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

94 Abstracts WeOrC1311-WeOrC1374

XIV International AIDS Conference

WeOrC 311 Comparing the results of HIV impact modelling
with reporting by the health information system:
lessons learned in Namibia
F.H. Van der Veen', N. Forster2. 1Family Health International, Windhoek,;
2Ministry of Health and Social Services, PO. Box 40455, Ausspannplatz,
Windhoek
Background: Namibia established in 1992 a surveillance system for HIV and
AIDS, including HIV sentinel sero-surveys, case reporting of HIV positive diagnosis, hospitalizations and death from AIDS. Reporting is entirely integrated into the
Health Information System (HIS). To model the impact of HIV on demographic indicatiors, a national all stakeholders agreed during a consensus meeting in 2000
to use the Spectrum program as this provides the most useful output for planning
purposes.
Methods: The outcome of modelling of the HIV epidemic was compared with
data from the existing computerised Namibian Health Information System (HIS),
covering the period from 1995 to 2000.
Results: The modelling outcome projects an increase of deaths from all causes
from 18,080 in 1995 to 28,220 in 2000 and an increase of deaths from AIDS from
1,440 in 1995 to 11,000 in 2000. The reported number of deaths in hospitals from
all causes increased from 6,402 in 1995 to 12,635 in 2000 and the reported number of hospital deaths from AIDS increased from 628 in 1995 to 3,388 in 2000.
The proportion of projected death from all causes occurring in hospitals increased
from 35% in the year 1995 to 44% in the year 2000. This increase is mainly a result of an increased proportion of children dying in hospitals from 27% to 35%.
AIDS is underreported as cause of death, particularly among children. According to projections, AIDS would cause 40% of all deaths in 2000, while only 28%
hospital deaths were reported to be caused by AIDS.
Conclusions: Despite a projected doubling of the total number of deaths from
all causes over the period from 1995 to 2000, hospitals have been able to cope,
increasing the coverage of services, in particular for terminally ill children. HIS
data provide valuable information on geographical distribution and service utilization, which are essential for monitoring of the health sector response to HIV and
cannot be obtained by modelling.
Presenting author: Fred van der veen, PRO. Box 40455, Ausspannplatz, Windhoek, Namibia, Tel.: [phone], Fax: [phone], E-mail: fred@fhi.
org.na
WeOrC13121 Strategies for reducing the burden of TB infection
and disease in high HIV prevalence populations:
modelling the impact of active case finding,
antiretrovirals and preventive therapy
E.L. Corbett1, C. Currie2, G.J. Churchyard3, B.G. Williams4. '1London School of
Hygiene and Tropical Medicine, BRTI, PO Box CY 1753, Causeway Harare,
United Kingdom; 2University of Southampton, Southampton, United Kingdom;
3Aurum Health Research Unit, Welkom, South Africa; 4World Health
Organisation, Geneva, Switzerland
Background: The HIV epidemic has greatly worsened TB control in Africa. Gold
miners are a closely studied population in whom new control strategies are being
considered.
Methods: We adapted existing mathematical models of TB, and simulated an
HIV epidemic to fit data on time trends of TB incidence and prevalence in HIV+ve
and HIV-ve gold miners, in order to investigate different approaches to improving
TB control.
Results: Several strategies led to major (>70%) reductions in projected TB incidence, but rapidity of onset and durability varied. The most successful in the
short-term was a single round of mass case-finding plus mass preventive therapy
(PT), which simultaneously reduced the burden of latent TB infection, prevalent
TB disease and TB transmission. Durability, however, was < 10yrs unless combined with ongoing measures. Without an initial round of mass PT, antiretroviral
therapy (ART), PT targeted to HIV+ve miners (targeted PT), and increased active
case-finding had slow onset, and were limited by uptake of HIV-testing. A potentially paradoxical effect of widespread ART was observed. If the net effect was to
increase absolute numbers in the mid or late stages of immunosuppression, for
example by intermittent therapy aimed at symptomatic individuals, then overall TB
incidence increased along with workforce HIV prevalence. If, however, normal immune function was maintained, then TB incidence decreased despite increasing
HIV prevalence.
Conclusions: New strategies capable of controlling TB in high HIV prevalence
areas are urgently needed. Our model suggests that combined approaches that
tackle both latent TB and ongoing TB transmission are likely to be most successful, particularly when not exclusively targeted to those of known HIV status.
Care is needed when introducing ART into high HIV prevalence populations, as
some strategies may have a deleterious impact on the control of HIV-associated
infections, such as TB.

Presenting author: Elizabeth Corbett, BRTI, PO Box CY 1753, Causeway,
Harare, Zimbabwe, Tel.: [phone], Fax: [phone], E-mail: elci
@ mweb.co.zw

WeOrC1313 Goals modeling in Lesotho: using a resource
allocation model to assess Lesotho's HIV/AIDS
budget and its impact on the country's epidemic
E.M. Gaillard', M.M. Moteetee2, N. Jaase3, M. Khobotle3, J. Stover4,
L. Bollinger4. 1The Futures Group International, PO Box 407139, Ft. Lauderdale,
FL 33340, Haiti; 2Lesotho AIDS Program Coordination Authority, Maseru,
Lesotho; 3Ministry of Development and Planning, Maseru, Lesotho; 4 The
Futures Group International, Glastonbury CT United States
Background: The Government of Lesotho initial estimates indicated that a National AIDS Strategic Budget of $1 billion, spread over three years (2001-2003),
would be required to reduce the spread of HIV In order to evaluate if this budget accurately reflected the country's needs and capacity, the Goals model was
applied.
Methods: An economic team of 2 economists and 1 CPA carefully reviewed the
budget and scrutinized the cost of each line item. The Goals Model was used to
distribute the budget by program component, and to estimate the impact of the
budget on the prevalence of HIV. As changes were made in the assumptions, the
model automatically updated the budget and the prevalence figures, thereby, ensuring the integrity of the estimates. The results were presented to the HIV/AIDS
task force, out of which a technical team of 9 participants was created to further
stream down the budget through a prioritization process. In addition, an inventory
of available funding in the country as well as an advocacy document, and various consensus-building meetings were planned with key stakeholders to ensure
consensus.
Results: The model estimated that to achieve a drop of 19% in HIV prevalence, a
budget of not less than $100 millions ($16 per capita) and not more that $275 millions ($47 per capita) would be required for the national response to the HIV-AIDS
epidemic in Lesotho. Consequently, a new budget was designed that more realistically reflected the country's goals and proposed activities. Conclusions: The
analysis reveals that without the economic team and without the Goals Model, it
would have been extremely difficult to optimize the allocations of resources while
measuring the impact of the budget on the prevalence of HIV. The task force
felt confident that the figures were backed up by assumptions that were clearly
defined, transparent and defendable at the donors meeting.
Presenting author: Eric Gaillard, PO Box 407139, Ft. Lauderdale, FL 33340,
United States, Tel.: [phone], Fax: [phone], E-mail: ericgaillard
@ksg01.harvard.edu
WeOrC1374 Estimated risk of human herpesvirus 8
transmission by blood transfusion, Uganda
W. Hladik', S. Dollard', R. Downing2, E. Nzaro3, P. Kataaha4, F Banage2,
R. Ransom2, J. Karon', P. Pellett', T. Dondero', E. Lackritz', J. Mermin2
I Centers for Disease Control and Prevention, 1600 Clifton Rd, MS E-46, Atlanta,
GA-30333, United States; 2Centers for Disease Control and Prevention,
Entebbe, Uganda; 3Mulago Hospital, Kampala, Uganda; 4Nakasero Blood Bank,
Kampala, Uganda
Background: Kaposi's sarcoma, the most frequent malignancy in AIDS patients,
is caused by human herpesvirus 8 (HHV8). It is unknown if HHV8 can be transmitted by blood transfusion; worldwide, blood donations are not screened for HHV8.
We evaluated the risk of HHV8 transmission by transfusion in a hospital in Kampala, Uganda.
Methods: We evaluated transfusion recipients and their linked blood donations,
from December 2000-October 2001. Informed consent was given by recipients
or their parents, and blood donors. We collected recipient blood samples prior to
transfusion and one, two, and every four weeks thereafter for six months. Specimens were tested for HHV8 antibodies by two peptide-based enzyme immunoassays based on the ORF65 and K8.1 antigens, and by an immunofluorescence
assay (IFA). Specimens reactive in any two assays or in the IFA were classified
as HHV8-positive, all others as HHV8-negative. PCR and serological testing is
ongoing. We compared HHV8 seroconversion in recipients of HHV8-positive and
negative blood.
Results: We evaluated 1811 patients who received 2687 blood units (72%
packed cells, 19% whole blood) from 1795 blood donors. Recipients' median
age was 1.8 years (range:0.1-89); 246 (14%) died. Median follow-up time was
150 days; 134 (7%) were lost to follow-up. At interim analysis, test results were
available for a median of 29 days post-transfusion (interquartile range:28-32). Of
141 transfused units tested, 69 (49%) were HHV8-positive. Of 119 recipients, 42
(35%) tested HHV8-positive and 77 (65%) tested HHV8-negative prior to transfusion. Of the 63 HHV8-negative recipients with follow-up test results, HHV-8 seroconversion was similar among those who received HHV8-seropositive (2/31) and
HHV8-seronegative transfusions (2/32, p=NS).
Conclusions: Blood donors and transfusion recipients in Uganda have high
HHV8 seroprevalence. Interim analysis revealed no excess HHV8 seroconversion among recipients of HHV8-positive blood.
Presenting author: Wolfgang Hladik, 1600 Clifton Rd, MS E-46, Atlanta, GA
30333, United States, Tel.: [phone], Fax: [phone], E-mail: wfh3
@cdc.gov