Article Text


O12.1 Using Social Determinants to Predict Neisseria Gonorrhoeae Infection Risk at the Census Tract-Level: Findings from the STD Surveillance Network (SSuN), United States
  1. M Stenger1,
  2. M Samuel2,
  3. H Weinstock1 SSuN Working Group
  1. 11US Centers for Disease Control & Prevention, Atlanta, GA, United States
  2. 2STD Control Branch, DCDC, CID, CA Dept. of Public Health, Richmond, CA, United States


Background Associations between social factors and health outcomes have been established for many diseases, including gonorrhoea. Inequalities in gonorrhoea are particularly pronounced; the incidence for non-Hispanic Blacks is 17 times the incidence for non-Hispanic Whites. A gap exists in the literature on how social determinants for this population can be used to target STD interventions.

Methods Health departments from 10 diverse states geocoded female gonorrhoea case reports from 2009–2011 to the census tract. We calculated three-year gonorrhoea incidence rates among females for 11,832 tracts in the study area. Tract-level social determinants were obtained from the American Community Survey. Female rates were used to preclude confounding by the unknown proportion of men-who-have-sex-with-men among male cases. Tracts were stratified as above or below the county-level mean proportion of non-Hispanic Black residents to minimise interactions between race and other factors, and subsequent analyses were limited to the stratum with a higher proportion of blacks (N = 4,345). Using hierarchical models and logistic regression, associations between female gonorrhoea incidence, household income, education and housing characteristics were examined. Factors found to be associated with gonorrhoea incidence were used to construct tract-level gonorrhoea risk indices.

Results The mean tract-level female incidence rate was 194/100,000 (range, 0–2,836) in this stratum. Gonorrhea incidence was correlated with: proportion of households with annual income below $20,000 (β = 0.7089, P < 0.0001), housing vacancy rate (β = 0.5519, p < 0.0001) and proportion of population with less than high school education (β = 0.3206, p < 0.0001). Risk indices based on these factors correctly predicted 77.3% of tracts with incidence in the highest quartile in this stratum while excluding 58.9% of tracts with lowest incidence.

Conclusions Area-based social factors provide more specific criteria than race alone and may be useful for better targeting disease prevention activities to areas at highest risk for gonorrhoea incidence, especially where street-level address information on reported cases is unavailable.

  • gonorrhea
  • social determinants

Statistics from

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.