Abstract Book Vol. 1 [International Conference on AIDS (16th: 2006: Toronto, Canada)]

Recommendations: Multi-method triangulated analyses have proven to be
a vital approach to evaluating the effectiveness of HIV prevention programs
in settings where more rigorous controlled approaches cannot be used.
The inclusion of a detailed 'exposure to intervention' module in surveillance
questionnaires is required for this purpose and is highly recommended.
TUPE0325
Routine testing for HIV in Botswana: results for the
first two years
T.W. Steen, F.d. Gomez, M.G. Anderson, W. Jimbo. Ministry of Health, AIDS
Prevention and Care, Gaborone, Botswana
Background: Routine HIV testing (RHT), without formal pretest counseling,
was introduced in Botswana in January 2004. Main objectives for this policy
change were to make more people aware of their HIV status, thereby
facilitating behavioral change and early assessment for treatment; and stigma
reduction. Main indications for RHT are: symptoms of HIV/AIDS, pregnancy,
STI, attendance for general medical examination. Patients who opt out, may
be offered pretest counseling. The public was informed of the policy change
through mass media and health education lessons. We report results from the
first two years of routine testing.
Methods: Testing is by ELISA or rapid tests, and PCR in infants. All health
facilities send summary monthly RHT reports to MOH for computer entry and
data analysis.
Results: In 2005, 177,831 persons were offered testing and 157,894 (88.8%)
were tested. 60,846 (42,235 females and 18,511 males) persons were tested
through RHT in 2004 vs 109,713 females and 48,181 males in 2005. The rate of
RHT in the population in 2004 was 3.6/100 population as compared to 10.5/100
in 2005. In 2004, 41.8% (40.3% of females vs 45.6% of males) tested positive
as compared to 31.5% (30.4% of females, 34.2% of males) testing positive
in 2005. Reasons for testing in 2005 were: Patient's wish 50.1%, pregnancy
24.7%, medical examination 6.5%, clinical suspicion 5.7%, STI 2.3%, rape
1.0%, TB 0.4%, needlestick injury 0.2%, other 9.1%. In 2005, 90.2% of
females versus 85.8% of males who were offered testing, accepted. Testing
uptake was lowest in the age group 15-19 years (87.0%), uptake for females
15-19 was 89.4% vs 75.6% for males.
Conclusions: Routine HIV testing in Botswana has been introduced without
major problems, and we think the main objectives have been attained. Testing
activity increased significantly from 2004 to 2005. However, the quality of
reporting needs to be improved.

TUPE0327
Disscusing accuracy of indicators used in sentinel
HIV surveillance in Rosario, Argentina
M. Chavero1, A. Tomisani2, C. Moyano2, R. Rail2. 'Secretary of Public Health,
Rosario AIDS Program, Rosario, Argentina, 2Secretary of Public Health,
Epidemiology, Rosario, Argentina
Background: Sentinel HIV Surveillance in pregnant women attending
antenatal clinics is one of the main and more expanded methods for estimating
trends in HIV epidemic and is used as an approximation of Prevalence among
the adult population. Traditionally, the positivity rate is measured through the
indicator "Prevalence in Pregnants'", defined as the proportion of HIV tests that
are positive among all the HIV tests carried out in pregnants, over a given
period of time.
Sentinel surveillance has been carried out since 1998 in Rosario city (1 million
inhabitants), where 40% of the approximately 16.000 annual deliveries occurs
at the two antenatal clinics of Municipal Health Public System (MHPS). As test
offering has increased year by year in MHPS and is performed twice in many
pregnants, it represents a problem for prevalence estimation by enlargening
the denominator. Besides that, for women diagnosed as HIV+ previously, test
is not performed again, diminishing numerator. Another indicator has been
developed and compared with the "traditional" one.
Methods: Two indicators were used and compared for the period 1998 -
2005:
a) Prevalence in pregnants: defined as the proportion of HIV tests that are
positive among all the HIV tests carried out in pregnants, anually.
b) Prevalence in parturients: defined as the proportion of parturients HIV+
among total number of parturients, anually.
Results: Indicator "a" has dropped from 0.53% in 1998 to 0.11% in 2005
(average: 0.21%).
Indicator "b" has remained relatively stable: 0.79% in 1998 and 0.82% in 2005
(average: 0.74%).
Conclusions: In many settings the indicator "a" may generate epidemiological
misconceptions by underestimating prevalence in pregnants and so on in the
adult population. Indicator "b" may achieve a better estimation of Prevalence
for epidemiological purposes. Traditional methods may not be "universally"
useful. Different methodological aproaches should be used according to each
setting conditions.
TI IDIflN Q

Tuesday
15 August
Paster
Exhibition
1

Feasibility of clinic-based follow-up of uninfected
TUPE0326 children born to HIV-infected women in the UK:
TUPEO326the children exposed to antiretroviral therapy study
Evaluation of HIV sentinel sero-surveillance system the children exposed to antiretroviral therapy study
in Thailand, 2005 (Chart)

S. Iamsirithaworn', R. Detels2. 'Bureau of Epidemiology, Department of
Disease Control, Ministry of Public Health, Nonthaburi, Thailand, 2University of
California at Los Angeles, Department of Epidemiology, Los Angeles, United
States
Background: Implemented countrywide for over 15 years, the HIV sentinel
sero-surveillance (HSS) system has provided important information on the
HIV situation and its dynamics in Thailand. The purpose of this study was to
evaluate the quality, data usefulness, and cost-effectiveness of the system.
Methods: Health officials in charge of the HSS system at the provincial
level were surveyed using mailed questionnaires, followed by focus group
discussions with a subset of officials. During June-July 2005, 25 of 76 provinces
were randomly selected for field observation of surveillance activities and cost
estimation. At the national level, key informants were interviewed. Surveillance
protocols and HIV prevalence data of each sentinel group from the previous 23
annual surveys were reviewed.
Results: 97% of surveyed officials indicated that HIV prevalence data have
been useful for monitoring the HIV/AIDS epidemic and AIDS program planning
and evaluation. During the early epidemic phase, high HIV prevalence in female
sex workers guided the 100%condom use program implementation, and the
declining prevalence trend from the surveillance data subsequently illustrated
the program's success. Annual HIV prevalence surveys of core sentinel groups,
including pregnant women attending antenatal clinics, blood donors, and female
sex workers, have been regularly conducted in most provinces. However, the
number of provinces reporting HIV prevalence among males seen at STD clinics
and injecting drug users at methadone clinics has greatly dropped (>50%)
in the past 5 years. Recent health care reform has impacted the system by
decreasing financial support and creating shortages of surveillance staff. The
annual cost of the system is estimated to be <1% of the total national HIV/
AIDS program budget.
Conclusions: The HSS system has proven to be cost-effective and a major
source of information for HIV epidemic monitoring, policy development, and
evaluation of control programs at both the provincial and country levels.
Recently, budget constraints have affected its sustainability.

C. Hankin', H. Lyall', J. Masters', C. Peckham', P. Tookey'. 'Institute of Child
Health, Centre for Paediatric Epidemiology and Biostatistics, London, United
Kingdom, 2Imperial College School of Medicine and St Mary's Hospital,
London, United Kingdom
Background: Most UK-born uninfected children born to HIV-infected
diagnosed women have been exposed to antiretroviral therapy (ART) in utero.
Possible adverse effects of exposure include prematurity and mitochondrial
abnormalities; animal studies suggest a risk of cancer. However, long-term
monitoring of apparently healthy uninfected children is challenging. The Chart
study was established to assess the feasibility of clinic-based follow-up.
Methods: Obstetric and paediatric HIV surveillance is conducted through the
National Study of HIV in Pregnancy and Childhood (NSHPC): health professionals
routinely report cases through active reporting schemes. In 2002-2005 notifying
health professionals enrolled uninfected children born 1996-2004 for annual
consented follow-up, using a standard questionnaire (Chart study).
Results: Of 2104 eligible children, 704 (33%) were enrolled; parents of 5%
declined, 3% had gone abroad, 22% had moved and were not contactable;
37% were not enrolled mainly because of lack of local clinic resources or
because the responsible clinician was unwilling to approach the family. Enrolled
and non-enrolled children were similar with respect to ART exposure, perinatal
outcomes and maternal demographics. At least 25% of enrolled children were
subsequently lost to follow-up within the study period because the family
moved without leaving contact details, or declined further involvement. Median
age at last follow-up was 23.5 months (range 5-106). Minor childhood ailments
were commonly reported; 51 children (7%) had more serious health problems,
eg: febrile seizures, sickle cell disease, congenital heart disease, autism,
cerebral palsy. Two children (neither enrolled) died by end 2005 (sudden infant
death, tuberculosis). There was no apparent association between specific ART
exposures and particular conditions, but numbers are small and follow-up
limited.
Conclusions: Individualised on-going clinic-based follow-up is not feasible. A
less intrusive follow-up strategy is being developed, based on linking NSHPC
data with other national sources of routinely collected morbidity and mortality
data, with appropriate safeguards regarding confidentiality.

XVI INTERNATIONAL AIDS CONFERENCE * 13-18 AUGUST 2006 * TORONTO CANADA * ABSTRACT BOOK VOLUME 1