|Home | About | Journals | Submit | Contact Us | Français|
All authors collaborated in conceptualizing ideas, interpreting findings, and writing and reviewing drafts of the article.
With a $3 billion investment by the federal government, the National Children’s Study (NCS) recently began recruitment. The NCS is a golden—and potentially missed—opportunity to study one of the most underrepresented populations in clinical research: pregnant women.
As the nation’s largest-ever study of children’s health, the NCS will examine the effects of the environment on children from before birth to 21 years of age, with participants sampled primarily through women during pregnancy. Thus the NCS presents a rare opportunity to study the health of women during and after pregnancy, in addition to the health of their children.
On both moral and policy grounds, we make the case for inclusion of women’s health outcomes in the NCS.
The year 2009 marks an important threshold for the advancement of children’s health. After nearly a decade of careful planning, recruitment has begun for the pilot phase of the National Children’s Study (NCS), the largest longitudinal study of children’s health ever conducted in the United States. Authorized by Congress in 2000 and led by a consortium of federal agencies, including the National Institutes of Child Health and Development, the National Institute of Environmental Health Sciences, the Centers for Disease Control and Prevention, and the US Environmental Protection Agency, the NCS will examine the effects of environmental influences on more than 100 000 children, with data collected from before birth to 21 years of age. At a cost of more than $3 billion, the study aims to gather research findings that will critically inform the “basis of child health guidance, interventions, and policy for generations to come.”1
Although the potential benefits are enormous, the NCS may also be one of the largest missed opportunities for health advancement in recent history: the opportunity to prospectively and systematically study the health of women during and following pregnancy, as well as the babies they bear.
Many of the 4 million women who give birth each year in the United States face conditions that require medical treatment. Approximately two thirds of women are prescribed at least one medication other than a vitamin or mineral supplement during pregnancy.2 Yet as clinicians well know, the evidence base for determining how to treat the medical conditions of pregnant women is distressingly poor. Indeed, some have argued that there is no group of patients for whom the evidence base is weaker.3 Only a dozen medications are approved by the Food and Drug Administration for use during pregnancy, and all are for gestation- or birth-related issues such as anesthesia or nausea.4 Any medicine taken to treat a woman’s illness during pregnancy—from hypertension5 to cancer,6 thromboembolism7 to asthma8—is used without data adequate to guide dosing, make decisions about safety, or inform differential decisions about which medicine to prescribe.
The cost of this ignorance is profound.9 The physiological changes of pregnancy affect the metabolism and activity of medications in dramatic and often unpredictable ways. For instance, the typical 30% to 40% increase in renal blood flow during pregnancy causes some medications to be cleared at much higher rates than in nonpregnant women.10 Increases in blood volume, decreases in gastric-emptying time, changes in the concentrations of sex hormones, alterations in liver enzymes, and the presence of a fetal–placental unit can all alter a drug’s pharmacokinetic and pharmacodynamic properties in ways that can profoundly affect efficacy. For example, a recent study revealed that amoxicillin, prescribed routinely to pregnant women and recommended by the American College of Obstetricians and Gynecologists as postexposure prophylaxis for anthrax during pregnancy, may in fact be metabolized so quickly as to prevent its reaching concentrations adequate to prevent anthrax.11 Similarly, pharmacokinetic measurements on pregnant women taking oral medication for diabetes revealed that the drug is metabolized and excreted so quickly that considerably higher doses than are currently recommended will likely be needed to treat them effectively.12
In the absence of reassuring data, clinicians and patients often undertreat or halt medications indicated for medical conditions that continue or emerge during pregnancy. But the failure to treat illness can also lead to significant harm to women and their fetuses—indeed, harm that easily can outweigh the possible risks that might accompany medication use. Poorly treated asthma is associated with adverse outcomes for women (preeclampsia, hemorrhage) and fetuses (growth restriction, prematurity); by contrast, women whose asthma is well controlled with medications have perinatal outcomes similar to comparable groups without asthma.13 Untreated major depression, too, is associated with worse outcomes for the pregnant woman and her fetus than is depression controlled with medication.14 And when an untreated woman dies from cancer after pregnancy, neonatal health concerns are one thing, the wrenching implications of life without a mother quite another.
One cause of the dearth of data is problematic enrollment practices. Despite a 1994 Institute of Medicine report recommending that pregnant women be “presumed eligible for participation in clinical studies,”15 many institutional review boards still regard pregnancy as a near-automatic cause for exclusion, even for studies that carry virtually no risks. Federal requirements for paternal consent in some situations—in conflict with rejection of such requirements by the Institute of Medicine15 and American College of Obstetricians and Gynecologists16—are an ongoing area of controversy. But a key barrier, of course, is that inclusion of pregnant women in clinical research brings with it genuine ethical complexities. It can be difficult to strike the appropriate balance between fetal protection, permissible trade-offs in maternal and fetal risks, and sound scientific methodology. Although some progress is being made with innovative study protocols, such as administration of medications immediately prior to delivery to evaluate absorbtion,17,18 a substantial core of knowledge for the adequate treatment of pregnant women will of necessity have to come from large observational studies that do not put fetuses at any additional risk.
This is just what the NCS represents—and in dramatic form. Here we have a very large study in which pregnant women are by design enrolled. Of the 100 000 children to be studied, researchers aim to enroll 25% prior to conception and a cumulative 90% during the first trimester of pregnancy. In practical terms, this means that researchers will be in regular contact with more than 90 000 women throughout the course of their pregnancies. Moreover, as researchers continue to collect data on the effects of the environment on the children that are born, they will stay in touch with these women for years to come. Such a cohort would yield an estimated 4000 women with diabetes, 4000 women with pregnancy-associated hypertension, 1000 women with chronic hypertension,19 12 000 women with depression,20 1000 women whose pregnancies were conceived with assisted reproductive technologies,21 4000 to 8000 women with asthma,22 and 2700 women with thyroid disease.23
Further, the NCS is positioned to address a range of issues regarding women’s medical needs that other observational studies under way are not. The Norwegians have launched a large observational Mother and Child Cohort Study—a title reflective of its goal: “to find causes of serious diseases in mothers and children.”24 Like the NCS, this study aims to enroll 100 000 women and infant pairs. Yet available data suggest that this study, although helpful, will not address the range of health issues and outcomes that could be followed by the NCS. For instance, enrollment in the Norwegian Cohort does not start until the middle trimester (around 17 weeks’ gestation): critical first-trimester health data exposures that the NCS has been so careful to include will thus not be collected. Moreover, data collection is completed when children are 6 (rather than 21) years of age. The Norwegian cohort data will also be derived from self-administered questionnaires and linked to medical records, as well as to blood and urine samples collected at enrollment and birth. By contrast, families enrolled in the NCS will participate in a minimum of 15 in-person visits; many more communications via telephone, computer, or questionnaire; and the collection of biological specimens and environmental sampling from home, school, and other environments. The NCS also has as a signature advantage in its sampling of individuals from the geographically and otherwise diverse US population.
The NCS represents, in short, a potential wealth of information that could dramatically improve the health prospects of women—during their pregnancies and in the years after they give birth—about which we currently know precious little.
As designed, however, the NCS will miss this golden opportunity. Although the study has been refined over time to include more complete information about the health status of pregnant women and pregnancy outcomes,25 virtually all data about pregnancy are collected only as predictors for fetal or pediatric outcomes, not as predictors for women’s health itself. Pregnancy outcomes to be recorded by the NCS are limited to preterm birth and structural congenital abnormalities, including subtle variations in morphogenesis, fetal growth restriction, and pregnancy loss. Strikingly, they include no maternal health outcomes. No data are to be collected on, for example, hemorrhage, transfusion, cardiovascular events such as heart attack or stroke, or pregnancy-induced hypertension or preeclampsia, a disease that is as common as it is poorly understood.
One maternal outcome being considered for potential inclusion relates to pelvic floor disorders and their relationship to delivery—an issue that, although important, is dwarfed by the serious and urgent nature of many medical conditions that afflict pregnant women. Perhaps most notably, although the study protocol already involves blood draws, it does not include collection of any data from those samples about the pharmacokinetics and pharmacodynamics of medications—data that are fundamental to determining the safe and effective dosing of required medications. In short, although the NCS will provide important and much-needed information about the effects of maternal medication use on the health of the fetus, absent is the other side of the coin: whether the medication is safe for the woman herself, whether it is effective in treating her underlying or emergent illness, and whether decisions to forgo medication or substitute older medications compromise her health.
Clearly, expanding the NCS to include these measures would involve additional effort and costs. Already, the NCS is under serious financial pressure: given the expanded geographical and demographic size of the study and challenges inherent to cost projections over 25 years, the original $3.2 billion approved for its duration will likely not be adequate to study all of the children’s outcomes desired. Adding further outcomes without expanded appropriations would make hard decisions about priorities even more difficult. Instead, the issue should be one of additional funding. Although money is always tight, it is hard to imagine a more compelling case for expenditure than the modest expansions required to capture critical information about women’s health. Given the advantages of piggybacking on the NCS and the valuable information that could result, the addition of maternal outcomes to the study would represent an enormous efficiency, especially compared with the cost of conducting a separate study focused solely on pregnant women.
Even more compelling than efficiency are issues of justice. Pregnant women have long been underrepresented in research; they and their needs have been substantially absent from social investments to advance medical knowledge.26 It is unjust for pregnant women to continue to benefit less than the rest of us from our enormous national effort to improve health through medical research. The NCS also raises specific issues of justice, given that enrollment of pregnant women is central to the success of the study. What was already unlikely—that a separate study of similar size geared toward measuring maternal outcomes would ever be funded in the United States—is even less likely now that investment in the NCS is under way.
So, what should be done? At a minimum, we strongly urge that the study’s core protocol be expanded to include two key components. First is collection of additional data, during pregnancy and around the time of birth, from the interviews and maternal chart reviews, already planned as part of the study, for purposes of addressing questions of maternal health in its own right. Because enrollment for the pilot phase of the study has just begun, such changes could be made before the core protocol is finalized in May 2010 without compromising sample size.
Second, we urge inclusion of opportunistic pharmacokinetic studies by taking advantage of the maternal blood draws already incorporated in the study’s core protocol. Blood drawn from women already on medications can be used to garner information about how drugs are metabolized in the pregnant body, across populations and trimesters. In fact, population pharmacokinetic studies could be done with limited additional expense, requiring only documentation of timing and dosage of medication use and timing of the blood draw. Such information would, at the very least, help to guide identification of drugs requiring intensive pharmacokinetic studies. Coupled with dose efficacy studies procured from nonpregnant populations, such studies could also provide long-sought information on how to effectively dose pregnant women with severe diseases.
Data from these modest additions would represent an extraordinary opportunity to address critically important questions. These include the near-term effect on women of taking, changing, or discontinuing antihypertensive and antidepressant medications during pregnancy, as well as the safety and kinetics of new antiepileptic drugs during pregnancy (often preferred in the nonpregnant population for more effective seizure control and fewer side effects), and of unfractionated and low-molecular-weight heparin for the treatment and prevention of venous thromboembolism, for which women are at a fourfold risk during pregnancy.27 They might even provide much-needed data about the management and outcomes of cerebrovascular events during pregnancy, for which we have only a handful of case reports.28
Indeed, these two efforts together—at very little additional cost—would result in a rich data set that could lead to critical improvements in the care of pregnant women. Failure to include them would be hard to justify.
Of course, more ambitious studies could lead to even more valuable data. Of particular note would be investigating the longer-term public health impact of treatment patterns and decisions on pregnant women. How do decisions to continue or forgo antidepressants during pregnancy affect psychiatric health over time? What are the long-term effects of using older drugs routinely prescribed to pregnant women instead of newer drugs now used for the nonpregnant population in treating chronic illnesses? What are the long-term effects on women of using assisted reproductive technologies? Because contact with women in the NCS will continue both for assessment of the child’s health and for surveillance for subsequent pregnancies, periodic documentation of women’s health outcomes at longer intervals (5, 10, 15 years) is feasible. As with all observational studies, the study design limits the variables that can be controlled; still, it is widely agreed that observational studies of such magnitude can yield invaluable information. Just as the NCS will help to define the role of a breadth of factors on children’s health, it has the potential to do the same for their mothers.
These more expansive efforts would take more expansive resources. Ideally, once women’s health advocates and others begin to appreciate just how much we can learn, for just how little, by adding these more ambitious outcomes to the NCS, political support and revenue streams might be forthcoming.
Whether modest or ambitious, expansion to include maternal outcomes will require additional funding. Such funding might involve creative collaboration across institutes with strong interests in specific diseases relevant to pregnancy, such as the National Heart, Lung, and Blood Institute; the National Institute of Diabetes and Digestive and Kidney Diseases; and the National Institute of Mental Health, together with the Office of Research on Women’s Health. It could also involve funding from industry, foundations, and advocacy organizations. Although as many key maternal outcomes as possible should be incorporated into the core protocol, consideration should also be given to the addition of adjunct studies, which are foreseen as a part of the broader efforts to enhance the value of the NCS. If expanding the NCS in these ways will be costly, not taking advantage of the NCS will be costlier still.
The NCS is certainly one of the most exciting research endeavors of the century—both for what it promises for understanding and promoting the health of children and for the opportunity it presents to advance the health of women who will gestate and parent them. The opportunity to address maternal outcomes is a watershed moment in the challenging history of research and treatment of illness in pregnant women. It is an opportunity we cannot afford to miss.
Anne Drapkin Lyerly, Department of Obstetrics and Gynecology and the Trent Center for Bioethics, Humanities, and History of Medicine, Duke University, Durham, NC.
Margaret Olivia Little, Kennedy Institute of Ethics, Georgetown University, Washington, DC.
Ruth R. Faden, Berman Institute of Bioethics, Johns Hopkins University, Baltimore, MD.