University of California, San Francisco Logo

University of California, San Francisco | About UCSF | Search UCSF | UCSF Medical Center

HIV testing rates and outcomes in a South African community, 2001-2006: implications for expanded screening policies
Global Health Sciences Literature Digest
Published January 4, 2010
Journal Article

April MD, Walensky RP, Chang Y, et al. HIV testing rates and outcomes in a South African community, 2001-2006: implications for expanded screening policies. J Acquir Immune Def Syndr 2009;51:310-6.

In Context

In 2007, the World Health Organization (WHO) issued recommendations for routine, opt-out HIV testing for adults in settings where HIV prevalence rates are estimated to be >1%.(1) Both voluntary counseling and testing (VCT) and opt-out, routine, provider-based testing at clinics and hospitals have been available to adults in a peri-urban area near Cape Town, South Africa since 2001. Routine, opt-out testing, without the requirement of signed consent, has been conducted among tuberculosis (TB) patients since 2001 and pregnant women since 2002. Expansion of testing to reach a greater proportion of the population is essential given the increased availability of antiretroviral therapy (ART) in South Africa. Information on testing rates over time in South Africa, a country with HIV prevalence estimated at 16% among adults, has been reported at the individual level, from persons accessing care, but not on a population level.(2,3,4) Examination of trends through 2006 will provide a baseline from which to determine the impact of the 2007 WHO HIV testing recommendations.


To measure trends in HIV testing, the WHO clinical stage at diagnosis, and rates of ART initiation.


A peri-urban African community near Cape Town, South Africa.

Study design

Serial cross-sectional analysis of multiple data sources.


Participants were: a) Persons residing in the region listed in census data in 2002, 2004, and 2006; b) patients tested at VCT, TB, and antenatal clinics (ANCs) from 2001 through 2006; and c) patients who received HIV-related medical care during the study period.


The rates, trends, and characteristics of persons who were tested and clinical indicators of HIV-positive patients.


Data were obtained from three sources. The community census was used to develop estimates of the number of adults aged >14 years each year from 2001 through 2006 by age and sex. The HIV testing registers at the local clinics and hospitals provided the number of testing encounters, the client identification number, name, date of birth, residence, medical record number, test acceptance, test result, and test site (clinic or hospital), and testing service (VCT, TB, ANC). Repeat test encounters by the same person were linked. Clinical information was obtained from patient medical records.

Observations of persons who previously tested HIV-positive that were missing age or residence were excluded. HIV testing rates for the entire adult population were measured by dividing the number of eligible adults tested each year by the annual population. These measurements were conducted separately for men, non-pregnant women, and pregnant women. Missing census data were random and therefore imputed based upon the distribution of available data. Adjusting for testing eligibility was done by subtracting the numbers of community residents receiving HIV-positive diagnoses each year from the population counts used in subsequent years. The frequencies of re-testing among men and non-pregnant women were compared using data among persons with previous HIV-negative tests. The linear trends in the probability of testing positive were done by restricting the analysis to the first HIV test received by each person from 2001 through 2006. The median CD4 count; WHO clinical stage at diagnosis; and rates of posttest counseling, referral, and treatment were determined using the data obtained from the medical records.

Tests for trends were done using the chi square test, analysis of variance, and linear regression. Survival analysis techniques were used to estimate the time to retesting and time to positive retest. The log-rank test was used to compare the time to retesting between men and women.


From 2001 through 2006, a total of 5006 adult community residents not previously diagnosed with HIV infection had 6322 test encounters, 1890 (30%) of which were provider initiated.

Acceptance rates averaged 94% and were similarly high across all sexes, age groups, test services, test venues, and years. Among first-time testers, 3125 (67%) initiated VCT encounters, 162 (3%) were TB patients, and 1385 (30%) were ANC users. The proportion of first-time tests administered to ANC users rose from 7% in 2001 to 40% in 2002 when testing of pregnant females became provider initiated. After 2002, this proportion declined, whereas that of first-time tests initiated by clients through VCT services rose from 55% in 2002 to 74% by 2006.

The annual testing rates increased from 4% in 2001 to 20% in 2006. Rates among pregnant females were substantially higher than were those for non-pregnant females and males from 2002 onward. The rise in rates among males and non-pregnant females increased after 2003, the year ART first became available. Males consistently tested at lower rates than non-pregnant females (P<0.001 for all years). Testing rates did not differ substantially by age.

There were 1171 retests (20% of total tests), and the proportion increased significantly from 2003 through 2006. Of 2214 adults who received negative results after first time testing-excluding females tested through ANCs-16% re-tested within one year, 30% within two years, and 53% within five years. The incidence of seroconversion among re-testers was 1.9% within one year of initial testing, 3.6% within two years, and 12.6% within five years.

Data from medical records of testers were obtained from 84% of patients. The median CD4 cell counts remained constant over time, with 36% of initial counts being >200 cells/mm3. Sixty-nine percent of patients met the WHO clinical criteria of stages 1-2 at diagnosis. Post-test counseling was received by more than 96% of persons testing each year. The proportion referred for care increased from 29% in 2006 to 67% in 2006 (P<0.001). Rates of ART initiation increased from zero in 2001 to 68% of treatment-eligible patients (P<0.001).


Uptake of HIV testing and retesting increased from 2001 through 2006 with the highest rates occurring among pregnant women. Provider-initiated, opt-out testing along with the availability of ART appears to have led to the substantial increase in testing. Despite high testing rates, a considerable proportion of persons are diagnosed late in the course of their disease.

Quality Rating

April and colleagues have conducted and reported on a high quality study. The data sources were appropriate and the data were of reasonably high quality. The study site and data sources led provided a representative sample.

Programmatic Implications

The choice of how to provide HIV testing has been of debate for many years. Over time, particularly after ART became widely available, the use of provider-based, opt-out testing has become increasingly acceptable. At the time of HIV diagnosis many persons already have evidence of substantial immunocompromise, however, suggesting delays in HIV testing. This delay may reflect acquisition of HIV years before the availability of ART and provider-initiated testing. The higher rates of testing from provided-initiated testing will likely lead to narrowing the time between infection and diagnosis. Efforts to provide testing and care to persons who do not routinely obtain medical care should be expanded.


  1. WHO/UNAIDS. Guidance on provider-initiated HIV testing and counselling in health facilities. Geneva, Switzerland: World Health Organization, Joint United Nations Programme on HIV/AIDS; 2007. Accessed September 10, 2008.
  2. Homsy J, Kalamya JN, Obonyo J, et al. Routine intrapartum HIV counseling and testing for prevention of mother-to-child transmission of HIV in a rural Ugandan hospital. J Acquir Immune Defic Syndr 2006;42:149-54.
  3. Zachariah R, Spielmann MP, Harries AD, et al. Voluntary counselling, HIV testing and sexual behaviour among patients with tuberculosis in a rural district of Malawi. Int J Tuberc Lung Dis 2003;7:65-71.
  4. Bassett IV, Giddy J, Nkera J, et al. Routine voluntary HIV testing in Durban, South Africa: the experience from an outpatient department. J Acquir Immune Defic Syndr 2007;46:181-6.