Utilization of Social Determinants Z Codes and Health System Characteristics, 2015-2024

ABSTRACT

Objectives: Social determinants of health (SDOH) Z codes help standardize the capture of social needs, such as adverse housing, economic, or social circumstances. Facility characteristics may lead to underuse of SDOH Z codes among more clinically complex individuals in resource-constrained safety-net organizations. In this study, we aim to identify health system characteristics associated with SDOH documentation.

Study Design: Repeated cross-sectional study using Epic Cosmos data from October 2015 to December 2024.

Methods: We investigated encounter-level use of Z codes across inpatient, outpatient, and emergency department encounters. Hospitals were categorized into quartiles by the proportion of encounters accounted for by Medicaid or uninsured patients, as a measure of social complexity. We also categorized hospitals into quartiles by patient volume.

Results: Z-code use remained low overall; less than 1% of all hospital encounters used SDOH Z codes across all hospital volumes. Hospitals in the highest social complexity quartile accounted for 33.4% of all encounters with Z codes (9,217,016), compared with 18.4% (5,069,460) for hospitals in the low social complexity quartile. Our data do not show a linear relationship between hospitals treating the most socially complex patients and Z-code utilization; facility-level variation in use is attributable to systems-level factors in how Z codes are used.

Conclusions: Our findings suggest a connection between facility characteristics and Z-code use, corroborating prior work indicating that certain types of hospitals are more likely to use Z codes. Therefore, targeted policies are needed to adequately measure SDOH and, subsequently, ensure that Z codes are entered in the electronic health record.

Am J Manag Care. 2026;32(10):540-542

doi:10.37765/ajmc.2026.90024

Clinical documentation of social determinants of health (SDOH) has been historically limited by a lack of standardized coding rubrics. In 2015, a set of codes (Z codes) was created to address this gap through uniform documentation of SDOH in the electronic health record (EHR).^1,2 Overall Z code adoption has increased over time, though concerns exist about uneven use across patient groups—in particular, underuse among more clinically complex individuals and those with less historical health care utilization.^3,4 Facility characteristics may drive analogous dynamics, with factors such as limited infrastructure, staffing, or technology systems prompting uneven code use. Such factors may be most salient to resource-constrained safety-net organizations and those serving more socially complex patient populations. However, how Z code use varies among hospitals remains unknown.

METHODS

This repeated cross-sectional study used Epic Cosmos data from October 2015 to December 2024. Epic Cosmos is a large nationwide, integrated EHR database representing 1884 hospitals, 42,400 clinics, and more than 300 million patients as of February 2026.⁵ The community represents patients from all 50 states and Washington, DC. Our primary outcome was encounter-level use of Z codes (Z55-Z65) across inpatient, outpatient, and emergency department encounters within hospitals. Hospitals were categorized into quartiles based on the proportion of encounters accounted for by Medicaid or uninsured patients, as a simplified measure of social complexity: low, medium, high, and highest. We also categorized hospitals into quartiles by patient volume: first (lowest volume), second, third, and fourth (highest volume). The study was determined to be non–human participant research due to the use of deidentified data from the Epic Cosmos database.

RESULTS

From 2015 to 2024, hospitals in the highest social complexity quartile accounted for 33.4% of all encounters with Z codes (9,217,016), compared with 18.4% (5,069,460) for hospitals in the low social complexity quartile (Figure 1). Out of all encounters in the highest social complexity quartile, only 0.33% had Z codes, followed by 0.16% and 0.20% in the high and medium social complexity quartiles. The percentage of encounters with Z codes in the low social complexity quartile was 0.16%.

During the same period, less than 1% of all hospital encounters used SDOH Z codes across all hospital volumes (Figure 2). Throughout the study period, first-quartile (lowest-volume) hospitals showed fluctuating Z-code use across quartiles of social complexity. Second-quartile hospitals showed that the highest quartile of socially complex patients consistently had more Z codes than low socially complex patients, with Z code use within social complexity quartiles remaining similar from 2015 to 2024. Third- and fourth-quartile hospitals showed an increase in Z code utilization over time, with the highest socially complex patients having more encounters with Z codes.

DISCUSSION

Z code use remained low overall but varied by hospital volume and proportion of socially complex patients. In the absence of barriers to documentation, we would expect an exact linear relationship between hospitals that treat the most socially complex patients, based on Medicaid and uninsured status, and those with the highest utilization of Z codes. However, we find that this relationship is more consistent for higher-volume hospitals, with greater variability in lower-volume hospitals (first- and second-quartile hospitals), thus highlighting Z code utilization that varies at the facility level. Overall, these findings suggest a connection between facility characteristics and Z code use, corroborating prior work indicating that certain types of hospitals are more likely to use Z codes.^3,6 Potential barriers to broader Z-code utilization, particularly among smaller hospitals,may include limited time and competing interests of providers, discomfort discussing social needs with patients, staffing, EHR integration, and lack of financial incentives.⁷

Limitations of this study include the repeated cross-sectional design, potential coding bias, a lack of hospital-level data, and very low uptake of SDOH Z codes. Epic Cosmos hospitals may overrepresent hospitals that are part of larger systems, limiting generalizability. Additionally, these data do not have outcomes related to Z code utilization. Although uptake was low, Epic Cosmos, as a large national database, offered the most comprehensive insight into SDOH Z code utilization. Our findings highlight the need for uniform screening and documentation standards for all health systems to improve adequate and accurate SDOH measurement nationwide.

Author Affiliations: Department of Plastic Surgery, UT Southwestern Medical Center (JIB), Dallas, TX; Program on Policy Evaluation and Learning (JIB, JHJ, MDD, JRC, JPS, JML), Dallas, TX; Department of Internal Medicine, UT Southwestern Medical Center (JIB, JHJ, MDD, JRC, JPS, JML), Dallas, TX; Department of Health Economics, Systems, and Policy, UT Southwestern Medical Center (JHJ, JPS, JML), Dallas, TX

Source of Funding: None

Author Disclosures: Dr Joo reports receiving royalties from Wolters Kluwer. Dr Liao reports service on the Medicare Payment Advisory Commission and receiving grants from the Patrick and Catherine Weldon Donaghue Medical Research Foundation. The remaining authors report no relationships or financial interests with any entity that would pose a conflict of interest regarding the subject matter of this article.

Authorship Information: Concept and design (JIB, JPS, JML); acquisition of data (MDD, JML); analysis and interpretation of data (JIB, JHJ, MDD, JRC, JPS, JML); drafting of the manuscript (JIB, JHJ, MDD, JRC); critical revision of the manuscript for important intellectual content (JIB, JHJ, JRC, JPS, JML); statistical analysis (MDD); and supervision (JIB, JML).

Address Correspondence to: Jessica I. Billig, MD, MSc, UT Southwestern Medical Center, 1801 Inwood Rd, Dallas, TX 75390. Email: Jessica.Billig@UTSouthwestern.edu.

REFERENCES

1. Centers for Medicare and Medicaid Services. ICD-10-CM official guidelines for coding and reporting: FY 2015. September 26, 2014. Accessed September 3, 2026. https://stacks.cdc.gov/view/cdc/35704

2. Improving the collection of social determinants of health (SDOH) data with ICD-10-CM Z codes. CMS. 2023. Accessed December 3, 2025. https://www.cms.gov/files/document/cms-2023-omh-Z-code-resource.pdf

3. Truong HP, Luke AA, Hammond G, Wadhera RK, Reidhead M, Joynt Maddox KE. Utilization of social determinants of health ICD-10 Z-codes among hospitalized patients in the United States, 2016-2017. Med Care. 2020;58(12):1037-1043. doi:10.1097/MLR.0000000000001418

4. Chatterjee P, Macneal E, Roberts ET. Measurement bias in documentation of social risk among Medicare beneficiaries. JAMA Health Forum. 2025;6(7):e251923. doi:10.1001/jamahealthforum.2025.1923

5. Cosmos Community. Epic Systems Corporation. Accessed September 3, 2026. https://www.epic.com/cosmos/participants/

6. Ryus CR, Janke AT, Granovsky RL, Granovsky MA. A national snapshot of social determinants of health documentation in emergency departments. West J Emerg Med. 2023;24(4):680-684. doi:10.5811/westjem.58149

7. Kepper MM, Walsh-Bailey C, Prusaczyk B, Zhao M, Herrick C, Foraker R. The adoption of social determinants of health documentation in clinical settings. Health Serv Res. 2023;58(1):67-77. doi:10.1111/1475-6773.14039