O2-S3.06 Area-based socioeconomic measures associated with female chlamydia and gonorrhoea, San Francisco, 2009

BackgroundSTD risk is influenced by individual-, network-, and community-level factors. Most research on STD risk has focused on individual-level factors, while limited data are available on community-level influences. In an ecologic analysis, we examined census tract-level factors associated with rates of female chlamydia (CT) and gonorrhoea (GC) in San Francisco in 2009.MethodsAll female CT and GC morbidity reported in 2009 to the San Francisco Department of Public Health was geocoded to census tract. Tract-specific disease rates per 100 000 females were calculated using 2009 population estimates. We used US Census data to examine area-based socioeconomic measures (ABSMs), calculated as the proportions of each census tract that were: lead by a female head of household, vacant homes, owner-occupied homes, living in the same residence for ≥5 years, persons living below poverty, households with ≥30% of income spent on rent, less than high school (HS) education, and receiving food stamps. We also examined the number of people per room in the residence and graffiti complaints per square mile. Poisson regression models were created to explore the relationship between ABSMs and tract-level CT and GC rates. ABSMs were categorised into quintiles, and RR were estimated that corresponded to the increase in CT or GC rate associated with a 1-quintile change in the ABSM.ResultsA total of 3267 CT and 1466 GC cases were geocoded to one of the 176 census tracts in San Francisco. The proportions of tracts with vacant units (RR=1.11, p<0.0001), with less than HS education (RR=0.91, p=0.0017), with female head of household (RR=1.40, p<0.0001), living below poverty (RR=1.14, p<0.0001), living in the same residence for ≥5 years (RR=0.92, p=0.0017), and receiving food stamps (RR=1.25, p<0.0001) were independently associated with female CT rates. In the analysis of female GC rates, the proportions of tracts with female head of household (RR=1.25, p=0.0026), living below poverty (RR=1.36, p-0.002), people residing per room (RR=0.84, p=0.023), and receiving food stamps (RR=1.77, p<0.0001) were associated with disease rates see Abstract O2-S3.06 table 1.Abstract O2-S3.06 Table 1Area-based socioeconomic measures and poisson regression modelsChlamydia rates, femalesGonorrhoea rates, femalesRR (95% CI)RR (95% CI)Economic deprivation % Below poverty line1.14 (1.08 to 1.21)1.36 (1.16 to 1.60) % Households with >=30% of income spent on rent––Concentrated disadvantage % Population 25+ with <12 years of education0.91 (0.85 to 0.96)– % Homes with >1 person per room–0.84 (0.72 to 0.98) % Receiving any food stamp benefits1.25 (1.16 to 1.35)1.77 (1.41 to 2.22)Neighbourhood stability and social cohesion % Homes occupied by owner–– % Vacant homes1.11 (1.06 to 1.16)– % Female-headed households1.40 (1.31 to 1.49)1.25 (1.08 to 1.45) % Same residence for >=5 years0.92 (0.87 to 0.97)-Removed from model (p>0.05).Note: Each Poisson regression model had the following dependent variable: 1) Chlamydia rates in females, 2) Gonorrhoea rates in females. Rate ratios (RR) represent the change in the STI rate associated with a quintile change in the predictor variable.ConclusionsSocioeconomic factors were associated with female CT and GC rates at the census tract level in San Francisco. Further exploration as to the potential etiologic role of community-level factors, as well as innovative means to modify the environment to improve sexual health, are warranted.