A Summary of ABC evidence Presentation to the PACHA 32904


1
A Summary of ABC evidencePresentation to the
PACHA 3/29/04
Edward C. Green Harvard Center for Population
and Development Studies
2
Most HIV is transmitted through sexual
intercourse. Having multiple sexual partners d
rives AIDS epidemics. If people did not have m
ultiple sexual partners epidemics would not
develop or once developed be sustained.
3
YetAIDS prevention programs do not address the
behavioral pattern of multiple partners except
in oblique ways. Programs funded by major dono
rs do not directly address this crucial factor
4
Two approaches to disease prevention
1. Risk- reduction or remedies interventions
(condoms treating STIs with drugs)
2. Risk avoidance (abstinence or delay of age of
first sex mutual monogamy).
This dichotomy is imperfect because reduction in
number of sexual partners would have to be
classified as risk reduction not avoidance.
Primary Behavior Change
5
ABC Strategy
6
AIDS prevention strategies were developed in the
United States with the American epidemic in
mind. When HIV infections are concentrated in
high risk (or core transmitter) groups (MSM CSW
IDU) as in America or Thailand it may make
sense to focus interventions on risk reduction
i.e.

• Condoms
• Treatment of STIs
• Provision of clean syringes

7
Yet perhaps this prevention model should not have
been exported to Africa intact without modifying
it for the epidemic pattern found in
(Sub-Saharan)
Africa
8
Prevention resources go to risk reduction to
medical devices and medicines and this is the
basis upon which we monitor and evaluate AIDS
prevention programs The 4 main areas of prevent
ion today are Condom social marketing VCT (
leading to drugs and/or condom use)
Treatment of STIs PMTCT (based on nevirapine)
9
Meanwhile A and B (risk avoidance)
interventions are largely ignored
or left to religious groups Yet these C and
D risk reduction strategies have not proven
very effective in Africa. Some examples of recen
t research follow showing that the condom
solution to AIDS is not working very well
10

-Hearst and Chen UNAIDS study 2003
inconsistent condom use does not
protect against HIV there are no definite e
xamples yet of generalized epidemics that have
been turned back by prevention programs based
primarily on condom promotion.
-UNAIDS multi-site study (condom levels found to
be non-determining of HIV infection levels)
-Weller and Davis 2002 consistent condom
effectiveness provided 80 in rate of HIV
infection -Shelton and Johnston 2001 there are o
nly 4.6 condoms per male per year in Africa
(Uganda is 4.0 below average)
-Which countries have highest condom rates See
next
11
Average number of condoms per male 15-49 in
African countries for which data are available
12
Kenya HIV Prevalence Condom Sales
HIV Prev
Condom Sales
13
Botswana HIV Prevalence Condom Sales
14
Cameroon HIV Prevalence Condom Sales
Condom Sales
Urban and Rural Prevalence
15

• Conclusions by PSI
• Condom sales tend to rise over time - but so does
  HIV prevalence
  
• condom sales increases alone have not led to
  reductions in incidence or prevalence in many
  countries.
  
• (perhaps not in any country-certainly not in any
  country with a generalized epidemic-TG)
  
• Other Conclusions there may not be any causal
  connection but 7-8 studies now show association
  between condom use and greater HIV infection
  levels than among non-users (probably because
  inconsistent use disinhibition false sense of
  security)

16
Uganda in 1993
Something different was happening something that
the experts had not predicted
HIV infection rates had started to decline yet
condom rates were too low to have had any
significant impact
17
Decline in National HIV Seroprevalence in Uganda
Based on 15 Sentinel Surveillance Sites
18
The prevention approach that was developed in
Africa
The ABC approach Abstain Be faithful or use
Condoms Uganda put strong emphasis on and res
ources into fidelity abstinence delay of debut
among youth who were the primary targets in AIDS
prevention. In spite of all we hear about ragin
g hormones behavior changed first and to the
greatest degree among age group 15-19. This group
also had the greatest decline in HIV prevalence.
19
Ugandas Early Response

• National response began in 1986 with bold
  leadership by President Museveni.
  
• The period 1986-91 is important since HIV
  incidence eventually prevalence peaked then.
  
• Condom social marketing didnt take off until
  mid-1990s. There was some condom promotion from
  the beginning but this approach was not favored
  by the President.

20
President Museveni on condoms
Just as we were offered the magic bullet in
the early 1940s we are now being offered the
condom for safe sex. We are being told that
only a thin piece of rubber stands between us and
the death of our continent. I feel that condoms
have a role to play as a means of protection
especially in couples who are HIV-positive but
they cannot become the main means of stemming the
tide of AIDS. President Museveni 1991
21
Pres. Musevenis approach
He put emphasis on persuading youth to delay sex
until they were married and those already
sexually active were urged to be faithful to one
partner only (zero-grazing).
When I had a chance I would shout at them h
e said. I used to say you are going to die if
you dont stop this. You are going to die!
22
Distinguishing features of the prevention
approach (all beginning 1986-7)
and in addition to an ABC approach

• Bold leadership at the highest level open
  discussion re. AIDS sexual behavior sounded
  the alarm
  
• AIDS preventive education in primary schools
  reaching children before they are sexually
  active
  
• Involvement of religious leaders
• Involvement of PLWHAs in AIDS prevention
• Fear arousal meant to engender risk perception
  and behavioral change
  
• Face-to-face open discussion about AIDS
  community involvement
  
• Major involvement and advancement of women and
  youth
  
• Fought AIDS-associated Stigma

23
Changes in sexual behavior among men in Uganda
WHO/GPA surveys 1989 1995
24
Changes in sexual behavior among women in Uganda
WHO/GPA surveys 1989 1995
25
Types of reported behavior change in UgandaDHS
1995
26
Types of reported behavior change in UgandaDHS
1995
27
Types of reported behavior change 3 Ugandan
Districts 2000-2001
28
Newsweek on Uganda
A cover story about AIDS in Africa in Newsweek
(1/17/00) points out that its not all doom and
gloom there is at least one success story to
learn from Uganda. In Ugandahealth workers
turned Protector condoms into must-have fashion
accessories simply by introducing a flashy new
package and a marketing slogan (So strong so
smooth). No other method of prevention was eve
n mentioned.
29
Actual Condom use in the general population of
Uganda
In 1995 about 6 of sexually active Ugandans
used a condom with some regularity according to
the Demographic and Health Survey.
By 2000 this rose to 11 of sexually active Uga
ndans or 8 of all Ugandans. However condom use
has become quite high among those who need them
most namely those relatively few who are still
having multiple partners (e.g. among CSW
clients 95)
30
Thailand
Thailand was the first country in Asia to docum
ent HIV epidemics among IDUs and among sex
workers and their clients. Between 1989-90
shocking prevalence figures emerged and were
allowed to be broadcast to the public e.g. 1989
findings that 44 of Chiang Mai brothel workers
were HIV Thailands national response to AIDS
began in 1989. Directed initially by Mechai
Viravaidya and the Population and Community
Development Association (PDA)
31
Thailand early national response
Primary target groups were adolescents CSWs
clients of CSWs and wives of men with multiple
partners (Pattalung and Bennett 19903)
Campaign used fear arousal messages to attract
attention to convey the core information that
AIDS is fatal but can be prevented
Condom use and partner reduction were the prima
ry behavior change messages. Abstinence/delay
became part of the message but not clear when
and to what degree.
32
Thai response
Rapid dissemination of information through
leadership structures down to grassroots
AIDS education was incorporated into school cur
ricula in 1990. Religious leaders were mobiliz
ed New Prime Minister (1991) became directly i
nvolved in AIDS control. Chaired the national
AIDS committee helped develop a 5-year plan.
Since HIV prev. declined sharply after m
id-1990s it is useful to look at behavioral
changes prior to this decline
33
Early results in Thailand
Comparison of national behavioral surveys in 1990
and 1993 just prior to drop in HIV prevalence
of men reporting premarital or extramarital
sex in last year fell from 28 to 15
of men visiting CSW from 22 to 10 and consi
stent condom use (reporting always using condoms
incommercial sex) rose from 36 to 71.
(Phoolcharoen et al 1998 AIDS)
By 1996 condom use rose to 97 in brothels an
d to 89 among indirect sex workers (e.g.
massage parlors) in Bangkok
34
Evidence of continuing behavioral change
(visits to female sex workers and condom
use at last visit) in 21 year old males
in the north.
Percent of sample
35
(No Transcript)
36
Parallels Thailand Uganda
An early and vigorous response
An indigenous strategy developed by both
governments before much foreign TA
3-4 years after national response began si
gnificant behavior change had occurred including
less casual sex
HIV prevalence had peaked.
There was also Open discussion Commu
nity mobilization Strong Political leadership
Explicit messages (no soft-pedaling of messages)
.

37
Replicability of Ugandas success
There seems to be no reason Ugandas ABC model
could not be replicated in other countries with
generalized epidemics. (re. the forcing the m
odel argument we donors have already forced a
universal risk reduction model everywhere
whatever the culture or epidemiological
pattern) Stoneburner and Low-Beer (2000) say t
he effect of HIV prevention interventions in
Uganda (particularly partner reduction) appears
to have had a similar impact as a potential
medical vaccine of 80 efficacy.)
38
ARVs and VCT
The big interest nowadays is in ARVs. VCT is the
entry point to treatment and it is also being
promoted as an effective tool of prevention.
Leading authorities are now saying we cant
expect behavior change unless people know their
sero-status. Yet there was a great deal of
behavior change in Uganda and Senegal and
elsewhere in the absence of VCT (and only 10 of
Ugandans have been tested in Uganda even today).
The evidence that VCT leads to behavior change a
nd lower HIV Prevalence is mixed. If The C in
VCT is based on ABC then it could lead to
positive behavior change