University of California, San Francisco Logo

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

Survival from 9 months of age among HIV-infected and uninfected Zambian children prior to the availability of antiretroviral therapy
Global Health Sciences Literature Digest
Published May 11, 2009
Journal Article

Sutcliffe CG, Scott S, Mugala N, et al. Survival from 9 months of age among HIV-infected and uninfected Zambian children prior to the availability of antiretroviral therapy. Clin Infect Dis 2008;47:837-44.

In Context

Although the rate of infant mortality is known to be very high in sub-Saharan Africa, longitudinal studies of infant mortality and vital statistics are rare.(1) With the expanded use of antiretroviral therapy (ART) in the region, information on mortality before ART is necessary for comparison with mortality after the introduction of ART.


To compare survival among HIV-infected and -uninfected infants


Lusaka, Zambia

Study Design

Observational cohort of the immunogenicity of measles vaccine in HIV-infected and -uninfected children


All HIV seropositive children and a random sample of HIV seronegative children aged 2-8 months who attended a public health clinic for routine vaccination from May 2000 through November 2002 and who had not been vaccinated against measles




Caregivers for participating infants were asked to return the infants for measles vaccination at 9 months of age. Follow-up of children was done at either one or three months after vaccination and was determined randomly. Follow-up of all children occurred at 15 and 27 months post-vaccination. A community health worker visited the child's home if the follow-up visit was not kept.

To obtain vital status, home visits were conducted every three months and continued until the child was at least three years old. The date of death was recorded for children who had died and verbal autopsies were performed using standard instruments at three months after death. Pediatricians reviewed the results from the verbal autopsy and determined up to four contributing causes of death for each decedent. One study physician assigned causes of death based upon the interpretations of the verbal autopsies. Children who died before receiving vaccine and who received verbal autopsies were included only in the analysis of the circumstances of death because HIV status was not obtained.

HIV serostatus was determined by antibody and RNA testing. Infants were classified as HIV-infected if HIV RNA was detected in plasma, as HIV seropositive but uninfected if antibodies were detected but RNA was negative, and as HIV seronegative if HIV antibodies were not detected. Mothers of seropositive infants were assumed to be infected. For infants who acquired HIV infection between follow-up visits, the date of HIV infection was estimated as the midpoint between the visit at which HIV was detected and the visit preceding it.

The Kaplan-Meier product limit test was used to estimate survival stratified by HIV serostatus and infection status. Age was the time axis because children entered the study at different ages. Children were followed until death or the last known contact with study staff. Mortality rates over the 27-month period were calculated by HIV infection and exposure status. Person-years were calculated from date of study entry until death, loss to follow-up, or censor. A Cox proportional hazards model was used to measure risk factors for mortality.


Of the 696 eligible children, 492 were vaccinated for measles and of these, 127 (26%) were HIV seronegative, 292 were seropositive but uninfected, and 73 (15%) were HIV infected. There were 32 additional HIV infections that occurred among the 292 children who were uninfected at entry.

There were 56 deaths during follow-up. Mortality was highest among the HIV-infected children; 39% of HIV-infected children had died by the end of the study compared with 5% of the seropositive but uninfected children and 1.6% of the seronegative children. Crude mortality was 290 deaths/1,000 person-years for HIV-infected children and 7.9 deaths/1,000 person-years for seronegative children. The crude mortality among HIV seropositive uninfected children was 25 deaths/1,000 person-years.

Survival at 24 and 36 months was 69% and 52%, respectively, for HIV-infected children, compared to 95% at both follow-up visits for HIV seropositive but uninfected children and 99% and 98%, respectively, for HIV seronegative children.

Independent predictors of mortality among the HIV-infected children were a clinic visit in the previous four weeks (adjusted hazards ratio [AHR], 4.6; 95% confidence interval [CI]: 1.5-13.5), hemoglobin level <8 g/dL at study entry (AHR, 4.4; 95% CI: 1.5-12.6), and CD4+ T lymphocyte percentage <15% at study entry (AHR, 3.2; 95% CI: 1.1-9.5).

The circumstances surrounding death was determined from verbal autopsies for six HIV seronegative children; 23 HIV seropositive, uninfected children; and 60 HIV-infected children. The duration of illness was longer for HIV-infected children (median 30 days) than for HIV seronegative and HIV seropositive but uninfected children (median 7 days and 8 days, respectively). The most common contributory causes of death were diarrhea, acute respiratory infection, malnutrition, tuberculosis, meningitis, and malaria. Multiple contributory causes of death were identified for two thirds of the children.


Before use of ART, approximately 50% of infants who were alive at nine months survived to three years, reinforcing the need to prevent mother-to-child transmission of HIV. These findings are consistent with other studies of infant mortality in sub-Saharan Africa.(2) It appears as well that mortality may be higher for HIV seropositive infected infants than for HIV-uninfected infants(2, 3) although the reasons for this are not clear.

Quality Rating

This study was of adequate quality. The sampling frame was representative of children in urban sub-Saharan Africa obtaining vaccination at public clinics but may not have been representative of children in rural areas of who may not receive routine care. Participation rate (proportion of eligible children who participated) was only 71%. HIV status was determined using standard testing methodology. The outcome was incompletely ascertained because of poor vital statistics in this area; and only 69% of participants were followed through the entire study period. Survival up to three years was measured, but longer term survival would have been helpful.

Programmatic Implications

This study provides additional evidence for the urgent need to expand programs to prevent mother-to-child transmission of HIV and to provide rapid treatment with ART for HIV-infected children. Expansion of ART programs and provision of nucleic acid-based tests for rapid HIV diagnosis in children will assist in identifying those in need of early treatment with the intent of improving survival.


  1. Little K, Thorne C, Luo C, et al. Disease progression in children with vertically-acquired HIV infection in sub-Saharan Africa: reviewing the need for HIV treatment. Curr HIV Res 2007;5:139-53.
  2. Taha TE, Kumwenda NI, Broadhead RL, et al. Mortality after the first year of life among human immunodeficiency virus type 1-infected and uninfected children. Pediatr Infect Dis J 1999;18:689-94.
  3. Brahmbhatt H, Kigozi G, Wabwire-Mangen F, et al. Mortality in HIV infected and uninfected children of HIV-infected and uninfected mothers in rural Uganda. J Acquir Immune Defic Syndr 2006;41:504-8.