Introduction

The scientific and ethical importance of including women of reproductive age in biomedical research is now widely acknowledged. Sex-based differences in disease mechanisms, treatments and interventions require inclusion of women to generate evidence for the care of female patients [1–3]. However, concerns regarding potential harms to fetuses of women who are or may become pregnant during the course of a trial have historically restricted women’s participation across biomedical research. While such restrictions are justified in some circumstances, they are often applied without robust consideration of maternal and fetal risks and benefits or consequences of exclusion. As a result, there are significant gaps in the evidence base for women’s health, and individual women may be unjustly denied the prospect of direct benefit from participation in some studies [4,5]. Though the inclusion of women and their interests in biomedical research has increased in recent years, robust inclusion remains elusive, as marked by enduring concerns about justice and fair access to benefits of investments in advancing biomedical knowledge [6]. Additionally, fetal protection concerns have also motivated contraception requirements among biomedical studies that do enroll reproductive-aged women, raising important questions about the appropriateness of these widespread requirements.

One pressing question is when, and under what conditions, women of reproductive potential should be required to use contraception as a precondition for enrollment in biomedical research. On the one hand, requirements for effective contraception have been considered an important way to prevent fetal exposure in biomedical research [7,8] and ensure women’s access to research where study interventions have the potential to cause fetal harm [9]. On the other hand, requirements for contraceptive use often persist independently of a woman’s actual potential for pregnancy, or in the absence of a meaningful assessment of the risks and benefits [10]. Further, provisions in U.S. Food and Drug Administration (FDA) guidance for the conduct of research with women of reproductive potential allow deferral of certain reproductive toxicity studies if requirements for contraception are instituted [8]; this raises the possibility that contraception requirements may at times be less about reducing theoretical fetal risks, and more about minimizing research and development (R&D) timelines and costs. These issues have raised concerns about whether current patterns of contraceptive requirements disrespect women’s autonomy, or unduly impose the risks and burdens of contraceptives to research participants [11,12], and unfairly impede access to research carrying the prospect of direct benefit. Additionally, concerns have been raised that the widespread use of hormonal contraception in biomedical research has problematically restricted our understanding of variations in pharmacokinetics associated with normal or natural menstrual cycles occurring in the absence of synthetic hormones [10].

In order to address these important concerns, improved guidance regarding contraception requirements in biomedical research is needed. There are currently no harmonized international regulatory guidelines for such requirements [13]. The U.S. FDA has recommended that “women be counseled about reliable use of contraception or abstinence … [but does not] specify the type of contraception used because FDA believes decisions of that nature are best left to the woman in consultation with her health care provider” [7]. U.S. regulations governing research with human subjects are silent on the use of contraception [14], and the Council for International Organizations of Medical Sciences (CIOMS) Ethical Guidelines for Research emphasizes the importance of access to contraception for women of childbearing potential in clinical studies, but does not address the ethical complexities of contraception requirements [9]. Individual Institutional Review Boards (IRBs) generally develop institution-specific guidance, which can vary widely. Additionally, limited suggested guidance has appeared in the academic literature; the two proposals that have been published used the FDA’s now-retired, letter-based labeling categories, and differ on several key points, including when, if ever, investigators may require abstinence or the use of contraception as an inclusion criterion [11,15].

It is important that researchers and policy makers take account of the diverse array of perspectives of women as they shape evolving regulations and best practices to address pregnancy, fertility, and the role of contraception in clinical trials [16]. In recent years, research subjects’ views have been recognized as a critical component to adequate research ethics analysis and oversight [17]. Although potential participants may not be knowledgeable about historical, institutional, or other factors that bear on policy options, they may well be aware of other morally relevant considerations that policy makers and scholars might otherwise overlook, including the ways in which their own values and life context affect how they view clinical trials when contraception is required.

Yet women’s views are largely missing from the discussion about contraception in research. We could identify no studies that asked women their views about the advantages and disadvantages of contraception requirements in research contexts, and only two studies of trial participants’ opinions that touched on the subject. A qualitative study in Uganda explored male and female participants’ understanding of the informed consent process for required contraceptives and reasons for noncompliance with contraception requirements in trials. While all respondents described being advised to avoid pregnancy during trial participation, factors reported to be associated with pregnancy included the belief that the investigational product presented low fetal risk, a need for children, male partner unwillingess to use condoms, and side effects of contraception [18]. A small survey of US women queried willingness to participate in hypothetical trials of drugs to reduce breast cancer risk, and found that requirements for the use of specific contraceptives, as opposed to women’s own choice of birth control method, reduced women’s willingness to participate [19]. Thus, debates among IRB members, researchers, regulators and policymakers have proceeded in the face of little empirical information about the views and concerns of those potentially most affected by such policies.

To address this void, we elicited women’s views on contraception requirements as part of a larger qualitative study conducted by the Pregnancy and HIV/AIDS: Seeking Equitable Study (PHASES) project [20]. Our study interviewed pregnant and recently pregnant women living with or at risk of HIV/AIDS, exploring their views about and experiences around clinical trial participation. The context of HIV/AIDS research is particularly relevant to the exploration of contraception requirements in clinical trials for three reasons. First, considerable research has been conducted with women living with or at risk for HIV/AIDS, and issues related to pregnancy and contraception have been a matter of debate [21–23]. Second, women living with or at risk for HIV/AIDS are disproportionately low-income and face a range of challenges regarding access and consent to research participation in health care settings; understanding particular challenges of low- and middle- income country (LMIC) contexts where HIV research is often conducted and to which findings will be most relevant is critical [18,24–26]. Third, given the international partnerships that characterize HIV/AIDS research, standard research practices need to be sensitive to implications they may have for critical research in international settings. We therefore conducted a study of pregnant women in both the US and in Malawi, a country with a high prevalence of HIV/AIDS and a robust research infrastructure.

Methods

The data for this analysis were collected as part of a larger project, PHASES, which aims to develop guidance for ethically acceptable research on HIV treatment and prevention during pregnancy. We conducted qualitative, in-depth interviews using a semi-structured guide, as described previously [20]. The data for this analysis, collected between August 2016 and April 2017, pertain to women’s opinions about common contraception requirements for women participating in clinical trials. Interviewers described the common practice in clinical trials of requiring women to use birth control in order to participate, highlighting that two forms of birth control are often mandated, and that male partner vasectomy may not count as a type of birth control [12,27]. If the woman does not agree to use birth control, then she cannot participate in the study, even if she does not currently have a sexual partner (see Box 1). The purpose of this approach was to surface participants’ initial reactions to this widespread requirement in biomedical research with women of reproductive potential, including their thoughts around when it would and wouldn’t be appropriate. Therefore, no potential justifications for the contraception requirement were provided to the participants. Interviewers asked women to discuss what they thought about this practice of requiring contraception with open-ended probing of advantages and disadvantages, asking them to reflect on requiring one versus two forms, as well as their overall opinion.

Written, informed consent was completed prior to the interview. Interviews in Malawi were conducted in Chichewa by one of two researchers who are native speakers; all U.S. interviews were conducted in English. Demographic questions were also completed. The research was approved by the Institutional Review Boards at the University of North Carolina at Chapel Hill and Johns Hopkins Bloomberg School of Public Health and the National Health Science Research Committee of Malawi.

Results

Participant characteristics

Participant demographics are presented in Table 1. Participant age ranged from 18 to 50 (mean age Malawi = 29.4, U.S. = 29.8), with the largest proportion of women in both countries in the 25–34 year age group. Seventy percent (n = 49) of U.S. participants were Black/African-American, and all women in Malawi were black Africans. Twenty-three percent of U.S. women were college graduates or higher, and 20% of participants in Malawi completed secondary school or higher. Thirty-six percent of U.S. women and 90% of Malawian women were married. It is worth noting that the concept of marriage in Malawi is colloquially considered to be broader than the explicit legal definition, and includes cohabitating couples who typically raise children together. In both countries, the majority of participants identified as Christians (67% in the US, 89% in Malawi), and 50% of participants were HIV-positive (35 in U.S. and 35 in Malawi).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Demographic characteristics of women participating in PHASES.

https://doi.org/10.1371/journal.pone.0216332.t001

More than half of women in both the U.S. (60%) and Malawi (58%) indicated overall support for the practice of requiring contraception in clinical trials; about a fifth opposed the requirement, and the remainder–also about a fifth–were unsure or said it depends on the circumstances (see Table 2). Those who were unsure or said it depends often believed the requirements should be determined based on the potential risks of exposure to the study drug if pregnancy were to occur during the course of the trial. Others felt a requirement for one type of birth control was reasonable, whereas requiring two forms placed an unnecessary burden on the woman and/or couple. In addition to their overall assessment, participants discussed what they saw as advantages and disadvantages of the practice. Their responses diverged both within and across countries, with several emergent themes being specific to one country or the other, as described below and summarized in Table 3. Many women, regardless of overall support for or opposition to a contraceptive requirement, articulated nuanced and multifaceted opinions, including the different contexts in which and reasons why they felt it to be appropriate or not.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Participants’ overall views by country on contraception requirements for clinical trial participation.

https://doi.org/10.1371/journal.pone.0216332.t002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Themes emergent across participants’ assessments of advantages and disadvantages of contraception requirements.

https://doi.org/10.1371/journal.pone.0216332.t003

Discussion

Our study is the first cross-national qualitative study to our knowledge of women’s views about contraception requirements for biomedical research participation. Among 140 women in the U.S and Malawi, a majority offered overall support for contraception requirements, while some women were opposed and others were unsure. Seven themes characterized the identified advantages and disadvantages of research-related contraception requirements generally, several of which were country-specific (see Table 3). Not only do our findings expand the conversation beyond considerations generally raised in debates about contraception requirements, but they highlight the ways that context shapes women’s experience–as well as the ethics–of requiring contraception in a clinical trial.

Specifically, while the proportion of participants in favor of or against contraception requirements were similar across the two countries, the reasons offered were not. Particularly notable is the contrast between Malawian and U.S. participants in their characterization of the burdens of contraception requirements. U.S. women tended to focus on biomedical concerns such as side effects and impact on fertility, and questioned whether requiring two forms of birth control was medically justified. In contrast, Malawian women focused on the social risks of birth control requirements, which could be profound: while the use of a contraceptive could lead to violations of trust in marital relations and suspicions of potential infidelity, it could also serve as a leverage point for women’s reproductive control.

Respondents in the U.S. and Malawi were more similar in the ways that they characterized the benefits of contraception requirements, though important differences remained. Many of the positives noted were related to contraception use in general, and not specific to a trial context. Women in both countries noted that contraception requirements could enhance reproductive control by reducing unintended pregnancy. If contraception was meaningfully accessible and its use supported outside the context of a study, it would seem that requiring contraception use by research participants would not in fact increase reproductive control. But contraceptives are not necessarily easily accessible for women in either setting, and their socio-cultural contexts may not support contraceptive use. For example, in the U.S., there may be financial constraints for women without any health insurance or with health insurance that does not cover all contraception types [29,30]. In the Malawian context, there may be cultural constraints such as a presumption that use of contraceptives implies infidelity in marriage, or structural constraints such as transport barriers, long wait-times, and limited contraception options. In both contexts, then, a research participation requirement that mandates and supplies contraception can be both useful to and welcomed by women who do not desire pregnancy. In the Malawi context, for example, contraception requirements for trial participation were seen as facilitating birth spacing, which has important health benefits for women and children alike.

To the extent that a research requirement for contraception is welcomed by women whose use of contraception is otherwise constrained, some may question whether a requirement for and provision of contraception could represent undue inducement for some participants. While contraceptives may be an ancillary benefit for those who desire them, undue inducement pertains only when an offer has a distorting impact on a research participant’s ability to rationally weigh costs and benefits of participation [31]. Nothing respondents in our study said indicate such a distorting influence; and none offered that access to contraception itself would change their mind in favor of participation. Further, given the unfortunate constraints on contraception use for some women outside of a research context, the reasons participants offered in favor of access are entirely rational.

Whether in favor of or against contraception requirements, our data highlight the importance of considering how socioeconomic and cultural context–gender, race, socioeconomic status, cultural considerations–shape women’s views about contraception. In recent years, discussions of ethics and reproduction have increasingly employed what is known as a “reproductive justice” framework. As opposed to a “reproductive rights” framework, which tends to focus on rights of access to contraception and abortion, the reproductive justice framework takes a broader view and considers the myriad issues that affect women’s reproductive lives. Centrally, the framework recognizes women’s interests in having (as well as not having) children and being able to raise them in a safe, healthy environment [32]. The concept of justice thus extends beyond ensuring legal access to means to prevent or discontinue pregnancy to ensuring fair access to resources, non-discriminatory policies, and institutional structures that enable women to maintain control over their bodies and lives, and would press questions not just about fair access to contraceptives but the ethics of requiring that women use them [33]. Critically, the reproductive justice framework also considers how gender, race, and socioeconomic status shape access to and experiences of reproductive health care, including within clinical research [34]. In keeping with the reproductive justice framework, the differential concerns of Malawian and U.S. women reflect the important impact that cultural context may have on women’s views about and experiences of contraception requirements in clinical studies. In particular, in Malawian contexts, where society has been described as largely patriarchal [35,36], contraception requirements may be seen as a way for women to work around gender power differentials when contraception is desired; or as a burden that may engender conflict in marital relationships. A reproductive justice framework would attend to both sets of concerns as well as the ways in which cultural norms around gender shape the experience of the requirements–both in terms of creating opportunities for empowerment through contraceptive access and introducing risk for women who desire study participation. As such, a reproductive justice framework would potentially entail a shift away from an emphasis on required use, and toward approaches that support women’s informed choices about use or non-use of contraceptives, as well as access to desired contraceptives in research contexts.

For the research community, contraception in clinical trials is motivated by various factors, often in combination. These include a desire to protect the fetus from research risks, reduce costs and product development timelines, protect from liability due to pregnancy exposures in clinical trials, and maintain the “statistical integrity” of studies that can be compromised by high pregnancy rates associated with study withdrawal [22,37]. Some women in both the U.S. and Malawi felt that researchers–by virtue of their authority as researchers–should be able to impose contraception requirements. However, many respondents’ reasons for favoring contraception requirements included the benefits of contraception generally, rather than being solely motivated by a desire to protect their potential offspring. Also, from our Malawi respondents we learned that contraception requirements entail burdens for women that may not have been previously appreciated. While a range of external considerations such as sponsor liability are likely to inform decisions about contraception requirements, our findings suggest at least four ways in which such requirements could be more responsive to and respectful of research participants.

First, contraception requirements should be sensitive to actual fetal risk assessments. In studies involving women, contraception requirements have tended to be presumed to be justified–and necessary–even where fetal risk is uncertain. For instance, women of childbearing age may be required to use highly effective methods of birth control and pregnancy tests when participating in early phase trials before non-clinical reproductive toxicology studies have been conducted [8]. But there are many circumstances where contraception is required in which the risk of harm to offspring is extremely low and well-characterized. As some of our U.S. participants indicated, whether contraception requirements are justified under these conditions is, at best, questionable. Given the burdens made vivid by our study, it may be more appropriate and respectful of women’s interests to require that researchers justify any requirement for contraception–and not only in terms of the likelihood of harm of early fetal exposure to a drug or biologic but also in terms of the likelihood of incident pregnancy. Factors to be considered should include sexual activity and orientation of participants, the potential burdens of contraceptives both in terms of biomedical and social risks, and the likelihood that women of childbearing age will use the intervention in clinical contexts and become pregnant.

Second, wherever possible, researchers should strongly consider providing access to contraception, even where pregnancy-associated risks of research are minimal and contraception requirements are unwarranted. There is increasing recognition that researchers have responsibilities to provide needed medical care that is beyond the narrow purview of their research (ancillary care) [38]. Access to contraception is widely recognized as important to women’s health and empowerment, and global disparities in such access is a topic of ongoing concern [39,40]. Given that many of the benefits that women reported from contraception requirements were actually benefits of having access to contraception that might otherwise be challenging to obtain, ensuring access to contraception as a form of ancillary care should be considered in research planning and implementation.

Third, when contraception requirements are deemed appropriate, they should be directed at optimizing effective pregnancy prevention rather than a blunt approach of mandating “two forms” of contraception absent clear evidence of improved protection. Interestingly, where prevention of pregnancy is a priority, “clinical guidelines and national data” have indicated that “using one effective method [of contraception] correctly and consistently” is the best way to avoid pregnancy in clinical settings, and that providing choice over modality increases use [41,42]. There is no reason to believe that the research context would be different. Additionally, individual factors, including if a woman exclusively has sex with women, is abstinent, or infertile should be explored and integrated into assessing the appropriateness of the requirements. Incorporating evidence-based contraceptive counseling that is responsive to women’s needs and life circumstances, and allowing a range of contraceptive options, may serve a dual purpose of respecting women’s autonomy as well as reducing unintended pregnancies in research contexts, while also potentially increasing trial eligibility, recruitment, and retention.

Finally, our findings may have implications for understanding the high rates of contraceptive failure in clinical studies [37,43,44]. Research on the psychology of contraceptive use indicates that women’s efforts to avoid pregnancy are based on a “benefit:burden ratio” and that contraceptive use depends on a range of evolving factors, including positive and negative views about becoming pregnant as well as perceptions of and experiences with specific contraception methods [38,45,46]. We are not aware of any studies on the psychology of contraceptive use where contraception was required in clinical research. One observational study cited misconceptions and incorrect use as contributing to unintended pregnancies in trials [47]. Non-adherence to study requirements for contraception may be high where motivations regarding pregnancy are at odds with study requirements, there is ambivalence about contraceptive methods or use, and social contexts may weigh heavily on decisions about pregnancy and contraception. In addition to addressing high rates of contraceptive failure, it may be prudent to be more deliberate and reasoned about when contraception should be required, and consider whether offering more contraceptive choices in the context of studies could lead to a more robust evidence base that meets the needs of women across their reproductive lives.

Several considerations inform interpretation and generalizability of our results. Our qualitative methodology was intended to identify and explore women’s opinions on contraception requirements in order to surface themes and concrete considerations that policy makers should be cognizant of when determining regulations and best practices. It was not intended to measure the prevalence of such views among women, or to assess women’s views about their appropriateness in specific study contexts with varying levels of potential risks to the fetus, both of which may further inform future policies. Our sample is not representative of all women at risk for or living with HIV in the U.S. and Malawi, or women of reproductive potential more generally. Our results cannot be generalized to the population of women who might be eligible for research participation.

Our study has several limitations. The sample for this study included pregnant women, for whom the contraception requirement would not be immediately applicable. How pregnancy may affect women’s views of contraception requirements while participating in research remains unclear. Additionally, a more representative sample of reproductive-aged women at risk for or living with HIV may include different proportions of women who are infertile, who do not desire pregnancy, or who exclusively have female sex partners, characteristics which are likely to influence views on clinical trial contraception requirements.

Future qualitative research with a complementary sample of women of reproductive age who are neither pregnant nor recently pregnant would be useful to explore if substantive variations in views related to these considerations exist. Additionally, research using quantitative methodologies would be beneficial to more clearly delineate the prevalence of and potential associations within these views. Further, while this particular set of questions referenced research generally and did not specify a particular health condition, respondents were primed to respond to HIV-related research with interview questions that preceded this one; their answers may have been influenced by these prior references. We did not ask women to consider specific types of contraceptives, nor did we ask them to reflect on how level of or uncertainty about fetal risk, or other potential justifications for contraception requirements, would affect their views. All would be important topics of future study.

Requiring a woman to use contraception in any circumstance is ethically complicated. Yet in the context of biomedical research with women of childbearing age, contraception requirements have often been viewed as necessary to the ethical conduct of research–or an allowable option for minimizing R&D timelines and costs by delaying developmental toxicity studies. Our study highlights the ways in which contraception in the context of research may be viewed as either valuable or burdensome to women, depending on social context, and underscores the ethical complexity of contraception requirements in clinical research. These findings highlight considerations important to developing ethical guidance for the provision or requirement of contraceptives in the context of clinical trials, and to robustly orienting such trials around the interests and lives of the women who will participate in them.

Acknowledgments

The authors gratefully acknowledge the women who participated in these interviews and shared their time, views, and experiences. We very much appreciate Dr. Francis Martinson at UNC Project Malawi and Dr. Claire Farel at the UNC Infectious Disease Clinic, for their support of this work. Warm thanks are also extended to Bernadette Harley-Boon, Altricia Belk, Patrice Ross-Johns, and Krissa Pecunes at Johns Hopkins University, and Karen Dorman and Katherine Barley at UNC Chapel Hill for their assistance in navigating the clinics and the participant recruitment process.