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

104 Abstracts WePeC6083-WePeC6086

XIV International AIDS Conference

HIV-1. We analyzed the relationship between mode of delivery [SCS, nonscheduled cesarean section (NSCS), and vaginal (V)] and postpartum HIV-1 disease
progression among North American women enrolled in WITS.
Methods: Disease progression (changes in CD4 percent (%), plasma HIV RNA,
CDC disease category, and death) in 1345 HIV-1-infected women was analyzed
according to mode of delivery (SCS: 108; NSCS: 192; V: 1045). General estimating equations were used to compare changes in adjusted mean CD4% and HIV-1
RNA counts by mode of delivery over the 18 months following delivery; Cox Proportional Hazard models assessed differences in time to first CDC Class C event
or death from any cause over the 10 years following delivery
Results: After adjusting for age, antiretroviral therapy, and antepartum CD4%,
women who underwent NSCS had a small, transient increase in CD4% from antepartum to 2 months postpartum that was not present in the SCS or V groups.
After that point, the rates of decline in CD4% were not significantly different across
the study groups (p=0.85). Neither category of c-section was associated with increased HIV-1 viral loads during the postpartum period. Over time, 155 women
progressed to a first Class C event and 93 died. When compared to vaginal delivery, cesarean delivery was not associated with progression to a CDC class C
event; relative risks (RR) were: SCS: 0.62 (95% confidence interval (CI), 0.31 -1.26, and NSCS: 0.69, CI=0.41-1.15. Similarly, cesarean delivery was not associated with an increased risk for all-cause mortality [RR for scheduled c-section
was 0.54, CI = 0.18-1.55, and for nonscheduled RR= 1.07, CI = 0.58-1.98].
Conclusion: These findings suggest that mode of delivery is not associated
with postpartum HIV-1 disease progression as assessed immunologically, virologically, or clinically.
Presenting author: Elena L. Navas-Nacher, School of Public Health, Division of
Epidemiology and Biostatistics, 808 N. Judson Avenue, Suite 4A, Evanston, Illinois 60202, United States, Tel.: [phone], Fax: [phone], E-mail:
elenanacher@ hotmail.com
WePeC6083 Death certificate as a marker of underregistration
of AIDS cases
K.R.V Lemos', J.G. Valente2. 'Rio de Janeiro State Health Departament, Rio
de Janeiro - RJ, Brazil; 2Rio de Janeiro State University, Rio de Janeiro, Brazil
Background: Studies have been performed in various countries to evaluate the
quality and degree of coverage of the Epidemiological Surveillance System. An
important indicator of flaws in the Epidemiological Surveillance System is the proportion of deaths from AIDS not registered previously as cases of the disease.
Method: This study aimed to estimate and evaluate the underregistration of AIDS
cases that evolved to death. From 1991 to 1995, the Mortality Information System registered 9,213 adult deaths from AIDS, and these were related by name,
date of birth, and date of death to 15,505 AIDS cases registered by the Disease
Registration Information System from 1982 to 1996.
Results: This procedure showed that 51.9% of AIDS deaths registered during
this period had not been registered previously as AIDS cases as they should
have been. Univariate and multivariate analysis showed that women had a greater
probability of underregistration than men (OR = 1.27). Lower probability of registration was observed in individuals with fewer years of schooling, and illiterates
had twice the probability of not being registered as compared to individuals with
university degrees. Deaths in health care facilities classified as private units had a
greater chance of not being registered than those occurring elsewhere, mainly in
facilities classified as public reference units (OR = 2.58). AIDS deaths in the city
of Rio de Janeiro had a greater chance of underregistration than those in other
cities (OR = 2.20).
Conclusion: Approximately 50% of AIDS cases that evolved to death developed
AIDS and died without being reported. Identification of the AIDS case through
the death certificate proved to be an important indicator of underregistration of
the disease, highlighting the importance of aggregating information available at
various levels.
*Article published in Cad. Saude Publica, Rio de Janeiro, 17(3): 607-616, MayJune, 2001.
Presenting author: Katia Lemos, Rua Capitao Rezende 35/308, Meier, Rio de
Janeiro - RJ, CEP: 20780-190, Brazil, Tel.: [phone], Fax: [phone], E-mail: [email protected]
WePeC6084 The impact of AIDS on early mortality. Potential
years of life lost in the state of Rio de Janeiro from
1990 to 1998
K.R.V. Lemos, M.A. Sole-Pla. Health Department of Rio de Janeiro State, Rio de
Janeiro - R J, Brazil
Background: Morbidity and Mortality data show that AIDS has been affecting
mostly young adults in Rio de Janeiro. A relevant indicator of premature mortality
can be the Loss of potential years of life (PYLL).
Methods: Premature mortality was estimated from PYLL of 1-70 year old per
sons. Investigated causes of death were AIDS, cerebrovascular disease, acute
myocardial infarcts, transport accidents, homicides, tuberculosis, as well as prostrate and breast cancers. The method consisted of totaling the number of deaths
in 5-year age brackets (from 1 to 70 years) and multiplying them by the years of
life remaining until age 70. Data referred to deaths occurring from 1990 to 1998
among residents of the State of Rio de Janeiro, as registered in the Mortality
Information System (MIS).

Results: From 1990 to 1998 there were a total of 12,865,299 PYLL-70 in the
State of Rio de Janeiro, of which 547,045 (4.3%) were due to AIDS. Among
women, the PYLL due to AIDS increased 206% between 1990 (5,394 years) and
1998 (16,507 years). The increase in PYLL due to AIDS was much smaller in men
(13%, from 32,055 in 1990 to 36,407 in 1998). In 1990, AIDS was the smallest
of the compared causes of PYLL among women, but by 1998, the PYLL due to
AIDS was larger than that due to homicides, breast cancer, transport accidents
and tuberculosis. In men, the PYLL due to AIDS was already larger than that due
to prostate cancer and tuberculosis in 1990, but, in 1998, it also became larger
than the PYLL due to transport accidents.
Conclusion: With a prime incidence in younger age groups, the share of AIDS
in early mortality can be observed through PYLL from the disease, which has
increased, especially among women. The rise in the number of early deaths
highlights the epidemic's transcendence, affecting segments of the population
in their productive and reproductive years and generating major economic and
social costs.
Presenting author: Katia Lemos, Rua Capitao Rezende 35/308 - Meier - Rio de
Janeiro - RJ - CEP: 20780-190 - Brazil, Brazil, Tel.: [phone], Fax: +55
21 2533-4226, E-mail: [email protected]
[WePeC6085 HIV incidence in Peace Corps Volunteers (PCVs)
and interventions to reduce transmission
A.R. Gerber, C.M. White, R.H. Banks, D. Prybylski, A. Picken, E. Von Arx,
S.M. Weinberg. Peace Corps, 1111 20th St. N. W, Washington, DC 20526,
United States
Background: To accomplish its mission of grassroots, person-to-person international development, combined with cross-cultural exchange, approximately 6,500
PCVs a year live and work in 80 countries worldwide (including Africa, Latin America, Central Asia, the Middle East, and Pacific Islands). All PCVs receive HIV
counseling and testing prior to beginning service. During pre-service training, all
PCVs receive HIV prevention education that includes a videotape of former PCVs
who became HIV-infected during service and who share their experiences in order
to serve as role models and provide HIV prevention messages. During service,
PCVs are cared for by medical officers who have been trained in interpersonal
HIV prevention counseling. HIV post-exposure prophylaxis is offered after sexual or needlestick exposures. At any time during service and at close-of-service,
voluntary HIV counseling and testing is available and encouraged.
Methods: Data collected include demographics (gender and age), country of service, and risk exposures (among seroconverters).
Results: Since 1993, 21 PCVs have become infected with HIV (59,000 PCVyears of observation). Overall incidence was 3.6 per 10,000 PCV-years. Incidence
in women (4.2 per 10,000 PCV-year) was greater than incidence in men (2.8).
Age-specific incidence was highest in 30-39 year olds (16.0). By region, the highest incidence (7.9) and greatest number (17 [81%]) occurred in Africa. Twenty
(95%) infections were acquired through unprotected sex.
Conclusions: PCVs serve in many countries in the developing world, are sexually active, and are at risk for HIV infection. Peace Corps promotes a strategy to
prevent HIV among PCVs through education, prevention counseling, and health
services. PCVs also share skills and conduct projects in HIV education and prevention with country counterparts. These interventions reduce but have not eliminated HIV infections among PCVs, and support worldwide HIV prevention efforts.
Presenting author: A. Russell Gerber, 1111 20th St. N.W., Washington, DC
20526, United States, Tel.: [phone], Fax: [phone], E-mail:
[email protected]
iWePeC6086 Interpreting neonatal seroprevalence HIV data:
Importance of differential fertility
S. Cliffe1, M.L. Newell', M. Cortina-Borja', A. Nicoll2. 'Institute of Child Health,
London, United Kingdom; 2Communicable Disease Surveillance Centre,
London, United Kingdom
Background: To explore differential HIV and fertility patterns amongst subgroups of women (born in sub-Saharan Africa (SSA), Injecting Drug Users (IDUs)
and Rest) to improve estimates of HIV infection among the general population in
Great Britain using neonatal seroprevalence data.
Methods: HIV risk was estimated using anonymous neonatal seroprevalence
data and national reports of diagnosed HIV infected pregnant women. Live birth
rates were estimated from routine population data and results from the first National Survey of Sexual Attitudes and Lifestyles (NATSAL), adjusting for area of
residence and age group. Results were validated using data from a prospective
European cohort of HIV infected pregnant women.
Results: HIV prevalence was higher in both women born in SSA (range 0.77 -2.16%) and IDUs (range 0.37-1.79%) compared to women categorised as Rest
(range 0.01%-0.04%). Risk of HIV by sub-group varied greatly according to UK
area of residence at time of delivery. Fertility was also higher in women born in

SSA and IDUs (OR1.30 and 2.06 respectively using population data and OR1.69
and 1.39 respectively using the European cohort data). Whilst sensitivity analyses indicated that live birth rates amongst IDUs could vary between 108.7-163.0
per 1000 women with small adjustments to survey data, drug users were consistently more likely to have a live birth than women born in SSA and the rest of the
population. IDUs were also more likely to have had a termination.
Conclusions: Results show that fertility differentials in population sub-groups at