HIV Prevalence in Voluntary Counseling and Testing Centers Compared With National HIV Serosurvey Data in Uganda

Abstract

To compare HIV prevalence from routine voluntary counseling and testing (VCT) data with a population-based serosurvey in Uganda and to assess the utility of VCT data as a supplemental data source for HIV surveillance. We analyzed HIV testing data from 75,640 unique VCT clients aged 15-59 years collected from August 2004 to January 2005 at 160 VCT sites. We excluded clients who reported illness as the reason for testing. During the same time period, 18,525 adults aged 15-59 years were tested for HIV in the Uganda HIV/AIDS Sero-Behavioral Survey (UHSBS). We compared UHSBS HIV prevalence with age-standardized VCT prevalence, overall and among stand-alone and facility-based VCT sites. HIV prevalence in urban areas was similar overall [UHSBS: 9.7%, 95% confidence interval (CI) 8.6 to 10.7; VCT: 10.1%, CI 9.8 to 10.5] and for both men (UHSBS: 6.3%, CI 4.9 to 7.6; VCT: 7.1, CI 6.6 to 7.5) and women (UHSBS: 12.2%, CI 10.6 to 13.7; VCT: 12.9%, CI 12.3 to 13.4). Urban prevalence from UHSBS (9.7%, CI 8.6 to 10.7), VCT stand-alone sites (10.3% CI 9.8 to 10.8), and VCT sites in health facility settings (10.0%, CI 9.5 to 10.4) was similar. However, in rural areas where VCT coverage is much lower than in urban areas (10% versus 31%), HIV prevalence was much higher among rural VCT clients (8.2%, CI 7.9% to 8.4%) than among rural UHSBS participants (5.2%, CI 4.8% to 5.5%). This resulted in overall higher HIV prevalence among all VCT clients (8.8%, CI 8.7 to 9.1) compared with all survey participants (5.9%, CI 5.6 to 6.2). After excluding clients who give illness as a reason for testing, VCT data may be used without further adjustment to monitor the HIV epidemic among urban Ugandans using either VCT data from stand-alone or health facility-based sites. However, monitoring rural and overall HIV prevalence using VCT data may not be appropriate until the uptake of VCT in rural areas is significantly improved or an adjustment factor is applied.