Background
The development of drugs for obstetric and non-obstetric illnesses for pregnant women is a slowly evolving process. Even though more than half of pregnant women take (prescription) medications during pregnancy for both obstetric and non-obstetric indications [
1,
2], there has always been a widespread reluctance to include pregnant women in clinical research due to, among other issues, potential harm to the fetus. Although sound data are unfortunately lacking, there are estimates that the total percentage of women who take medications during pregnancy, either prescribed or over-the-counter, may currently be as high as 64–90% [
2‐
5]. Common medications include painkillers, antibiotics, asthma, sleep and anti-nausea medications [
6].
If drugs are tested in pregnant women, studies usually concern investigator-initiated studies of long-existing and long-used medications (that were previously approved for non-pregnant conditions) that are now tested for effectiveness during pregnancy and labor, such as a low-dose aspirin to prevent spontaneous preterm labor. The results of these studies seldom lead to registrations for new indications during pregnancy, but at best to evidence for off-label use. Innovative drugs for pregnant women are rarely developed. As refraining from taking medication during pregnancy could also harm the mother and the fetus, in the past decades regulators, bioethicists and researchers seemed to have reached consensus that the inclusion of pregnant women in research should be promoted [
7‐
12]. Extrapolation of data from studies conducted in men and non-pregnant women is often uncertain, as pregnancy alters the way that drugs are metabolized by the body and act on the body in a fashion difficult to predict from the pharmacokinetics and pharmacodynamics of non-pregnant groups [
1,
11,
13,
14]. Risk-benefit profiles are likely to differ as well [
8]. Gathering conclusive data in order to develop effective treatments for pregnant women with acute or chronic non-obstetric illnesses, as well as innovative medications for obstetric illnesses, therefore, requires research in pregnant women.
The poor evidence base for drug use in pregnancy is widely regarded as unfair [
9]. Already in 1993 the Council for International Organizations of Medical Sciences claimed that the exclusion of pregnant women as a class is unjust [
12], and in 1994, the Office of Research on Women’s Health (ORWH) of the Department of Health and Human Services (DHHS) in the United States endorsed the view that pregnant women are to be presumed eligible for participation in clinical research and stated that pregnant women ought to be “fairly enrolled” in clinical research. This view was later supported by regulatory agencies (US Food and Drug Administration (FDA) and the European Medicines Agency (EMA)) [
15‐
17], the US Institute of Medicine [
18], and by many individual bioethicists. Despite this longstanding consensus on the need to include pregnant women in clinical research, the situation has not significantly changed since 1994. Exclusion of pregnant women from research is still common practice [
19,
20]. A recent review demonstrated that between 1960 and 2013 only about 1% of pharmacokinetic clinical trials were conducted for pregnant women, and the ones that were undertaken had a strong focus on acute labor and delivery issues [
21]. Not surprisingly, a 2011 study on all medications approved by the FDA from 1980 to 2010 found that 91% of the medications approved for use by adults did not have sufficient data on safety, efficacy and fetal risk of medication taken during pregnancy [
22]. At the same time, the number of pregnant women who take medications, as well as the number of medications that these pregnant women take, has increased [
6,
21].
Evidently, even after the awareness of “fair enrollment,” pregnant women remain poorly represented. Among the different reasons for the continuous underrepresentation is the problem that guidelines are ambiguous with respect to if, and when, pregnant women should be included in clinical research and what renders their inclusion fair [
23‐
26]. Many scholars and guidelines currently take the position that fairness comes down to the demand to justify the exclusion of pregnant women from research unless there are compelling “scientific reasons” for their exclusion [
9,
25,
27,
28]. It is questionable whether this approach to fairness renders research with women fair, since it has now transformed from the one extreme (no inclusion) to having to justify exclusion except when scientific reasons exist. Furthermore, apart from clear-cut cases, such as shown teratogenicity in preclinical studies or unfavorable high risks for the pregnant woman or the fetus, it is unclear what constitutes a scientifically compelling reason to exclude pregnant women. The National Institutes of Health’s
Policy and Guidelines on the Inclusion of Women and Minorities as Subjects in Clinical Research (2001 amendment) is currently the most elaborated guidance document to clarify this “scientific reason” in relation to clinical research in women [
29]. Nevertheless, we will argue below that this document has methodological and ethical shortcomings when applied to
pregnant women. Therefore, the aim of this paper is to analyze and evaluate when research with pregnant women can be considered as fair and what constitutes scientific reasons for exclusion.
Methods
We will first perform a conceptual ethical analysis of fair inclusion and then look at fair inclusion from an integrated ethical and methodological perspective by applying the National Institutes of Health (NIH) Policy document to pregnant women.
It is important to note that we assume that scientific- and justice-based reasons are highly integrated and in principle not easy to distinguish. If research is not designed in a scientifically rigorous manner, participants may unnecessarily be exposed to research risks [
30]. We will focus primarily on phase III drug research and we assume that a phase III trial is always preceded by sufficient phase I and phase II trials in pregnant women in order to obtain safety and dosing data to be able to expect that the drug is, and will remain, safe enough in pregnant women, and that, therefore, the risk of serious adverse effects is low. We will not touch upon the level of evidence needed to be able to conduct trials in pregnant women, nor on timing of trials in pregnant women and hence also not on trial designs and models that may speed up knowledge generation in this field. Finally, although our paper focuses primarily on the interests of pregnant women, the findings may also be of relevance to other underrepresented groups including breastfeeding women.
Discussion
Fair inclusion of pregnant women means (1) that pregnant women who are eligible are not excluded solely for being pregnant and (2) that the research interests of pregnant women are prioritized, meaning that they ought to receive substantially more attention. The first component of fair inclusion should not be mistaken for routine inclusion in virtually every trial. Fair inclusion has methodological limitations and exclusion can be justified for scientific reasons. We have described three scenarios that outline where scientific considerations should be taken into account. In scenario 1, it is known that intervention effects for pregnant women differ from those for non-pregnant women. We recommend that pregnant women in this scenario should not be included in phase III drug research that consists of non-pregnant women, but to initiate separate trials for pregnant women during phase III or to conduct phase IV and post-marketing studies.
Alternatively, we know that no differences exist (scenario 2), or we are uncertain whether differences exist (scenario 3). In scenario 2, when we know that there are no differences, it may be best to conduct post-marketing studies or to establish registries, such as the pREGnant registry that has been developed by the Netherlands Pharmacovigilance Center Lareb [
40]. Also, when we assume rather than know that there are no differences, we should refer back to the default of scenario 1. In scenario 3, when there is no sufficient prior information, which will in most instances be the case, it may be preferable to return to scenario 1 and to conduct separate trials in pregnant women based on scientific and precautionary considerations. If there is prior information but the information does not indicate either differences or no differences, the inclusion of pregnant women should be sufficient, which explicitly should not mean just enrolling only a few pregnant women in a trial. In this scenario, sufficiency boils down to representativeness in terms of the proportion of pregnant and non-pregnant women or to actually oversampling of pregnant women, depending on the actual research question.
Regarding the second component of fair inclusion, our paper has shown that fair inclusion cannot, and should not, be realized at the moment of ethical review of already designed research projects, but rather that fair inclusion requires a joint effort. Due to the current vagueness of the demand to justify exclusion unless scientific reasons exist and the ambiguity as to the level at which and the actors at whom fair inclusion is directed, no group or institution seems to make fair inclusion its sincere priority.
At present, it seems that fair inclusion only comes into play at the moment of ethical review of already designed individual research projects. However, our paper has demonstrated that the establishment of separate trials has to be realized at the earliest phases of research with pregnant women and that the demand to justify the exclusion of pregnant women cannot be bestowed upon individual researchers and research ethics committees, since protocols are not easily adjusted once researchers have planned their study methods and budgets may be restricted. Additionally, researchers that may be willing to include more pregnant women or to develop separate trials will need extra budget to do so. And thus funders and scientific advisory councils must see it as their priority to promote research with pregnant women and to facilitate the research infrastructure [
18]. In this respect, it will also be important to pay more attention to in vitro studies, that currently hardly distinguish between sexes in cell lines and hence contribute to the poor pre-clinical evidence base for drugs in (pregnant) women.
Moreover, in order to develop truly innovative medications for pregnant women, we cannot rely on investigator-initiated research only and we have to look at pharmaceutical companies. Pharmaceutical companies may be asked to substantially invest in sex-specific dosage or medications, yet, with the costs involved in research and development on this topic, together with additional packaging, marketing and liability fears, they may, understandably, be reluctant. Their additional risk is that an alternative company will claim equal effectiveness for both men and women for their compound, which may be preferred by physicians and society. The marketing campaign for sex-specific medications could turn out to be detrimental. Nevertheless, this year Ferring Pharmaceuticals launched NOCDURNA with gender-specific doses tailored to men and women. The success of this compound and the success of the gender-specific strategy is to be determined in the coming years.
In addition, the integrated analysis of fair inclusion has demonstrated that in most cases it will be essential to establish separate trials or registries and this is typically an activity that necessitates the involvement of authorities, such as national pharmacovigilance centers or regulatory authorities such as the FDA and EMA. However, although the role of the FDA and EMA is regulatory and they may guide the directions, they cannot require of pharmaceutical companies to conduct separate trials in (pregnant) women, unless it is laid down in a regulation or directive such as the EU regulation, comparable to research with children [
41]. Similar to the Paediatric Regulation in Europe with a Paediatric Committee and the requirements for Paediatric Investigation Plans (PIPs) for marketing approval, the EMA could establish a pregnancy committee and require pregnancy investigation plans if the drug can potentially be used by pregnant women.
Additional stakeholder groups are journal editors and pregnant women themselves. Journal editors could for instance require subgroup analyses from researchers who submit papers to their journal. Currently, this requirement is still a rarity and does not apply to the conduct of separate trials. Pregnant women could associate in patient groups which, in other medical fields, such as the field of orphan diseases or pediatric research, has had success in stimulating drug development. Without patient groups, radical breakthroughs can only be initiated by others than those whose interests are at stake.
In sum, although it is beyond the scope of this paper to conclusively state whose responsibility it is to ensure corrective justice and to prioritize the health interests of pregnant women in research, our paper shows that fair inclusion of pregnant women in research must primarily be seen as a joint responsibility to further the evidence base for drug use in pregnant women.
Conclusions
The demand to justify the exclusion of pregnant women from research is not only essential for reasons of equity but also for reasons of corrective justice. Since scientific knowledge on the effects of treatments for the health needs of pregnant women is relatively underrepresented, fair inclusion implies that intensive stimulation of research in this population is justified. Fairness does not imply that pregnant women should be included in virtually every research project. Inclusion of only a few pregnant women in a population of women will not help to determine the effectiveness and safety of a treatment in pregnant women. If pregnant women are included it should be done representatively or they should be oversampled in order to be able to determine a difference in intervention effects between groups of pregnant and non-pregnant women. In the few cases where we may be certain that there are no differences between pregnant and non-pregnant women, we should conduct post-marketing studies or arrange the establishment of registries. But, since evidence is typically limited for the treatment of health conditions that affect pregnant women, we either know, or otherwise have to assume, that pregnant women differ from other subpopulations. Separate trials may then be preferable. The current vagueness of the demand to justify exclusion unless scientific reasons exist seems to indicate that fair inclusion only comes into play at the moment of ethical review of already-designed individual research projects. However, fair inclusion is not only an obligation for individual researchers and research ethics committees. The development of separate trials has to be realized at the earliest phases of research with pregnant women. In addition to researchers and research ethics committees, scientific advisory councils, funders, drug regulatory agencies, pharmaceutical companies, journal editors and others all have a joint responsibility to further the evidence base for drug use in pregnant women.