University of California, San Francisco Logo

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

Home > Global Health Literature Digest > Birth diagnosis of HIV
Birth diagnosis of HIV infection on infants to reduce infant mortality and monitor for elimination of mother-to-child transmission
Global Health Sciences Literature Digest
Published May 13, 2013
Journal Article

Lilian RR, Kalk E, Technau KG, Sherman GG. Birth diagnosis of HIV infection on infants to reduce infant mortality and monitor for elimination of mother-to-child transmission. Pediatr Infect Dis J. 2013 Oct;32(10):1080-5.


To identify ways to optimize early infant diagnosis (EID) to improve outcomes for HIV-exposed infants.


Academic hospital in Johannesburg, South Africa, specialized in HIV care and prevention of mother-to-child HIV transmission (PMTCT) services.

Study Design

Prospective observational cohort study.


HIV-exposed infants and their mothers. Eligible mothers were age ≥18 years, HIV-infected, aware of their HIV status before delivery and planning to attend the hospital for postnatal PMTCT care.

Intervention (Predictor Variable)

Periodic dried blood spot (DBS) testing from birth in the HIV-exposed infants.

Main Outcome Measures

Proportion of infants for whom HIV status was determined by age six weeks; proportion of HIV-infected infants initiating antiretroviral therapy (ART); loss to follow-up (LTFU).


Mother-infant pairs were recruited from August 2008 to July 2010. Follow-up continued until December 2010. All mother-infant pairs received routine PMTCT care per South Africa's PMTCT guidelines (primarily the 2008 guidelines,(1) as the 2010 guidelines(2) were released in April 2010). The 2008 guidelines advised that women with CD4 count ≤200 cells/µL should initiate ART, and that women with higher CD4 counts should receive zidovudine (AZT) from 28 weeks gestation, with single-dose nevirapine (sdNVP) given at delivery. Infants received sdNVP at birth and seven days of AZT. Infants whose mothers had received less than four weeks of antenatal AZT received AZT for 28 days. Infants were to be exclusively formula-fed. The 2010 revision of the guidelines stipulated that HIV-infected women were ART-eligible with CD4 counts ≤350 cells/µL. Those with higher counts were given AZT from 14 weeks gestation. Infants received NVP for six weeks, and were to be exclusively breastfed.

All infants were to undergo polymerase chain reaction (PCR) testing for HIV DNA at age six weeks. Results of this test for each infant were communicated to caregivers when the infant received immunizations at age 10 weeks. Infants testing HIV-negative were discharged from the study. Infants testing HIV-positive received baseline viral load tests and CD4 count tests and were referred for ART initiation.

Medical practitioners examined each infant at birth, and at ages two weeks, four weeks, six weeks and 10 weeks, collecting and storing DBS samples at each examination. One standard PCR assay and two newer assays were used to determine whether HIV-infected infants could be identified earlier than age six weeks. Infants were classified as having been infected in utero when two assays found positive results in samples from birth. Infants were classified as intrapartum-infected when results were negative at birth but positive on two different assays by age six weeks. HIV-infected infants receiving care at the hospital were followed up until the end of the study in December 2010. Infants were considered to have remained in care if they attended the hospital "in the last nine months of the study," though it is not clear whether this means they attended at least once during the last nine months, or continued attending all the way through the last nine months.

When an infant was LTFU before the PCR testing at age six weeks, investigators searched the National Health Laboratory Service's Laboratory Information System (LIS) to ascertain whether the infant had received a PCR test at a different facility, and if so, what was the result. If surnames and birth dates matched, investigators assigned an HIV status to the infant based on the LIS data. At the study's conclusion, all birth DBS samples from infants with unknown HIV status were tested to identify any other infants infected in utero. Investigators also tested all available week four DBS samples.


Over two years of the study, 848 mother-infant pairs were enrolled, including nine sets of twins. Ten withdrew during the study. There was a net total of 829 mothers and 838 infants. Of the 804 women whose nationality was known, 346 (43%) were from Zimbabwe. HIV serostatus was obtained for 710 (85%) infants, 606 (72%) of whom received PCR testing at the hospital, with 85 (10%) tested at other facilities. An HIV status was determined for 19 LTFU infants through testing of DBS samples. In the remaining 128 LTFU infants, only infection in utero could be ruled out through birth DBS samples, as week six DBS and PCR results were not available.

A total of 38 (5.3%) HIV-infected infants were identified among those for whom HIV status could be determined. Most (29/38, 76.3%) were infected in utero. Thirty (79%) HIV-infected infants received PCR testing at age six weeks. Twenty-three infants initiated ART at a median age of 16 weeks, but only 14 were in care a median of 68 weeks later and four had died. The infants who initiated ART did so at a median age of 16 weeks. The median delay between PCR testing and ART initiation was 10 weeks. The median delay between receipt of PCR results and ART initiation was six weeks. Fourteen (54%) infants achieved virologic suppression at a median age of 36.5 weeks. A median of 18.4 weeks elapsed from ART initiation to virologic suppression.

Most (565/606, 93%) infants accessed routine diagnostic testing at the hospital and later returned to receive test results. Twenty-six (4.3%) infants were thus identified as being HIV-infected. Of these, 25 (96%) received PCR results and 23 (88%) began ART. Two of the three infants who did not initiate treatment were LTFU, and one moved out of the area before ART initiation. Eight (21%) of the HIV-infected infants were identified through testing of DBS samples. These eight infants had been infected in utero. Their mothers were younger, more likely to be from Zimbabwe and to have accessed less optimal antenatal care. Two of these infants died.

In addition to the 26 HIV-infected infants who were diagnosed at the hospital, four infants defaulted from the study and were diagnosed at other facilities. One of these infants died at age four weeks. One returned to the hospital for treatment at age 20 months and remained in care, and the other two were LTFU. The 23 infants who began ART were followed for a median of 51.6 weeks (range 0.0-112.1 weeks). Ten (43%) infants were not in care at study's end. Two infants of these had achieved viral suppression and were then LTFU, five were LTFU with viral load unknown, and three died soon after beginning ART. Of the 13 infants remaining in care, 12 had achieved viral suppression and one was close to achieving suppression.


The authors conclude that PCR-based EID testing should ideally begin at birth so that all perinatally infected infants are identified and early infant mortality is reduced. Optimally-timed additional testing later in infancy and efforts to retain infants in care will be needed, especially for infants who have initiated ART. Testing from birth will also enhance monitoring of MTCT elimination.(3)

Risk of Bias

The risk of bias in this non-randomized observational study is moderate to high. The study population included a high proportion of foreign nationals, so it may not have been representative of the general South African HIV-infected female population. Since women were aware of their HIV-infected status antenatally (and not just at delivery), they may have been more likely to return for care.(4) It was not possible to determine the HIV status of 15% of enrolled infants.

In Context

In order to reduce HIV-related early infant mortality, infants must be diagnosed earlier, and begin ART sooner.(5) Rates of EID would be much higher if PCR testing were done at birth, rather than waiting to age six weeks. However, this could create additional strain on already-overburdened health systems in low-resource settings, especially since the diagnosis would require confirmatory testing so infants do not initiate ART unnecessarily.

Programmatic Implications

The World Health Organization (WHO) recommends that HIV-exposed infants undergo PCR testing between the ages of four and six weeks.(6) Clinicians should follow WHO guidelines. If feasible, periodic DBS testing from birth could be implemented both to measure PMTCT rates as well as to diagnose infants even earlier. New WHO guidance is expected in mid-2013.


  1. South African National Department of Health. Policy and Guidelines for the Implementation of the PMTCT Programme. Pretoria: National Department of Health; 2008. [accessed 25 April 2013]
  2. South African National Department of Health. Clinical Guidelines: PMTCT (Prevention of Mother-to-Child Transmission). Pretoria: National Department of Health; 2010. [accessed 25 April 2013]
  3. Joint United Nations Programme on HIV/AIDS (UNAIDS). Countdown to Zero: Global plan towards the elimination of new HIV infections among children by 2015 and keeping their mothers alive. [accessed 6 May 2013].
  4. Technau KG. Can a routine peri-partum HIV counselling and testing service for women improve access to HIV prevention, early testing and treatment of children? Dissertation: Faculty of Health Sciences. Johannesburg (South Africa): University of the Witwatersrand, 2009. [accessed 6 May 2013].
  5. Violari A, Cotton MF, Gibb DM, Babiker AG, Steyn J, Madhi SA, Jean-Philippe P, McIntyre JA; CHER Study Team. Early antiretroviral therapy and mortality among HIV-infected infants. N Engl J Med. 2008 Nov 20;359(21):2233-44.
  6. World Health Organization (WHO). Antiretroviral therapy for HIV infection in infants and children: Towards universal access. [accessed 6 May 2013]