Home Search Abstracts Browse Sessions Program Committee View Session E-mail Abstract Author

 

 




Session 42 Oral Abstracts
The Evolving HIV Epidemic: Risk Behavior, Incidence, and Prevalence
Friday, 4 - 6 pm
Presentation Time: 5:00 pm
Ballroom A


170
Spatial Monitoring of Incidence: Using HIV Biomarkers to Identify Clusters
Kristen Hampton*1, C Pilcher1, S Fiscus1, E Foust2, W Messer1, M Serre1, J McPherson2, D Williams2, R Ashby2, T Nguyen1, B Stalzer1, J Harris2, A Cachafeiro1, J Eron, Jr1, and P Leone1,2
1Univ of North Carolina at Chapel Hill, Ctr for AIDS Res, USA and 2North Carolina Dept of Hlth and Human Svcs, Raleigh, USA

Background:  Mapping of disease incidence has successfully been used to investigate case-clustering and assess transmission patterns of other sexually transmitted diseases but not for HIV. North Carolina’s Screening and Tracing Active Transmission (STAT) program uses added laboratory testing (NAAT for EIA/Western blot—specimens and a less-sensitive EIA for EIA+ specimens) to identify cases of incident HIV in the public HIV testing population and to estimate the time of transmission for each case. We examined and mapped VCT records to detect areas with significantly elevated HIV incidence.

Methods:  All records from publicly funded VCT in North Carolina from 1994 to 2004 were collected in a study database, de-linked from names or individual locator information but including demographic information as well as zip code of residence. Records with out-of-state, null or invalid zip codes were excluded. Approximate date of infection was calculated for each acute (transmission date < 16 days) or “recent” (td < 200 days) case of HIV infection detected in the 2002–2004 study period. Observed incidence was calculated for each ZIP code, per number of clients tested. Expected rates were calculated using pre-2002, 8-year historical prevalent HIV rates for each ZIP code, adjusted by the ratio of incident:antibody positive cases observed statewide in 2002 to 2004.  

Results:  Of VCT surveillance records meeting the study criteria, 1,076,143 of 1,107,847 (97%) total tests, 6563 of 6825 (96%) total HIV cases and 217 of 226 (96%) incident cases were able to be geocoded to a residential-delivery weighted zip code of residence. Observed incidence was greater than expected in 96 of 739 (13%) ZIP codes, with 32 (33%) of these ZIP codes having significantly greater than expected incidence (p < 0.05). Of the ZIP codes with significantly greater than expected rates, 22 (69%) defined or surrounded urban areas; 10 distinct geographical clusters (high transmission areas) were clearly identifiable from these data.  

Conclusions:  The STAT system for geomapping makes it possible for North Carolina to passively monitor HIV transmission events and detect localized high transmission areas based on significantly raised incidence. Geomapping is a powerful tool for better understanding transmission dynamics and clustering of HIV infections, and can be used to guide resource allocation at the state level.

Keywords: Surveillance; Clustering; Mapping