1425 - Impact of Social Determinants on Racial Differences in Carbapenem-Resistant Enterobacteriaceae Incidence, Atlanta, 2012-2018

Background: Public health prioritizes addressing social determinants of health to promote health equity.  We hypothesized that social determinants of health, including poverty, are associated with racial disparity in the incidence of carbapenem-resistant Enterobacteriaceae (CRE).

Methods: The Georgia Emerging Infections Program conducted CDC-funded, active population-based CRE surveillance in metropolitan Atlanta (2017 population: 3.9 million) from 2012-2018. CRE cases were defined as Atlanta residents with a urine or normally sterile specimen growing E. coli, Klebsiella spp., or Enterobacter spp. resistant to ≥ 1 carbapenems (excluding ertapenem) and all third generation cephalosporins tested. Poverty, education and insurance levels by census tract of residence were obtained from the US Census Bureau’s 2017 American Community Survey. Race and end-stage renal disease (ESRD) were determined from chart review, and primary care provider (PCP) shortage area was obtained from the Health Resources and Services Administration. Age-adjusted incidence rate ratios were individually calculated using direct age standardization. Covariates were considered for inclusion in a multivariable Poisson regression model for the expected rate of CRE.

Results: Adjusting for age, CRE incidence was three times higher in blacks than whites. Higher CRE incidence was also observed among cases assigned > 40% below poverty level, > 15% below high school education, > 10% uninsured, and in a PCP shortage area (Table 1). CRE incidence was 58 times higher among ESRD cases than non-ESRD cases. In the multivariable model (Figure 1) addition of education, poverty or ESRD (p< 0.001), but not PCP access (p = 0.61) and insurance status (p=0.19), significantly reduced the racial difference in CRE incidence compared to race and age alone.  Although controlling for age and either education, poverty level or ESRD reduced CRE among blacks, CRE incidence in blacks remained double that of whites.Figure 1. Comparison of race CRE incidence rate ratio adjusting for age alone (red line) to adjusting for age and individual social determinants (blue bars)"
Table 1. Social Determinants Distribution and Age-Adjusted CRE Incidence Rate Ratio (N = 378)


Conclusion: Poverty level, ESRD and education only partially account for the racial differences seen in CRE incidence. While ESRD suggests a possible biologic component, persistent racial differences indicate the need for targeted public health interventions to address social determinants of health.