A small number of behavioral interventions to increase use of dual methods have been tested. One trial [57
], randomized black and Latina adolescent girls on HC either to a video, STI/HIV counseling based on Project RESPECT [52
], both the video and counseling, or usual care. The video was designed to promote dual methods and increase perceived vulnerability to HIV. At a three-month follow-up, the women who had received both interventions were significantly more likely to report having used a condom during their most recent sexual encounter, compared with each of the other three groups. Unfortunately, this result was nonsignificant at the 12-month follow-up. This study suffered from severe attrition, with only 55% of participants returning for the three-month follow-up and 49% for the 12-month follow-up. This low-retention rate renders interpretation of the findings difficult.
An analysis of data from the Project RESPECT trial, a condom-focused multisite intervention study conducted in STI clinics, examined uptake of condoms by women in the trial who were using HC [58
]. Among these women, condom use increased as a result of the intervention, and most women remained on their HC regimens. Thus, in response to a condom-focused intervention, many women became dual-methods users, although most did not use condoms consistently (i.e., were alternating dual method users).
A more recent study tested a computer-based intervention designed to encourage dual methods use among at-risk women [59
]. Participants were at risk for either pregnancy or STI and were enrolled irrespective of type of contraception used. Women were randomized to receive this intervention or general contraceptive information. Based on self-report, the intervention increased the ever-use of dual methods. While the confidence interval of the unadjusted OR (1.38) contained 1.00, when adjusted for baseline differences using a propensity score, this effect became significant (although alternating versus simultaneous use was not assessed). Differences between treatment arms for consistent condom use, STI, unplanned pregnancy, and individual STIs were all nonsignificant. The authors interpreted these findings to suggest that the use of dual methods was not sustained long enough to prevent STI and pregnancy. Use of dual methods was predicted by higher education level, substance use, and use of either hormonal contraceptives or male condoms at baseline.
Finally, a recent study tested a provider-delivered intervention designed to promote dual method use by providing counseling both about STI risk and pregnancy risk, compared with a standard of care in which only pregnancy was addressed [60
]. In both conditions, patients chose their preferred method without specific encouragement from the counselor. The primary outcome, rather than use of dual methods per se, was the number of sex acts unprotected by a male or female condom. The intervention group reported 3 fewer unprotected acts than the standard of care group, a difference that approached significance.
In summary, the research on interventions to promote dual-method use is mixed. Condom-focused interventions that focus attention on the threats of HIV and STI are successful in increasing condom use, even when women are on HC prior to the intervention. Other studies have yielded mixed results—that is, self-reported outcomes at odds with biological ones—and nonsustained or null results. These results suggest that dual methods counseling frequently fails to achieve its desired outcome, consistent and sustained use of HC and condoms at every act of intercourse. Since condoms protect very well against STI, and quite well against pregnancy, and given the availability of EC should condom failure occur, recommending consistent use of condoms may be more effective at preventing STI than recommending the use of dual methods. Moreover, if the method used must protect against the outcome of greatest concern for the woman, and if that outcome is pregnancy, she may be less tempted to abandon condom use if that is her only source of protection. On the other hand, if she is protected from pregnancy by hormonal contraceptives, she may be more tempted to forego condoms, particularly under male resistance. However, I believe that this hypothesis should be tested empirically, by comparing a dual-use recommendation with a condom-only one. I would hypothesize that women in the condom-only arm would acquire fewer STIs, but may need to use EC more often and may even become pregnant more often. In addition, characteristics of successful dual-method users could be identified, in a message-controlled context.
Would such a trial be feasible? Would young women be willing to be randomized to the form of protection that they would use for a prolonged period? Recent evidence suggests that the answer to this question is “yes.” A feasibility study to assess the acceptability of an RCT to examine whether the use of hormonal contraceptives creates biological vulnerability to STIs has been conducted [61
]. In this study, potential at-risk participants were asked whether they would be willing to be randomized to hormonal contraceptives or to an IUD. They were also asked to provide urine or endocervical swabs for STI testing. Overall, about 70% of participants said that they would be willing to participate in this trial, indicating that women are willing to be randomized, and suggesting that such a trial is indeed feasible.
Would such a trial be ethical? Of course, informed consent would be obtained, and participants will understand that they can cease participation in the trial (or switch arms?) without consequence. To ensure that participants (and parents/guardians?) were fully informed and willing to be randomized to either intervention, investigators would explain to participants the advantages and disadvantages of condoms and HC as tools for pregnancy and STI prevention.
The ethical sticking point appears to be, for some, randomizing participants to receive a less effective method of contraception. However, the overall risks and benefits to study participants must be weighed to determine whether equipoise exists between study arms. On the one hand, the literature indicates that the condoms-only group would receive a less effective pregnancy prevention message than the dual-method (standard of care) arm. Therefore the apparent risk of unintended pregnancy would be increased in the condoms-only arm. But on the other hand, the condoms-only message may be more effective than the dual methods message for preventing STIs, including HIV. (Could be noted here or elsewhere the negative health consequences of STI acquisition among teenagers.) The potential benefit of STI avoidance could balance or outweigh the increased risk of failed contraception. Moreover, participants would be counseled on emergency contraceptive use and given access to this method in cases of failed contraception. This counseling and access would reduce the risk of unintended pregnancy among all study participants and mitigate the increased risk in the condoms-only arm. Although the recommendation to adolescents and young women at risk for unplanned pregnancy and STI is to use both a hormonal contraceptive to prevent pregnancy and male condoms to prevent STI, there is evidence to suggest that this approach is failing, and if this is so, that a significant, unnecessary burden of STI, including HIV, may be the result. This hypothesis should be tested empirically, using a rigorous design.