HIV prevalence fell sharply
among children admitted to one of South Africa’s largest hospitals in 2009 and
2010, but remained high at 19.3%, researchers from Chris Hani Baragwanath (CHB) hospital in Soweto report in the advance online edition of the Journal of Acquired Immune Deficiency
In the 15 years preceding
2009, HIV prevalence among children admitted to the hospital had remained remarkably
constant, peaking at 31.7% in 2005. This indicated the extremely high rate of
vertical HIV transmission in South Africa prior to the implementation of
up-to-date interventions to reduce it.
However, the persistently high prevalence indicates that
– despite improvements in the efficacy of interventions to prevent mother-to-child transmission
– huge numbers of pregnant women with HIV in South Africa were still failing to receive effective treatment and care that could prevent their child from acquiring HIV.
In additional, these four
independent surveillance studies undertaken in 1996, 2005, 2007 and 2011 in the
paediatric wards of CHB hospital, show HIV-related death rates declining in the
latter three periods from 24 (2005) to 12 (2007) and 12.3% (2010/2011).
While pneumonia was the most
common cause of death throughout, deaths attributable to tuberculosis (TB)
steadily increased from 18 to 26.3% and 44% in 2005, 2007 and 2010-11,
respectively. TB continues to be an important co-infection in HIV-infected
These results show an
encouraging trend, but with close to one-fifth of the approximate 6000
admissions in 2010/2011 to the CHB paediatric wards HIV-related, improved
– as well as prevention of TB disease
– remains critical.
In 2009, South Africa had an
estimated 330,000 HIV-infected children (over 13% of children infected
worldwide) and as many as one in three deaths among children under the age of
five is estimated to be HIV-related.
Tracking HIV prevalence of
children admitted to hospital has been used as an indication of the effect of
HIV on health services for children. At CHB hospital this has been evaluated on
and off for about 20 years. Twenty-three children were diagnosed with HIV between May
1989 and April 1990. From 1990 to 1996 HIV-related paediatric admissions
increased from 1% to close to 30%, reflecting the rapid increase of HIV infection among
pregnant women. During this same period, in-hospital paediatric HIV-related
death rates increased by 42%.
provision for adults and children was introduced in 2004 by the South African Department of Health.
While uptake has been slow, South Africa now has the largest ART programme in
the world, with an estimated 54% paediatric ART coverage in 2010. Evaluation of
the effectiveness in 2010 of the national PMTCT programme showed that 31.4% of
infants were HIV-exposed, while the MTCT rate was 3.5% in these infants at 4 to 8
weeks of age.
In light of these programming
improvements, the authors chose to describe the effect on HIV prevalence and in-patient
death rates among children admitted to CHB hospital.
The hospital serves a
population of 1.4 million in Soweto, Johannesburg, in the Gauteng province of
South Africa. Close to 6000 children,aged up to 15 years, are admitted every
Methods among the
surveillance studies differed. For the most recent
– from 1 August 2010 to 31
– children were enrolled prospectively from one of four general
From 1 October to 31 December
2007, a cross-sectional retrospective review of all children admitted to all
four wards was undertaken.
The 2005 study was part of a
larger sentinel surveillance study to monitor the effect of HIV on heath
services in Gauteng Province. Information was collected for all patients
admitted in four hospitals over a 4 to 6 week period in April and May 2005. CHB
hospital was one of the sites and children were enrolled from all four wards.
From 1 July to 31 December
1996, children under the age of five admitted to one ward at CHB hospital were enrolled.
The results show an
encouraging trend: bothHIV-related paediatric hospital admissions and
overall death rates decreased. Such progress, note the authors, is reflective of improved
PMTCT programmes and ART coverage.
In addition, over the time
period, new vaccines were introduced into the South African immunisation
programme against influenza, pneumonia and diarrhoea. While less effective in
HIV-infected children, they have shown efficacy in reducing the burden of these
The authors suggest increased
death and disease due to TB may reflect an increase in TB prevalence, caused by increased household exposure or the increased risk for TB immune reconstitution
inflammatory syndrome (IRIS) in children taking antiretroviral therapy in more recent times. While there are improved tools for TB diagnosis,
diagnostic methods for paediatric TB have not changed, so this cannot explain
the rise in TB-related deaths in later time periods.
The authors stress that “efforts
to prevent TB disease and death should focus on the use of isoniazid preventive
therapy, early diagnosis and treatment of TB”.
Death rates declined among
HIV-infected children. While there was no significant change in death rates
among HIV-negative children, they were consistently lower than in their
HIV-infected counterparts: 11.2 (65/565) and 24% (43/179) in 2005; 6
(91/1510) and 12% (53/440) in 2007; and 4.2 (18/429) and 12/3% in 2010-11,
Children under six months are
especially vulnerable to HIV-related death. This study showed a decrease both
in death rates and absolute numbers admitted among this age group in 2010-11:
66.7 (18/27, 2005), 70 (28/40, 2007) and 44.4% (4//9, 2010-2011). This finding leads the authors to “cautiously
anticipate a reduction in infant and under five mortality rates…to attain MDG4
of a two thirds reduction in under five mortality by 2015”.
The median age of children
with HIV admitted to hospital increased in 2010-11: from 9.13 months (IQR: 3.6-28.8) in
2005 and 10 months (IQR: 3.0-44.5, p>0.10) in 2007 to 18 months (6.2-69.8,
p=0.048) in 2010-11. The increase in median age is explained, the authors note,
by the continued expansion of PMTCT programmes resulting in fewer infants
Limitations include the
absence of a uniform surveillance system at the hospital
– so all surveys used
different methodologies, making any direct comparisons difficult.
Budgetary constraints meant
fewer staff were available to get informed consent from caregivers in 2010-11,
resulting in the smallest sample size of the studies. The authors did not
believe this resulted in bias.
CHB hospital is a large, urban,
academic hospital in a well-resourced province so these findings may not be generalisable
to lesser-resourced or rural settings.
The authors conclude that “even
though results from the PMTCT programme are reassuring, HIV is a preventable
condition in children, and most cases should be successfully prevented…A high
index of suspicion for HIV-infection should be maintained and routine HIV
screening of all children presenting at health services should increase in
order to diagnose all infants and older children. With continued effort, South
Africa can regain some ground in attaining the MDG4 target and substantially
reduce new HIV infections and HIV-related deaths among children.”