most evaluated programs target adolescents, especially adolescent girls, knowledge about how to reduce unintended pregnancy among adult women and their partners is exceedingly limited; (3) there is insufficient evidence to determine whether "abstinence-only" programs for young adolescents are effective, but encouraging results are being reported by programs with more complex messages stressing both abstinence and contraceptive use once sexual activity has begun; (4) few evaluated programs actually provide contraceptive supplies, which may help to explain the small effects of many programs; (5) only mixed success has been reported from programs trying to prevent rapid repeat pregnancies among adolescents and young women; and (6) virtually none of the evaluated programs attempt to influence the surrounding community environment shaping sexual activity and contraceptive use.
The design of these new research and demonstration programs should also reflect four additional themes. First, unintended pregnancies derive in roughly equal proportions from couples who report some use of contraception, however imperfect, and from couples who report no use of contraception at all at the time of conception. Although many individuals move back and forth between these two states over time, it may nonetheless be useful to develop specific strategies for each group, especially for the very high-risk group of nonusers. For example, the former group may benefit particularly from ongoing support and special attention to developing better skills in contraceptive use, whereas the latter group may require a greater focus on underlying psychodynamics and couple interaction. Second, available data suggest that multifaceted programs to reduce unintended pregnancy are particularly effective—i.e., programs that include the actual provision of contraceptive supplies, as well as information, education, case management and follow-up, ongoing support, explicit attention to underlying attitudinal and motivational issues, and specific training in contraceptive negotiation and skills.
A third theme that should shape these new programs is the need to develop and test out new ways to involve men more deeply in the issue of pregnancy prevention and contraception. Although there is ever more talk about this idea, little investment in program-based research has been made to investigate the effectiveness of various strategies. Some advocate punitive approaches in order to force boys and men to "act responsibly," whereas others are convinced that carrots, not sticks, are needed. Research can help to develop effective interventions, particularly if experimental interventions address men's different ages, life stages, and cultural and personal preferences.
A fourth theme that these programs should explore is how to build community support for contraception. Although contraceptive use is ultimately a personal matter, community values and the surrounding culture clearly shape the actions of individuals and couples. Accordingly, at least some demonstration programs should target both the community and the individual, and some might also work exclusively at the community level. This approach has been used