The 2007 National Antenatal Human Immunodeficiency Virus survey in South Africa reported an antenatal HIV prevalence of 28% . Voluntary counselling and testing (VCT) for HIV in South African primary healthcare clinics is offered to pregnant women at their first antenatal visit and despite national recommendations , repeat HIV testing of seronegative women at subsequent antenatal visits is rarely implemented and seldom reported. Undetected infections with low HIV antibody levels and newly acquired infections subsequent to the first HIV test and during a pregnancy may largely be missed if VCT is only offered to women at their first or at one antenatal visit. Women who are HIV seronegative in the early stages of pregnancy are at continued risk of seroconverting during pregnancy and breastfeeding .
Therefore, this study was aimed at using the repeat HIV testing strategy in pregnancy as an opportunity to determine the incidence of HIV during pregnancy in a country with a high antenatal HIV prevalence and describe the potential risk factors for seroconversion in pregnancy.
This study was nested in a prevention of mother-to-child transmission (PMTCT) program at one rural and two urban health facilities in each of three provinces in South Africa, Mpumalanga, Eastern Cape and Free State between July 2006 and April 2007. The PMTCT program included pretest counselling, HIV testing by rapid tests and posttest counselling for pregnant women. If women tested HIV positive, they were provided the single-dose nevirapine regimen and infant feeding counselling to prevent mother-to-child transmission (MTCT) of HIV . If women tested HIV negative at their first antenatal visit, posttest counselling included implications of a negative result, an explanation of the ‘window period’ and women were advised on the need for a repeat HIV test at least 12 weeks later. Women's HIV status was documented in their antenatal carrier cards in the form of a code to maintain confidentiality.
For the study, pregnant women were screened at a subsequent antenatal visit between 36 and 40 weeks' gestation, and women with a previously documented HIV-negative status were enrolled for HIV retesting. To ascertain whether the potentially eligible women had tested HIV negative more than 12 weeks ago, the date of initial testing was obtained from the antenatal record. A written informed consent was obtained at the subsequent antenatal visit for participation in the retesting study. Following pretest counselling at the study visit, HIV testing continued using two rapid tests (Determine HIV-1/2, Abbott Laboratories, Abbott Park, Illinois, USA; and First Response HIV-1/HIV-2 WB, PMC Medical Pty Ltd., Daman, India) according to a serial algorithm. A dried blood spot (DBS) was collected and tested for enzyme-linked immunosorbent assay (ELISA) as confirmation of HIV status.
Data obtained from antenatal records included gestational age at the first antenatal visit, date of initial HIV testing and the number of pregnancies (including index pregnancy). Study participants self-administered a short structured standard questionnaire that required their educational level and marital status, as well as frequency of sexual activity, number of sexual partners and condom use prior to and during the index pregnancy.
Stata version 10 (StataCorp, College Station, Texas, USA) was used to analyse the data. Categorical risk factors for seroconversion were assessed using Pearson's χ2 (degrees of freedom = 2) test in the total sample and stratified by urban or rural grouping. A P value of less than 0.05 was considered statistically significant. Binary logistic regression models were used to assess the independent effects of risk factors for seroconversion.
Regulatory approvals for the study were obtained from the Biomedical Research Ethics Committee, University of KwaZulu-Natal and provincial health authorities.
Between June 2006 and June 2007, 5233 pregnant women registered for antenatal care and 4715 (90.1%) accessed HIV VCT services offered as part of the PMTCT program at the selected primary health facilities in the three provinces. Among the 4715 women who received pretest counselling, 3742 (79.4%) accepted HIV testing. One thousand and fifty-seven women (28.2%) tested HIV positive at their first visit.
Among the 2685 women who tested HIV negative at their first visit, 2377 (88.5%) were referred for participation in the repeat testing study, 1099 (46.2%) and 1278 (53.4%) were retested during pregnancy at urban and rural health facilities, respectively. The median age was 24 years (range 16–47 years), the majority (61.2%) were single, among these 72% were not living with their sexual partner and 75% attained secondary level of education (grades 8–12).
The median gestational age at the first visit was 22 weeks (range 6–41), and the majority (98%, n = 2329) first attended a clinic at 28 weeks of gestational age or less. HIV-negative women were retested at a median gestational age of 36 weeks, 61% (1450) at an age of 30–36 weeks and 39% (927) at an age of more than 36 weeks.
Seventy-two (3%) of the 2377 pregnant women were HIV positive at the repeat test at an average of 24 weeks (range 12–25 weeks) after testing negative at their first visit (679 woman years of exposure) yielding an incidence rate of HIV of 10.7/100 pregnant-woman-years (PWY) [95% confidence interval (CI) 8.2–13.1]. HIV incidence in pregnancy was higher but not statistically significant at the urban (39 HIV infections) facilities as compared with the rural (33 HIV infections) facilities (12.4/100 PWY; 95% CI 8.5–16.3 versus 9.1/100 PWY; 95% CI 5.9–12.1). The incidence risk ratio of urban vs. rural infections was 1.37 (95% CI 0.86–2.17), P = 0.182.
Demographic risk factors
None of the demographic factors was significantly associated with HIV incidence, but several trends were noted. The HIV incidence was at least two-fold higher among the 25–29 and 30–34-year-old groups (3.8 and 4.5%, respectively) as compared with the less than 20-year-old group (1.9%). HIV incidence was lower among women with tertiary education (0.5%) than those with lesser education (3%); single (3.4%) or divorced women were more likely to be at risk for HIV when compared with married women (1.5%, P = 0.017). The HIV incidence was higher among older women in rural areas and younger women in urban areas (Table 1).
Behavioral risk factors
Almost 50% of women had never used a condom prior to this pregnancy, and fewer women (40%) used a condom during the pregnancy. Condom usage did not differ between women who seroconverted and those who remained HIV negative (Table 2). At least 20% of women in both groups reportedly had at least two serial sexual partners prior to this pregnancy. Eighty-five percent of the women remained sexually active during pregnancy, though frequency was reduced among all pregnant women (n = 2377) independent of HIV status.
Three percent of pregnant women were HIV positive at the repeat test at an average of 24 weeks following the first test. HIV incidence in pregnancy was higher but not statistically significant in the urban as compared with rural populations. This study also revealed that the HIV incidence was at least more than two-fold higher among the 25–29 and 30–34-year-old women compared with the teenagers. A significantly higher proportion of women who remained HIV-negative had tertiary education or were married. This study revealed that the majority (98%) of South African women first attended a clinic at at least 28 weeks of gestation therefore offering, at a minimum, two additional opportunities for HIV retesting and PMTCT prophylaxis if required.
HIV incidence during pregnancy has been reported in a limited number of Sub-Saharan studies and range between 2.3/100 person years in Uganda and 4.8/100 person years in Zimbabwe [3,4]. The HIV incidence rates in Rakai, Uganda was 2.3/100 person years during pregnancy and 1.3/100 person years during breastfeeding . Using an analysis model in Zimbabwe, the HIV incidence ranged from 1.8 to 3%, and seroconversion is assumed to occur for 4 months during pregnancy and 6 months during breastfeeding .
Data from Tanzania and South Africa suggest that approximately 3 and 5.2% of pregnant women who are initially HIV seronegative, seroconvert prior to delivery, respectively [5,6]. Other studies in Brazil and the United States reported HIV incidences of 0.2/100 pregnant women per annum and 5.3/1000 person years, respectively [7,8]. In the MIRIAD study conducted in six United States cities from 2001 to 2005, 11% of the HIV-infected women were shown to have primary infection based on a two-test approach at subsequent intervals, with a minimum of 12 weeks between .
The HIV incidence among the South Africa pregnant women in our study is higher than that of the general population reported in a recent study by Rehle et al. . The HIV incidence in our pregnant population could be higher if women seroconverted prior to their first antenatal visit and hence were not included in this study. Rehle et al. also presented a higher HIV incidence in the pregnant population as compared with their nonpregnant counterparts. Similarly, Gray et al.  reported a higher HIV incidence among Ugandan pregnant women (2.1%) as compared with nonpregnant lactating women (1.2%). Gray et al. speculate that the higher HIV incidence rate in pregnant women as compared with their nonpregnant counterparts (1.1/100 person years) is independent of behavioral risk and is due to hormonal changes affecting genital tract mucosa or immune response during pregnancy . Our study findings support the lack of behavioral association with seroconversion in pregnancy. Condom use, multiple sex partners and frequency of sexual activity were not significantly different between the HIV-negative and seroconverting pregnant women. Our study, however, had its limitations in further elucidating the association of biological risk factors to include the infectiousness of the sexual partners of pregnant women and vulnerability of women themselves.
Although sexual behavior could be influenced by demographic factors such as age and marital status, both characteristics were strong indicators for HIV susceptibility independent of sexual behavior in our pregnant population. Similar trends of decrease in HIV prevalence in the younger age groups were observed in the South African antenatal survey of 2007 and Malawi between 1999 and 2004 [1,10]. The authors also noted that the trend in HIV prevalence showed a decline in specific age groups of 15–19 and 20–24 years, but a significant incline in the older age groups (30–34 and 35–39 years). A decrease in the incidence of HIV among young primigravid women in India from 2.2/100 person years (95% CI 1.6–3.0) in 2002 to 0.73/100 person years (95% CI 0.5–1.0) in 2006 was attributed to a proportional decrease in high-risk behavior among young married men thereby reducing risk of HIV transmission to their families . Shisana et al.  also demonstrated that the risk of HIV infection was significantly higher among unmarried compared with married people when sexual behavior factors were controlled for (odds ratio, 0.55; 95% CI 0.47–0.66).
The HIV incidence in pregnant women in this setting of high HIV prevalence is about four times higher than in women in the general population. Pregnant women continue to represent a vulnerable population, and HIV retesting late in pregnancy or labor offers an additional opportunity to prevent MTCT, prevent further horizontal transmission in the community as well as ensure continued care for women who seroconvert during pregnancy. In light of the above findings, public health programs need revisiting to ensure that HIV retesting in pregnancy is implemented, reinforce the need for continued education on prevention during pregnancy and extend the use of female-controlled prevention methods such as microbicides when proven well tolerated and efficacious during pregnancy.
The authors wish to thank the Centre for Disease Control, South Africa represented by Latasha Treger for the financial support of this study.
We acknowledge the support from the Provincial Head of Department, Management of the Maternal and Women's Health Departments, District Managers, and HAST Coordinators in Eastern Cape, Mpumalanga and Free State provinces. We greatly appreciate assistance from management and staff of the antenatal clinics, Global Laboratories, Durban for the laboratory support, Mr Samkelo Booi for assistance with the data management, Research Teams at each facility and Prof. S.S.S. Karim (Caprisa-UKZN) for editorial suggestions.
1. Department of Health, 2008. National HIV and Syphilis Prevalence Survey in South Africa
2. Department of Health, 2008. National, PMTCT. Guidelines, South Africa
3. Gray R, Li X, Kigozi G, Serwadda D, Brahmbhatt H, Wabwire-Mangen F, et al
. Increased risk of incident HIV during pregnancy in Rakai, Uganda: a prospective study. Lancet 2005; 366:1182–1188.
4. Dube S, Boily M, Mugurungi O, Mahomva A, Chikhata F, Gregson S. Estimating vertically acquired HIV infections and the impact of the prevention of mother-to-child transmission program in Zimbabwe: insights from decision analysis models. J AIDS 2008; 48:72–81.
5. Urassa W, Kaaya S, Mwakagile D, O'Brien M, Antelman G, Hunter D, et al
. Evidence of a substantial decline in prevalence of HIV-1 infection among pregnant women: data from 1995 to 2003 in Dar es Salaam,Tanzania. Scand J Publ Health 2006; 34:272–278.
6. Rehle T, Shisana O, Pillay V, Zuma K, Puren A, Parker W. National HIV incidence measures: new insights into the South African epidemic. S Afr Med J 2007; 97:194–199.
7. Nielsen-Saines K, Melo M, Varella I, Fonseca R, Lira R, Turella ML, et al. Primary HIV-1 infection during pregnancy: high rate of HIV-1 MTCT in a cohort of patients in southern Brazil. Retrovirology
2008, 5(Suppl 1)
8. Birkhead E, Kowalski S, Hackel W, Pulver R, Glaros B. The contribution of infection acquired during pregnancy to residual mother-to-child transmission rates in New York State
[abstract no. WePE0271]. International Conference of AIDS
; 13–18 Aug 2006; p. 16.
9. Nesheim S, Jamieson DJ, Danner SP, Danner SP, Maupin R, O'Sullivan MJ, et al
. Primary human immunodeficiency virus infection during pregnancy detected by repeat testing. Am J Obstet Gynecol 2007; 197:149e1–149e5.
10. Crampin AC, Jahn A, Kondowe M, Ngwira BM, Hemmings J, Glynn JR, et al
. Use of antenatal clinic surveillance to assess the effect of sexual behavior on HIV prevalence in young women in Karonga District, Malawi. J Acquir Immune Defic Syndr 2008; 48:196–202.
11. Gupte N, Sastry J, Brookmeyer R, Phadke MA, Bhosale RA, Bollinger RC. Declining HIV infection rates among recently married primigravid women in Pune, India. J Acquir Immune Defic Syndr 2007; 45:570–573.
12. Shisana O, Zungu-Dirwayi N, Toefy Y, Simbayi LC, Zuma K, Malik S. Marital status and risk of HIV infection in South Africa. S Afr Med J 2004; 94:537–543.