Primary Prevention 229HIV infections by 50% in the intervention group. Moreover,
after adjusting for baseline risk factors, incident STDs were
seven times less frequent among men assigned to the inter-
vention than the combined controls; however, contrary to ear-
lier “ndings (e.g., Kelly, Murphy, Sikkema, et al., 1997), there
was no diffusion of the intervention to adjacent barracks.
Summary
Three midstream approaches have been evaluated to date: in-
dividual and small group approaches, communitywide ap-
proaches, and media campaigns. Individual and small group
approaches are often guided by social-cognitive theoretical
frameworks and emphasize education; skills training (e.g.,
observation, modeling, rehearsal, feedback, reinforcement);
cognitive modi“cation (e.g., outcome expectancies, self-
ef“cacy beliefs); and social reinforcement for instituting
behavior changes. Consistent with a meta-analytic integra-
tion of this literature (Kalichman, Carey, & Johnson, 1996),
our observation is that such interventions consistently reduce
short-term risk behavior and, in a few studies, are also asso-
ciated with a lowered incidence of new STDs. Evidence of
the long-term effectiveness of these interventions is needed,
and the potential reach of such programs is limited due to
their cost.
A second midstream approach involves community-based
interventions, often in”uenced by social dif fusion theory.
Fewer studies of this approach have been completed, but
these also provide evidence of reductions in risk behavior at
the community level. As with small group approaches, the ef-
fect sizes from these studies tend to be modest. Finally, media
approaches have the widest reach but might be expected to
have the least short-term impact. However, media approaches
promise to prime communities and individuals for more in-
tensive risk-reduction efforts, can create social norms that
are supportive of risk reduction, and might serve as back-
ground •boostersŽ of more intensive risk-reduction strate-
gies. Indeed, several upstream approaches completed in
smaller countries demonstrate the value of combining face-
to-face interventions, with media and policy changes.
Upstream Approaches
Public policy interventions including governmental, institu-
tional, and organizational actions directed at entire popula-
tions rarely have been implemented in the United States.
One upstream approach that has been evaluated is needle
exchange programs, which is regarded as a •population-
levelŽ approach (Des Jarlais, 2000). The results from needle
exchange programs has been mixed. Three studies conducted
in Canada and the Netherlands have reported that needle ex-
change has no effect on HIV incidence (Bruneau et al., 1997;
Strathdee et al., 1997; van Ameijden, van den Hoek, van
Haastrecht, & Coutinho, 1992) but two U.S. studies have
found that HIV incidence is reduced by needle exchange pro-
grams (Des Jarlais et al., 1996; Heimer, Kaplan, Khoshnood,
Jariwala, & Cadman, 1993). Interpreting these mixed “nd-
ings is challenging, but most experts agree that (a) measuring
the effects of needle exchange on HIV incidence is compli-
cated, (b) needle exchange programs do serve as a link to
drug treatment and as platforms for the delivery of other
needed clinical services, and (c) needle exchange programs
are not enough but can serve as one part of a useful compre-
hensive approach to risk reduction for IDUs (Des Jarlais,
2000; Vlahov, 2000). There is no evidence that needle ex-
change programs lead to increased drug use, a concern that
has been expressed.
Other countries have implemented and evaluated large
national HIV prevention programs. For example, Nelson
et al. (1996) evaluated Thailand•s •100% Condom Pro-
gram,Ž a policy started in response to the high seropreva-
lence of HIV among sex workers. This program began in
1990 and mandated condom use during commercial sex.
Nelson et al. followed cohorts of young men who were con-
scripted into the army in 1991, 1993, and 1995. In the 1991
and 1993 cohorts, the prevalence of HIV infection was 10%
to 12% but by 1995, it had decreased to 7%. Over the study
period, the proportion of men who reported having sexual
relations with a sex worker also declined from 82% to 64%.
From 1991 to 1995, condom use during the most recent
sexual contacts with sex workers increased from 61% to
92%; and in 1995, 15% percent of men had a history of a
STD, as compared with 42% in 1991. Subsequent reports
from this research group have documented the dramatic
decrease in the incidence rates of STDs, including HIV in-
fection, among young men in military service in Thailand
(Celentano et al., 1998). Such programs could be imple-
mented by other countries experiencing epidemics of hetero-
sexually transmitted HIV and are likely to decrease the
spread of the epidemic.
Switzerland provides an example of a national AIDS
prevention strategy in the developed world (Dubois-Arber,
Jeannin, Konings, & Paccaud, 1997; Dubois-Arber, Jeannin,
& Spencer, 1999). The intervention strategy involved three
components: (a) a general publication education campaign
that used the national media (print, TV, radio, posters) and
public relations events to provide accurate information, to
promote condom use outside of stable monogamous rela-
tionships, to encourage monogamy, and to promote toler-
ance regarding infected persons; and (b) risk reduction