis, it is still important. This is because it takes a new look at SDOH in the Medicare population based on assigned Z codes. Key findings from the report include:
The five most commonly assigned Z codes were for:
- Problems related to living alone.
- Disappearance and death of family member.
- Problems in relationship with spouse or partner.
- “Other specified problems related to psychosocial circumstances.”
It is interesting that some of the most commonly assigned Z codes relate to familial/interpersonal relationships or psychological factors, rather than a SDOH.
Other key findings included that:
- People assigned Z codes often had chronic disease; for example, 72% had hypertension and 53% had depression.
- 25% of the Medicare beneficiaries assigned a Z code were dually eligible for Medicaid.
- 35% of the Medicare beneficiaries assigned a Z code were under age 65.
- For each of the commonly assigned Z codes, except homelessness, women were more commonly assigned a Z code than men.
- Minority groups — African Americans, Latino, Asian Americans, and American Indian/Alaska Natives — were proportionately assigned Z codes more than whites, but not universally so across codes.
The report has a number of very significant limitations, including:
- It is built on 2017 Medicare data, and Z code usage has likely changed since that time.
- It excludes Medicare Advantage plan members (more than a third of the Medicare beneficiaries), using only data from fee-for-service Medicare.
- Only 1.4% of Medicare fee-for-service beneficiaries had any Z code assigned.
This last limitation is probably the report’s most important finding — most health care providers are not assigning Z codes no matter how obvious the SDOH concern of their patient. This, in turn, spurs interesting questions about the appropriateness of medical professionals assessing non-medical concerns, the complex and curious titles of several codes, and the lack of incentives for health care providers to bother with any of this. Indeed, most medical students receive no formal preclinical training on assessing non-medical factors that contribute to patients’ health.
However, even with all of these limitations fully acknowledged, CMS has done the health care community a considerable service by releasing this report. While the report confirms earlier findings that Z code usage is very low, it also provides a touchstone for debating whether and how to insert the letter Z into SDOH.
[Note: The authors are actively working with health plans and actuarial partners to identify scenarios in which SDOH interventions are likely to be both clinically and economically effective.]