Abstract Book Vol. 2 [International Conference on AIDS (14th: 2002: Barcelona, Spain)]

438 Abstracts ThPpC2139-ThPpC2142

XIV International AIDS Conference

Cervical samples for HIV1-DNA and -RNA were taken during labour as well as
oral samples for HIV1-DNA from the neonate. Three hundred seventy five placentas from HIV1 infected women were processed as well as 277 HIV1 non-infected
controls.
Results: Malaria parasites were found in 4.3 % (12/277) of the HIV1 non-infected
and in 9.0 % (25/277) of the HIV1 infected placenta's (p= 0.027). Placental
malaria was not associated with HIV1 mother-to-child transmission, neither with in
utero infection (2.9 versus 5.6%, p=0.509) nor with peripartal HIV1 transmission
(11.5 versus 15.4%, p=0.6). Cervical HIV1-DNA (60.0 versus 56.2 %, p=0.745)
and -RNA shedding (20.0 versus 25.0%, p=0.620) was not significantly different
when placental malaria was present or not.
Conclusions: Placental malaria was more common in HIV1 infected women, but
was not identified as a risk factor for mother-to-child transmission of HIV1 nor for
cervical viral shedding.
Presenting author: Ingrid Inion, Arthur Sterckstraat 19, 2600 Berchem, Belgium,
E-mail: [email protected]
ThPpC2139 Child survival and mother's HIV status in a rural
Ugandan cohort
J.S. Nakiingil, J.A.G. Whitworth1, A. Ruberantwari1, J. Busingye',
S.M. Mbulaiteye2, B. Zaba3. 'MedicalResearch CouncilProgramme on AI/DS in
Uganda/Uganda Virus Research Insitute, M Entebbe, Uganda; 2 Viral
Epidemiology Branch, National Cancer Institute, Rockvile, United States;
3 Centre for Population Studies, London Schoolof Hygiene and Tropical
Medicine, London, United Kingdom
Background: Estimates of the effect of maternal HIV status on child survival are
usually based on studies of hospital births, and thus subject to selection biases.
We analyse the relationships between child mortality and maternal survival and
HIV status in a population cohort in rural SW Uganda.
Methods: Annual surveillance of approximately 10,000 individuals collects data
on births and deaths, as well as testing serum samples for HIV antibodies. Linked
mother and child records are analysed and Hazard Ratios (HR) for child deaths
are computed using a piecewise exponential hazard model with time-varying covariates.
Results: 3,727 children were born between 1989 and 2000, of whom 415 died
during 14,110 child years of follow-up. Mother's HIV status at birth was ascertained unambiguously for 3004 (81%) of children, of whom 218 (7.3%) were born
to HIV positive mothers.
Infant mortality risk was 53 per 1,000 live births to HIV negative mothers, (95%
confidence limits 45-62), for HIV positive mothers it was 225 (174-289). Child
mortality risks up to age 5 (CMR) were 114 (101-128) and 313 (251-385) for HIV
negative and positive mothers. 14% of child mortality was attributable to maternal
HIV infection.
CMR among living mothers was 128 (117-141), and 571 (403-750) for mothers
who died.
After controlling for child's age and sex, factors that contributed significantly and
independently to child mortality were: mother HIV positive [HR = 3.2, p<0.001];
mother's terminal illness or death [HR=3.2, p<0.001]; teenage motherhood
[HR=1.7 p<0.001]; twinning [HR=2.0, p=0.003] and mother's absence [HR=1.7,
p=0.006].
Conclusions: Maternal survival and HIV status are strong predictors of child survival. Orphans and children of severely ill mothers need support regardless of
maternal HIV status. Retrospective reports of child survival obtained from living
mothers under-estimate HIV-related child mortality.
Presenting author: Jessica Sendegeya Nakiyingi, Medical Research Council
Programme on AIDS in Uganda, and Uganda Virus Research Institute, PO.Box
49, Entebbe, Uganda, Tel.: [phone]/320272, Fax: [phone], Email: [email protected]
ThPpC2140 Monitoring the situation of orphans, a global
assessment
C.R. Monasch, R. Venu. UNICEF New York, NY United States
Issues: Over 10.4 million children under age 15 have lost their mother or both
parents to AIDS. The number of children orphaned is forecast to more than double by 2010. The main commitment from the UN Special Session on AIDS related
to orphans is '... to provide a supportive environment for orphans... and ensure
that they have access to education and to health services on an equal basis with
other children'. Little international comparable data has been available to monitor
this goal. Data on orphanhood is being collected through the Demographic Health
Surveys (DHS). However no info on their well being is collected because orphans
are excluded in the DHS (the questions are directed to mothers). To get a better understanding of their situation UNICEF interviewed in it's Multiple-Indicator
Cluster Survey (MICS) the caretakers of orphaned children.
Description: MICS is a nationally representative, pop.-based survey. In 2000/01,

over 65 countries implemented a MICS. For each of these countries child well
being indicators can be compared for orphans and non-orphans living in households.
Lessons learned: The composition of the households in which orphans live differs from one country to the other depending on culture, norms, and gender roles
in each country. Eg., in Zimbabwe/Ghana, most orphans live with their grandparents (mainly female-headed households). In Zambia/Cameroon, the major
ity grow up with other relatives (male-headed households). Orphans are found
to have less schooling and are more involved in child labour. New orphan/nonorphan statistics on child well being indicators will be presented for a large number of countries.
Recommendation: The loss of one or both parents has a significant impact on
children. The impact of culture and societal norms on children's living arrangements need to be taken into consideration in programming. It is recommended
that in future the situation of orphans is being monitored through other household
surveys as well.
Presenting author: roeland monasch, three un plaza, unicef house room 492,
new york, ny 10017, United States, Tel.: [phone], Fax: [phone],
E-mail: [email protected]
ThPpC2141 The longterm impact of HIV and orphanhood on the
mortality and well-being of children in rural Malawi
A.C. Crampin1, S. Floyd2, J.R. Glynn2, N.J. Madise3, A. Nyondo',
M.M. Khondowe1, C.L. Njoka, H. Kanyongoloka1, B. Ngwira, B. Zaba2,
P.E.M. Fine2. 'Karonga Prevention Study, Chilumba, Malawi;2London School of
Hygiene & Tropical Medicine, London, United Kingdom; 3University of
Southampton, Southampton, United Kingdom
Background: Few published studies in sub-Saharan Africa have recorded mortality or morbidity of children by HIV status of the mother beyond 2 years, and
none beyond three years. We assessed the influence of HIV status of the mother
and of orphanhood on the longterm mortality and morbidity of children.
Methods: We conducted a retrospective cohort study with more than 10 years
of followup. From population-based surveys in Karonga District, Malawi in the
1980s, 197 individuals were identified as HIV positive. These individuals and 396
age-sex matched HIV negative individuals and their spouses and offspring were
sought in 1998-2000.
Results: All but 11 of the index cases were traced, identifying 2520 offspring.
Among those with HIV positive mothers, infant mortality was 27% (95% confidence interval 19-38), under 5 mortality 46% (36-58) and under 10 mortality 49%
(38-61). The corresponding figures for those with HIV negative mothers were 11%
(9-13), 16% (13-19) and 17%(14-20). After adjusting for socio-demographic factors, the hazard ratio for death associated with having an HIV positive mother was
3.0 (1.7-5.1) for infants and 5.5 (2.6-11.6) between 1 and 5 years. Death of HIV
positive mothers, but not of HIV negative mothers or of fathers, was associated
with increased mortality Among survivors, the HIV status of the mother was not
associated with stunting, being underweight, or reported symptoms.
Conclusions: Children of HIV positive mothers had greatly increased mortality at
least up to 5 years. Parental death was only associated with increased mortality
for those with HIV positive mothers. This could be due to the greater risk of HIV
infection and of faster progression in children of mothers with more advanced HIV
disease. The lack of evidence of an impact of orphanhood in other groups may be
due to small numbers, but suggests that extended family support systems were
able to cope.
Presenting author: Sian Floyd, London School of Hygiene and Tropical
Medicine, Keppel St, London, WC1E 7HT, United Kingdom, Tel.: [phone]
7888, Fax: [phone], E-mail: [email protected]
ThPpC2142 Linking hospital HIV sero-survey data to discharge
records: impact of HIV/AIDS at Lacor Hospital,
North Uganda (1992-1998)
S. Accorsi1, M. Fabiani1, G. Rizzardini2, M. Lukwiya3, M.G. Dente1, C. Opira3,
R. Aleni3, S. Declich 1. 'Laboratory of Epidemiology and Biostatistics,Istituto
Superiore di Sanita, Rome, Italy; 2Infectious Diseases Unit, Circolo di Busto
Arsizio Hospital, Varese, Italy; 3St. Mary's Hospital Lacor, Gulu, Uganda
Background: Monitoring the impact of HIV/AIDS on health services is crucial to
develop sustainable strategies for delivering adequate health care. This study is
aimed at evaluating the impact of HIV/AIDS on the medical ward of St. Mary's
Hospital Lacor, North Uganda, in the 1992-1 998 period.
Methods: We performed a combined analysis of data from a cross-sectional HIV
serosurvey and hospital discharge records routinely collected at Lacor Hospital.
The one-month HIV serosurvey was conducted in March 1999 and included all
the 353 consenting patients aged over 12 years who were admitted to the medical ward. The hospital discharge records concerning 17,173 patients aged over
12 years who were admitted to the medical ward in the 1992-98 period were also
analysed.
Results: HIV prevalence in the medical ward was 42.0% (95% Cl: 36.7%-47.4%)
in 1999: 52.6% (Cl: 44.4%-60.7%), 44.6% (Cl: 35.6%-53.9%) and 13.2% (Cl:
6.2%-23.6%) in the general medicine, TB, and cancer units, respectively. All the
clinical diagnoses of AIDS and cryptococcal meningitis were confirmed by a positive HIV test, whereas the HIV prevalence among patients admitted for Kaposi's
sarcoma was 90.0%. Among patients admitted for pneumonia, enteritis, malaria

and TB, the disease-specific HIV prevalence ranged from 45.0% to 65.2%. The
in-hospital mortality was higher among HIV+ patients (15.3%; Cl: 9.7%-22.5%)
than among HIV- patients (3.1%; CI: 1.1%-6.6%), also after adjusting for age and
sex (Adjusted Odds Ratio=6.1; CI: 2.3-15.9). Combining data from hospital discharge records and the HIV sero-survey, the estimated percentage of bed days
attributable to HIV+ patients did not show evident trend over time in the period
1992-98, ranging from 35.0% in 1995 to 40.6% in 1997.