Chapter  14:  Management of Chronic Hypertension During Pregnancy: Evidence Report/Technology Assessment Number 14

Objectives

Of the nearly 4 million women giving birth in the United States each year, between 1 and 5 percent suffer from chronic hypertension. Chronic hypertension is associated with serious maternal and fetal complications, including superimposed preeclampsia, fetal growth restriction, premature delivery, placental abruption, and stillbirth. There is uncertainty and practice variability in monitoring and treatment strategies for chronic hypertension in pregnancy. This evidence report addresses several questions related to management of chronic hypertension. It summarizes research evidence concerning the magnitude of maternal and fetal risks associated with chronic hypertension, risks and benefits of antihypertensive agents before and during pregnancy, risks and benefits of aspirin and nonpharmacological therapies during pregnancy, and the role of various monitoring strategies in detecting fetal complications of chronic hypertension.

Search Strategy

English and non-English research literature was identified from 16 electronic databases, references of pertinent articles and reviews, a primary text that routinely systematically reviews and categorizes teratogenicity risks, and technical experts. In general, electronic bibliographical sources were searched from 1947 or from their inception to February 1999. Four general search strategies were used to identify evidence relevant to: (1) efficacy, (2) harms, (3) blood pressure risks, and (4) monitoring techniques.

Selection Criteria

Selection criteria varied according to the specific question addressed. For questions concerning treatment efficacy, the investigators of this evidence report selected randomized trials of antihypertensive agents or aspirin compared with either placebo or usual care that included a population of pregnant women with chronic hypertension and that reported clinical maternal and/or fetal outcomes. For the question regarding harm associated with antihypertensive therapy, we selected case reports, case-control and cohort studies, randomized trials, and surveillance studies that reported adverse maternal and/or fetal outcomes. To estimate the magnitude of risk associated with chronic hypertension, we examined case-control and cohort studies that compared maternal and fetal outcomes in women with chronic hypertension compared with those in either a general obstetrical population or pregnant women without chronic hypertension. For the question concerning monitoring techniques, we sought case series, cohort studies, and randomized trials that reported clinical perinatal morbidity or mortality outcomes.

Data Collection and Analysis

Two or more persons independently screened the 5,558 titles and abstracts identified in the searches. Two reviewers abstracted articles meeting inclusion criteria except for articles addressing adverse effects, which were reviewed by only one person. Data were synthesized descriptively, emphasizing methodological characteristics of the studies such as populations enrolled, definitions of selection and outcome criteria, interventions and comparisons, and study designs. Because of concerns about heterogeneity in study populations and interventions, quantitative methods were not used to combine trial results. Random effects methods were used to estimate summary odds ratios or risk of perinatal mortality and abruption associated with chronic hypertension.

Main Results

• Benefits of treating chronic hypertension before conception. There was no evidence that addressed the effect of blood pressure control before conception on fetal outcomes. With regard to maternal outcomes, evidence from randomized trials involving nonpregnant women 30 to 54 years of age showed that approximately 250 (95 percent confidence interval 158 to 1,606) such women with mild to moderate hypertension need to be treated for 5 years to prevent one cardiovascular event such as stroke or myocardial infarction. Much larger numbers of women younger than age 30 (approximately 8,000) would need to be treated for 1 year to prevent a cardiovascular event.

• Benefit of treating chronic hypertension during pregnancy. There was insufficient evidence to prove or disprove moderate to large clinical effects of antihypertensive agents on perinatal outcomes.

• Adverse effects of antihypertensive drugs. The quality of evidence addressing this question was poor. The best-established adverse effect of antihypertensive agents in pregnancy was renal failure associated with use of the angiotensin-converting enzyme inhibitors. There was evidence suggesting that atenolol used early in pregnancy may be associated with small-for-gestational-age fetuses.

• Effects of nonpharmacological interventions. There was no evidence to address this question.

• Optimum levels for initiating therapy and risk of chronic hypertension in pregnancy. There was insufficient evidence to identify an optimum blood pressure at which treatment should be initiated and thereafter maintained. Chronic hypertension in pregnancy tripled the risk of perinatal mortality and doubled the risk of placental abruption. The risks of preeclampsia and small-for-gestational-age infants also were increased. Increased risk was evident even in the absence of superimposed preeclampsia.

• Effect of aspirin. A single trial showed low-dose aspirin begun before 26 weeks gestational age neither reduced perinatal morbidity and mortality nor increased maternal risks in women with chronic hypertension. Although the trial was of moderate size, small increases or reductions in benefits or risks could have been missed.

• Effectiveness of monitoring strategies. There was insufficient evidence to evaluate the effectiveness of any particular monitoring test or sequence of tests for women with chronic hypertension in pregnancy.

Conclusions

Despite the burden of illness and costs imposed by chronic hypertension in pregnancy, evidence to date remains scant and provides little direction for clinicians. Epidemiological data demonstrate increased risks of perinatal morbidity and mortality in pregnant women with mild to moderate chronic hypertension. Treatment with antihypertensive agents has not been proven to lower those risks, though the evidence base is too small to rule out moderate to large effects on perinatal mortality, preeclampsia, and intrauterine growth retardation. Data on adverse effects of drug treatment is scant; the data on angiotensin-converting enzyme inhibitors suggest that their adverse effects are substantially greater than for other drugs. Moderate to large benefits of low dose aspirin begun early in pregnancy for women with chronic hypertension have been disproved. Nonpharmacological treatments remain unevaluated, as do monitoring strategies that are frequently used in pregnancies complicated by chronic hypertension.

Many important clinical issues faced by clinicians who care for pregnant women with chronic hypertension remain unresolved. Research in this area is critical. More pregnancies will be complicated by chronic hypertension as the trend continues for women to delay childbearing to older ages. Multicenter collaborative studies are needed with sufficient power to detect differences in outcomes such as perinatal mortality, preeclampsia, and small-for-gestational-age infants. Information from such trials can be augmented by well-designed surveillance systems to monitor outcomes and safety data from clinical practice. Finally, comparative studies with clinical outcomes are sorely needed to assess benefits, costs, and harms of various fetal monitoring strategies for the pregnant woman with chronic hypertension.

This document is in the public domain and may be used and reprinted without permission except those copyrighted materials noted for which further reproduction is prohibited without the specific permission of copyright holders.

Suggested Citation

Mulrow CD, Chiquette E, Ferrer RL, et al. Management of chronic hypertension during pregnancy. Evidence Report/Technology Assessment No. 14 (Prepared by the San Antonio Evidence-based Practice Center based at the University of Texas Health Science Center at San Antonio under Contract No. 290-97-0012). AHRQ Publication No. 00-E011. Rockville, MD: Agency for Healthcare Research and Quality. August 2000.

Overview

Chronic hypertension, defined as hypertension diagnosed before pregnancy or before 20 weeks' gestation, complicates from 1 percent to 5 percent of all pregnancies. The incidence is expected to rise as the demographic trend towards childbearing at older ages continues. Chronic hypertension in pregnancy is associated with serious maternal and fetal complications, including superimposed preeclampsia, fetal growth retardation, premature delivery, placental abruption, and stillbirth. Superimposed preeclampsia accounts for much of the increased risk of complications. The complications have significant economic impact, including costs of treating sick mothers and neonates and costs of intensive antenatal monitoring aimed at early detection of complications.

This review was proposed to the Agency for Healthcare Research and Quality by the American College of Obstetricians and Gynecologists because of perceived widespread uncertainty about the best management of chronic hypertension during pregnancy. The review is a systematic review of existing literature and not a guideline. The particular focus of the review was prespecified as management of mild to moderate chronic hypertension (blood pressure less than 170/110 mmHg) for several reasons. First, the Joint National Committee for Prevention, Detection, Evaluation, and Treatment of High Blood Pressure currently recommends nonpharmacological treatment for women with mild to moderate hypertension and no concomitant cardiovascular risk factors. Actual treatment practice, however, varies in younger, low-risk women. Second, antihypertensive agents are used widely in pregnancies complicated by mild to moderate chronic hypertension, despite unclear tradeoffs between potential benefits and harms. Multiple agents are available, and various guidelines recommend different agents as either contraindicated or drugs of choice. Third, recent trials have provided new data regarding effects of aspirin in preventing preeclampsia. Fourth, there is wide variation in practice regarding antenatal monitoring strategies applied to women with chronic hypertension and uncertainty about benefits and harms for the mother and fetus.

Reporting the Evidence

The purpose of this report is to help clinicians make informed choices about therapeutic interventions for pregnant women with chronic hypertension and to aid organizations developing guidelines for the treatment of this condition. The report evaluates evidence regarding risks of chronic hypertension, benefits and adverse effects of pharmacotherapy (antihypertensives and aspirin), nonpharmacological treatment, and monitoring techniques.

An expert multidisciplinary panel formulated 10 specific questions that address key diagnostic and treatment decisions faced by clinicians who provide care for pregnant women with mild to moderate chronic hypertension.

Question 1: What is appropriate management of women with chronic hypertension before pregnancy?

Question 2: Do antihypertensive agents for the treatment of mild to moderate chronic hypertension (<170/110 mmHg) during pregnancy improve maternal and perinatal outcomes? Are there subsets of women, such as those with diabetes or renal disease, for whom treatment of mild to moderate chronic hypertension during pregnancy is warranted?

Question 3: Is pharmacological treatment of mild to moderate chronic hypertension during pregnancy harmful to mothers, fetuses, and infants? If harmful, what is the type and magnitude of specific harm for mothers, fetuses, or infants?

Question 4: Are particular antihypertensive agents more effective or harmful than others in treating mild to moderate chronic hypertension during pregnancy?

Question 5: Does nonpharmacological treatment of mild to moderate chronic hypertension during pregnancy improve maternal and perinatal outcomes?

Question 6: Is nonpharmacological treatment as efficacious as pharmacological treatment to improve maternal and perinatal outcomes in women with chronic hypertension?

Question 7: Does a combination of pharmacological and nonpharmacological treatment improve maternal and perinatal outcomes over either treatment alone?

Question 8: What is an appropriate blood pressure level at which to treat chronic hypertension during pregnancy and when should therapy be initiated? What is an appropriate blood pressure level at which to maintain treatment?

Question 9: Is aspirin beneficial in preventing maternal and fetal complications in pregnant women with mild to moderate chronic hypertension?

Question 10: Is the use of special fetal monitoring techniques (biophysical profiles, Doppler velocimetry, nonstress tests, contraction stress tests, fundal measurements, amniotic fluid index, ultrasound fetal biometry, fetal movement counting) and strategies (priority ordering of testing, number of tests, or timing of tests) beneficial or harmful to mothers and fetuses? Are there particular subsets of women for whom special monitoring techniques are warranted?

The primary outcomes of interest were perinatal mortality, growth restriction, preterm birth, abruption, and superimposed preeclampsia. Secondary outcomes included Apgar scores, maternal complications, rate of cesarean delivery, and length of gestation. We anticipated that data would be absent or scant for some of the questions and that specific gaps in the currently available evidence base would be identified.

Methodology

English and non-English language literature was identified from multiple sources, including several electronic databases (e.g., MEDLINE, EMBASE, Biological Abstracts, CINAHL, Cochrane Controlled Trial Registry and Pregnancy and Childbirth Database, REPROTOX, and TERIS), pertinent textbooks, and references from retrieved articles and technical experts. Databases were searched back to 1947 or to their inception. Four overarching search strategies, based on the questions' population of interest, pertinent intervention or exposure, and relevant study designs, were used.

Two or more persons independently screened the 5,558 titles and abstracts identified in the searches. Two reviewers abstracted articles meeting inclusion criteria except for articles addressing adverse effects, which were reviewed by only one person. Data were synthesized descriptively, emphasizing methodological characteristics of the studies such as populations enrolled, definitions of selection and outcome criteria, interventions and comparisons, and study designs. Relationships between clinical outcomes, participant characteristics, and methodological characteristics were examined in evidence tables and graphical summaries such as forest plots and L'Abbe plots. Because of concerns about heterogeneity in study populations and interventions, quantitative methods were not used to combine trial results. We used random effects methods to estimate summary odds ratios or risk of perinatal mortality and abruption associated with chronic hypertension.

Findings

Future Research

Clinicians must grapple with a number of important decisions in caring for pregnant women with mild to moderate chronic hypertension: Is any special management necessary preconception? Should antihypertensive therapy be prescribed? If antihypertensive therapy is prescribed, should a specific drug be given? If so, at what blood pressure should it be started and to what target titrated? Should specific agents be avoided? If a woman with hypertension who is already well controlled with a particular antihypertensive agent becomes pregnant, should another agent be substituted? Are any nonpharmacological interventions of benefit? Is aspirin more beneficial than harmful? If used, what dose is appropriate and when should it be initiated and discontinued? How intensively should women with mild to moderate chronic hypertension be monitored for complications and with what tests?

This evidence report shows that these clinically salient questions are not well addressed with rigorously designed research. A pervasive problem is that the evidence base on chronic hypertension in pregnancy is small. There are few studies, and available studies typically have small numbers of participants and low power to detect moderate or sometimes large effects for important outcomes. A potpourri of women with different "high risk" obstetrical conditions has been studied, which complicates interpretation of results and alters precision of outcome measurements. Potential adverse effects of many antihypertensive drugs in pregnancy are either poorly established or unclearly quantified because of selection biases and coincidental occurrences that are reported in case reports and surveillance studies. Virtually no relevant research data with important outcomes are available to guide selection of fetal monitoring strategies in pregnant women with chronic hypertension.

Advancement of clinical knowledge regarding management of chronic hypertension during pregnancy requires a multipronged approach.

• A better understanding of current practice, including its motivations and its variations, is needed. Racial disparities in management approaches warrant study, as well as whether disparities are due to varying patient factors such as access and preferences or to varying provider approaches.

• Benefits and harms of commonly used, but unproven, therapies need to be tested in studies that are large, collaborative, multicenter, and population based and that enroll women with clearly established mild to moderate chronic hypertension. To detect moderate (20 percent) relative risk reductions in preeclampsia, intrauterine growth retardation, and perinatal death with adequate power (80 percent), randomized trials are needed with enrollments of approximately 1,000, 3,000, and 6,000 women with chronic hypertension, respectively. Trials testing important outcomes such as preterm birth, neonatal intensive care utilization, or combined outcomes could be smaller.

• Because efficacy of pharmacological treatment is not proven, therapy begun early in the course of pregnancy should be compared with placebo as well as among alternative commonly used drugs.

• Advancement of knowledge concerning incidences and risks of adverse effects requires more and better surveillance systems that routinely monitor adverse events and numbers of women exposed to particular agents. Population-based case control studies and more large multicenter cohort studies that give careful attention to both selection and reporting biases also will help elucidate adverse effects.

• Finally, to establish appropriate and cost-effective methods of monitoring women with chronic hypertension during pregnancy, large trials are needed that compare alternative strategies and use clinically important outcomes. In the absence of such trials, creative studies with case control designs and careful control for confounding factors may be helpful.

Purpose of Report and Target Audience

The purpose of this evidence report is to provide evidence-based information for health care providers regarding management and outcomes in women with chronic hypertension during pregnancy. Issues addressed include: benefits and risks of antihypertensive drugs to lower blood pressure in women with mild to moderate hypertension before and during pregnancy; appropriate blood pressure levels at which to initiate treatment; benefits and risks of special fetal monitoring techniques; and maternal and fetal benefits and risks of low-dose aspirin. We anticipate the report will be valuable to clinicians and patients desiring evidence for informed choices about pharmacological therapies and to organizations developing clinical guidelines for the management of chronic hypertension in pregnancy.

An expert multidisciplinary panel refined the focus of the report by identifying several important relevant questions. The guiding principles were to address questions with the potential to develop new information not addressed in previous literature synthesis, relevance to clinicians' making treatment decisions, and relevance to policymakers' developing guidelines. The resulting list of questions (given in Chapter 3) was long and broad based. It was anticipated that data would be sparse or even absent for some questions but that important gaps in evidence would be highlighted.

Definition and Classification of Chronic Hypertension

In the United States, the Joint National Committee for Prevention, Detection, Evaluation, and Treatment of High Blood Pressure classifies hypertension in adults into three stages: stage one (systolic BP of 140-159 mmHg or diastolic BP of 90-99 mmHg), stage two (systolic BP of 160-179 mmHg or diastolic BP of 100-109 mmHg), and stage three (systolic BP >180 mmHg or diastolic BP >110 mmHg). In general, the 5^th Korotkoff phase is recommended for diastolic measurements, except when the 5^th phase falls to 0 and the 4^th phase is used instead. For therapeutic decisions, patients with hypertension are stratified according to presence or absence of other cardiovascular risk factors (e.g., age, dyslipidemia, tobacco use, diabetes, family history of early cardiovascular disease, obesity, sedentary lifestyle), known clinical cardiovascular disease, and target organ damage.¹

There are no universally accepted criteria for the diagnosis of chronic hypertension in pregnancy. In the United States and in this report, the diagnosis of chronic hypertension in pregnancy is based on either a history of hypertension before pregnancy or the presence of hypertension before 20 weeks gestation.² Such hypertension can be characterized as mild to moderate or severe. For this report, mild to moderate hypertension is defined as a systolic blood pressure of 140-169 mmHg or a diastolic pressure of 90-109 mmHg; severe hypertension is defined as a systolic pressure of at least 170 mmHg or a diastolic pressure of at least 110 mmHg. The report focuses on management of mild to moderate chronic hypertension during pregnancy. Severe hypertension is not addressed because it is less common than mild to moderate hypertension and there is less uncertainty regarding whether to treat women with severe hypertension.

Natural History of Blood Pressure in Pregnancy

Mean blood pressure during pregnancy has a curvilinear pattern. It begins to decrease by the end of the first trimester with the lowest readings occurring around 16 to 18 weeks. Blood pressure then rises continuously to 36 weeks followed by a plateau until delivery.^{3,4, 5} Older pregnant women typically have slightly higher blood pressures than younger pregnant women.⁵ Multiple studies show that higher mean arterial blood pressure during the second trimester is positively associated with gestational hypertension.^6-19

Epidemiology

Hypertension during pregnancy, including chronic hypertension, is the most common medical disorder in pregnancy. The exact incidence of chronic hypertension during pregnancy in the United States is unknown; reported rates have ranged from 1 to 5 percent.¹ Using data from the National Hospital Discharge Survey, Samadi, Mayberry, Zaidi, et al. reported an incidence of 1.3 percent in 12 million women who delivered in-hospital during the years 1988 to 1992.²⁰ The incidence of chronic hypertension during pregnancy was 2.5 percent in African Americans and 1 percent in other racial groups.²⁰

The prevalence of hypertension in women of reproductive age varies according to age, race, ethnicity, and body mass index.²¹ According to data derived from the Third National Health and Nutrition Examination Survey, 1988-1991, prevalence of chronic hypertension in women 18 to 29 years old was 2.0 percent for African Americans, 0.6 percent for Caucasians, and 1.0 percent for Mexican Americans. For women 30 to 39 years old, the respective rates were 22.3 percent, 4.6 percent, and 6.2 percent. In women 40 to 49 years old, the prevalence was 30.5 percent for African Americans, 12.7 percent for Caucasians, and 10.6 percent for Mexican Americans.²¹ U.S. natality statistics showed the total number of pregnancies in women ages 35 or older is increasing.²² The birth rate for women 40 to 44 years old increased by 56 percent from 1980 to 1993, perhaps because of marriage or remarriage later in life, or because women delay pregnancy until they complete educational or career goals.²³ Given the current trend of childbearing at an older age, it is expected that the incidence of chronic hypertension in pregnancy will continue to rise. During the new millennium, and estimating a prevalence of chronic hypertension during pregnancy of 2 percent, at least 80,000 pregnant women with chronic hypertension per year will be seen in the United States.

Burden of Illness, Risks, and Economic Impact

The costs of managing women with chronic hypertension in pregnancy are high. Treatment costs include not only prenatal care and delivery, but also the expense of multiple laboratory tests, antenatal fetal evaluation testing, health care visits to control hypertension, prenatal hospitalization in some instances, and prolonged antenatal hospitalization. Pregnancies complicated by chronic hypertension are associated with increased rates of adverse outcomes such as fetal deaths, fetal growth retardation, small-for-gestational-age (SGA) infants, premature delivery, and abruption.^2,24 The frequency of these complications is particularly increased in women with superimposed preeclampsia, severe hypertension,²⁵ and long-standing hypertension and in those with preexisting cardiovascular and renal disease.^26,27 The combination of superimposed preeclampsia and abruption are leading causes of disseminated intravascular coagulopathy, acute renal failure, pulmonary edema, and intracerebral hemorrhage in pregnant women.²⁸ The direct and indirect costs attributable to such complications are extremely high considering acute and long-term costs for infants born prior to 34 weeks' gestation and costs attributable to management and long-term care of women with the above complications.

Current Management and Variation in Practice

Preconception management of mild to moderate hypertension varies in younger, low-risk women. Currently, the Joint National Committee for Prevention, Detection, Evaluation and Treatment of High Blood Pressure recommends nonpharmacological treatment for young women without concomitant cardiovascular risk factors. Antihypertensive agents are used widely in pregnancies complicated by mild to moderate chronic hypertension, despite unclear tradeoffs between potential benefits and harms. Multiple agents are available and various guidelines recommend different agents as either contraindicated or drugs of choice. For example, guidelines in the United States recommend pharmacological treatment for pregnant women with diastolic blood pressures >100 mmHg and cite methyldopa as the drug of choice and angiotensin-converting enzyme (ACE) inhibitors as contraindicated. Guidelines in Canada recommend pharmacological treatment for blood pressures >140-150/90-95 mmHg. Listed drugs of choice include methyldopa, labetalol, pindolol, oxprenolol, and nifedipine. ACE inhibitors and angiotensin receptor blockers (ARBs) are cited as drug classes to avoid. In Australia, treatment is recommended for blood pressures >159/89 mmHg. Preferred drugs are methyldopa, labetalol, oxprenolol, and clonidine, whereas both ACE inhibitors and diuretics are cited as drug classes worth avoiding.

Various nonpharmacological treatments have been recommended such as bed rest, limited activity, diet modification such as salt and protein intake restriction, stress reduction, and cessation of smoking and alcohol. Recent trials have provided new data regarding effects of aspirin in preventing preeclampsia. There is variation in practice regarding antenatal monitoring strategies applied to women with chronic hypertension and uncertainty about benefits and harms for the mother and fetus. Recommended fetal evaluations have included serial ultrasound measurements of fetal growth, nonstress testing, biophysical profile, or Doppler flow velocimetry measurements.

Challenges in Interpreting Research Evidence

Several problems emerge in a review of the literature regarding management of mild to moderate chronic hypertension in pregnancy. Varying criteria have been used to classify severity of hypertension. In some studies, patients with all forms of hypertension during pregnancy, including chronic, gestational, and preeclampsia, have been considered together. Some studies have included women with hypertension detected at various gestational ages up to 32 weeks. Other studies have included and mixed women with markedly varying underlying baseline risks for adverse outcomes, such as women with no risk factors other than hypertension, women with multiple risk factors including diabetes, and women without hypertension but with another risk factor such as renal disease.

Definitions of superimposed preeclampsia have included one or more of the following: exacerbation of hypertension, development of edema, elevation in uric acid levels, and appearance of proteinuria. Exacerbation of hypertension in women receiving antihypertensive medications has been assessed differently and has not always taken into account frequency of dosing or the total dose used. Some studies have included "important" edema and "critical" levels of uric acid in their definitions of preeclampsia whereas others have not, since women with silent renal disease and chronic hypertension may have edema and elevated uric acid early in pregnancy. Many studies have included proteinuria in their definitions without taking into account that baseline proteinuria may confound predicating diagnosis on appearance of proteinuria.

Questions Addressed in Evidence Report

Table 1. Questions and selection criteria grouped by (more...)

Table 1. Questions and selection criteria grouped by search strategies

Search 1 for efficacy data and randomized controlled trials
Pertinent questions	Selection criteria
What is appropriate antihypertensive management of women with chronic hypertension before pregnancy? (Question 1) Do antihypertensive agents for the treatment of mild to moderate chronic hypertension (<170/110 mmHg) during pregnancy improve maternal and perinatal outcomes? Are there subsets of women, such as those with diabetes or renal disease, for whom treatment of mild to moderate chronic hypertension during pregnancy is warranted? (Question 2)	RCT or systematic review of RCTs Participants: women of childbearing age or pregnant with "mild to moderate" chronic hypertension Antihypertensive pharmacological therapy Control group: placebo, usual care, or another agent Maternal and/or fetal morbidity or mortality outcome
Does nonpharmacological treatment of mild to moderate chronic hypertension during pregnancy improve maternal and perinatal outcomes? (Question 5) Is nonpharmacological treatment as efficacious as pharmacological treatment to improve maternal and perinatal outcomes? (Question 6)	RCT Pregnant participants with "mild to moderate" chronic hypertension Nonpharmacological therapy Control group: placebo, usual care, or pharmacological therapy Maternal and/or fetal morbidity or mortality outcome
Does a combination of pharmacological and nonpharmacological treatment improve maternal and perinatal outcomes over either treatment alone? (Question 7)	RCT Pregnant participants with "mild to moderate" chronic hypertension Pharmacological plus nonpharmacological therapy Control group: pharmacological or nonpharmacological therapy Maternal and/or fetal morbidity or mortality outcome
Is aspirin beneficial in preventing maternal and fetal complications in pregnant women with mild to moderate chronic hypertension? (Question 9)	RCT Pregnant participants with "mild to moderate" chronic hypertension Aspirin therapy Control group: placebo or usual care Maternal and/or fetal morbidity or mortality outcome
Search 2 for data about harms
Pertinent questions	Selection criteria
Is pharmacological treatment of mild to moderate chronic hypertension during pregnancy harmful to mothers, fetuses, and infants? If harmful, what is the type and magnitude of specific harms? (Question 3) Are particular antihypertensive agents more effective or harmful than others in treating mild to moderate chronic hypertension during pregnancy? (Question 4)	Case report, case control, cohort, surveillance studies, RCTs, systematic reviews Pregnant or nongravid participants Antihypertensive agent Control group: not required except for efficacy part of questions (search 1) Teratogenic or clinical fetal or maternal outcome
Search 3 for data about blood pressure risks and optimum treatment levels
Pertinent question	Selection criteria
What is an appropriate blood pressure level at which to treat chronic hypertension during pregnancy and when? What is an appropriate blood pressure level at which to maintain treatment? (Question 8)	Case control or cohort study, RCT Pregnant participants Chronic hypertension Control group: general OB population, women with normotension, preeclampsia, or varying hypertension severity
Search 4 for data about benefits and harms of special fetal monitoring techniques
Pertinent question	Selection criteria
Is the use of special fetal monitoring techniques (biophysical profiles, Doppler velocimetry, nonstress tests, contraction stress tests, fundal measurements, amniotic fluid index, ultrasound fetal biometry, fetal movement counting) and strategies (priority ordering of testing, number of tests or timing of tests) beneficial or harmful to mothers and fetuses? Are there particular subsets of women for whom special monitoring techniques are warranted? (Question 10)	Case series, cohort study or RCT Pregnant participants with chronic hypertension Control group: not required, but any accepted Individual monitoring techniques Perinatal morbidity or mortality outcomes

RCT = randomized controlled trial; OB = obstetric

Literature Search and Selection Methods

Sources and Search Methods

Table 2. Electronic sources searched

Table 2. Electronic sources searched

Electronic database	Description
Biological Abstracts	Contains references to life science journal literature from 1972 to the present. Coverage is international and includes agriculture, biochemistry, biomedicine, biotechnology, botany, ecology, microbiology, pharmacology, and zoology.
Canadian Coordinating Office for Health Technology Assessment (http://www.ccohta.ca/)	Contains information about ongoing internal and commissioned projects.
CINAHL (Cumulative Index to Nursing and Allied Health Literature)	Includes citations from over 500 biomedical and popular sources, including NLN and ANA publications. Covers publications from 1982 to the present and is considered the premier nursing database.
Cochrane Library (http://www.cochrane.org) DARE (Database of Reviews of Effectiveness) (http://www.york.ac.uk/inst/crd/ ) The Cochrane Controlled Trials Registry The Cochrane Hypertension Review Group Specialized Registry The Cochrane Pregnancy and Childbirth Database	Contains references of randomized controlled trials and systematic reviews identified from electronic bibliographic sources and hand searching of multiple journals and symposia or meeting proceedings.
Conseil d'Evaluation des Technologies de la Sante (http://www.msss.gouv.qc.ca)	Has results of health technology assessments.
EMBASE	Contains biomedical and pharmaceutical citations and is considered the premier biomedical database in Europe.
FedRIP (Federal Research in Progress)	Contains information about ongoing federally funded research projects in physical sciences, life sciences, and engineering.
Health Services Utilization and Research Commission (http://www.ccohta.ca/main-f.html)	Contains information about factors affecting use of health services and health services resources and about the effectiveness of health procedures, practices, and technologies.
HealthSTAR (Health Services, Technology, Administration, and Research)	Contains citations to published literature in health services, technology, administration, and research.
Institute for Clinical Evaluative Sciences (ICES) (http://www.ices.on.ca/)	Includes information about ICES clinical and medical research plus selected articles from Informed, ICES newsletter.
MEDLINE	Indexes almost 4,000 international biomedical journals from 1966 to the present. Includes references from Index Medicus, International Nursing Index, and Index to Dental Literature and is considered the premier biomedical database in the United States.
Motherisk Program (http://motherisk.org/)	Contains evidence-based information about potential risks to the fetus or infant from exposure to drugs, chemicals, diseases, radiation, and environmental agents.
National Institute of Maternal and Child Health and Development publications and clearinghouse (http://www.nmchc.org/)	Contains a complete inventory of clearinghouse holdings, new titles, and a searchable subject and title database.
NHS CRD HTA (National Health Service Centre for Reviews and Dissemination, Health Technology Assessment) (http://nhscrd.york.ac.uk/welcome.html)	Contains abstracts produced by International Network of Agencies for Health Technology Assessment and other health care technology agencies.
REPROTOX (http://reprotox.org/)	Contains assessments on potentially harmful effects of environmental exposure to chemicals and physical agents on human pregnancy, reproduction, and development.
TERIS: Teratogen Information System (http://www.mdx.com/)	A component of the REPRORISK System produced by MICROMEDEX. Contains information on the possible teratogenic effects of drugs and environmental agents on pregnant women.

NLN = National League for Nursing; ANA = American Nurses Association

English and non-English citations were identified from the 16 electronic databases cited inTable 2, references of pertinent articles and reviews, and technical experts. Searches regarding harms were supplemented with information from a primary text that routinely systematically reviews and categorizes teratogenicity risks.²⁹ Serious adverse effects of antihypertensive agents that have been observed in the general nongravid population were identified from two routinely updated textbooks.^30,31

Four general search strategies were used to identify evidence relevant to (1) efficacy, (2) harms, (3) blood pressure risks, and (4) monitoring techniques. Detailed descriptions of the searches are presented in Appendix A. In general, bibliographical sources were searched from 1947 (or their inception) to February 1999. Bibliographical searches identified 6,228 records of which 670 were duplicate citations.

Selection Processes

Table 1 presents selection criteria for specific review questions arranged by type of search strategies that were used. Criteria specified the types of participants, interventions, control groups, outcomes, and study designs that were deemed appropriate. Two reviewers screened title/abstracts for preliminary eligibility, except for titles/abstracts of harm searches, which were screened only by one person. There were 4,201 records that could be excluded with certainty at screening. Most of these either did not involve pregnant women with chronic hypertension or

Figure 1. Flow diagram of selection process

   Figure 1. Flow diagram of selection process

did not meet design inclusion criteria. Fourteen records that could not be excluded with certainty by reviewing titles/abstracts were not retrievable (Figure 1). Because they were primarily published in foreign journals, they were unavailable either through the interlibrary loan system in the United States or through the Cochrane Pregnancy and Childbirth Group.

At least two independent reviewers screened the full text of each of the 1,343 articles retrieved to determine the final selection. Of these, 215 met selection criteria. Details of the screening and selection process for specific questions are presented in the results section of the report.

Data Abstraction Process

Two independent persons with clinical and methodological expertise abstracted data from trials that were identified in the efficacy searches. They were not blinded either to study title or to author's names. Items related to the internal validity of studies that were assessed included adequacy of randomization (method and concealment of assignment), whether the trial was single or double blind, cointerventions, and the number of dropouts. Disagreements in abstractions were resolved by consensus. Consensus abstractions were filed electronically to enable easy updating.

One physician with methodological expertise reviewed data about risks associated with blood pressure levels. One pharmacotherapist, with expertise in methodology, abstracted studies addressing adverse effects. Items abstracted from studies are recorded in the pertinent evidence tables.

Unpublished Data

Although some of the bibliographic sources included studies that were unpublished, no such studies met selection criteria. Some published studies met selection criteria but did not report critical design features or outcome data specific to women with chronic hypertension. In such instances, authors were contacted and requested to provide such information, if possible.

Data Analysis Process

Data were synthesized descriptively, emphasizing methodological characteristics of the studies such as populations enrolled, definitions of selection and outcome criteria, sample sizes, adequacy of randomization process, interventions and comparisons, cointerventions, biases in outcome assessment or intervention administration, and study designs. Relationships between clinical outcomes, participant characteristics, and methodological characteristics were examined in evidence tables and graphical summaries such as forest plots and L'Abbe plots.

The absolute risk reduction or mean difference was used as our measure of effect size from randomized controlled trials. Because of concerns about heterogeneity in study populations and interventions, quantitative summary methods were not used to combine any trial results.

Since most of the nonrandomized studies used retrospective case-control or cohort studies, the odds ratio (OR) was used to estimate risk for adverse outcomes associated with chronic hypertension. Random effects methods were used to estimate summary ORs or risk of perinatal mortality and abruption associated with chronic hypertension. Studies evaluating risks differed with respect to study design, country in which the study was conducted, and whether the reported OR was adjusted for potential confounding variables. Subgroup analyses and meta-regression were used to evaluate whether risk estimates systematically varied with respect to the above mentioned characteristics.

Question 1: Hypertension Management Before Pregnancy

What is the appropriate management of women with chronic hypertension before pregnancy?

Precise estimates of the clinical benefits of antihypertensive therapy in women younger than age 30 with mild to moderate hypertension are not available. Three randomized trials involving 8,565 nongravid women with mild to moderate hypertension, ages 30 to 54, show that approximately 259 (confidence interval [CI] 158 to 1,606) such women would need to be treated for 5 years to prevent a fatal or nonfatal cardiovascular event. No clinical outcome data from trials in nongravid younger women treated for similar or shorter intervals were found. Whether particular blood pressure management strategies that are used prior to conception have beneficial or adverse effects on conception or pregnancy outcomes has not been studied in trials.

Description of Trial Data Relevant to Preconception Management

Table 3. Morbidity and mortality results of antihypertensive (more...)

Table 3. Morbidity and mortality results of antihypertensive trials for 8,565 women ages 30-54

Study	Trial Design	Women (N)	Intervention	Pooled Data
Hypertension,Detection,and Followup Program	Not blinded DBP >90	Control: 1,568 Treated: 1,540	Thiazide diuretic vs. usual care Followup: 5 years	Fatal and nonfatal CVA: RR 0.58 (CI 0.36 to 0.67) NNT: 264 (CI 174 to 1,369) Fatal and nonfatal coronary events: RR 0.80 (CI 0.56 to 2.86) Fatal and nonfatal cardiovascular events: RR 0.73 (CI 0.54 to 0.97) NNT: 259 (CI 158 to 1,606) All deaths:2 RR 0.92 (CI 0.68 to 2.79)
The Australian Therapeutic Trial in Mild Hypertension	Double blinded DBP >95-110 SBP >200	Control: 416 Treated: 445	Thiazide diuretic vs. placebo Followup: 4 years
Medical Research Council Trial	Single blinded SBP <200 DBP 90-109	Control: 2,277 Treated: 2,319	Thiazide diuretic or beta-blocker vs. placebo Followup: 4.9 years

RR = relative risk; CI = confidence interval; NNT = number needed to treat; CVA = cerebrovascular accident; DBP = diastolic blood pressure; SBP = systolic blood pressure

A recent systematic review summarized available trial data regarding cardiovascular morbidity and mortality outcomes in women.³⁴ A pooled population of 23,000 women was described. Most were not of childbearing age; 8,565 were ages 30 to 54. These younger women were studied in three trials that are summarized in Table 3 ^{35,36, 37} In general, trial participants had mild to moderate hypertension and those assigned to active therapy received either thiazide diuretics or a beta-blocker as initial therapy. One trial used a usual-care control group, and only one was designed as a double-blind placebo comparison.

Question 2: Outcomes With Antihypertensive Agents

Do antihypertensive agents for the treatment of mild to moderate chronic hypertension (<170/110 mmHg) during pregnancy improve maternal and perinatal outcomes? Are there subsets of women, such as those with diabetes or renal disease, for whom treatment of mild to moderate chronic hypertension during pregnancy is warranted?

Data from randomized controlled trials are too scant to either prove or disprove clinically significant benefits from treating mild to moderate chronic hypertension during pregnancy. There are no randomized controlled trial data addressing effects of antihypertensive treatment in pregnant women with chronic hypertension and other risk factors such as diabetes and preexisting renal disease.

Description of Relevant Evidence

Figure 2. Selection process for antihypertensive treatment (more...)

   Figure 2. Selection process for antihypertensive treatment trials

^a Refers to publications, not trials, and includes trials with subgroup of chronic hypertension but without subgroup results.

RCT = randomized controlled trial; CHTN = chronic hypertension

Details of the search results are depicted graphically in Figure 2. Seven randomized trials, involving 595 patients, included only women with mild to moderate chronic hypertension (hypertension diagnosed prepregnancy or before the 20^th week of pregnancy, with blood pressure lower than 170/110 mmHg), and compared antihypertensive treatment with no antihypertensive treatment. An additional six trials included subgroups of 460 women with mild to moderate chronic hypertension. In four of these subgroups, results were reported in publications; in two, the data was obtained by personal correspondence.

Table 4. Drug dosages and secondary regimens from antihypertensive (more...)

Table 4. Drug dosages and secondary regimens from antihypertensive trials in chronic hypertension in pregnancy

Study	Year	Drug	Initial daily dose (mg)	Max. daily dose (mg)	Secondary treatment	Criteria for secondary Rx
45	1968	Bendrofluazide + Methyldopa (combination)	5 500	10 2000	Yes, but not described	"Ineffective" primary Rx
38	1976	Methyldopa	Not given		Yes, but not described	BP >170/110
46	1979	Hydralazine Methyldopa HCTZ (3 combinations)	75 750 50	250 2000 50
41	1981	Hydralazine Methyldopa	75 750		Yes, but not described	"Drug failure"
47	1984	Diuretics	Not given		Methyldopa	BP >160/110
49	1985	Metoprolol + Hydralazine (combination)	100 50	200 150	Yes, but not described	DBP >110
42	1987	Methyldopa	750	2000
39	1990	Atenolol	50	200
50	1990	Methyldopa Labetalol	750 300	4000 2400	Hydralazine	BP >140/90
40	1995	Isradipine SR	10
44	1996	Pindolol	10	20	Hydralazine	DBP >100
43	1997	Ketanserin a	40	80	Methyldopa	DBP>90
48	1998	Nifedipine SR	20	80	Beta-blocker	"M.D. judgment"

HCTZ = hydrochlorothiazide; BP = blood pressure; DBP = diastolic blood pressure; SR = sustained release

Table 5. Methodological characteristics of antihypertensive (more...)

Table 5. Methodological characteristics of antihypertensive trials

Reference	Age Mean	Gest Age Wk Mean	Blinded	Allocation Concealment	Dropouts Percent	Exclusion
Arias 197946	32	15.5	N	ND	0	Nulliparous target organ damage-"Major obstetrical problems"
Welt 198141	ND	21.3	S	ND	0	Multiple gestation insulin req. DM
Sibai 198447	28	< 13 wk	N	ND	0	ND
Weitz 198742	24	<34 wk	D	ND	0	Proteinuria Multiple gestations
Butters 199039	ND	16	D	ND	12	Contraindication to beta-blocker
Sibai 199050	30	11	N	PA	12	ND
Steyn 199743	32	15	D	CA	0	Proteinuria, multiple gestations "Major defects" on sonogram ECG abnormal in mother
Leather 196645	ND	<20	N	ND	ND	ND
Parazzini 199848	31	24	N	CA	6.5 9.2	Chronic dz/ fetal malformation
Wide-Swensson 199540	29	34	D	PA	5.9	Multiple gestations
Redman 197638 (<28 wk)	28	21	N	ND	2.4	"Major obstetrical risks"
Hogstedt 198549	29	30	N	ND	4	Proteinuria Mult. preg DM/ asthma/heart dz/ psych. d/o
Hirsch 199644	33	25	S	PA	10	ND

ND = not described; N = no; S = single; D = double; T = triple; ND = not described; PA = possibly adequate; CA = clearly adequate; NA = not adequate; ECG = electrocardiogram; DM = diabetes mellitus; dz = heart disease; d/o = psychiatric disorder

Table 6. Total number of participants treated with (more...)

Table 6. Total number of participants treated with particular antihypertensive agents in randomized trials

Methyldopa	207
Methyldopa + diuretic	23
Metoprolol + hydralazine	18
Hydralazine	9
Labetalol	86
Pindolol	15
Atenolol	15
Nifedipine SR	66
Isradipine	7
Ketanserin (+ aspirin)	69
Diuretics	10

SR = sustained release

Of the 13 trials, five were conducted in the United States, three in the United Kingdom, two in Sweden, one in Italy, one in Israel, and one in South Africa. Three were supported by pharmaceutical firms.^{38,39, 40} Six were placebo controlled,^39-44 and seven had a no-treatment group. Table 4 shows medication regimens used in the trials, Table 6 shows methodological characteristics of the trials, and Table 6 shows numbers of patients who were assigned to a particular drug. Eleven different drug regimens were used in the trials. The most commonly studied drug, methyldopa, was given to just over 200 subjects. In all but two trials, antihypertensive agents were given as monotherapy. One study used methyldopa and bendrofluazide in combination,⁴⁵ whereas another trial⁴⁶ used three different two- and three-drug combinations of methyldopa, hydralazine, and hydrochlorothiazide but analyzed the results of all three regimens together. Six studies applied secondary treatments if blood pressure response to the primary study drug was suboptimal. In one study, methyldopa was substituted for the study drug;⁴³ in other studies, additional drugs were added, such as methyldopa,⁴⁷ hydralazine,⁴⁴ beta-blockers,⁴⁸ or unspecified other drugs.^45,49

Table 7. Baseline blood pressure and response to treatment (more...)

Table 7. Baseline blood pressure and response to treatment in antihypertensive trials of chronic hypertension in pregnancy^a

Study	Year	Initial BP (weeks gest. age)	BP method	Korotkoff phase for DBP	Position BP measured	MAP Entry	MAP 2nd trimester	MAP 3rd trimester
45	1968	<20	NS	NS	NS	93.8 b 93.4 b		88.4 c 83.0 c
38a d	1976	<28	Mercury	IV	L lateral
46	1979	<20	Mercury	NS	Sitting	97.11 100.6	98.61 96.64	102.09 99.55
41	1981	<26	Mercury	NS	Sitting	100.1	96.4	94.7
47	1984	<14	NS	NS	NS	105 102	95 98	108 106
49a d	1985	<37	Mercury	V	L lateral
42	1987	<34	Mercury	NS	NS	97.6 106.8	109 95	105 95
39	1990	12-24	NS	V	NS	106.7 105.3		98.3 e 93.3 e
50	1990	6-13	NS	IV	Sitting	108.3 107 106.7	97 90.7 92.3	100 93.7 96
40	1995	25-37	Mercury	IV	Supine	113.5 113.6		119.3 114.8
44	1996	<35	NS	IV	Supine
43	1997	12-20	Auto	IV	Supine
48a d	1998	12-34	NS	NS	NS

a

For each study, the first line of entries represents the blood pressures in the control group; the second and subsequent lines represent the blood pressures in the treatment groups.

b

Diastolic blood pressures

c

Mean of all diastolic blood pressures during treatment

d

Blood pressures not available separately for chronic hypertension subgroup

e

Blood pressures given were measured on drug therapy at unspecified time

NS = not stated; MAP = mean arterial pressure; BP = blood pressure; DBP = diastolic blood pressure

Blood pressure criteria for inclusion were relatively consistent across the trials (Table 7). All the studies used a diastolic blood pressure threshold for inclusion: 80, 85, 90, or 95 mmHg. Six required a systolic blood pressure threshold of at least 140 mmHg and one a systolic threshold of 130 mmHg. The mean arterial pressure calculated as (systolic blood pressure [SBP] + 2 diastolic blood pressure [DBP]/3) at entry into the studies varied within a range of 100 to 113 mmHg. Mean maternal age ranged between 24 and 33 years. The mean gestational age at entry into the trials varied from 11 to 34 weeks. Most participants were recruited from obstetrical clinics. No trials focused on women with chronic hypertension and other risk factors such as diabetes or preexisting renal disease.

The trials examined a wide variety of outcomes including perinatal death (11), preeclampsia (4), SGA infants (8), intrauterine growth retardation (2), low 5-minute Apgar score (4), rate of cesarean delivery (3), length of gestation (7), and birthweight (11). Some outcomes such as preeclampsia and SGA were defined differently across trials.

Nine trials enrolled fewer than 60 subjects. Only one had a randomization mechanism of allocation concealment that was clearly adequate.⁴³ Dropout rates were less than 20 percent in all trials. There was heterogeneity in the gestational age at which drug treatment was initiated. Treatment often was started well into the second trimester, which may be too late to prevent early pathophysiological events leading to preeclampsia. Whether chronic hypertension or superimposed preeclampsia was really the entity being treated was not always clear.

Trial Findings

Figure 3. Perinatal death results of trials given as (more...)

   Figure 3. Perinatal death results of trials given as risk differences between antihypertensive treatment groups and control groups

Figure 4. Preeclampsia results of trials given as risk (more...)

   Figure 4. Preeclampsia results of trials given as risk differences between antihypertensive treatment groups and control groups

In six trials, antihypertensive therapy showed no benefit in terms of reduced perinatal mortality. In three trials,^38,43,45 a trend towards lower perinatal mortality with treatment was observed (Figure 3). Only one of the four trials that assessed the incidence of preeclampsia demonstrated a risk reduction from treatment.⁴³ Steyn and Odendaal⁴³ found that treatment with ketanserin + aspirin lowered the incidence of preeclampsia when compared with aspirin alone (27.5 vs. 4.3 percent). In the other three trials,^42,47,50 the risk of preeclampsia was not significantly decreased by antihypertensive treatment (Figure 4).

Figure 5. Birthweights of treatment and control groups (more...)

   Figure 5. Birthweights of treatment and control groups

Birthweights of treatment and control groups from randomized trials of antihypertensive treatment vs. placebo or no treatment for chronic hypertension during pregnancy. Groups from the same study are indicated by symbols.

Figure 6. Small-for-gestational age results of trials (more...)

   Figure 6. Small-for-gestational age results of trials given as risk differences between antihypertensive treatment groups and control groups

Butters, Kennedy, and Rubin³⁹ demonstrated a markedly lower average birthweight in their atenolol group than in their placebo group (2620 vs. 3530 g). The other antihypertensive trials all assessed either birthweight or risk of SGA infant (or both), and none found significant differences between treated and untreated patients (Figures 5,6). No significant differences in gestational age at delivery were demonstrated in the 6 trials in which that outcome was assessed.

Question 3: Potential Harm From Pharmacological Treatment

Is pharmacological treatment of mild to moderate chronic hypertension during pregnancy harmful to mothers, fetuses, and infants? If harmful, what is the type and magnitude of specific harm for mothers, fetuses, or infants?

Clinical trials enrolling women with either chronic hypertension or other hypertensive disorders of pregnancy have not detected serious antihypertensive-related adverse events in mothers or fetuses. However, the total number of women in these trials is insufficient to provide confidence that these agents do not cause harm when used in pregnancy. Methyldopa, the most commonly prescribed drug for hypertension in pregnancy, has been used in just over 500 women in clinical trials, and only two other drugs, labetalol and diuretics, have been used in more than 200 women.

Because the clinical trial experience with antihypertensives in pregnancy is so limited, most of the evidence found on possible adverse effects came from sources other than clinical trials, such as surveillance studies and case reports. Their designs severely limit ability to calculate the absolute risk of adverse events or to unravel causal relationships.

Methyldopa, the antihypertensive often recommended as a first-line agent for treating chronic hypertension in pregnancy, appears to be safe for the fetus. The most serious risk for the mother is hepatitis. This occurs in the nongravid population at a rate of 1 to 10/100,000 treated persons; whether the risk is modified by pregnancy is unknown.

Other antihypertensive classes often recommended for chronic hypertension in pregnancy include beta-blockers, alpha/beta-blockers, diuretics, and calcium channel blockers. Theoretical concerns about diuretics' effect on plasma volume in pregnancy were not supported by a large epidemiological study. Both randomized trials and epidemiological studies have raised concerns about an association between beta-blockers (and alpha/beta-blockers) and fetal growth restriction, though the evidence is conflicting. The risk of intrauterine growth retardation may be increased if treatment with this class of drugs begins early in pregnancy. Some calcium channel blockers appear safe in pregnancy, but the published experience with this class of agents is the most limited of all the classes discussed above and drugs in this class are very heterogeneous.

Second and third trimester exposure to ACE inhibitors is associated with fetal renal failure. Though ARBs have not been used in pregnancy, their mechanisms of action are similar and they may cause similar adverse effects.

Description of Relevant Evidence

The following section addresses the potential maternal and fetal harmful effects of antihypertensive agents during pregnancy. It focuses on serious adverse reactions, defined as any event that is fatal, life-threatening, or significantly disabling or that requires hospitalization or results in congenital anomalies. This section does not attempt to elucidate the association between antihypertensive agents and mortality, as this was discussed in Question 2. Results are presented for the following antihypertensive agents: methyldopa, calcium channel blockers, beta-blockers, alpha/beta-blockers, ACE inhibitors, ARBs, diuretics, hydralazine, and clonidine. Of these, methyldopa was recommended as the first-line choice for drug therapy in nine of nine recent reviews of hypertension in pregnancy.^1,55-62

Figure 7. Selection process for studies assessing adverse (more...)

   Figure 7. Selection process for studies assessing adverse effects

^a 62 articles were identified from the treatment search as possibly reporting serious adverse reactions or successful uneventful exposure to antihypertensive during pregnancy.

^b 127 original trials were identified. One trial may have multiple publications.

CHTN = chronic hypertension

Details of the search results are depicted graphically in Figure 7. There were 127 original reports of maternal, fetal, or neonatal serious adverse reactions or reports of successful uneventful use of antihypertensives during pregnancy. Of these, only 13 randomized controlled trials and 4 uncontrolled trials were conducted in pregnant women with mild to moderate chronic hypertension. One-hundred and ten studies (48 randomized controlled trials, 42 case reports, 6 case controls, 12 case series, and 2 surveillance studies) described experience in treating hypertensive diseases of pregnancy (other than preexisting mild to moderate hypertension) with antihypertensive agents.

In reviewing the evidence on adverse events, it is important to remember that there is great variation in the frequency with which the listed antihypertensives are used in pregnancy, so that the effective denominator for the adverse events described in case reports varies by several orders of magnitude from commonly prescribed agents to rarely prescribed agents. Scarce evidence of harm from methyldopa, therefore, deserves much more interpretive weight than does scarce evidence of harm from drugs such as nitrendipine or oxprenolol.

Trial Findings

Table 8. Teratogenicity risk

Table 8. Teratogenicity risk

FDAa	Briggs	Australia	N (1st trimester exposure)	Evidence of birth defect in animal model	Description of Defect
Angiotensin-converting enzyme inhibitors
C b	D	D	80	Yes	Oligohydramnios, neonatal anuria, persistent patent ductus arteriosis, fetal death, skull bone anomalies
C	NR	D	NR	Yes	Increase mortality and renal anomalies noted in animals
Acebutolol
C	C	C	125	No
Atenolol
D	D	D	144	No	Possibly associated with hypospadias
Labetalol
C	C	D	115	No
Metoprolol
C	C	C	53	No
Oxprenolol
C	C	C	NR	NR	Not applicable
Pindolol
C	B	C	NR	No
Diltiazem
C	C	C	37	Yes	Cardiovascular defects
Nifedipine
C	C	C	71	Yes
Verapamil
C	C	C	108	Yes
Clonidine
C	C	C	60	No
Diuretics (hydrochlorothiazide)
B	D	C	567	No
Hydralazine
C	C	C	40	Yes
Methyldopa
C	C	C	339	No

a

Denotes categories of risk as defined by different groups or agencies; see Tables 11-13 for complete definitions.

b

FDA category D for the 2nd and 3rd trimester

NR = not reported; FDA = Food and Drug Administration

Table 9. Serious fetal and neonatal adverse reactions (more...)

Table 9. Serious fetal and neonatal adverse reactions

Type	Consistency	Estimate of magnitude
Angiotensin-converting enzyme inhibitors
Fetal growth retardation	<10 case reports	Unknown
Pulmonary hypoplasia	<10 case reports	Unknown
Patent ductus arteriosus	<10 case reports	Unknown
Respiratory distress syndrome	>10 case reports	Unknown
Angiotensin receptor blockers
Data lacking	Not applicable	Not applicable
Acebutolol
No serious adverse reactions	RCTs and controlled trials but overall sample size <100	Not applicable
Atenolol
Intrauterine growth restriction	RCTs, appears related to time of administration, dose and length of therapy	Unknown
Retroperitoneal fibromatosis	1 case report	Unknown
Labetalol
Intrauterine growth retardation	Conflicting RCTs (N>300)	Unknown
Severe hypoglycemia, hypotension, pericardial effusion and myocardial hypertrophy	<10 case reports	Unknown
Metoprolol
None reported	Controlled trials (N=150)	Not applicable
Oxprenolol
None reported	RCT (N=224)	Not applicable
Pindolol
None reported	RCTs (n=200)	Not applicable
Calcium channel blockers
None reported	Mostly RCTs (n=200)	Not applicable
Clonidine
None reported	RCT (n=72)	Not applicable
Behavioral changes	Case series	Unknown
Diuretics
Thrombocytopenia	>10 case reports	Unknown
Neuroblastoma	Case controls (2)	Odds ratio 4.1 to 5.75
Deafness	Case control	Not given
No serious adverse reactions	9 RCTs (n>5,000)	Not applicable
Hydralazine
Premature atrial contractions	<10 case reports	Unknown
Transient thrombocytopenia	<10 case reports	Unknown
Lupus-like syndrome	<10 case reports	Unknown
Methyldopa
Decreased head circumference	1 RCT not replicated in 18 RCTs	Not given
Excessive tremors	<10 case reports	Unknown

RCT = randomized controlled trial

Table 10. Maternal serious adverse reactions during (more...)

Table 10. Maternal serious adverse reactions during pregnancy

Type	Consistency	Estimate of magnitude
ACE inhibitors
Data lacking	Not applicable	Not applicable
ARBs
Data lacking	Not applicable	Not applicable
Acebutolol
Data lacking	Not applicable	Not applicable
Atenolol
Data lacking	Not applicable	Not applicable
Labetalol
Data lacking	Not applicable	Not applicable
Metoprolol
Data lacking	Not applicable	Not applicable
Oxprenolol
Data lacking	Not applicable	Not applicable
Pindolol
Data lacking	Not applicable	Not applicable
Calcium channel blockers
Myocardial infarction	<10 case reports	Unknown
Severe hypotension	<10 case reports	Unknown
Neuromuscular blockade	<10 case reports (drug interaction between nifedipine and magnesium)	Unknown
Pseudo-obstruction of the colon	<10 case reports (drug interaction between nifedipine and magnesium)	Unknown
Clonidine
Data lacking	Not applicable	Not applicable
Diuretics
Pancreatitis	<10 case reports	Unknown
Ototoxicity	<10 case reports	Unknown
None	9 RCTs (n=5,000)	Not applicable
Hydralazine
Lupus-like syndrome	<10 case reports	Unknown
Hepatitis	<10 case reports	Unknown
Methyldopa
Hepatitis	<10 case reports	Unknown

ACE = angiotension-converting enzyme; ARB = angiotension receptor blocker; RCT = randomized controlled trial

Table 11. U.S. Food and Drug Administration pregnancy (more...)

Table 11. U.S. Food and Drug Administration pregnancy categories definitions¹³⁰

Pregnancy risk categories

A. Controlled studies in women fail to demonstrate a risk to the fetus in the first trimester, and the possibility of fetal harm appears remote.

B. Animal studies do not indicate a risk to the fetus and there are no controlled human studies, or animal studies do show an adverse effect on the fetus but well-controlled studies in pregnant women have failed to demonstrate a risk to the fetus.

C. Studies have shown that the drug exerts animal teratogenic or embryocidal effects, but there are no controlled studies in women, or no studies are available in either animals or women.

D. Positive evidence of human fetal risk exists, but benefits in certain situations (e.g., life-threatening situations or serious diseases for which safer drugs cannot be used or are ineffective) may make use of the drug acceptable despite its risks.

X. Studies in animals or humans have demonstrated fetal abnormalities or there is evidence of fetal risk based on human experience, or both, and the risk clearly outweighs any possible benefit.

Table 12. Australian pregnancy categories definitions (more...)

Table 12. Australian pregnancy categories definitions

A. Drugs that have been taken by a large number of pregnant women and women of childbearing age without any proven increase in the frequency of malfunctions or other direct or indirect harmful effects on the fetus having been observed.

B1. Drugs that have been taken by only a limited number of pregnant women and women of childbearing age without an increase in the frequency of malformation or other direct or indirect harmful effects on the human fetus having been observed.

B2. Drugs that have been taken by only a limited number of pregnant women and women of childbearing age, without an increase in the frequency of malformation or other direct or indirect harmful effects on the human fetus having been observed. Studies in animals 1 are inadequate or may be lacking, but available data show no evidence of an increased occurrence of fetal damage.

B3. Drugs that have been taken by only a limited number of pregnant women and women of childbearing age without an increase in the frequency of malformation or other direct or indirect harmful effects on the human fetus having been observed. Studies in animals have shown evidence of an increased occurrence of fetal damage.

C. Drugs that, owing to their pharmacological effects, have caused or may be suspected of causing harmful effects on the human fetus or neonate without causing malformations. These effects may be reversible. Accompanying texts should be consulted for further details.

D. Drugs that have caused, are suspected to have caused, or may be expected to cause an increased incidence of human fetal malformations or irreversible damage. These drugs may also have adverse pharmacological effects. Accompanying texts should be consulted for further details.

J. Drugs that have such a high risk of causing permanent damage to the fetus that they should not be used in pregnancy or when there is a possibility of pregnancy.

Note: For drugs in the B1, B2, and B3 categories, human data are lacking or inadequate and subcategorization is therefore based on available animal data. The allocation of a B category does not imply greater safety than the C category. Drugs in category D are not absolutely contraindicated in pregnancy (e.g., anticonvulsants). Moreover, in some cases the D category has been assigned on the basis of "suspicion."

Table 13. Briggs, Freeman, and Yaffee pregnancy categories (more...)

Table 13. Briggs, Freeman, and Yaffee pregnancy categories definitions²⁹

Category A: Controlled studies in women fail to demonstrate a risk to the fetus in the first trimester (and there is no evidence of a risk in later trimesters), and the possibility of fetal harm appears remote.

Category B: Either animal reproduction studies have not demonstrated a fetal risk but there are no controlled studies in pregnant women or animal reproduction studies have shown an adverse effect (other than a decrease in fertility) that was not confirmed in controlled studies in women in the first trimester (and there is no evidence of a risk in later trimesters).

Category C: Either studies in animals have revealed adverse effects on the fetus (teratogenic or embryocidal or other) and there are no controlled studies in women or studies in women and animals are not available. Drugs should be given only if the potential benefit justifies the potential risk to the fetus.

Category D: There is positive evidence of human fetal risk, but the benefits from use in pregnant women may be acceptable despite the risk (e.g., if the drug is needed in a life-threatening situation or for a serious disease for which safer drugs cannot be used or are ineffective).

Category X: Studies in animals or human beings have demonstrated fetal abnormalities or there is evidence of fetal risk based on human experience or both, and the risk of the use of the drug in pregnant women clearly outweighs any possible benefit. The drug is contraindicated in women who are or may become pregnant.

Tables 8, 9, and 10 contain a summary of the findings on teratogenicity, fetal and neonatal serious adverse reactions, and maternal serious adverse reactions, respectively. (See Tables 11, 12, and 13 for complete category definitions.)

Central Alpha-Agonists

Methyldopa

Teratogenicity. The published information related to first trimester methyldopa exposure is limited to 339 pregnant women treated for hypertension. No significant increased risk of congenital anomalies was reported.^29,63

Fetal and neonatal adverse reactions. More than 130 women with mild to moderate chronic hypertension during pregnancy were treated with methyldopa in four randomized controlled trials.^{38,41,42,50,64-69} No fetal and neonatal serious adverse reactions were reported. A prospective study comparing methyldopa with acebutolol for the treatment of mild to moderate chronic hypertension in 20 pregnant women reported no serious adverse outcomes.⁷⁰ A small case series of seven women treated early in pregnancy with methyldopa for chronic hypertension reported no detrimental fetal or neonatal effect of the drug.⁷¹

Similarly, no clinically significant adverse reactions were reported in 15 additional randomized controlled trials and two case series in which more than 500 pregnant women with pregnancy-induced hypertension (mild to severe) were treated with methyldopa.^71-91 In one randomized controlled trial, including approximately 50 hypertensive women treated with methyldopa before 20 weeks of gestation, a small but statistically significant decrease in head circumference was observed.⁶⁸

Figure 7. Selection process for studies assessing adverse effects

No difference in head size or cognitive development was apparent several years later in the children exposed to methyldopa in utero.^64,69,92 These findings were not replicated in other trials.

Anecdotal reports of excessive tremors unresponsive to calcium, magnesium, or pyridoxine have been observed in newborns after maternal exposure to methyldopa.^93,94

Maternal serious adverse reactions. In nongravid populations, 20 percent to 40 percent of patients develop a positive Coomb's Test, of which 0.1 percent to 0.2 percent will experience hemolytic anemia. A positive Coomb's Test is rare with treatment before 6 months. One case report of a positive Coomb's test during pregnancy has been published; signs of hemolytic anemia were not found in the mother or newborn.^95,96 Methyldopa-induced hepatitis can occur at anytime during therapy but is rare. It is estimated that 1 to 10 persons will develop hepatitis for every 100,000 persons exposed to methyldopa.^30,31 Three instances of hepatitis, one fatal, have been reported in women treated with methyldopa during pregnancy.^97,98,99

Methyldopa, in nongravid population, has been associated with drug-induced parkinsonism, lichenoid eruptions, agranulocytosis, cardiac dysrhythmias, autoimmune thrombocytopenia, and pancreatitis in published case reports.³¹

Clonidine

Teratogenicity. There is one case report of a newborn with multiple birth defects exposed throughout gestation to clonidine.¹⁰⁰ Three major birth defects (5.1 percent) were reported from 59 pregnancies when clonidine was administered during the first trimester in the Michigan Medicaid Surveillance study.¹⁰¹ The incidence is not significantly greater than the expected incidence for major birth defects in the general population.

Fetal and neonatal adverse events. No controlled trials of mild to moderate chronic hypertension during pregnancy treated with clonidine were found. No fetal or neonatal serious adverse drug reactions were reported in 72 women with other hypertensive diseases enrolled late in pregnancy in two randomized controlled trials.^102,103 An excess of hyperactivity and sleep disturbances was reported in 22 children with a mean age of 6 years, prenatally exposed to clonidine.¹⁰⁴ This adverse effect also has been reported in animal models.⁶³

Maternal serious adverse reactions. No reports of maternal serious adverse reactions were found. Clonidine, in nongravid population, has been associated with drug-induced lupus erythematosus, hepatitis, and potentially fatal drug withdrawal rebound hypertension.³¹

Beta-Blockers

The Sixth Joint National Committee for Prevention, Detection, Evaluation, and Treatment of High Blood Pressure has recommended beta-blockers as first-line treatment for uncomplicated essential hypertension. For that reason, beta-blockers are likely to be used in women of childbearing potential and may be continued during pregnancy. This section describes teratogenicity risk and neonatal and maternal serious adverse reactions reported during administration of beta-blockers in pregnancy.

Teratogenicity. Atenolol is the beta-blocker with the most evidence addressing teratogenicity. Twelve (11.4 percent) of 105 pregnant women registered in the Michigan Medicaid Surveillance Program²⁹ gave birth to infants with major birth defects. A detailed evaluation from Briggs, Freeman, and Yaffee²⁹ indicates that an association may exist between first trimester exposure to atenolol and hypospadias. However, the authors caution about the interpretation of the association, as concomitant drug use and mother's underlying pathological conditions may have an impact on the outcomes observed.²⁹ Other case series did not find this association. A small prospective followup study of five pregnant women treated with atenolol during the first trimester for essential hypertension did not find evidence of teratogenicity.^105,106 Similarly, a case series of 60 women treated with beta-blockers (atenolol, pindolol, labetalol, or propanolol) of whom 8 women were on a beta-blocker agent during the first trimester, did not find evidence of teratogenicity.¹⁰⁷ Evidence of teratogenicity also was not found in a case series of 125 pregnancies exposed to beta-blockers (acebutolol, n=56; pindolol, n=38; atenolol, n=21), in which four (3.2 percent) malformations were reported; none was in women exposed to atenolol.^107,108

No congenital anomalies from first trimester exposure to metoprolol have been reported. The relative safety of metoprolol during organogenesis is further supported by data on 52 newborns in the Michigan Medicaid Surveillance Study and one case report of exposure throughout gestation.^29,63

No studies reporting experience in using acebutolol, oxprenolol, or pindolol during the first trimester were found.²⁹ There are case reports of arthrogryposis in a newborn exposed to acebutolol after the 24^th week of gestation^107,108 and meconium ileus, cleft palate, and vesico ureteral reflux in neonates exposed in utero to pindolol after the 32^nd, 29^th, and 1^st week of gestation, respectively.^107,108

Fetal and neonatal adverse events. Several studies suggest that beginning atenolol therapy early in pregnancy is associated with lower birthweight. In a pharmacosurveillance study of 76 women who started atenolol during the first 20 weeks of pregnancy, lower birthweight was reported with atenolol compared with no drugs or other antihypertensives (2190 g compared with 2780 g in the drug-free group).¹⁰⁹ Likewise, in a case series of 125 pregnancies exposed to beta-blockers (acebutolol, n=56; pindolol, n=38; atenolol, n=31), the birthweight of neonates exposed to atenolol was significantly lower than newborns exposed to pindolol or acebutolol.^107,108

Clinical trial experience with long-term use of atenolol to manage mild to moderate chronic hypertension during pregnancy is limited to 15 women^39,110 in whom atenolol treatment (50 to 200 mg/day) began early in the second trimester and continued until delivery. Newborns from the atenolol-treated group had a significantly lower birthweight than the control group, suggestive of intrauterine growth retardation (2670 g compared with 3470 g, respectively). By 12 months of age, however, there was no significant difference in weight between the two groups.¹¹¹

Other studies have not demonstrated growth restriction from atenolol. Four randomized controlled trials were published in which 147 pregnant women were treated late in pregnancy with atenolol for mild to moderate pregnancy-induced hypertension.^111-116 Two trials (n=37) compared atenolol with pindolol, and one trial compared atenolol with verapamil. None of these trials found a clinically significant difference in birthweight or other serious adverse metabolic effects, nor did a case series of 12 pregnant women in which treatment with atenolol (50-100 mg) for mild to moderate chronic hypertension was started in the third trimester.¹¹⁷ The discrepancies in these studies may be explained by differences in length of therapy, gestational age at which therapy was initiated, and dose administered.

Other adverse events reported with atenolol include an anecdotal report of a retroperitoneal fibromatosis tumor in a newborn exposed to atenolol 100 mg from the second month of gestation to delivery¹¹⁸ and clinical signs of beta-blockade (bradycardia, hypotension) in a newborn exposed in utero to atenolol.¹¹⁹

Fewer data are available on the use of other beta-blockers in pregnancy. In controlled trials of pregnancy-induced hypertension, including more than 150 women treated with metoprolol late in pregnancy, no serious adverse reactions during labor, or the fetal metabolic, cardiac status, or fetal weight, were reported.^{49,78,120,121, 122} Similarly, no serious fetal or neonatal adverse reactions were reported in 189 women treated with metoprolol, alone or in combination with hydralazine or diuretics late in pregnancy.^123,124 No randomized controlled trials reporting treatment of mild to moderate chronic hypertension with metoprolol, acebutolol, or oxprenolol were found. No serious fetal adverse events occurred in a prospective study comparing acebutolol with methyldopa for the treatment of mild to moderate chronic hypertension in 20 pregnant women⁷⁰ or in a randomized controlled trial comparing acebutolol with methyldopa and labetalol in 21 pregnant women with mild or moderate pregnancy-induced hypertension.⁷²

A randomized controlled trial in 30 pregnant women with mild to moderate chronic hypertension comparing pindolol (n=15) with placebo reported no serious adverse outcomes.⁴⁴ In 170 pregnant women treated with pindolol for mild to moderate pregnancy-induced hypertension enrolled late in pregnancy in six randomized controlled trials and one cohort study,^{84,112,113,116,125, 126, 127} no serious adverse reactions associated with the use of pindolol were reported.

Oxprenolol was compared with placebo and methyldopa in three randomized controlled trials for the treatment of mild to moderately severe pregnancy-induced hypertension.^{79,80, 81,128,129} No serious adverse reactions were reported after third trimester exposure to oxprenolol in 224 pregnant women with hypertensive diseases.

Maternal serious adverse reactions. No reports of maternal serious adverse reactions during treatment with beta-blockers were found.

Generally, beta-blockers are well-tolerated in nongravid populations. Cutaneous vasculitis, atenolol-induced lupus erythematosus, tissue necrosis, seizures, sclerosing peritonitis, and retroperitoneal fibrosis have been described in case reports.^31,130 Hepatitis with relapse on rechallenge has been reported with acebutolol.³¹

Severe bradycardia (5 percent) has been reported with metoprolol, particularly in patients with preexisting atrioventricular (AV) node disease or preexisting AV block caused by drugs or other factors. Despite its relative ₁blocking selectivity, 1 percent of patients exposed to metoprolol at doses greater than 100 mg may experience bronchoconstriction, dyspnea, and wheezing. Isolated case reports of hepatic necrosis (following 2 weeks of therapy in one case) and retroperitoneal fibrosis have been reported after metoprolol exposure.³¹

Oxprenolol has been associated with retroperitoneal fibrosis, sclerosing peritonitis, and thrombocytopenia in nongravid population.³¹ Pindolol, in nongravid population, has been associated with drug-induced lupus erythematosus in nongravid population.³¹

Combined Alpha/Beta-Blockers

Labetalol

Teratogenicity. In the Michigan Medicaid surveillance study, only 29 newborns were exposed in utero to labetalol. Four (13.8 percent) major birth defects were reported.²⁹ A detailed evaluation by Briggs, Freeman, and Yaffee indicates that a possible teratogenicity risk exists, but the authors cautioned that the number of exposures is small and it may be difficult to differentiate drug effect from the mother's underlying severe medical condition.

In a randomized controlled trial, 86 women with mild to moderate chronic hypertension were enrolled between 6 and 13 weeks of gestation to be treated with labetalol.⁵⁰ None of the newborns exposed in utero to labetalol had major birth defects.

Fetal and neonatal adverse events. No fetal and neonatal serious adverse reactions were reported after long-term use of labetalol in 86 pregnant women with mild to moderate chronic hypertension enrolled early in pregnancy in a randomized controlled trial.⁵⁰ Nonparenteral labetalol has been studied in 500 pregnant women in clinical trials of the treatment of mild to severe pregnancy-induced hypertension.^{72-77,131-140} No serious fetal or neonatal serious adverse reactions were found to occur more frequently in those administered labetalol than in the control groups (no treatment, methyldopa, or hydralazine). However, three randomized controlled trials have described intrauterine growth retardation in newborns of 195 pregnant women treated with labetalol during the second and third trimesters.^131-135,137

Serious hypoglycemia and severe hypotension episodes resulting in fetal deaths also have been reported.^141-147 Most cases cited were after parenteral administration of labetalol late in pregnancy to control severe hypertension and may not be relevant to the treatment of mild to moderate chronic hypertension. Isolated findings of pericardial effusion and myocardial hypertrophy were reported in infants exposed to labetalol in utero.¹⁴⁸

Maternal serious adverse reactions. No reports of serious adverse reaction during pregnancy were found.

Labetalol like other beta-blockers is well tolerated in nonpregnant populations. Fatal hepatic necrosis (after 2 months of therapy) has been reported.³¹ Also, the agent has been associated with drug-induced systemic lupus erythematosus.³¹

Calcium Channel Blockers

Among the newer antihypertensive drugs, calcium channel blockers have become popular in the treatment of uncomplicated hypertension and are likely to be used in women of childbearing potential.

Teratogenicity. Premarketing animal studies consistently have shown digital and limb defects. Many of the processes of embryogenesis appear to be calcium dependent in animals and may explain animal findings. The teratogenicity risk knowledge in humans is limited to 211 first trimester exposures.^{29,51,149,150, 151}

The Michigan Medicaid Surveillance Study reports 140 first trimester exposures to nifedipine (n=37), diltiazem (n=27) and verapamil (n=76), resulting in an estimated major birth defect incidence of 5 percent, 14.8 percent, and 1.3 percent, respectively. None of the major birth defects observed in diltiazem-exposed newborns was a limb defect. Cardiovascular defects were observed in 2 of 27 drug-exposed infants.

In a prospective, multicenter cohort study of 78 women exposed to calcium channel blockers (mostly nifedipine and verapamil), there were 2 children with limb defects (estimated incidence 3 percent). In both cases, the pregnancies were exposed to other agents as well; one pregnancy was in a woman with poorly controlled diabetes.

An anecdotal report found that verapamil exposure during pregnancy to treat supraventricular tachycardia was associated with congenital hypertrophic cardiomyopathy.^149,150

Fetal and neonatal adverse events. Only two randomized controlled trials comparing a calcium channel blocker (nifedipine or isradipine) with no treatment or placebo included a subgroup of pregnant women with mild to moderate chronic hypertension.^40,48 However, no results for the subgroup were given. In the overall sample, congenital malformations were observed in both treated and control groups at a similar frequency. In controlled trials of pregnancy-induced hypertension, more than 300 women were treated with a calcium channel blocker: nifedipine (n=180); nitrendipine, a drug not available in the United States, (n=12); isradipine (n=21); verapamil (n=22); nicardipine (n=50); and amlodipine (n=50).^{40,48,85,87,90,116,120,152-155} None of the trials reported fetal anomalies or serious adverse outcomes. No controlled trials or case reports were found that addressed diltiazem therapy during pregnancy. Studies using felodipine and isradipine were not found but were not specifically targeted in the search.

Maternal serious adverse reactions. Sublingual nifedipine was associated with adverse maternal outcomes such as myocardial infarction and severe hypotension resulting in fetal distress.^29,156,157 Isolated reports documented serious adverse reactions (neuromuscular blockade and acute pseudo-obstruction of the colon) following the concomitant use of nifedipine and magnesium.^{158, 159, 160}

Other significant adverse reactions were observed with the calcium channel blockers in nongravid populations: agranulocytosis¹³⁰ and thrombocytopenia (nifedipine and diltiazem); dysrhythmia such as asystole, ventricular fibrillation, and electromechanical dissociation (verapamil, diltiazem, and nifedipine); myocardial ischemia or infarction (short-acting nifedipine); myoclonus with an incidence lower than 0.1 percent (verapamil, diltiazem, and nifedipine); hepatotoxicity such as cholestasis and hepatitis (verapamil, diltiazem, and nifedipine); acute renal failure with an incidence lower than 0.1 percent (diltiazem, nifedipine); Stevens-Johnson syndrome (verapamil, diltiazem, and nifedipine in 18 case reports); and systemic lupus erythematosus (diltiazem). Calcium channel blockers were associated with an increased risk of cancer in several observational studies, though the results were not replicated by the World Health Organization.¹³⁰

Diuretics

Teratogenicity. The majority of data relating to diuretics during pregnancy is limited to thiazides such as chlorthalidone, hydrochlorothiazide (HCTZ), and chlorothiazide. Birth defects (mostly congenital dislocation of the hip) were observed in newborns of 20 of 233 pregnant women (8.6 percent) who were exposed to chlorthalidone during the first trimester in the Collaborative Perinatal Project.⁶³ There was no increased risk of birth defects in the 107 newborns of women treated with HCTZ in this cohort. Nor was there an increased risk in the 50 of 99 women exposed to HCTZ in the Boston Collaborative Drug Surveillance Program.¹⁶¹ The Michigan Medicaid Surveillance Study reported 635 first-trimester exposures to chlorothiazide (n=20), chlorthalidone (n=48), and HCTZ (n=67).²⁹ Estimated major birth defect incidences were 10 percent, 4.2 percent, and 4.2 percent, respectively. From these reports, it appears that HCTZ may be the least teratogenic of the thiazide diuretics. Teratogenicity data on another diuretic, furosemide, is limited to 350 pregnant women exposed during their first trimester in the Michigan Medicaid Surveillance Study.²⁹ Eighteen major birth defects (5.1 percent) were observed. Hypospadias accounted for the slight increase above the expected number of birth defects.

One randomized controlled trial assessed safety of diuretics (not specified) used throughout pregnancy for the management of mild to moderate chronic hypertension. Twenty pregnant women were randomized before 13 weeks of gestation to either continuing or discontinuing drug therapy during pregnancy. The authors reported no major birth defects.⁴⁷

Fetal and neonatal adverse effects. Only one randomized controlled trial assessed the safety in continuing diuretics as therapy for chronic hypertension during pregnancy. No serious maternal, fetal, or neonatal adverse drug reactions were reported, but there were only 10 women in the treatment arm.⁴⁷

Reports of thrombocytopenia in 11 newborns (2 neonatal deaths) exposed to thiazides in utero were found.^162-169 In contrast, a case control study found no increased risk of thrombocytopenia after 3 weeks of diuretics therapy late in pregnancy.¹⁷⁰ A meta-analysis of nine RCTs reviewing diuretic exposure of more than 5,000 pregnant women with or without hypertension described only one fatal thrombocytopenia in a newborn; no cases of neuroblastoma were reported.^166,171-180

Two case control studies found a significant association between risk of neuroblastoma and in utero exposure to diuretics for the treatment of high blood pressure (OR 4.1 95 percent CI 1.0 to 16.9 and OR 5.75 [p<.0.001]).^181,182 The association between thrombocytopenia, neuroblastoma, and in utero exposure to diuretics cannot be ruled out with certainty, but the potential incidence for these serious adverse reactions appears very low (<0.02 percent).

Other isolated adverse events reported with diuretics include a case report of fetal bradycardia associated with a combination of thiazide and reserpine¹⁸³ and a case control study that found a significant association between newborn deafness and in utero exposure to furosemide (no OR given).¹⁸⁴

Diuretics prevent normal volume expansion later in pregnancy. The resulting depletion in plasma volume may decrease uterine perfusion and lead to adverse fetal outcomes. However, these theoretical concerns were not supported in randomized controlled trials for the treatment of mild to moderate chronic hypertension or in one large review of nine randomized controlled trials (RCT)s.

Maternal serious adverse reactions. Seven cases of pancreatitis (two with fatal hemorrhagic pancreatitis) in pregnant women exposed to diuretics have been published,^{166,185, 186, 187} although no cases of maternal pancreatitis were reported in a meta-analysis of more than 5,000 pregnant women exposed to diuretics during pregnancy.^{166,171-179,188} A case report of diuretic-induced ototoxicity in a young pregnant woman treated with furosemide 80 to 180 mg and ethacrynic acid 100 to 300 mg during delivery and postpartum for 5 days was published.¹⁸⁴ Thiazide diuretics have been associated with agranulocytosis and thrombocytopenia in nonpregnant individuals.³¹

Vasodilators

Hydralazine

Teratogenicity. Hydralazine is most commonly used during the second and third trimesters to control elevated blood pressure seen in preeclampsia and eclampsia. For these reasons, first trimester experience is limited. The Collaborative Perinatal Project found no birth defects in neonates of eight women exposed to hydralazine during their first trimester. The frequency of major birth defects was not significantly increased in 40 newborns exposed to hydralazine during the first trimester in the Michigan Medicaid Surveillance Study.¹⁰¹

In a case series, no birth defects were observed in pregnant women with moderately severe chronic hypertension treated with hydralazine and propanolol early in pregnancy.¹⁸⁹

Fetal and neonatal adverse events. No serious adverse reactions were reported in approximately 20 pregnant women randomized to oral hydralazine for treatment of mild to moderate chronic hypertension.^41,46 Two additional trials failed to detect serious adverse reactions. In controlled trials, oral hydralazine for management of mild to severe pregnancy-induced hypertension did not result in serious adverse reactions.^{49,125,138,190} There are case reports of premature labor,¹⁰¹ neonatal transient thrombocytopenia,¹⁹¹ and a lupus-like syndrome associated with hydralazine use.¹⁹²

Maternal serious adverse reactions. One case of a lupus-like syndrome associated with parenteral hydralazine therapy late in pregnancy was located; the symptoms resolved after discontinuation of the drug.¹⁹² Six cases of hydralazine-induced hepatitis during pregnancy were found.^193,194

Thrombocytopenia, pancytopenia, cholecystitis, and glomerulonephritis have been described in nongravid populations during hydralazine therapy. Long-term therapy with hydralazine has been associated with a syndrome resembling systemic lupus erythematosus in up to 74 percent of subjects exposed to the drug. More than 20 cases of hydralazine-induced hepatitis have been reported.³¹

Angiotensin-Converting Enzyme Inhibitors

Teratogenicity. The majority of data relevant to use of ACE inhibitors during pregnancy are limited to captopril, enalapril, and lisinopril; but reports suggest that teratogenicity risk is similar throughout the drug class. It appears dependent on timing of initial administration, dose, and duration of exposure. The Michigan Medicaid Surveillance Study ²⁹ included 141 newborns who had been exposed to ACE inhibitors (86 exposed to captopril, 40 to enalapril, and 15 to lisinopril) during the first trimester. The estimated incidence of major birth defects was 7.0 percent.²⁹ A review has summarized 85 case reports published before 1989 of pregnancies exposed to captopril and enalapril.¹⁹⁵ Neonatal anuria was reported in 15 percent of the pregnancies exposed to ACE inhibitors after 16 weeks of gestation. Oligohydramnios was reported in 14 percent of the pregnancies and appeared more common with higher doses of ACE inhibitors. Since the 1989 review, 8 additional case reports appeared, describing 29 adverse outcomes related to fetal kidney development after second and third trimester exposure to ACE inhibitors.^196-203 Three cases of severe skull hypoplasia were attributed to use of ACE inhibitors during pregnancy.^204,205 This complication has been attributed to the agents' deleterious effect on amniotic fluid, peripheral perfusion, and fetal blood pressure.²⁰⁶

Fetal hypotension and persistent inhibition of the renin-angiotensin system have been proposed as the underlying mechanisms for the observed fetopathy. Because fetal tubular function begins at the 10^th embryonic week, exposure during the first trimester could be safe. In fact, limited published reports confirm the relative safety of ACE inhibitors during first trimester use only. Administration of an ACE inhibitors during the first trimester only did not result in teratogenic effects in 80 cases.^{54,207, 208, 209}

Fetal and neonatal adverse events. Several case reports associated ACE inhibitors used during second and third trimester to fetal growth retardation, hypotension, pulmonary hypoplasia, respiratory distress syndrome, and patent ductus arteriosus.^29,63 A detailed review of 85 case reports of ACE inhibitors-exposed pregnancy published before 1989 supported the assumption of increased risk for neonatal pulmonary hypoplasia (3 cases found) but did not confirm the suggested increased risk of respiratory distress syndrome or persistent patent ductus arteriosus after considering weeks of gestation at birth.

No published reports were found describing fetal or neonatal outcomes with the use of quinapril, ramipril, trandolapril, fosinopril, cilazapril, and perindopril.

Maternal serious adverse reactions during pregnancy. No published reports of maternal serious adverse reaction were found.

ACE inhibitors are widely used to treat chronic hypertension in the nongravid population and are the preferred agent for hypertensive individuals with renal insufficiency and diabetes. In general, this antihypertensive class is well tolerated. ACE inhibitors have been shown to cause acute renal failure (0.1 to 0.2 percent). Particularly at risk are subjects with altered renal function, hypovolemia, and bilateral renal artery stenosis. ACE inhibitors also have been associated with drug-induced agranulocytosis, Guillain-Barré neuropathy, and hyponatremia-induced delirium.³¹ Cholestatic jaundice due to captopril was reported, and the risk is estimated to be 0.09 per 1,000 patients.³¹ Acute hepatitis also has been described with enalapril.

Angiotensin receptor blockers

Because these agents act on the same physiologic pathway as ACE inhibitors (the renin-angiotensin system) and they appeared on the market after the fetal adverse events associated with ACE inhibitors were well known, they have not been used to treat pregnant women.

Teratogenicity. No published reports on the teratogenicity of this class of agents were found.

Fetal and neonatal adverse events. No published reports of first trimester or gestational exposure were found.

Maternal serious adverse reactions. No published reports of maternal serious adverse reactions during pregnancy were found.

Safety and tolerability data from controlled trials in the general population suggest that the side effect profile of ARBs is similar to placebo. Limited long-term experience exists with these agents. Case reports of purpura, acute psychosis with paranoid delusion, pancreatis, hepatotoxicity, and angioedema have been published.

Limitations

Most of the evidence on harm associated with antihypertensives in pregnancy is limited to case reports. The interpretation of these reports is difficult for several reasons. As mentioned previously, the actual incidence of the serious adverse reactions reported are unknown. Two factors complicate the estimation of the incidence: (1) it is likely that the number of published case reports is an underestimate of the actual existing number of subjects experiencing the reported adverse reaction, and (2) it is impossible to ascertain the exact number of women exposed to antihypertensives during pregnancy. Also, case reports do not establish causality. This limitation is amplified in drug exposure occurring during pregnancy, as rechallenge and information related to previous exposure is nonexistent. The condition for which pregnant women are treated with antihypertensives is usually serious and can be partially responsible for the adverse fetal and neonatal outcomes. In most cases, antihypertensives were administered as part of a multidrug regimen.

Question 4: Effects of Specific Antihypertensive Agents

Are particular antihypertensive agents more effective or harmful than others in treating mild to moderate chronic hypertension during pregnancy?

As previously noted, trial data are inadequate to prove or disprove benefits of treating mild to moderate chronic hypertension during pregnancy. Evidence on direct comparisons of antihypertensive drugs in trials is even more limited. Until further data become available, selection of a specific antihypertensive, if one is used at all, must be based on a weighing of the relative risks and benefits of antihypertensives derived from various evidence sources. This section presents a balance sheet summarizing the benefits and harms of antihypertensive agents either recommended or proscribed in recent guidelines and reviews of chronic hypertension in pregnancy. In the absence of demonstrated benefit from any of the drugs, their safety record becomes paramount. Methyldopa and thiazide diuretics appear to be safe in pregnancy, given their long clinical use and few reported adverse events. There is evidence that the beta-blocker, atenolol, given early in pregnancy may be associated with intrauterine growth restriction. There is no evidence of major adverse events associated with calcium channel blockers, but the clinical experience with this drug class is much more limited. Angiotensin-converting enzyme inhibitors given in the second or third trimester cause fetal renal failure. As the same complication may also occur with angiotensin II receptor blockers, this class of agents has been avoided in pregnancy.

Trial Findings

Data on head-to-head comparisons of antihypertensive agents were very limited. Two trials compared different antihypertensive agents in pregnant women with chronic hypertension.^41,50 The first trial involved 300 women and compared methyldopa with labetalol and placebo.⁵⁰ No significant differences between the treatment arms in neonatal mortality, birthweight, or Apgar scores were reported. The second trial involved 21 women and compared methyldopa with hydralazine and placebo. No statistically significant differences in pregnancy outcomes were noted. Neither of the trials had sufficient power to detect even moderate to large differences.

Table 14. Balance sheet of benefits and harms of antihypertensive (more...)

Table 14. Balance sheet of benefits and harms of antihypertensive agents

Agent or Class	Benefits	Harms	Clinical experience in pregnancy
Methyldopa	Fetal: Insufficient evidence to rule out large effect on perinatal morbidity or mortality. Maternal: Insufficient evidence to rule out large effect on maternal morbidity.	Fetal: Evidence of no major adverse events. Maternal: Hepatitis (est. 1-10 per 100,000 in nongravid population)	Large
Beta-blockers Alpha/beta- blockers	Fetal: Insufficient evidence to rule out large effect on perinatal morbidity or mortality. Maternal: Insufficient evidence to rule out large effect on maternal morbidity.	Fetal: Limited evidence of possible IUGR with atenolol used early in pregnancy. Maternal: Evidence of no major adverse events.	Large (beta-blockers) Small (alpha/beta blockers)
Diuretics	Fetal: Insufficient evidence to rule out large effect on perinatal morbidity or mortality. Maternal: Insufficient evidence to rule out large effect on maternal morbidity.	Fetal: Evidence of no major adverse events. Maternal: Evidence of no major adverse events.	Large
Calcium channel blockers	Fetal: Insufficient evidence to rule out large effect on perinatal morbidity or mortality. Maternal: Insufficient evidence to rule out large effect on maternal morbidity.	Fetal: Very limited evidence of no major adverse events. Maternal: Very limited evidence of no major adverse events.	Small
Hydralazine	Fetal: Insufficient evidence to rule out large effect on perinatal morbidity or mortality. Maternal: Insufficient evidence to rule out large effect on maternal morbidity.	Fetal: Evidence of no major adverse events. Maternal: Evidence of no major adverse events.	Moderate (for chronic hypertension)
ACE inhibitors ARB	Fetal: No evidence. Maternal: No evidence.	Fetal: Risk of fetal renal failure if used in 2nd or 3rd trimester. Maternal: No evidence.	Small None

IUGR = intrauterine growth retardation; ACE = angiotensin-converting enzyme; ARB = angiotensin receptor blocker

Until data to directly address the relative efficacy of antihypertensives in pregnancy are available, the choice of drug will depend on an assessment of the relative benefits and harms of the different agents. Table 14 provides a summary of the efficacy and harm data summarized in Questions 2 and 3. The antihypertensives in the table are those that are specifically recommended or proscribed in recent reviews and guidelines.

Question 5: Nonpharmacological Treatment and Outcomes

Does nonpharmacological treatment of mild to moderate chronic hypertension during pregnancy improve maternal and perinatal outcomes?

No pertinent randomized controlled trials were found that addressed this question.

Question 6: Nonpharmacological vs. Pharmacological Treatment

Is nonpharmacological treatment as efficacious as pharmacological treatment to improve maternal and perinatal outcomes in women with chronic hypertension?

No pertinent randomized controlled trials were found that addressed this question.

Question 7: Combination Treatment

Does a combination of pharmacological and nonpharmacological treatment improve maternal and perinatal outcomes over either treatment alone?

No pertinent randomized controlled trials were found that addressed this question.

Question 8: Blood Pressure Levels and Treatment

What is an appropriate blood pressure level at which to treat chronic hypertension during pregnancy and when should therapy be initiated? What is an appropriate blood pressure level at which to maintain treatment?

Multiple large cohort and case-control studies from multiple countries consistently show increased risk of the following in pregnant women with mild to moderate diastolic hypertension: perinatal mortality, abruption, low birthweight, intrauterine growth retardation, and preeclampsia. Risks of the adverse outcomes appear increased in pregnant women with chronic hypertension, regardless of whether or not preeclampsia is superimposed. Risks are apparent with DBP levels greater than 85 to 90 mmHg, and increase with higher levels of diastolic blood pressure. None of the studies evaluates risks controlled for treatment cointerventions. The studies do not directly address questions of whether antihypertensive drug treatment is more beneficial than harmful or evaluates appropriate blood pressure levels at which to initiate or maintain treatment.

Description of Relevant Evidence

Figure 8. Selection process for studies addressing (more...)

   Figure 8. Selection process for studies addressing blood pressure level to initiate and maintain treatment

Details of the search for evidence are presented graphically in Figure 8. From 1,774 records identified in the search, 46 met selection criteria specified in Table 1.^{7,20,25, 26, 27,56,123,213-255} Each of the studies provided information about either maternal or fetal risks associated with chronic hypertension during pregnancy. Although definitions of chronic hypertension varied among studies, most defined chronic hypertension as known hypertension prior to pregnancy and/or blood pressures of >140/90 mmHg occurring before 20 weeks' gestation. The Korotkoff stage used to determine DBP usually was not stated; therefore, biases in overestimating blood pressure with use of phase IV instead of phase V could not be ascertained.

Most studies used retrospective cohort or case-control designs. Only four were prospective cohorts.^{7,243,249,252} Two-thirds of the studies evaluated effects of single variables (e.g., presence or absence of chronic hypertension) on outcomes, without taking into account other important confounders such as age of mother, parity status, sociodemographic factors, smoking, prior history of preeclampsia, or potential cointerventions such as pharmacological or nonpharmacological treatment. Eleven studies used multivariate techniques to try to adjust for some of these factors.^{27,217,218,237, 238, 239,241,243,252,254,255}

There were 24 studies from North America, 12 from Europe or Scandinavia, and 4 from developing countries (Africa, China, Brazil). Sample sizes varied as follows: <1,000 participants, 18 studies; >1000 but <10,000 participants, 9 studies; >10,000 but <100,000 participants, 13 studies; >100,000 but less than 1,000,000 participants, 3 studies; and >1 million participants, 3 studies. The large studies often used birth registry or hospital record data to assess outcomes in general obstetric populations. The smaller studies were typically case series of high-risk patients or small case-control studies. Thirty-two of the studies were published in the 1990's whereas 8 were published before 1980 and included patients who had been studied in the 1950's.

Trial Findings

Figure 9. Perinatal death outcome results given as (more...)

   Figure 9. Perinatal death outcome results given as the relative risk (log scale) associated with chronic hypertension compared with either normotensive or general obstetrical populations

Findings of the individual studies are detailed in evidence tables that follow the glossary. Results of studies that reported perinatal mortality outcomes in chronic hypertensive mothers compared with outcomes in normotensive mothers or general obstetrics populations are presented in Figure 9. Every study consistently shows increased risk of perinatal death associated with chronic hypertension. The random effects summary odds ratio for these studies is 3.4 (95 percent CI 3.0 to 3.7).

Table 15. Summary random effects odds ratios of perinatal (more...)

Table 15. Summary random effects odds ratios of perinatal mortality risks derived from studies comparing outcomes in women with chronic hypertension with women who were either normotensive or from a general obstetrical population

Study Type	Summary OR	95 Percent CI	Heterogeneity test
North American	3.2	2.6 to 3.9	P=0.10 ( 4 studies)
Other countries	3.5	2.5 to 4.9	P=0.98 ( 6 studies)
Prospective cohort	2.9	0.4 to 22.0	P=0.72 ( 2 studies)
Retrospective cohort	2.9	2.4 to 3.5	P=0.39 (4 studies)
Case-control	3.6	3.2 to 4.1	P=0.92 (4 studies)
All studies	3.4	3.0 to 3.7	P=0.64 (10 studies)

OR = odds ratio; CI = confidence interval

(Of note: In 1997, the overall incidence of infant deaths in the United States was 7.2 per 1,000 live births.) Summary odds ratios for North American studies, studies from other countries, prospective cohort studies, all cohort studies, and case-control studies are shown inTable 15. In every instance, chronic hypertension was associated with increased perinatal mortality. Three studies confirmed that more severe maternal hypertension is associated with greater perinatal mortality than in mild to moderate hypertension.^221,227,247

Figure 10. Abruption outcome results given as relative (more...)

   Figure 10. Abruption outcome results given as relative risk (log scale) associated with chronic hypertension compared with normotensive or general obstetrical populations

Results of studies that reported abruption occurring in chronic hypertensive mothers compared with normotensive mothers or the general obstetrics populations are shown inFigure 10. Studies generally show increased risk of abruption associated with chronic hypertension, although one study suggests hypertension is associated with decreased risk.²⁴³ The summary OR for risk of abruption is 2.0 (95% CI 1.1 to 3.7). Of note, four studies provide risks of abruption associated with chronic hypertension adjusted for sociodemographic factors, smoking status, and sometimes parity. These adjusted risk estimates were 2.3 (95 percent CI 1.5 to 3.6),²¹⁷ 1.4 (95 percent CI 0.5 to 3.6),²⁴⁰ 1.81 (95 percent CI 1.07 to 3.05),²⁴¹ and 0.5 (95 percent CI unavailable, published article gives incorrect value).²⁴³ One large retrospective cohort study and a retrospective analysis of a multicenter trial show that chronic hypertension with superimposed preeclampsia is associated with higher risks of abruption than chronic hypertension alone.^27,239

All ^{20,221,223,242,243} but one ⁵⁶ of the studies that reported associated risks of preeclampsia with chronic hypertension compared with no hypertension found positive associations, though risk estimates varied, probably because of varying definitions of preeclampsia. Several studies evaluated risks of prematurity,^216,237 SGA,^{56,224,227,232,243,252} low birthweight,^{26,223,225,234,248,255} or intrauterine growth restriction ^{20,25,216,221,249} associated with chronic hypertension compared with either the general obstetrics population or normotensive pregnant women. In all but two instances, chronic hypertension was associated with increased risks of these outcomes.^224,234

One large retrospective analysis of a prospectively maintained perinatal database in the United States examined maternal mortality associated with chronic hypertension.²¹⁶ In this study, chronic hypertension was defined as blood pressure >140/90 mmHg occurring earlier than the 20^th week of gestation or prepregnancy. The reported maternal mortality in chronic hypertensive pregnant women was 230/100,000 live births. Mortality in normotensive pregnant women was 10.6/100,000 live births. Mortality in women with pregnancy-induced hypertension was 160/100,000. Mortality by severity of hypertension and superimposed eclampsia was not given.

Question 9: Benefits of Aspirin

Is aspirin beneficial in preventing maternal and fetal complications in pregnant women with mild to moderate chronic hypertension?

One double-blind placebo controlled randomized trial involving 774 women with chronic hypertension found that low-dose aspirin, 60 mg daily, begun before 26 weeks gestational age did not significantly reduce risk of preeclampsia (risk difference 0.01, 95 percent CI -0.05 to 0.08), intrauterine growth retardation (risk difference 0.01, 95 percent -0.03 to 0.06) or perinatal mortality (risk difference -0.02, 95 percent CI -0.05 to 0.01). Low-dose aspirin was not associated with significant increases in abruption, postpartum hemorrhage, or neonatal intraventricular hemorrhage. The trial could have missed small reductions in the incidence of preeclampsia (10 to 20 percent), moderate reductions (30 percent) in intrauterine growth restriction, and large reductions in perinatal mortality (50 percent), as well as small increases in risks of maternal bleeding. Six other trials of low-dose aspirin that provided subgroup data regarding women with chronic hypertension are generally consistent with the large trial's findings, but there are 11 trials from which such subgroup data were not obtained.

Description of Relevant Evidence

Figure 11. Selection process for aspirin prophylaxis (more...)

   Figure 11. Selection process for aspirin prophylaxis studies

The literature search identified 46 randomized controlled trials evaluating benefits of aspirin in pregnant women. The process of selection that identified 18 trials meeting the prespecified selection criteria is depicted in Figure 11. These 18 trials involved pregnant participants with "mild to moderate" chronic hypertension and compared aspirin with either placebo or usual care. All had maternal and/or fetal morbidity or mortality outcomes. Trials that did not meet prespecified selection criteria were generally trials conducted in populations that did not include chronic hypertensives or did not have morbidity or mortality outcomes.

The 18 trials were conducted in countries around the globe including Australia, Brazil, Finland, France, Israel, Italy, the United Kingdom, the United States, West Indies, and Zimbabwe. Only one of the trials stratified randomization on the basis of maternal chronic hypertension; it included 774 such women.²⁶¹ The remaining trials included women with chronic hypertension as a subgroup of eligible "high risk" women. "High risk" women were defined in various ways and included women with diabetes, renal disease, previous complicated pregnancy, history of preeclampsia, or multifetal gestation. Subgroup data for 699 women with chronic hypertension were obtained from six of the trials^229,262-266 but were unobtainable for more than 2,500 women with chronic hypertension from reports or requests to authors of 11 trials.^256,267-276

Table 16. Summary of aspirin trials

Table 16. Summary of aspirin trials

Study

Date

Country

Participant numbers CHTN/total

BP definition

Aspirin dose (mg)

Comparision group

Gestational age at aspirin initiation (wks)

Outcomes measured and reported for CHTN group

Preterm delivery

IUGR

Perinatal death

Pre-eclampsia

Gestational age at birth

Birthweight

261 a

1998

USA

763/763

SBP >140 DBP >90

60

Placebo

20+ 4

262 a

1998

Jamaica

54/6,275

SBP >140 DBP >90

60

Placebo

<20

229 a

1998

Zimbabwe

38/250

Not given

75

Placebo

20-28

263

1996

Brazil

473/1,009

CHTN detected before or during pregnancy

60

Placebo

12-32

264 a

1989

Italy

11/33

SBP >140 DBP >90

60 + atenolol + hydralazine

Placebo

12

265

1994

Pakistan

100/200

SBP >140 DBP >90

150

Placebo + antihypertensive

Not given

266

1996

Columbia

23/ not given

ACOG criteria

100

Placebo

12

a

Denotes that unpublished data were obtained from investigators and included in this report.

CHTN = chronic hypertension; IUGR = intrauterine growth retardation; SBP = systolic blood pressure; DBP = diastolic blood pressure; ACOG = American College of Obstetricians and Gynecologists

Characteristics of the trials are shown in Table 16. In the trial that stratified randomization based on hypertension, chronic hypertension was defined as: documentation of antihypertensive drug therapy by medical records or a blood pressure while sitting of 140/90 mmHg or higher taken on two occasions at least 4 hours apart, either before pregnancy or before the 26^th week of pregnancy.²⁶¹ Trials with subgroup data either did not report their definition of chronic hypertension or defined it as blood pressure over 140/90 mmHg, with few further details given. Aspirin was used as the single therapy in doses of 50 to 150 mg daily except for one study that combined aspirin therapy with antihypertensive therapy.²⁶⁴ In the trial that stratified randomization based on chronic hypertension,²⁶¹ treatment was begun within 24 weeks of gestation. Trials with subgroup data began aspirin before 12 weeks gestation,^264,266 less than 20 weeks gestation,²⁶² or over 20 weeks^229,263or did not report timing of initiation of therapy.²⁶⁵ Aspirin treatment was stopped shortly before or at time of delivery.

Table 17. Design characteristics of aspirin trials

Table 17. Design characteristics of aspirin trials

Study	Randomization concealment	Blinding	Drop Outs	A priori hypothesis for chronic hypertension
261	Clearly adequate	Double	Yes, not specified	Yes
262	Clearly adequate	Double	Yes, not specified	Not clear
229	Clearly adequate	Double	Yes, not specified	Not clear
263	Possibly adequate	Double	18/473	Not clear
264	Inadequate	Single	0	Not clear
265	Not described	Single	0	Not clear
266	Not described	Double	Not specified	Not clear

Important quality parameters of the trials are shown in Table 17. The trial that stratified randomization based on hypertension was well designed and conducted, whereas several of the other trials were single blind or did not report randomization concealment. The trial with stratified randomization had a stated a priori hypothesis and sample size calculation specifically for women with chronic hypertension. Trials with subgroup data usually did not have clearly stated a priori hypotheses regarding chronic hypertension and either did not give sample size calculations or did not clearly base them on numbers of women with chronic hypertension.

Trial Findings

Figure 12. Rates of preeclampsia observed in chronic (more...)

   Figure 12. Rates of preeclampsia observed in chronic hypertensive vs. all high-risk participants in various aspirin trials^a

Figure 12 shows the rates of preeclampsia in women with chronic hypertension who were not treated with aspirin and in other "high risk" women who were also not treated with aspirin. These rates were calculated from the control groups of the relevant trials. In general, rates of preeclampsia were higher in women with chronic hypertension than for "high risk" women. These findings show that different populations have been studied in the various trials and participants with chronic hypertension typically had higher risk than the mixed "high-risk groups."

The figure also displays markedly varying rates of preeclampsia in chronic hypertensive participants in the trials. The three trials^229,261,262 that defined chronic hypertension as occurring early in pregnancy or known prior to pregnancy showed rates of preeclampsia of approximately 20 percent to 25 percent. One of the studies that had a much higher rate of preeclampsia was extremely small (n=11) and single blind.²⁶⁴ The other study with a high rate of preeclampsia included women with both chronic hypertension and prior pregnancies complicated by hypertension.²⁶⁵ The large study with the particularly low rate of preeclampsia (7 percent) recruited patients up to 32 weeks of gestation and included patients with probable pregnancy-induced hypertension within the definition of chronic hypertension.²⁶³

Figure 13. Preeclampsia results of trials: Risk differences (more...)

   Figure 13. Preeclampsia results of trials: Risk differences between aspirin prophylaxis groups vs. control groups

Outliers regarding preeclampsia results are further examined in Figure 13.^229,261-266 In this analysis, the larger of the two trials with very high rates of preeclampsia ²⁶⁵ was confirmed as having significantly different results from the remaining trials.

Figure 14a. Intrauterine growth retardation results (more...)

   Figure 14a. Intrauterine growth retardation results of trials: Risk differences between aspirin prophylaxis groups vs. control groups

Figure 14b. Small-for-gestational-age results of trials: (more...)

   Figure 14b. Small-for-gestational-age results of trials: Risk differences between aspirin prophylaxis groups vs. control groups

Figure 14c. Birthweight results of trials: Mean difference (more...)

   Figure 14c. Birthweight results of trials: Mean difference between aspirin prophylaxis groups vs. control groups

Figure 15. Perinatal death results of trials: Risk (more...)

   Figure 15. Perinatal death results of trials: Risk differences between aspirin prophylaxis groups vs. control groups

Figure 16. Preterm delivery results of trials: Risk (more...)

   Figure 16. Preterm delivery results of trials: Risk differences between aspirin prophylaxis groups vs. control groups

Trial findings relevant to women with chronic hypertension are presented in forest plots for several maternal/fetal/newborn outcomes (Figures 14, 15). Except for the one study already confirmed as an outlier,²⁶⁵ data show neither statistically significant positive nor negative effects on outcomes. Two trials had borderline significant reductions in preterm delivery with aspirin compared with placebo (Figure 16).^261,263 In one, aspirin was associated with a slightly longer (average 2 days, 38.1 vs. 37.8 weeks) gestation period than placebo.²⁶³ Posthoc power analyses of the trial that stratified randomization showed 80 percent power (alpha=0.05) to detect about a 50 percent RR reduction in perinatal mortality, about a 10 percent to 20 percent RR reduction in preeclampsia, and about a 30 percent RR reduction in intrauterine growth retardation.

Question 10: Use of Special Monitoring Techniques

Is the use of special fetal monitoring techniques (biophysical profiles, Doppler velocimetry, nonstress tests, contraction stress tests, fundal measurements, amniotic fluid index, ultrasound fetal biometry, fetal movement counting) and strategies (priority ordering of testing, number of tests, or timing of tests) beneficial or harmful to mothers and fetuses? Are there particular subsets of women for whom special monitoring techniques are warranted?

There are no data from methodologically acceptable studies to address either the benefits or the harms of various monitoring strategies for pregnant women with chronic hyptertension.

Background and Lack of Evidence

Women with chronic hypertension in pregnancy are often monitored intensively for the appearance of maternal and fetal complications, though the benefits, harms, and marginal gains of such monitoring are not clear. Monitoring techniques include serial ultrasonography for fetal growth, Doppler velocimetry of the umbilical or uterine arteries, nonstress tests, biophysical profiles, and biochemical tests such as plasma urate. Important issues are: (1) the diagnostic accuracy of the tests for detecting particular maternal and fetal complications, (2) the efficacy of the monitoring techniques for preventing perinatal morbidity and mortality, and (3) the most effective timing, repeat intervals, and sequencing of tests. For this report, technical experts were specifically interested in whether there was evidence of clinical benefits or harms of monitoring techniques in women with chronic hypertension. Thus, literature related solely to diagnostic accuracy of a monitoring technique compared with a reference standard was not reviewed.

Figure 17. Selection process for studies addressing (more...)

   Figure 17. Selection process for studies addressing fetal/maternal monitoring techniques

RCT = randomized controlled trial

Searches identified 939 records addressing a monitoring technique. Of these, none met prespecified selection criteria of having pregnant participants with chronic hypertension and clinical perinatal morbidity or mortality outcomes (Figure 17). Rather, most of the studies were small case series in other groups of pregnant women and examined accuracy of a monitoring technique compared with another standard.

Benefits of Treating Hypertension Preconception

Precise estimates of clinical benefits of antihypertensive therapy in young women with mild to moderate hypertension are not available. Three randomized trials involving 8,565 nongravid women with mild to moderate hypertension, ages 30 to 54, show approximately 259 (CI 158 to 1,606) such women need to be treated for 5 years to prevent a fatal or nonfatal cardiovascular event. No clinical outcome data from trials in younger women are available, and absolute benefits are likely smaller the younger the woman and the shorter the treatment. Assuming an annual risk of any adverse cardiovascular event of less than 0.5 percent (a safe assumption for most women of childbearing age with mild to moderate chronic hypertension) and assuming that RR reductions established in trials are relatively stable, approximately 8,000 women need to be treated annually to prevent one cardiovascular event (95 percent CI 2,500 to 50,000).

Benefits of Treating Hypertension During Pregnancy

Data were too scant to either prove or disprove clinical improvements of at least 20 percent when mild to moderate chronic hypertension during pregnancy was treated.

Adverse Effects of Antihypertensive Agents

The methodological quality of research evidence addressing adverse effects of antihypertensive drug therapy in pregnant women is weak. Establishing causation in pregnancy with dechallenge/rechallenge tests is not feasible, and large clinical trials enrolling pregnant women have not been done. Regardless, several antihypertensive agents have been associated with specific adverse events.

• ACE inhibitors used in the second or third trimester have caused renal dysfunction in the fetus, leading to oligohydramnios and anuria, and have been associated with pulmonary hypoplasia, growth retardation, and a unique hypoplasia of the fetal skull.

• Atenolol has been associated with fetal growth retardation in several studies, especially when the drug was started early in pregnancy. The causal nature of this association has been difficult to disentangle; multiple agents have been used in many of the studies and effects due to underlying maternal conditions vs. effects due to drugs usually were not clear. Labetalol, used late in pregnancy, has been associated with intrauterine growth retardation in three randomized trials, though other studies have not confirmed this. Metoprolol, pindolol, and oxprenolol have not been associated with intrauterine growth retardation, although experience with these drugs is limited. Overall, there is insufficient evidence to establish whether certain beta-blockers are safer than others and whether apparent associations with intrauterine growth retardation are a result of an adverse drug effect or a complication of hypertensive disease.

• Despite theoretical concerns of the effect of diuretics on plasma volume, no increased risk of fetal adverse events were reported in several RCTs. Diuretics were not associated with an increased risk of fetal adverse events in several randomized trials.

• Methyldopa has not been associated with any pattern of fetal anomalies. The risk of hepatitis in nonpregnant women is estimated at 1 to 10 per 100,000 persons exposed to methyldopa; it has not been established whether this risk is different in pregnancy.

• In anecdotal reports, nifedipine has been associated with severe neuromuscular blockade in cases where it was given with magnesium and with hypotension resulting in fetal distress when short-acting nifedipine was given alone.

• Calcium channel blockers for the treatment of pregnancy-induced hypertension have not been associated with serious fetal or neonatal outcomes. Caution is advised on interpreting the safety of calcium channel blockers as the evidence is limited.

Effects of Nonpharmacological Interventions

There were no randomized trials in pregnant women with mild to moderate chronic hypertension that evaluated effects of interventions such as diet, bed rest, dietary supplements, hydrotherapy, biofeedback, relaxation, smoking cessation, rhubarb, and fish oil.

Optimum Levels for Initiating Therapy and Risks of Elevated Blood Pressure

Optimum blood pressure for initiating and maintaining treatment could not be gleaned from the studies that evaluated risks associated with chronic hypertension. Though studies were limited by many confounding factors, they consistently showed chronic hypertension was associated with approximately threefold increases in risk of perinatal mortality and approximately twofold increases in risks of abruption. Consistent increased risks of preeclampsia and of smaller babies also were observed. Risks were higher in women with more severe hypertension. Increased fetal risks were apparent even without superimposed preeclampsia.

Effects of Low-Dose Aspirin

Low-dose aspirin, 60 mg daily, begun before 26 weeks gestational age, did not significantly reduce preeclampsia, intrauterine growth retardation, and perinatal mortality in the only double-blind trial designed specifically for pregnant women with chronic hypertension. Nor did aspirin significantly increase abruption, postpartum hemorrhage, and neonatal ventricular hemorrhage. Although the trial was of moderate size, small (20 percent improvements) clinical benefits such as prevention of intrauterine growth retardation or decreased perinatal mortality and small (20 percent) increases in risks such as hemorrhages could have been missed.

Special Monitoring Techniques

Most studies addressing monitoring techniques were small case series without clinical outcomes. None provided useful information regarding benefits and harms of fetal monitoring in pregnant women with chronic hypertension with techniques such as biophysical profiles, Doppler velocimetry, nonstress tests, contraction stress tests, fundal measurements, amniotic fluid index, ultrasound fetal biometry, and fetal movement counting.

Summary Implications

Despite the burden of illness and costs imposed by chronic hypertension in pregnancy, evidence to date remains scant and provides little direction for clinicians. Epidemiological data demonstrate increased risks of perinatal morbidity and mortality in pregnant women with mild to moderate chronic hypertension. However, treatment with either antihypertensives or aspirin has not been proven to lower those risks, though the evidence base remains too small to rule out moderate to large effects of antihypertensive agents on perinatal mortality, preeclampsia, and intrauterine growth retardation. Data on adverse effects of drug treatment are scant; the data on ACE inhibitors suggest that their adverse effects are substantially greater than for other drugs. Nonpharmacological treatments remain unevaluated, as do monitoring strategies that are frequently used in pregnancies complicated by chronic hypertension.

Many important clinical issues faced by clinicians who care for pregnant women with chronic hypertension remain unresolved. Research in this area is crucial. More pregnancies will be complicated by chronic hypertension as the trend continues for women to delay childbearing to older ages. Multicenter collaborative studies are needed with sufficient power to detect differences in outcomes such as perinatal mortality, preeclampsia, and SGA infants. Information from such trials can be augmented by well-designed surveillance systems to monitor outcomes and safety data from clinical practice.

Efficacy Data and Randomized Controlled Trials

Data About Harms

Data About Blood Pressure Risks and Optimum Treatment Levels

Data About Benefits and Harms of Special Fetal Monitoring Techniques

Technical Experts

Our panel of technical experts played an active role throughout the preparation of this report, particularly in these areas:

• Selecting the research questions,

• Suggesting the types of data to be abstracted from pertinent studies,

• Guiding the selection of relevant data to include in the evidence tables,

• Reviewing the draft report, and

• Suggesting effective methods to disseminate the final report to health professionals, policymakers and consumers.

• Baha M. Sibai, MD (Specific Content Expert)

• Professor and Chief of Maternal-Fetal Medicine

• University of Tennessee Medical Group, Inc

• Representing: Society of Gynecologist Investigations

• Robert Louis Ferrer, MD (General Content Expert)

• Assistant Professor

• Department of Family Practice

• The University of Texas Health Science Center at San Antonio

• Gerald Briggs, B Pharm

• Department of Pharmacy

• Careline Pharmacist, Women's Hospital

• Long Beach Memorial Medical Center

• Representing: American College of Clinical Pharmacists

• George H. Davis, DO, FACOOG

• Thomas Jefferson University Hospital

• Department of Obstetrics and Gynecology

• Division of Maternal-Fetal Medicine

• Jefferson Medical College

• Representing: American College of Osteopathic Obstetricians and Gynecologists

• Lelia Duley, MD, MSc, MRCOG

• Clinical Coordinator

• Institute of Health Sciences and Center for Statistics in Medicine

• Oxford, UK

• Representing: The Cochrane Collaboration's Pregnancy and Childbirth Review Group

• Thomas R. Easterling, MD

• Department of Obstetrics and Gynecology

• University of Washington Medical Center

• Seattle, WA

• Representing: Society for Maternal-Fetal Medicine

• Mary Ann Faucher, CNM, MS, MPH

• Faculty, Parkland School of Nurse-Midwifery

• Faculty Associate, University of Texas, Southwestern Medical Center

• Department of Obstetrics and Gynecology

• Dallas, TX

• Representing: American College of Nurse-Midwives

• Barbara Hatcher, PhD, MPH, RN

• Director, Scientific and Professional Affairs

• American Public Health Association

• Representing: Healthy Mothers, Healthy Babies Coalition

• William J. Hueston, MD

• Chair, Department of Family Medicine

• Medical University of South Carolina

• Representing: American Academy of Family Physicians

• Richard Lee, MD

• Professor of Medicine

• State University of New York at Buffalo

• Representing: American College of Physicians

• Evelyne Rey, MD, MSc

• Head - Obstetrical Medicine Unit

• Department of Obstetrics and Gynecology

• Sainte-Justine Hospital

• Montreal, Québec

• Representing: Canadian Hypertension Society

• Kathleen R. Stevens, RN, EdD, FAAN

• Professor

• Department of Family Nursing Care

• The University of Texas Health Science Center at San Antonio

• Representing: American Nurses Association

• Isabelle Wilkens, MD

• Associate Professor

• Department of Obstetrics and Gynecology

• Director of Maternal Fetal Medicine

• Representing: American College of Obstetricians and Gynecologists

• Stanley Zinberg, MD, MS, FACOG

• Vice President, Practice Activities

• American College of Obstetricians and Gynecologists

• Representing: American College of Obstetricians and Gynecologists

Peer Reviewers

Thirty-nine reviewers offered feedback on the draft report. Their names were obtained from the following professional groups in order to represent a variety of backgrounds and viewpoints:

• American Academy of Nurse Practitioners

• American College of Cardiology

• American College of Clinical Pharmacy

• American College of Nurse Midwives

• American College of Obstetricians and Gynecologists

• American College of Osteopathic Obstetricians and Gynecologists

• American College of Physicians

• Cochrane Collaboration's Menstrual Disorders and Subfertility Collaborative Review Group

• Society of Obstetrics and Gynecology of Canada

• Linda A. Barbour, MD

• Associate Professor of Medicine

• University of Colorado School of Medicine

• Denver, CO

• Dawn Bell, PharmD, BCPS

• Assistant Professor of Clinical Pharmacy

• West Virginia University

• Morgantown, WV

• Andrea Coffee, PharmD, BCPS

• Department of Obstetrics and Gynecology

• Scott & White Memorial Hospital and Clinic

• Temple, TX

• Michal Coleman, Sc.D, DABR

• ATL Ultrasound

• Bothell, WA

• Gary Cunningham, MD

• Professor and Chairman, Department of Obstetrics and Gynecology

• University of Texas Southwestern Medical Center

• Dallas, TX

• Larry C. Gilstrap III, MD

• Professor and Chairman, Department of Obstetrics and Gynecology

• University of Texas-Houston Health Science Center

• Houston, TX

• John C. Hauth, MD

• Professor and Director

• Center for Research in Women's Health

• University of Alabama

• Birmingham, AL

• Erin Keely, MD

• Associate Professor, Department of Medicine, Obstetrics and Gynecology

• Ottawa General Hospital

• Ottawa, Canada

• Tekoa King, CNM, MPH

• Assistant Professor, Department of Obstetrics, Gynecology and Reproductive Services

• University of California at San Francisco

• San Francisco, CA

• Gayle Olson Koutrouvelis, MD

• Assistant Professor, Department of Obstetrics and Gynecology

• University of Texas Medical Branch at Galveston

• Galveston, TX

• Line Leduc, MD

• Associate Clinical Professor, Department of Obstetrics and Gynecology

• Sainte-Justine Hospital

• Montreal, Canada

• Ronald J. Librizzi, DO

• Vice-Chairman, Department of Obstetrics and Gynecology

• Thomas Jefferson University

• Philadelphia, PA

• Lynna Littleton, PhD, RNC

• Associate Professor of Clinical Nursing

• Director, Perinatal Nurse Practitioner Program

• University of Texas Houston Health Science Center School of Nursing

• Houston TX

• Michael D. Lockshin, MD

• Professor of Medicine

• Director, Barbara Volcker Center for Women and Rheumatic Disease Hospital for Special Surgery

• College of Medicine, Cornell University

• New York, NY

• Scott N. MacGregor, DO

• Northwestern University

• Evanston, IL

• Kay F. McFarland, MD

• Professor of Medicine

• University of South Carolina School of Medicine

• Columbia, SC

• Jean-Marie Moutquin, MD, MSc

• Chairman, Department of Obstetrics & Gynecology University of Sherbrooke, Centre Régional de Fleurimont

• Sherbrooke Qc, Canada

• Patricia A. Murphy, CNM, DrPH

• Department of Obstretrics and Gynecology

• Columbia University College of Physicians and Surgeons

• New York, NY

• Julie Oki, PharmD, BCPS

• Associate Professor of Medicine

• University of Missouri-Kansas City

• Kansas City, MO

• Marie-Eileen Onieal, MMHS, RN, CPNP

• President, AANP

• Health Policy Coordinator, Bureau of Health Quality Management

• Massachusetts Department of Health

• Boston, MA

• Mark S. Paller, MD

• Professor of Medicine

• University of Minnesota

• Minneapolis, MN

• Jeffrey Pickard, MD

• Assistant Professor of Medicine

• University of Colorado and HealthONE Affiliated Hospitals

• Denver, CO

• Raymond O. Powrie, MD

• Assistant Professor

• Brown University School of Medicine

• Women's and Infant's Hospital

• Providence, RI

• George Saade, MD

• Associate Professor of Obstetrics and Gynecology

• University of Texas Medical Branch at Galveston

• Galveston, TX

• Myron H. Weinberger, MD

• Professor of Medicine

• Director, Hypertension Research Center

• Indiana University

• Indianapolis, IN

• Deborah Woolley, RN, CNM, PhD

• University of Chicago

• Women's Care Center

• Chicago Lying-In-Hospital

• Oak Park, IL

• Bruce Young, MD

• Silverman Professor of Obstetrics and Gynecology

• Director, Obstetrical Services and Maternal Fetal Medicine

• NYU Medical Center

• New York, NY

Authors Who Provided Additional Data

Some articles included in this report had relevant missing data from their publications. We contacted the authors requesting this information. Our heartfelt thanks to those who responded:

• Romano Nkumbwa Byaruhanga, MD

• St. Francis Hospital, Nsambya

• Kampala, Uganda

• D.H. Wide-Swensson, MD

• University Hospital

• Lund, Denmark

• Stellan Hogstedt, MD

• Vasteras Hospital

• Vasteras, Sweden

• Prof. Gerard Breart

• Institut National de la Sante et de la Recherche Medicale

• Paris, France

• R.G. Priest, MD

• St. Mary's Hospital

• London, England

• Klas Wichman, MD

• University Hospital

• Linkoping, Sweden

• H.J. Odendaal, MD

• University of Stellanbosch and Tygerbey Hospital

• Parowvallei, Cape Province, South Africa

• Peter Zimmermann, MD, PhD

• Obstetrics and Gynecology

• Hollola, Finland

• J. Hachicha, MD

• Centre Hospitalier Universitaire

• Service de Medicine Interne et de Nêphrologie

• Tunisie

• Patricio Lopez-Jaramillo, MD

• Facultad de Ciencias Medicas Iquique y Sodiro

• Unidad de Metabolismo Mineral

• Quito, Ecuador

• M. Renier, MD

• University Hospital of Antwerp

• Edegem, Belgium

• Eileen Gallery, MD

• Royal North Shore Hospital

• St. Leonards, Australia

• Steve Caritis, MD

• Magee Women's Hospital

• Pittsburgh, PA

• USA

• D.W. Steyn, MD

• University of Stellenbosch Tygerberg Hospital

• Tygerberg, South Africa

• P.F. Plouin, MD

• Hôpital Broussais

• Paris, France

San Antonio Evidence-based Practice Center Partners

San Antonio Evidence-based Practice Center Staff

• Statistician

• John E. Cornell, PhD

• Librarian

• Molly Harris, MA, MLS

• Technical Writer

• Karen Stamm

• Administration

• Annie Almanza

• Linn Morgan

• Computer Resources

• Dave Mullins

• Yvan Negovetic

• Peer Review Coordinator

• Sherry Parmer, C-PA

• Abstraction Form Development

• Christine Aguilar, MD, MPH