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

XIV International AIDS Conference

Abstracts WePeC6191-WePeC6194 129

United States, Tel.: +[phone], Fax: +[phone], E-mail: tbabikian06g
@ univ.llu.edu
WePeC6191 Behavioural surveillance among intravenous drug
users (IDUs) out-of-treatment in Barcelona (Spain),
1993-2000
K. Perez1, A. Rodes2, M. Meroho3, J. Casabona 4. 'Centre for Epidemiological
Studies on AIDS of Catalonia - Hospital Universitari Germans Trias i Pujol,
Center for Epidemiological Studies on AIDS of Catalonia (CEESCAT), Hospital
Universitari Germans Trias i Pujol, Ctra de Canyet sn, 08916 Badalona, Spain;
2Secretaria del Plan Nacional sobre Sida, Madrid, Spain; 3Ambit Prevenci6,
Barcelona, Spain; 4CEESCAT Badalona, Spain
Background: Catalonia, a Spanish autonomous region with 6 million inhabitants,
had in 2000 an AIDS incidence rate of 81.8 per 106. In 1993 behavioural monitoring of intravenous drug users (IDUs) was introduced as part of its integrated
HIV/AIDS surveillance system. This paper aims to compare HIV prevalence and
risk behaviours of four crossectional surveys among IDUs out of treatment carried out in 1993 (n = 200), 1996 (n = 300), 1998 (n = 214) and 2000 (n=300) in
Barcelona.
Methods: Subjects were street recruited by trained ex-IDUs interviewers. After
informed consent, subjects responded to a structured interview, adapted from
the WHO collaborative group, which explored risk behaviours in the previous six
months. Volunteered saliva samples were also anonymously collected to test for
anti-HIV antibodies. When comparisons are statistically significant are presented
in the same order 1993, 1996, 1998 and 2000 respectively.
Results: Of the 1014 participants, 75% were over 25 years old, and 70% were
men. Although an overall reduction of direct sharing was observed over the four
surveys (30,2%, 10,7%, 8,4% and 5% accepted used syringes daily or weekly,
p<0,001), indirect sharing as front/back loading increased (67,5%, 72,3%, 78,5%
and 77,6% p=0,028). Consistent condom use increased with occasional partners
(36,1%, 63,4%, 50% and 64,8% p=0,001) but remained low around 28% with
steady partners. Nearly half of the participants had a non-injector steady partner.
HIV Prevalence was high in the four point surveys, showing a decrease only from
1996 to 1998 but stabilized in 2000 (51%, 50%, 39,7% and 39,3%).
Conclusions: While direct sharing decreased over the years, probably due in
part to harm reduction interventions, indirect sharing, unprotected sex, especially
with steady partners, and HIV prevalence are is still very high. The role of IDUs in
the current heterosexual epidemic needs to be evaluated and specifically tailored
interventions to different sub-groups are needed.
Presenting author: Katherine Perez, Center for Epidemiological Studies on AIDS
of Catalonia (CEESCAT), Hospital Universitari Germans Trias i Pujol, Ctra de
Canyet sn, 08916 Badalona, Spain, Tel.: [phone], Fax: [phone]
91, E-mail: [email protected]
WePeC6192 Programmatic use of behavioural surveillance in
Central America
M.M.G. Melendez. PASMO, 6a. Avenida 16-32, zona 19, Colonia San Francisco
I, Guatemala, Guatemala, Guatemala
Background: Two rounds of behavioral surveillance surveys for Female Sex
Workers (FSW) were conducted in Central America (Guatemala, El Salvador,
Honduras, Nicaragua and Costa Rica) in 1997 and 2000, with the objective of
measuring the impact of (Panamerican Social Marketing Organization) PASMO's
HIV prevention activities and to inform future program design.
The sample consisted of 2000 respondents, 400 from the capital city of each
country. Given the difficulty of defining a universe, convenience sampling methodology was used, based on mapping of sites where FSW congregate (brothels,
bars and streets) and randomly selecting a previously estimated number of respondents from randomly selected sites in each study location.
Results: Analyzed for the region (and all data adjusted for sample differences
in age, education level, and location), access to information on condoms from
TV increased from 50% to 70% (p value=.000). Reported risky vaginal sexual
activity (without using a condom) in last month, showed statistically significant
decrease (p values =.000). Reported condom use in the last encounter with any
client, showed statistically significant change from 71% to 77% (p value = 0.000).
Reported condom use in the last encounter with her currently partner, showed no
statistically significant change from 27% to 21% (p value = 0.008), but showed
an increase in the risky sexual activity with their partners. The data set for 2000
was cross tabulated for CSW participating in PASMO interpersonal activities and
those who had not, and correct condom use skills were demonstrated by 46% of
participants compared with 29% (p value=.000) for non-participants. Data analysis by country showed considerable variation.
Conclusions: Activities done by PASMO need to be scaled up. Special focusing
will be necessary for their currently partners. Qualitative research is needed to
identify motivational factors to incorporate findings of all research in the future
modules.

Presenting author: Giovanni Melendez, 6a. Avenida 16-32, zona 19, Colonia
San Francisco I, Guatemala, Guatemala, Guatemala, Tel.: +502 366 - 1557 / 9,
Fax: +502 366 - 1567, E-mail: [email protected]

WePeC6193 ISurveillance of sexual risk behavior among HIV
testers in Santos, Brazil: 1996-1999
K. Alves1, M.E. Falcao2, T.M.S. Vieira2, H. Hernandez3, M.M. Caseiro4,
R.S. Diaz4, M.C.A. Sucupira5, G. Rutherford3, K. Page-Shafer3. 'Schoolof
Public Health, Univ California, 1626 Pierce st, #420, San Francisco, CA, 94115,
United States; 2Centro de Testagem e Aconselhamento, Santos, SR Brazil;
3Center for AIDS Prevention Studies, Univ of California, San Francisco, CA,
United States; 4Federal Univ of S&o Paulo, Sao Paulo, SR Brazil; 5lnstituto de
Ciencias Biomedicas-Sao Paulo Univ, Sao Paulo, SP, Brazil
Background: Santos, Brazil has consistently reported high prevalence of HIV
and AIDS incidence and has been the focus of intense prevention and intervention campaigns. The anonymous testing site (ATS) in Santos serves multiple highrisk populations and has a substantial prevalence and incidence of HIV among its
clients. We assessed changes in HIV risk behavior in ATS clients from 1996 to
1999.
Methods: We abstracted demographic and self-reported HIV risk behaviors from
pre-HIV test counseling forms from 7,275 subjects presenting to the ATS. We analyzed the proportion of clients reporting more than 2 sexual partners over the
last 12 months and inconsistent (sometimes or never) condom use.
Results: HIV risk behavior was stable from 1996 to 1998. In 1999 there was a
sharp decline in number of sexual partners and inconsistent condoms use. That
trend is present in both males and females and by sexual orientation. There were
no differences when data were stratified by history of previous HIV testing.
1996 1997 1998 1999
2+ Partners 50.3% 49.8% 47.2% 28.1%
Inconsistent condoms use 74.5% 76.5% 78.1% 69.9%
Both 34.5% 35.2% 34.3% 16.6%
Conclusions: Report of high-risk behavior for HIV infection had been relatively
stable from 1996 to 1998; however, in 1999 it decreased significantly This trend
occurred in females, male heterosexuals and male homosexuals. It may represent a widespread decrease in HIV risk as a result of sustained Brazilian HIV
prevention efforts including prevention campaigns; testing, counseling and harm
reduction. While there may be underreporting of risk behavior at ATS sites, our
analyses found no significant explanations for this sudden trend. These data are
notable since many countries that have made antiretrovirals available have observed increases in risk behavior. Further behavioral surveillance will help to determine if risk reduction is maintained.
Presenting author: Katia Alves, 1626 Pierce st., #420, San Francisco, CA,
94115, United States, Tel.: [phone], Fax: [phone], E-mail:
[email protected]
WePeC6194 Comparison of computer-assisted interviewing
(CAI) modes for behavioral surveillance among
interviewees in two rural villages in Zimbabwe
D. Kasprzyk1, G.B. Woelk2, D.E. Montahol, N.I.H. NIMH Collaborative HIV/STD
Prevention Trial3. 1Battelle, CPHRE, Battelle, CPHRE, 4500 Sand Point Way
N.E., Suite 100, Seattle, WA, 98105-3949, United States; 2Department of
Community Medicine, Medical School, University of Zimbabwe, Harare,
Zimbabwe; 3National Institute of Mental Health, Washington, DC, United States
Background: Self-administered questionnaires (SAQs) increase reporting of
sensitive behaviors; ie, sexual, drug, alcohol use. Audio Computer Assisted SelfInterviews (ACASI) have been promoted as an SAQ that does not rely on reading
level. Methods research shows individuals, especially youth, reveal riskier behaviors using ACASI. Most ACASI methods studies have been conducted in the
US; few in developing countries. This study assessed method acceptability and
whether ACASI or Computer Assisted Personal Interviews (CAPI) engendered
greater revelation of sensitive behaviors among individuals in rural villages in Zimbabwe.
Methods: Interviewers enumerated households for residents aged 16-30; randomly selected a respondent; gave informed consent. 81 individuals were interviewed twice in a 2-3 day period and randomly assigned to receive ACASI or CAPI
first. Interviewers then reinterviewed each with the opposite interview mode. Individuals assessed acceptability of each mode.
Results: Comparison of ACASI and CAPI responses to sensitive questions: lifetime, sexual partners in last 12/3 months, partner type, condom use, STD/HIV
testing, commercial sex, and alcohol use showed no significant differences between interview mode. All individuals, regardless of mode, revealed the same frequency of sensitive behaviors. Significant differences only occurred in two questions. Individuals revealed more STD symptoms using ACASI, and more injection
exposure in CAPI. Debriefings showed high acceptability for both modes of interviewing, but ACASI was felt to be more private.
Conclusions: The feasibility of conducting CAIs in rural villages in Zimbabwe was
established. Even though individuals felt ACASI was more private, they revealed

most sensitive behaviors at equal frequency using CAPI. Skilled interviewers establishing rapport can achieve similar results to ACASI. Rural villagers liked the
CAIs. Our study showed CAIs are feasible and well liked by rural Zimbabweans.
Presenting author: Danuta Kasprzyk, Battelle, CPHRE, 4500 Sand Point Way
N.E., Suite 100, Seattle, WA, 98105-3949, United States, Tel.: [phone],
Fax: [phone], E-mail: [email protected]