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

152 Abstracts WeOrD1353-WePpD2103

XIV International AIDS Conference

54%), whilst more than 1 in 4 RAs were followed by the development of medical and non-medical interventions, or policy impact and change (25/83; 30%).
Furthermore, response development following RA on substance use has been
identified in 50 countries.
Conclusions: RAs have achieved global spread and can facilitate important public health outcomes. However, to realise RA's potential on a wider scale, investment has to be made in the approach's evidence and knowledge base: in short,
there is a need to introduce a culture of learning and reflection into a methodology
primarily premised on rapidity and pragmatism.
Presenting author: Chris Fitch, Imperial College, Reynolds Building (3rd floor),
Charing Cross Campus, St Dunstans Road, London, United Kingdom, Tel.: +44
207 594 0822, Fax: [phone], E-mail: [email protected]
WeOrD 353 Measuring relationship power: adaptation of the
SRPS for South Africa
R.K. Jewkes, M. Nduna, RN. Jama, J.B. Levin. Medical Research Council,
Gender & Health Group, Medical Research Council, Private Bag X385, Pretoria
0001, South Africa
Introduction: HIV preventive interventions are increasingly focusing on gender
dynamics in relationships, creating new demands to develop measures for evaluation tools. Pulerwitz et al (2000) developed a 23-item Sexual Relationship Power
Scale (sub-scales: relationship control and decision-making dominance) in the
United States. This paper describes its adaptation and testing in South Africa
Methods: Based on our knowledge of gender issues in South African relationships, the relationship control sub-scale was used alone. Four out of its1l5 items
were removed and one added. The resulting 12-item scale was tested on young
people (70 male and 51 female) in a pilot evaluation of the 'Stepping Stones'
intervention with pre- and post-intervention interviews.
Results: The sub-scale performed well on reliability testing. Cronbach's alpha for
the sample of men was 0.57 and for women was 0.813. The mean score for the
scale for women was 39.84 (SD 8.45) and for men was 43.07 (SD 5.51) out of a
maximum of 48. A higher score denoted more equitable relationships. At baseline
women who had used a condom on last sexual intercourse had a higher mean
score than non-condom users (45.6 v. 35.9, p<0.0001). The same association
was found for men (44.1 v. 41.0, p=0.027).
The intervention was effective in increasing women's power in their relationship
for women participants (p=0.022), but the change for men was not significant
(p=0.34). This may have resulted from men exaggerating the equitability of their
relationships at baseline.
The 5 women and 7 men who reported using a condom during last sex on follow
up, but not at baseline, improved their sexual relationship power score more than
those who did not use a condom (for women 4.8 v 2.1 and for men 2.3 v. -1.5).
This change was non-significant (p=0.109), perhaps due to the small numbers.
Conclusion: The adapted scale was a useful tool and associations between
greater gender equity and condom use were as expected.
Presenting author: Rachel Jewkes, Gender & Health Group, Medical Research
Council, Private Bag X385, Pretoria 0001, South Africa, Tel.: [phone],
Fax: [phone], E-mail: [email protected]
WeOrD1354 Informed consent in a phase II trial of the potential
microbicide, CarraguardTM: lessons learned
B. Friedland1, C.E. McGrory1, S. Magwaza2, M. Marumo3, A. de Kock2,
M.H. Sebola3, N. Coetzee4. 1Population Council, Population Council, 1 Dag
Hammarskjold Plaza, New York, New York 10017, United States; 2fUniversity of
Cape Town, Cape Town, South Africa; 3Medunsa, Medunsa, South Africa;
4University of Cape Town, Caape Town, South Africa
Background: Ensuring informed consent and voluntary participation in research
is challenging, particularly when conducting complex trials in vulnerable populations. To address this, the informed consent form and procedures were evaluated
before, during, and after a Phase II expanded safety trial of the potential microbicide, Carraguard, in South Africa.
Methods: Independent researchers conducted focus groups and in-depth interviews using guides developed with Population Council researchers. Before the
trial, the consent form was pilot-tested in the recruitment communities. Study staff
then made significant revisions to the form and developed tools to aid comprehension (flip chart, recruitment script). Once the trial began, further research led the
staff to modify existing materials and develop a study booklet with illustrations to
further explain the study.
Results: Although the original consent form was written at an 8th-grade reading level, respondents had difficulty reading it and were overwhelmed by the
amount of information. Even after significant revision of the form (simpler language; clearer lay out), many respondents could not articulate the primary purpose of the study (safety vs. effectiveness). After the booklet was introduced,
nearly all women understood the study purpose and some difficult research con
cepts, though "placebo" was still confusing to many Participants also reported
that one-on-one counseling sessions aided understanding and reaffirmed voluntary participation.
Conclusion: Informed consent is not a one-time event. Making informed consent
meaningful requires a range of approaches and tools. Researchers must invest
in ongoing assessment of the informed consent process and adapt it to meet the
needs of study participants.

Presenting author: Barbara Friedland, Population Council, 1 Dag Hammarskjold
Plaza, New York, New York 10017, United States, Tel.: [phone], Fax: [phone], E-mail: bfriedland @ popcouncil.org
WeOrD1355 Community, research and ethics: Participatory
approaches to assessing HIV and STD burden in
urban India
S. Sivaram1, A.K. Srikrishnan2, K.G. Murgavel2, K. Balakrishnan2, K.H. Mayer3,
D.D. Celentanol, S. Solomon2, U.S.A. NIMH Collaborative HIV/STD Prevention
Trial4. 1Johns Hopkins University Bloomberg School of Public Health, 205 Lee
Street, Apt 513, Gaithersburg, MD 20877, United States; 2YRG CARE, Chennai,
India; 3Brown University Providence, United States; 4USA, Bethesda, United
States
Background: Assessing the burden of HIV and sexually transmitted diseases
(STDs) in a community is an key step in any community-based effort to prevent
the spread of HIV However, the challenge is to conduct assessment while protecting the human rights of the participants and providing follow-up care in the
event a person is positive with HIV. This paper describes how we addressed this
challenge as part of a large multi-site randomized controlled HIV prevention trial
in the city of Chennai in south India.
Methods: We conducted assessment activities in 30 trial communities through
health fairs. In these activities, we measured both behavioral and biological
risks. Community leaders and volunteers helped organize these camps. Sixty-five
adults from each community between the ages of 18 and 40 years were randomly
selected from a project census list. We administered informed consent to all selected participants. A week to ten days after the camp, all participants and health
camp attendees who provided biological samples were given their test results in
a confidential manner. We also facilitated follow-up care and referrals.
Results: Of the 1950 selected respondents, 1839 persons were given health
cards to attend the fairs. Of these, 91% of the participants took part in data collection activities. Forty three percent of the participants were men and 56% were
women. Of the men, 99% provided biological samples - serum and urine. Of
the women, 98% provided serum and 77% provided vaginal swabs. Community
participation was facilitated by repeated visits to the community and offering clarifications in the informed consent process.
Conclusions: Informed consent is a multi-step process and constant interaction
with the community before assessment not only achieves higher participation but
also increases credibility of the study. These are important precursors to planning
the trial and designing the intervention.
Presenting author: Sudha Sivaram, 205 Lee Street, Apt 513, Gaithersburg, MD
20877, United States, Tel.: [phone], Fax: [phone], E-mail: [email protected]
WePpD2103 Decline in HIV risk behavior among injecting drug
users (IDUs) in the AIDSVAX~B/E vaccine trial in
Bangkok, Thailand
S. Vanichsenil, F. van Griensven2, B. Phasithiphol3, PR Pitisuttithum4,
D. Kitayaporn4, K. Orelind5, J. Tappero2, K. Choopanyal. 'Bangkok Vaccine
Evaluation Group, BVEG office, Taksin Hospital, Bangkok, Thailand; 2Thai
MOPH - US CDC Collaboration, Bangkok, Thailand; 3Bangkok Metropolitan
Administration, Bangkok, Thailand; 4Mahidol University, Bangkok, Thailand;
5VaxGen, Inc., Brisbane, CA, United States
Background: Concern exists that participation in HIV vaccine efficacy trials may
lead to increased risk behavior as a result of undue beliefs about protection from
HIV infection. We evaluated changes in risk behavior among IDUs participating
in the AIDSVAX~B/E vaccine trial in Bangkok, Thailand, in relation to participant
characteristics and beliefs about study-arm assignment and vaccine efficacy.
Methods: From March 1999 through August 2000, 2,545 IDUs were enrolled.
Participant characteristics, beliefs and risk behavior data were collected at baseline, and every 6 months thereafter. Risk reduction counseling is provided at every
study visit. Generalized estimation logistic regression analysis was used to study
trends in risk behavior and associations with participant characteristics and beliefs.
Results: 93.4% of participants were male, the median age was 26 years and
95.0% had at least primary education. At baseline, 62.3% received methadone
detoxification and 29.7% maintenance. From baseline to 12-month follow-up, reported injection drug use decreased from 93.8% to 66.6% (p <0.001) and reported needle sharing decreased from 33.0% to 17.5% (p <0.001). Participants
receiving methadone treatment were more likely to report injection drug use than
those who were not (p <0.001); however, if injection drug use was reported,
those receiving no methadone treatment were more likely to report needle sharing. Younger age and injecting less than weekly at baseline were associated with
more rapid declines in injection drug use (p <0.01). Beliefs about study-arm assignment and vaccine efficacy, educational level and gender were not related to
levels of risk behavior and behavioral change over time.

Conclusion: Injection drug use and needle sharing decreased significantly over
the first 12 months of the trial. No evidence was found of increased risk behavior
in relation to beliefs about study-arm assignment and vaccine efficacy.
Presenting author: Suphak Vanichseni, BVEG office, Taksin Hospital,
543 Somdet Chaopraya Rd., Klongsan, Bangkok, 10600, Thailand, Tel.:
[phone], Fax: [phone], E-mail: kachitc @ dmc.inet.co.th