Abstracts Vol. 1 [International Conference on AIDS (11th: 1996: Vancouver, Canada)]

Track C: Epidemiology & Public Health
Mo.C.1575
FUTURE IMPACT OF HIV-ASSOCIATED DEMENTIA AND DEPRESSION
Starace F, van den Boom F, Sherr L, Gorna R, Reinking DP. Inst. of Psychiatry Naples, Italy,
Netherlands Red Cross Amsterdam, Netherlands, RFHSM London, UK,THT London, UK,
Netherlands Inst. of Mental Health, Netherlands
Issue: HIV-associated dementia and depression are two of the most imnpairing mental health
conditions, with a high impact on individuals, their social networks, public health and health
care providers.
Project: By linking data from prevalence studies on HIV-associated dementia and depression
with epidemiological projections, an indication of future impact ' termns of costs and burden on health care will be given. Published studies have been selected according to a
defined set of methodological requirements such as group sizes and diagnostic criteria.
Results: The review of the literature showed: I) HIV-associated dem entia prevalence rates
ranging between 6.5% and 7.5%; 2) HIV-associated depression prevalence rates ranging
between 4.0% and I 1.0%; 3) the current prevalence of major depression is consistently
higher in the selected samples than in the general population, suggesting a vulnerability to
affective disorders in persons with at risk behaviours.Within the framework of the
Multinational Scenario Project, epidemiological projections will be linked with data on neuropsychiatric morbidity and, wherever possible, with cost calculations.
Lessons Learned: Notwithstanding the highly debilitating impact of HIlVassociated neuropsychiatric disorders, mental health issues are hardly ever taken into account in epidemiological projections and costing studies.The Multinational AIDS Scenario Project provides a
methodological framework within which this can occurr.
Fabrizio Starace - Via G. Santacroce 19/E - 80 I 29 Napoli, Italy Fax: + 39 8 I 5786295
email: F.Starace@dAgora. Sth.
Mo.C. 1576
PROJECTIONS FOR THE HIV EPIDEMIC IN SCOTLAND 1995-1999.
RaabGillian*, Allardice G+,Yang SY*, McMenamin J+, Goldberg D'. Napier University,
Edinburgh, Scotland. +SCIEH, Ruchill Hospital, Glasgow, Scotland.
Objective: To project the nurmbers of HIV positive people who will need health care to the
year 1999 in Scotland.
Methods: The evidence for HIV infection in Scotland in the 1980's indicated that the largest
group of infected people were intra-venous drug users (IDUs), as well as a smaller number
of men infected via homosexual intercourse (MSM).The 1990s have shown a decrease in
the number of new reports of HIV infection in IDUs, a leveling of the rate of new reports
for MSM and a mnodest rise in newly reported infections attributed to heterosexual intercourse (MSW). These results are, of course, subject to reporting biases. Reported AIDS
cases appear to have plateaued in the early 1990s for the first two groups but are still
increasing for the MSW group. Up to December 1995 there were a total of 732 reported
AIDS cases in Scotland.
In addition to these data, further information about the HIV epidemic is available from various sources. These include unlinked anonymous testing (UAT) schemes in various settings,
information from the testing laboratories on the number of people under imunological surveillance, data on newly acquired infections from records of repeat tests and information on
deaths in HIV positive people fr'om record linkage of test results with death registrations.
This paper will describe a Bayesian methodology that has been developed that allows a
range of possible prevalence estimates to be used in setting the parameters for the number
of past infections and used as a model for the HIV epidemic in Scotland.
Results and Conclusions: The model shows that there is evidence of continuing spread of
infection in all three risk groups in the I1990s.This is lowest in the IDU risk group, most evident in MSM group and most uncertain in the MSW group.This uncertainty is a consequence both of the later emergence of infection in this group and also of the high proportion of unknown infections detected by UATThe model projects that the number of incident AIDS cases in Scotland will rise from 140 in 1995 to 155 in 1999, and prevalent AIDS
cases firom 240 in 1995 to 270 in 1999.
Gillian Raab, Napier University, Sighthill Court, Edinburgh, EH4I I BN, Scotland.
Tel: [phone]; Fax: [phone]; e-mail: [email protected]
Mo.C. 1577
7 YEARS OF SENTINEL SURVEILLANCE OF HIV2 AND HIVI INFECTION:THE USE
OF AN HIV I MODEL TO ESTIMATE THE SPREAD OF THE EPIDEMIC
Wade, A.*, Sokal, D.**, Diaw P.O.*, Dia, F.L.*, Dieng Sarr, A.*, Diallo, A.A*., Ndoye, I.***, CoIlSeck, A.*, Diakhate, L.*. Hanne, A.A.*, Mboup, S.*. * University Cheikhr Anta [)IOPR Dakar; **
Family Health International, USA; *** IHS, Dakar
Objectives:To use 7 years (1989-95) of sentinel surveillance data to establish prolections
for both viruses in a HIV2 endemic area;To give decision makers a tool to evaluate the
future impact of HIV/AIDS and to plan preventive actions.
Methods: A meta-analysis on main epidemiologic findings was conducted and correction
fhctors on gender: HIV (%HIV2 HIVI +0.2) and urban/rural (%R-0.3 U) distribution
were calculated.We modifed a number of factors, including the progression rate from HIV
to AIDS and the perinatal transmission rate f-om mother to children for HIV2.This
approach seemed most accurate in accoun ing for HIV2 specitcities.
Results: Seroprevalence varies from 0.5% in Dakar to 1.6% in Ziguinchor amonig pregnant
women and from I 6.4% to 39% among commercial female sex workers in the same locations. Our model suggests that ftom 1994 to 1998, the number of persons infected in the
general population will increase by I 62%, AIDS cases by I 85% and deaths by I 58%. Our
projections revealed a variation in seroprevalence from 0.13% to 2.88% in 5 years among
adults (15 49 years). In 1997, 606 13 individuals will be infected by HIV.
Conclusions: Existing models of HIV I progression can be modited for use with HIV2 if taking into account the natural history of this virus.Those data are helpfil to build an eftcient
health information system and will allow decision makers to improve their understanding of
the AIDS crisis and better monitor the dynamics of the epidemcs.
Pape Ousmane DIAW Laboratoire de Bacteriologie-Virologie HALD BP 73 I 8 Dakar

Senegal.Tel.: [phone] Fax: [phone]

Mo.C.1575 - Mo.C.1581
Mo.C.1578
THE USE OF MODELLING IN PLANNING FOR THE HIV/AIDS EPIDEMIC:
A SOUTHERN AFRICAN EXPERIENCE
Mason, BL.*,Wood, G.D.*,Whiteside, A.**. *Mason & Wood cc *55Economic Research
Unit, University of Natal, South Africa
Issues: It is important for governments and service organisations to plan for the impact of
HIV/AIDS in the future. Computer models can be used to project the course of the
HIV/AIDS epidemics and hence to provide a basis from which planning may be carried out.
However these models require input data as a basis from which the course of the epidemic
is simulated and also depend on the informed use of several assumptions about HIV/AIDS.
Project: Two AIDS impact projection studies were carried out in the Southern African
region. In the first study the researchers used computer modelling as the basis from which to
provide an assessment of the socio-ecornomic impact of AIDS for the Ministry of Economic
Planning and Development in Swaziland. In the second study the researchers made use of
computer modelling in their assessment of the impact of AIDS on planning for theTown and
Regional Planning Commission of the province of KwaZulu-Natal, South Africa.
Results: In both studies the impact of AIDS on demographics (eg. population growth rates
and dependency ratios), the demand for services, and consequent costs from household to
national level were significant. Both studies provided useful information and direction for
planning in these regions, and denionstrate the value of modelling the AIDS epidemic.
IH owever in the Swaziland study unreliable input data from sentinel surveys resulted in the
specific projections for this region being questionable as to their accuracy.
Lessons learned: Computer modelling is a useful tool in providing projections of the
AIDS epidermic for use in impact prediction and planning. However these models depend
on input data and on several assumptions and therefore must be used with caution and
responsibility.
Ms B.L.Mason, 2 I Wilkes Rd Prestbury Pietermaritzburg 3201 South Africa
tel: +27 33 I 4411234 fix: [phone] email: [email protected]
Mo.C. 1579
GENDERING THE EU CONCERTED ACTION FOR MULTINATIONAL SCENARIO
ANALYSIS ON THE IMPACT OF HIV/AIDS ON SOCIETY
Gorna R, Sherr L, Reinking DP, Starace F, van den Boom F. THT London, RFHSM London,
Netherlands Inst of Mental Health, Inst of Psychiatry, Naples, Italy, Netherlands Red Cross
Amsterdam.
Issues: Current models of AIDS Policy are based on the fact that the first groups affected
were men and this informed the understanding that the male is the norm.The growing
numbers of women with HIV infection and the corresponding available data must be fed
into future scenarios to ensure that women and girl's issues are adequately understood,
anticipated and costed.This branch of the Scenario Analysis aims to integrate a thorough
gender analysis within an ongoing European Union (EU) multinational project to assess past
and future epidemiological, economic and social impacts of HIV and AIDS on society
Project: Data on the following issues is collated and fed into the matrices for establishing a
baseline analysis of HIV prevention interventions, systemic impacts (family) and neuropsychological and quality of life impacts:- current and projected epidemiology; differentials in
male to female, female to male and female to fetus HIV transmission; availability, uptake
and costs of interventions to reduce the risks of HIV transmission during pregnancy
(including HIV testing in ante-natal and abortion centres, rates of termination, medical interventions, such as zidovudine & Caesarean Section); through sex with men (including HIV
prevention mechanisms, and effective inititiatives); care needs of HIV-positive women,
including costs of prophylaxes, gynaecological care (related to findings from Natural History
studies), late presentation of illness; care needs of HIV- affected adolescents and children
(including costings of child care, family clinics, psycho-social implications).The analyses begin
from a state of the art review of research findings conducted in all industrialized countries,
with particular attention paid to the EU.
Lessons Learned: The data emerging reveals the limitation of male based models in terms
of understanding and providing for women and families. Future planning may have to
increase the focus and extent of models to accommodate the needs of women and
respond to gender issues as they affect relationships, sexual negotiation, health care uptake,
procreation, mental health, and hospital costs.
Dick Reinking, PO Box 5103, 3502 JC Utrecht,The Netherlands, tel. [phone],
fax. [phone] II
Mo.C. 1580
PROJECTING HIV/AIDS IMPACTS IN LOCAL REGIONS, 1996-2006
Rattner, Edward, Harris, Carl M. Dept. of Operations Research/Engineering, George Mason
University Faiifax,VA 22030, USA
'[his talk will present a HIV/AIDS case forecasting model for ssnall, local regions (like cities
and counties) oh a large geographic entity (like the United States or Canada) where reasonably reliable data on the epidemic are available.The approach has been successfully applied
to the Northern Virginia suburban areas of Washington, DC, using a combination of Virginia
state data and U.S. national tles fror the Centers for Disease Control and Prevention.
Additional preliminary data analyses have been done for the states of Maryland, West
Virginia and Delaware. Special strategies and their implications will also be discussed for
extending this form of pattern analysis to smaller nations.
Edward Rattner, 9308 CoronadoTerrace, Fairfax,VA 22031, USATel. [phone]
Fax: [phone]
Mo.C.1581
STRATEGIES FOR PROPHYLAXIS AGAINST PNEUMOCYSTIS PNEUMONIA
IN PERSONS WITH AIDS WHO BECOME SENSITIZED TO TRIMETHOPRIM -
SULFAMETHOXAZOLE:A COST-EFFECTIVENESS ANALYSIS
Coyne H. Srti1tRobert P, Wenrnberg D. Department of Medicine, Maine Medical Centet
Portland, Maine.

Objective: To determine the most cost - effective strategy of prophylaxis against pneumocystis pneumronia for persons with AIDS and a history of adverse events attributable to
trrirethoprimn- sulfamethoxazole.,,O
0"
U
0
C
r
(1)
U
Q)
Q/)
cc
U)
cc
U
b
cO
a)
c
4
15