Date
6-26-2025
Department
School of Health Sciences
Degree
Doctor of Philosophy in Health Sciences (PhD)
Chair
Keith Pelletier
Keywords
social determinants of health, ICD-10 Z-codes, emergency department utilization, Medicaid beneficiaries, healthcare disparities
Disciplines
Health Information Technology | Medicine and Health Sciences
Recommended Citation
Gomez, Sofia L., "Evaluating the Utilization of ICD-10 Z-Codes for Documenting Social Determinants of Health and the Association with Frequent Emergency Department Visits Among Medicaid Beneficiaries" (2025). Doctoral Dissertations and Projects. 7135.
https://digitalcommons.liberty.edu/doctoral/7135
Abstract
This study assesses the use of ICD-10 Z-codes for documenting social determinants of health (SDoH) and examines their relationship with frequent emergency department (ED) visits among Arizona Medicaid-eligible beneficiaries. This retrospective cohort study utilizes claims data from a major Arizona Medicaid payor to explore the frequency of ICD-10 Z-codes use for documenting SDoH, associations between variables, and factors most strongly linked to higher ED utilization. The study sample (N = 53,368) includes Medicaid beneficiaries with ED visits in 2023, including 47,153 individuals without documented SDoH and 6,215 individuals with documented SDoH. Descriptive and statistical analyses, such as binomial tests, chi-square tests of independence, and binary logistic regression, were conducted. Findings indicate that 11.65% of beneficiaries had documented SDoH ICD-10 Z-codes, significantly higher than the CMS-reported national average of 1.59% (p < .001). Chi-square analysis revealed a significant association between documented SDoH and frequent ED utilization (χ²(1) = 143.97, p < 0.001), with beneficiaries having documented SDoH more likely to experience frequent ED visits (≥ 4 visits). Logistic regression identified housing and economic problems (Z59; OR = 1.231, p = 0.022) and psychosocial circumstances (Z64; OR = 2.668, p < .001) as the strongest predictors of frequent ED visits. Despite statistical significance, the regression model’s predictive power was low (Nagelkerke R² = 0.006), indicating that additional factors contribute to frequent ED utilization. These findings emphasize the importance of addressing social risk factors and integrating SDoH documentation into care management strategies. Future studies should explore additional factors to better understand relationships influencing ED utilization.
