Wet nursing increases risk of HIV infection among babies

Pat Sidley
Johannesburg
South African children using public hospitals are exposed to HIV infection because of a lack
of infection control and because some babies are breast fed by women who are not their mothers
and who are HIV positive, a new study says.
The study was conducted by the country's Human Sciences Research Council, Medical
Research Council, and others for the Nelson Mandela Foundation.
It was commissioned after a previous study by the same group in 2002 found an unusually
high proportion of children between the ages of 2 and 9 years infected with HIV. The suspicion had
been voiced at the time that infection during these years may have been caused by particularly high
rates of sexual abuse of children in South Africa. However, the study's authors, Dr Olive Shisana
and colleagues, say that for ethical and legal reasons they did not look into the possibility of sexual
abuse among the children in the study. They have instead concentrated on the possibility that HIV
infection was contracted in hospitals.

The study was conducted at several public hospitals in the Free State province and looked at
dental facilities, maternity wards, and paediatric wards. In all these settings the researchers found
that HIV could potentially be contracted.
Nearly 30% of the milk to be fed to babies tested positive for HIV viral RNA. Almost half
(47%) of the instruments to be used clinically on children and a quarter of instruments to be used in
children's mouths and gums had traces of blood on them.
The new finding of wet nursing as a mode of HIV transmission in South Africa followed
from the finding that 1.7% of children in the study were breast fed by women who weren't their
mothers. HIV positive children were 17 times more likely than HIV negative children to have been
breast fed by a woman other than their mother

Reducing the risk of transmitting HIV from mother to child in

Africa
Trish Groves takes you through a quality improvement report about HIV services in Zimbabwe
Abstract
Problem—Zimbabwe has one of the highest rates of HIV seroprevalence in the world. In 2001 only
4% of women and children in need of services for prevention of mother to child transmission of
HIV were receiving them.

Design—Pilot implementation of the first programme for prevention of mother to child

transmission of HIV in rural Zimbabwe.

Setting—120 bed district hospital in Buhera district (285 000 inhabitants), Manicaland, Zimbabwe.

Key measures for improvement—Programme uptake indicators monitored for 18 months; impact
of policy evaluated by assessing up-scaling of programme.
Strategies for change—Voluntary counselling and testing services for HIV were provided in the
hospital antenatal clinic.Women identified as HIV positive and informed of their serostatus and
their newborn were offered a single dose antiretroviral treatment of nevirapine; mother-child pairs
were followed up through routine health services. Nursing staff and social workers were trained,
and community mobilisation was conducted.

Effects of change—No services for prevention of mother to child transmission of HIV were
available at baseline. Within 18 months, 2298 pregnant women had received pretest counselling,
and the acceptance of HIV testing reached 93.0%. Of all 2137 women who had an HIV test, 1588
(74.3%) returned to collect their result; 326 of the 437 HIV positive women diagnosed had post-test
counselling, and 104 (24%) mother-child pairs received nevirapine prophylaxis.

Lessons learnt—Minimum staffing, an enhanced training programme, and the involvement of

district health authorities are needed for the implementation and successful integration of services
for prevention of mother to child transmission of HIV. Voluntary counselling and testing services
are important entry points for HIV prevention and care and for referral to community networks and
medical HIV care services. A district approach is critical to extend programmes for prevention of
mother to child transmission of HIV in rural settings. The lessons learnt from this pilot programme
have contributed to the design of the national expansion strategy for prevention of mother to child
transmission of HIV in Zimbabwe.
In 2003, about five million people worldwide became infected with HIV. About 800 000
were children, of whom 90% caught the virus from their mothers during pregnancy, birth, or breast
feeding. Zimbabwe has one of the highest rates of HIV infection in the world and is also facing
political turmoil and getting poorer by the day. Nearly a quarter of the country's pregnant women
have HIV, and only 1 in 25 of the women and children at risk of transmission get preventive care.
The paper describes how the Murambinda Mission Hospital and some of its local clinics set up an
18 month project to try to reduce mother to child transmission of HIV in one district of Zimbabwe.
The project was supported by the Zimbabwean Ministry of Health and Child Welfare, and free
nevirapine (an anti-HIV drug) was supplied by its manufacturers, Boehringer Ingelheim. Given
Zimbabwe's very limited funds for health care, the ministry needed to know whether the project was
doing any good and whether it was worth extending to other parts of the country. Researchers from
Zimbabwe's ministry, an international AIDS charity, and a French public health institute worked
together to evaluate the project and now they have described it in this quality improvement report.
A quality improvement report is a special type of paper that describes how people try to
change and improve health services. Projects aimed at improving the quality of health care often
comprise many interventions with repeated cycles of measurement, change, further measurement,
and further change. The standard way to assess changes in health care is by audit. Audit is a way to
find out whether people are doing what they are supposed to be doing—for instance, to see whether
patients with diabetes are having their blood pressure checked regularly. One definition of audit is
an evaluation that establishes the extent to which a process, programme, or service conforms to a set
of standards. Audit usually has three main phases—baseline audit; feedback, education, and change
to ensure that ideal standards are being followed more closely; and reaudit. This is called the audit
cycle, and it often has to be repeated several times.
Audit is fine as a process, but it can be difficult to write up because the traditional structure of
scientific articles (introduction, methods, results, and discussion) gives only limited information
about quite a complicated process. More importantly, this formal structure can emphasise the wrong
parts of a quality improvement programme. Readers may learn more from the methods and from the
interventions that did not work than from the results and the successful interventions. And they
often need to know as much about the background to a new service as they do about the service
itself. Quality improvement reports solve these problems and ensure that readers get the important
messages about how to implementchange in health care (box).
Quality improvement reports should be able to answer the key questions, “Did this lead to
improvement for patients?” and, “What have you learnt and achieved, and how will you take this
forward?”
The programme introduced, for the first time, a service aimed at preventing HIV infection
among babies in Zimbabwe. The ground was prepared by taking on new staff, training nurses, and
social workers in HIV counselling, and telling local communities about the new service. Pregnant
women attending the Murambinda Mission Hospital and its district clinics for antenatal care were
offered counselling and rapid HIV testing; a single dose of nevirapine for HIV positive women and
their newborn babies; counselling and support to encourage exclusive breast feeding for six months;
and regular follow up care of mothers and babies and antibiotics to prevent bacterial infections.
Fathers and other male partners were also encouraged to come along for HIV testing.
The researchers observed and interviewed the programme's staff to see how they felt and
coped, counted how many women accepted counselling and HIV testing, counted how many
mothers and babies took anti-HIV drugs, and worked out how many babies avoided infection with
HIV. They also interviewed women to find out why some turned down counselling and why many
male partners refused HIV testing.
Before the programme started, Zimbabwe had no services for HIV counselling and testing
among pregnant women. Within 18 months, well over 2000 women had been counselled, more than
90% had agreed to HIV testing, and more than three quarters of those tested had come back for their
results. Of the 437 women who were HIV positive, three quarters had further counselling about
treatment and breast feeding but only 104 mother and baby pairs had anti-HIV treatment. Perhaps
most importantly for the future, the information, education, andcommunication about HIV increased
the desire of other family members and the community in general to have access to HIV counselling
and testing services.
When this paper was published in the BMJ, a reader, James Shelton from Washington, sent
this rapid response about it to bmj.com: “This is clearly a very significant undertaking, beginning
with 2298 women counselled (and presumably more approached for counselling.) But with only
104 women receiving nevirapine, assuming an infection rate [among the babies] of
about 15%, and that the nevirapine would reduce the infection rate by about 50%—that means only
about 8 infections [were] averted. Even this assumes that the drug is actually properly
taken. Perhaps I am missing something, and these efforts might lead to something more. But this
cannot even make a dent in the projected 60% child mortality from HIV/AIDS in Zimbabwe.”
This is valid criticism. But does this mean the preventive programme should be stopped? I
do not think so. As Dr Shelton says, the programme may lead to something more. Indeed, the
authors say at the end of their paper that Zimbabwe's district health authorities are planning to roll
out a follow up package that includes clinical and psychosocial support, more training of health
professionals, continued decentralisation of services, and improved transport to allow community
based follow up. In addition, other and possibly more effective drugs for preventing mother to child
transmission of HIV are being considered.
At the least, this programme has ensured that more than 2000 women in one area of
Zimbabwe have tried to avoid passing on HIV to their babies, something they could not have
attempted before the programme began because help was not available. Perhaps this paper should
have been called a quality establishing report.
A randomised controlled trial is the most scientifically strong design for studying the effect
of an intervention, whether it is a treatment, a type of education, or a service. But it would not have
been ethical to screen pregnant women for HIV and allocate them randomly (by chance) to either
receive or be denied help that might prevent HIV in their babies.
If these researchers had studied two similar groups of pregnant women—some who were
offered preventive care and some who were not—we might be fairly sure that any subsequent
differences between the groups were due to the package of counselling, testing, and treatment. But
there may have been some other difference between the two groups of women that made one group
more likely to agree to counselling and treatment, and this would have biased the results.
Randomising the women to one group or the other would have minimised the risk of bias and made
us much more certain that the new service had done some good.
This quality improvement report cannot prove, therefore, that the Murambinda Mission
Hospital programme prevented any cases of HIV among the babies born in the area in the 18
months of the study. But, given that the ministry wanted to set up the service anyway, it is good that
the researchers were able to monitor the programme's first year and a half and describe carefully
what happened.
The strengths of a study like this are the clear description of the situation before the
programme started, the explicit statement of the measures that might show improved quality of care,
and the thorough reporting of what happened. This quality improvement report could help other
hospitals to set up similar programmes and, just as importantly, help other researchers to study the
uptake of HIV counselling and testing among pregnant women in other low income countries.