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

Tuesday, July 9, 1996

Tu.C.561 -Tu.C.571

Conclusion: The national data, though crude, indicates that Canadian women who inject
drugs or have sex partners at risk for HIV are increasingly becoming infected.There is a
need to better define trends, risk factors and geographic shifts of HIV among Canadian
women to help target prevention and care programmes.
Dr Lee Y Lior Field Epidemiologist, Division of HIV/AIDS & STD; LCDC, Health Canada PL
0202A Tunney's Pasture, Ottawa Ontario Tel: 613-94 I-3156: Fax: [phone];
email: [email protected]
Tu.C.56 I1
NONINJECTION DRUG USE AND SEXUAL RISK BEHAVIOURS OF WOMEN WITH
HIV AND AIDS - UNITED STATES
Kacanek, Deborah, Diaz,T Ward. JW and the Supplement to HIV and AIDS Project Group.
Centers for Disease Control and Prevention, Atlanta, GA, USA
Objective: Crack use is associated with an increased risk of HIV transmission, especially
among women. However; little is known about the behaviours associated with the use of
other noninjection drugs. We examined the sexual behaviours associated with crack and
other noninjection drug use among HIV-infected women who do not inject drugs.
Methods: We analyzed data from 2270 questionnaires administered to women >18 years
of age with HIV or AIDS reported to 12 state or local health departments between June I,
1990 and june 22, 1995.Women were asked about noninjection drug use in the 5 years
before the interview.
Results: Of all women interviewed, 488 (21%) used noninjection drugs but did not inject
drugs.Thirty eight percent of women who acquired their HIV infection through heterosexual transmission used noninjection drugs. Among noninjection drug users, 206 (42%) ever
used crack, I138 (28%) only used marijuana, 65 (I13%) primarily used marijuana, 44 (9%) primarily snorted cocaine, 7 (2%) primarily smoked heroin, and 28 (6%) used other or
unknown drugs. Many women reported only one male sex partner in the past year (range:
4 1% of crack users to 59% of primarily cocaine users).Thirty-six percent of crack users,
I 1% of primarily marijuana users, and 7% of primarily cocaine users had received money
for sex in the past five years (<5% of those using other noninjection drugs accepted money
for sex). Additionally 66% of crack users, 37% of primarily cocaine users, 37% of marijuana
only users, and 4 1% of primarily marijuana users had had a sexually transmitted disease
(STD); however; none of the primarily heroin users had had an STD.
Conclusions: As many as one third of HIV-infected women exposed to HIV through heterosexual contact use noninjection drugs. Crack users have the highest rates of risky sexual
behaviour; however, a substantial portion of women who use other noninjection drugs also
engage in risky sexual behaviours. HIV preventive and health care services for women must
address other noninjection drug use in addition to crack and emphasize sexual risk reduction to prevent acquisition and further transmission of HIV
Deborah Kacanek, I 600 Clifton Road, MS-E-47, Atlanta, GA 30333 USA.
Tel: [phone], Fax: [phone]
Tu.C.562
EVALUATE HIV INCIDENCE FROM SERIAL PREVALENCE DATA IN PREGNANT
WOMEN, PARIS AREA, FRANCE.
Couturier Elisabeth*, BatterV*, Brossard Y**, Six C*, Larsen C*, Larsen M*, Henrion R***,
Brunet JB*. * European Centre for the Epidemiological Monitoring of AIDS, Saint-Maurice;
** CHP; ** Port-Royal, France
Objective: To examine HIV trends in pregnant women between I 991I and 1995 in the Paris
area, France.
Methods: Unlinked HIV antibody testing was carried out during the same 4-week period in
I99 1, 1993 and 1995 among women at outcome of pregnancy (delivery, abortions, ectopic
pregnancy). Sampling was exhaustive for all women in all medical settings. Recorded variables were pregnancy outcome, exact age, country of birth. HIV prevalence was analysed
by 5 birth-year cohorts.
Results: HIV prevalence by outcome was similar across the years. Overall, HIV prevalence
was 3.5 times higher in women who underwent abortions than in those who delivered
(0.94% vs 0.27%). No difference in HIV prevalence was seen between women born in
France and in North Africa. HIV prevalences in women born in sub-Saharan African countries or in the Caribbean were 4 to 8 times higher than in women born in France.The table
represents HIV prevalences by birth-year cohort (age group based on age in I995).

Results: Out of 2,236 person-years (PYS) of follow-up, 75 HIV seroconverters were identified (incidence of HIV=3.4/100 PYS). Compared to women aged 30+ years, those aged
<20 years were at increased risk of HIV [RR=2.3, 95% CI (I. I-4.9)].There was no association between level of education and risk of HIV infection.The incidence of HIV infection
among married and single women was 2.8/100 PYS and 6.5/100 PYS respectively
Compared to women reporting one sex partner in the last I year, those reporting > 3
partners had higher risk [RR=4.7 (2.5-9. I)]. However, the incidence of HIV infection was
high (- 3/100 PYS) even among those reporting only one sex partner Women reporting
that their spouse had sex with other women during the last I year were at increased risk
[RR 1.8 (0.9-3.6)] compared to those not reporting similar history HIV infection was independently associated with gonorrhoea [RR-4.2 (I.9-9.4)], candidiasis [RR-2.2 (I.3-3.7)],
and having an uncircumcised husband [RR=4.2 (I.3-13.7)]. Users of contraceptive pills
[RR=1.4 (0.6-2.8)], IUD [RR=0.9 (0.5-2.0)], and injectable contraceptives [RR-tI.3 (0.6-2.8)]
were not at increased risk of HIV infection.
Conclusions: HIV is a major public health problem in this population. Our finding confirms
that sexual behavior, STDs and lack of circumcision are important independent HIV risk factors. Contraceptive use is not associated with a significant increase of risk of HIV infection.
Saidi H. Kapiga, Harvard School of Public Health, 665 Huntington Ave, Bldg I, no. 805,
Boston, MA 021 15
Tu.C.570
MODELLING THE SPREAD OF HIV IN IDU IN EUROPE WITH STOCHASTIC
SIMULATIONS OF FRIEND AND STRANGER NETWORKS.
Kretzschmar, Miriam*,Wiessing L.G.**. *lnstitut for Medizinische Biometrie, University of
Ttbingen,Tbingen, Germany, **Nat. Inst. of Public Health & Environmental Protection,
Bilthoven,The Netherlands.
Objective: To relate observed sero-prevalence levels in populations of intravenous drug
users (idu) in Europe to risk behavior and investigate the risk of future rise in prevalence of
HIV; to study the effect of reduction in risk behaviour on HIV-incidence in populations with
high sero-prevalence levels.
Methods: Simulation studies with an individual based stochastic network model, which
describes needle sharing contacts. A distinction was made between risk behaviour in long
lasting friend relationships and incidental stranger contacts.The model parameters were
estimated from behaviour surveys among idu in the Netherlands. Sexual behavior was not
included, because transmission between sex partners was not observed.The average transmission probability per contact was assumed to be 0.0 I.The course of the epidemic after
the introduction of an index case or after a change in risk behavior was observed over 10
years under various assumptions on contact patterns and infectivity
Results: With contact patterns as presently observed in Dutch idu's and constant infectivity
the rise in prevalence was extremely slow with a large stochastic variability Once the
prevalence had reached a certain level, it remained high for a long time even after a change
in risk behaviour.Variable infectivity with a peak of high infectiousness at the beginning of an
infection can lead to a rapid initial rise of prevalence. Spread through the relatively stable
network of friends and buddies contributed to a larger extent to the epidemic than transmission in needle sharing between strangers. Risk reduction in stranger relationships was
less effective in reducing HIV-incidence than was risk reduction in friend relationships.
Conclusions: For present levels of risk behaviour observed in the Netherlands a future rise
in HIV-prevalence cannot be excluded. A plateauing of prevalence as observed in most idu
populations is not necessarily an indication of a change in risk behavior, but may be due to
the variability in infectivity. Prevention education should pay attention to risks of transmission in the close social environment of idu.
M. Kretzschrnat Institut fOr Medizinische Biometrie, University ofTubingen,Westbahnhofstr.
55, D-72070 Tubingen, Germany Telephone: [phone]; Fax: +49-7071- I 295219;
email: [email protected]
Tu.C.57 I
ESTIMATING LEVELS AND TRENDS OF THE HIV/AIDS EPIDEMIC USING
MORTALITY DATA
Verdecchia A.*, Mariotto A.**, Conti S.*, Rosenberg PS.~. *Istituto Superiore di Sanita,
Roma, Italy; ** National Institutes of Health, Bethesda, MD, USA
Objective:To provide a method for reconstructing HIV infection epidemic fr-om mortality
data in large countries where collecting surveillance data on AIDS is unpractical or not
feasable, i.e. most Asian, African and Latin American countries. Furthermore, the method
may prove useful in the United States where the 1993 expansion of the AIDS surveillance
definition has distorted trends in AIDS incidence.
Methods: An age, period and cohort back-calculation method, already developed and
applied to reconstruct the epidemic of HIV infection in Italy is adapted to use mortality
data, instead of AIDS counts, as input data.
Results: A reconstruction of AIDS/HIV epidemic was obtained in Italy by using mortality
data 1983 - 1991. Estimated incidence of AIDS satisfactorily reproduced the distribution of
notified AIDS cases in Italy HIV infection was estimated to peak in 1986 with about 14,000
new infections. Also, prevalence of HIV infection in I99 I was very close to the value of
50,000, previously estimated using back-calculation. Back-calculation estimates for the United
States based on AIDS incidence and mortality data will be compared.
Conclusions: Mortality data were proved to produce unbiased estimates of HIV/AIDS epidemic size both in Italy and in the United States. Estimates are unaffected by changes in the
disease definition criteria. Conversely mortality data do not allow estimates by risk category
and for recent years, due to fairly long delay of reporting involved. In countries in which systematic and complete registration of AIDS diagnosis is problematic or even not feasible,
mortality data which are far more easily available can offer a valid alternative to support
surveillance activities.
A.Verdecchia,Viale Regina Elena, 299, 0016 I Roma, ItalyTel: [phone]:
Fax: [phone]; email: [email protected]

O
Q)
0
V
c
I3
C
0
Q)
cOC
0
U
Ce
0
cCQ)
c
2
X
252

Year of Age in 1995
birth (years)
> 1970 < 25
1966-1970 25-29
1961-1965 30-34
< 1960 > 35
Total

1991
% N
0.00 674
0.51 2515
0.52 3995
0.28 _ 4248
0.41 I 11432

1993
% N
0.60 1172
0.63 2994
0.55 3476
0.48 2897
0.56 10539

1995
% N_
0.44 2063
0.62 3404
0.51 3126
0.33 2132
0.49 10725

Conclusions: No significant trends in HIV prevalence (overall and by outcoe) were
observed in women at outcome of pregnancy and HIV prevalences by age group were stable. Analysis by birth-year cohort suggests that the highest HIV incidence was in the
youngest age groups (<29 years). Data on fertility and mortality in HIV infected women will
be used to give a better estimate of HIV incidence in childbearing women.
V. Batter: European Centre for the Epidemiological Monitoring of AIDS, 14 rue du Val d'Osne,
94415 Saint-Maurice Cedex, FranceTel: (33-I) 43 96 65 45; Fax: (33-I) 43 96 50 81;
Email: [email protected]
Tu.C.563
SEXUALLY TRANSMITTED DISEASES (STDS), CONTRACEPTIVE USE,AND RISK OF HIV
INFECTION AMONG WOMEN IN DAR-ES-SALAAM,TANZANIA- A PROSPECTIVE STUDY.
Kapiga, Saidi, Lyamuya EF, Lwihula G, Hunter DJ*. Muhimbili Medical Centre, Dar-es-Salaam,
Tanzania and * Harvard School of Public Health, Boston, MA.
/-Objectives: To determine the incidence of HIV infection in Dar-es-Salaan and assess its
association with other STDs, contraceptive practice and sexual behavior.
Methods: HIV-ve women (N= I 370) were enrolled in a prospective cohort study at 3 family planning clinics. During the follow-up period (I0/92 to 07/95), information about sexual
behavio, contraceptive use, and HIV and STD incidence was collected.