Having personally devoted considerable time and energy to efforts to increase women’s access to emergency contraception, we have been disappointed that none of the interventions designed to increase the use of emergency contraceptive pills has significantly reduced rates of unintended pregnancy.1,2 A constructive reaction to this disappointment requires that we identify new directions and approaches to reducing rates of unintended pregnancy that merit increased research attention.

Because contraceptives are used to reduce the risk of “unintended” pregnancy, we must first examine how pregnancy intentions are assessed. Common measurements and conceptualizations of unintended pregnancy have significant limitations.3 For example, a third of pregnancies attributed to “contraceptive failures” in the 1995 National Survey of Family Growth could be classified as “intended.”4 And although up to a third of women attending urgent care clinics are ambivalent about getting pregnant,5 researchers and policymakers often dichotomize pregnancy intentions, assigning them to categories that are not meaningful to many women.6 Not surprisingly, the commonly used, retrospective classification of pregnancies as unwanted, mistimed or intended has not been consistently associated with differences in infant or child health outcomes.7 Therefore, developing and validating a measure of pregnancy intention that is predictive of health outcomes is a research priority, and its use could enable policymakers to monitor relevant trends and differentials meaningfully. Ideally, such a measure would be prospective, so that clinicians could appropriately tailor contraceptive and preconception messages in a way that would improve population-level pregnancy outcomes.

Similarly, there is a need to better understand what women consider a contraceptive “emergency.” Despite initial projections that widespread use of emergency contraceptive pills would dramatically reduce the number of unintended pregnancies,8 interventions designed to increase access to and use of this method have not done so.1,2 Sobering research provides further proof of the limited understanding of how women identify contraceptive emergencies. For example, women attending urgent care clinics for reasons other than obtaining emergency contraception were identified by researchers as “needing emergency contraceptive pills” because the women stated that they had had unprotected sex within the past five days, were trying to avoid pregnancy, and had no personal or religious objection to the method.9 All of these women received information about and a free packet of the pills, yet when contacted six months later, few reported having taken them.10 Therefore, finding ways to enable women to identify a contraceptive emergency is another research priority.

For women who are able to identify a contraceptive emergency, there is a need to develop ways to streamline the provision of methods that are more effective than emergency contraceptive pills. Given that unprotected sex is rarely a onetime occurrence,9 emergency contraception strategies that provide ongoing protection from pregnancy may have a greater effect at the population level. IUDs containing copper are a highly effective method of emergency contraception11,12 and offer the added benefit of long-term protection from pregnancy. In one study, an estimated 15% of women seeking emergency contraceptive pills would have been interested in having a same-day placement of an IUD if the option were available.13 Further research is needed to identify barriers to same-day placement when a contraceptive emergency is identified. In addition, providers need to learn how to remove barriers to same-day provision of IUDs and implants in nonemergency situations.

It is clear that increased use of IUDs and implants for regular contraception could significantly reduce the incidence of unintended pregnancy.14 Indeed, in the United Kingdom, increasing use of these highly effective reversible (HER) contraceptives is an explicit national goal.15 However, relatively few UK women have chosen to use these methods,16 despite their availability, free of charge, through the National Health Service. Qualitative research suggests that providing simple, clear and empowering information about ease of use, reversibility, lack of effect on weight gain and women’s positive experiences may increase interest in these methods.17 The brief scripted introduction to HER methods, developed by the Contraceptive CHOICE Project in St. Louis, appears to have been successful in promoting the uptake of these contraceptives. In this project, women who wish to avoid getting pregnant for at least a year and who want to initiate a new method of reversible contraception are provided with the method of their choice at no cost for three years. Of the first 5,000 women recruited, 67% chose a HER method (45% the LNG-IUS, 12% implants and 10% the copper IUD).18 Because of this project’s design, the effect of the scripted introduction cannot be disentangled from the effect of removing any financial barrier, but it is clear that the combined intervention led to a dramatic increase in the use of HER methods. Hence, further research is needed to explore the relative contributions of structured counseling versus financial constraints on contraceptive decision making.

Another priority area for research is that of contraception in the postpartum period, given the global need to support breast-feeding. The World Health Organization’s medical eligibility criteria for contraceptive use*19–22 state that for breast-feeding women, the theoretical and proven risks of using progestin-only pills, implants and injectables before six weeks postpartum usually outweigh the advantages; however, in the United States and the United Kingdom, these methods are frequently initiated immediately postpartum. Despite this common clinical practice, there are few studies on some methods and no studies on others when administered immediately postpartum, and the quality of available studies is poor or fair.23 Postpartum women need evidence-based guidance on how progestin-only methods may affect mood, sleep, weight loss and milk supply. A randomized controlled trial found that mothers who underwent immediate postplacental insertion of a levonorgestrel-releasing IUD were significantly less likely than mothers who delayed getting the device to be breast-feeding six months later.24 Given the benefits of lactation to both infant and maternal health,25,26 and the fact that the LNG-IUS carries a much lower systemic dose of progestin than the injectable, further study of how timing of initiation of progestin-only contraceptives affects women’s abilities to meet lactation recommendations27 is urgently needed.

Finally, given the growing prevalence of obesity28 and the fact that obesity increases the risk of adverse pregnancy outcomes,29 a better understanding is required of whether very heavy or obese women have an elevated risk of experiencing pill failure, especially when using the lowest dose formulations. Because existing studies of contraceptive failure among heavy or obese women have significant limitations, prospective clinical trial data on this topic are needed.30 In addition, we call for further research to assess whether obese women face an increased risk of emergency contraceptive pill failure.

We hope that new research focused on these priorities will allow us, 10 years from now, to look back and report greater success in preventing undesired pregnancy than we have seen with the last decade’s investment in emergency contraception.