Background: Sexually transmitted diseases (STDs), including AIDS, are transmitted by behaviors that are affected by social contexts. To advance our understanding of these factors beyond county-level studies of census data we need data on social phenomena that are not available in the Census and analysis at levels other than the county. Such data are available in the Project on Human Development in Chicago Neighborhoods (PHDCN). Data on reported STDs geocoded to census tracts are also available from the Chicago City Department of Public Health.
Study Aims :
The aim of this study is to determine the effect of neighborhood characteristics, as measured in the PHDCN study, on rates of STDs and AIDS in Chicago. There are several interrelated questions to be addressed when achieving this aim: (1) What are the most appropriate geographical units for community level studies of STDs? (2) What is the appropriate timing lag between neighborhood characteristics and STD outcomes? And (3) What is the relation between neighborhood characteristics, as measured in the PHDCN study, and census variables found to be associated with STD rates at the county level in previous studies? Population: All residents of the city of Chicago in the years 1995-1998. Methods: Data sources include: publicly available PHDCN data (collected in 1994-95); reported cases of gonorrhea, chlamydia, syphilis, and AIDS aggregated to census tracts for the years 1995-1998; Census data for 1990 and available interpolated data for 1995-98; empirically collected data on public STD treatment facilities in Chicago. Data will be merged and analyzed at the level of census tract (825 census tracts) and neighborhood (343 clusters of related census tracts). The principle analysis will be multivariable linear regression with three types of STD outcomes: absolute rates; rate changes from 1995-98; and racial differences in rates. Independent variables will include those shown to be associated with STD rates in previous studies, and variables from the PHDCN data such as social cohesion, collective efficacy, social disorganization, and physical structure of neighborhoods. The models will be informed by bivariate analyses of: relations between the independent variables; time lags between the neighborhood characteristics and STDs; and assessment of the strength of correlations at the two geographical units of analysis. The final model will include corrections for autocorrelation between the geographical units. ? ?
| Thomas, James C; Torrone, Elizabeth A; Browning, Christopher R (2010) Neighborhood factors affecting rates of sexually transmitted diseases in Chicago. J Urban Health 87:102-12 |
