| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
ORIGINAL RESEARCH |
From the Epidemiology Branch, National Institute of Child Health and Human Development, NIH, Bethesda, Maryland; and the Magee Womens Research Institute and Department of Obstetrics, Gynecology and Reproductive Sciences, University of Pittsburgh, Pittsburgh, Pennsylvania.
Address reprint requests to: Jun Zhang, PhD, MD Epidemiology Branch National Institute of Child Health and Human Development National Institutes of Health Building 6100, Room 7B03 Bethesda, MD 20892 E-mail: jim_zhang{at}nih.gov
 |
Abstract |
|---|
Top
 Abstract Materials and MethodsResultsDiscussionReferences |
|---|
Methods: We studied 9133 singleton nulliparous pregnancies with early prenatal care from the Collaborative Perinatal Project, a large cohort study conducted between 1959 and 1965. Definitions from five different groups were evaluated. Severe maternal and perinatal morbidity and mortality were used as the outcome measurements. Sensitivity, specificity, and positive predictive value for outcomes were compared across various definitions.
Results: Blood pressure alone had very poor discriminatory power to predict adverse outcomes. Positive predictive values of adverse outcomes by the diagnosis of preeclampsia were 18–20% based on antepartum and intrapartum blood pressures and 22–36% based on antepartum blood pressure only. Mild hypertension occurring for the first time in labor and isolated mild systolic hypertension were not associated with adverse outcomes. Similarly, an increase in diastolic blood pressure of 15 mmHg that did not achieve an absolute value of 90 mmHg did not predict adverse outcome.
Conclusion: Neither blood pressure nor blood pressure and proteinuria are accurate predictors of severe adverse maternal and perinatal outcomes. Mild hypertension occurring for the first time in labor and isolated mild systolic hypertension should not be considered indicators for hypertensive disorders in pregnancy in a research definition.
Hypertension in pregnancy, often defined as blood pressure (BP) reaching 140/90 mmHg,1–5 is a common complication affecting maternal and fetal health. Currently, five definitions are widely used, most of which rely primarily on diastolic BP. However, the origin of these thresholds has been poorly documented and careful validation of these definitions has never been done. The purpose of this study was to examine quantitatively the validity of several different definitions of hypertension in pregnancy. In addition, we addressed the following questions: 1) Is mild hypertension with or without proteinuria occurring for the first time in labor or delivery associated with adverse pregnancy outcomes? 2) Besides diastolic BP, does systolic BP contribute additional information to the definition of hypertension in pregnancy? 3) Does a rise in diastolic BP above 15 mmHg but below 90 mmHg affect pregnancy outcomes?
|
Materials and Methods |
|---|
|
Top
AbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Blood pressures were recorded at entry, during each prenatal visit, during labor and delivery, and postpartum. Korotkoff phase 4 (muffling) or phase 5 (disappearance) was used for diastolic BP.7 Random urine samples were tested for albumin at each prenatal visit. A validation study in which information on BP and urinary albumin was checked against that in the original medical records showed remarkable accuracy.7 In that study, investigators selected 772 recordings suspected of error because of wide deviations from the sequence of BPs recorded in that patient during the course of pregnancy. The percentage of error for these BP readings was 1.8%. In a random sample of urinary albumin data, the percentage of error was 0.08%. Therefore, the current data appear reliable for the purpose of our study.
A total of 55,908 pregnancies were included in the project. We restricted our analyses to nulliparas with singleton pregnancies who had their first prenatal visits at or before 22 weeks, had at least three prenatal visits, and gave birth between 20 and 45 weeks inclusive (n = 9484). We excluded births with potentially lethal malformations or births before 34 weeks with a spontaneous onset of labor because these births had little to do with hypertension but did contribute a substantial proportion of adverse perinatal outcomes. These restrictions and exclusions yielded 9133 women with 210,109 systolic and 213,421 diastolic BP readings and 106,102 urinary albumin test results.
We defined BP and proteinuria as follows:
The predictive value of the definitions of hypertension in pregnancy was assessed by whether women classified as hypertensive had unfavorable maternal or perinatal outcomes. Table 1
lists several major adverse outcomes that can be caused by hypertension in pregnancy. Given the low incidence of these events, we combined them into one overall outcome, ie, women who had any of these events were classified as having an adverse outcome. Receiver-operating characteristic curves were used to identify cutoff points of systolic and diastolic BPs as thresholds for defining hypertension in pregnancy.8 We calculated the sensitivity, specificity, and positive predictive value based on five common definitions.1–5 In the present study, sensitivity is the percentage of women with adverse outcome who also had hypertension; specificity is the percentage of women without adverse outcome who did not have hypertension; and positive predictive value is the percentage of women with hypertension who developed the outcome. These analyses were done separately in all women and in those with proteinuria. Clinical diagnosis of preeclampsia is sometimes made with findings in addition to BP and proteinuria, but we focused on BP and proteinuria because of the limited information on other biochemical signs.
|
|
Results |
|---|
|
Top
AbstractMaterials and MethodsResultsDiscussionReferences |
|---|
shows that women with adverse outcomes were slightly older, had lower socioeconomic status, were more likely to smoke during pregnancy, and weighed a bit less than women without adverse outcomes. Table 3 presents the BP and urinary albumin characteristics during pregnancy. A total of 1.6% and 1.0% of the subjects had baseline systolic BP and diastolic BP over 140 or 90 mmHg, respectively. Thirty percent of the total population had the highest systolic BP of at least 140 mmHg (versus 5% in antepartum), whereas 38% had the highest diastolic BP of at least 90 mmHg (versus 11% in antepartum). Twenty-two percent had a rise in systolic BP of at least 30 mmHg (versus 2% in antepartum), whereas 60% had a rise in diastolic BP of at least 15 mmHg (versus 21% in antepartum). Approximately 11% of women had proteinuria during pregnancy, but 2% had proteinuria even before 23 weeks’ gestation.
|
|
shows the receiver-operating characteristic curves of the highest systolic and diastolic BPs and the last antepartum systolic and diastolic BPs. Most of the curves are close to the diagonal line, which approximates a null association, ie, none of these measurements could distinguish between true positive and false positive with desirable accuracy.
|
shows that definitions by ACOG,1 the International Society for the Study of Hypertension in Pregnancy,2 and the World Health Organization5 gave similar results. The definition by the National High Blood Pressure Education Program3 tended to overdiagnose the syndrome and thus had high sensitivity but low specificity. Nevertheless, all the definitions had similar positive predictive values, ie, women classified as preeclamptic based on all BP measurements had 18–20% chance to have severe adverse outcome and those classified based on antepartum only BP had 22–36% chance. Given that the incidence of adverse outcomes was 15% in the overall population, these definitions appear to have very poor positive predictive value.
|
Anecdotal evidence suggests that mild hypertension occurring for the first time in labor at term might be attributed to anxiety, physical exertion, or medication and carries little risk of perinatal mortality and morbidity. In this regard, we selected 4796 women who never had diastolic BP over 90 mmHg throughout antepartum period and had the last prenatal visit within 2 weeks before labor or delivery. We assigned them to four groups according to diastolic BP (diastolic BP of at least 90 mmHg at least twice or diastolic BP of at least 110 mmHg at least once) and proteinuria status during labor or delivery. Table 5 shows that mild hypertension and preeclampsia occurring for the first time in labor were associated with longer gestation and higher birth-weight and had no direct impact on perinatal health. Therefore, the following analyses concern primarily antepartum BPs.
|
A rise in diastolic BP of at least 15 mmHg over baseline was used as a criterion for diagnosing hypertension in pregnancy.3 However, it is uncertain whether a rise in diastolic BP of at least 15 mmHg but with the diastolic BP remaining below 90 mmHg is associated with adverse perinatal outcomes. Table 6 compares women with various rises in antepartum diastolic BP and status of proteinuria. A rise in antepartum diastolic BP over 15 mmHg but with the diastolic BP remaining below 90 mmHg was not associated with increased risk of adverse perinatal outcomes, especially for those without proteinuria.
|
|
Discussion |
|---|
|
Top
AbstractMaterials and MethodsResultsDiscussionReferences |
|---|
The explanation for this inadequacy may lie in the common assumption that BP and proteinuria reflect the underlying pathophysiology of preeclampsia. If we assume that eclampsia represents the ultimate entity, anecdotal evidence repeatedly proved that high BP and severe proteinuria are neither sufficient nor necessary conditions. Further, BP and proteinuria can be affected by a variety of factors and underlying conditions. They may be convenient measurements but are not specific enough to define this disorder.
Problems might further occur secondary to the definition of hypertension. Hypertension during pregnancy is commonly defined as diastolic BP of at least 90 mmHg on two occasions at least 4 hours apart, but a recent well-conducted clinical trial showed that 80% of hypertensive BPs were obtained during labor and delivery,10 ie, most diagnoses of gestational hypertension and preeclampsia were based on intrapartum hypertension. Consequently, inclusion and exclusion of intrapartum BP have resulted in substantial differences in the incidence of hypertension (Table 4
). Our analysis and empirical observations indicate that mild hypertension occurring for the first time in labor is not associated with adverse outcomes. However, three recent multi-center studies conducted in the United States10–12 included antepartum and intrapartum BPs in the diagnosis of hypertension. Although the interpretation of proteinuria in voided samples obtained after membrane rupture might be questioned, two of the recent clinical trials allowed proteinuria from voided intrapartum samples.11,12
Several caveats should be noted. The data for this study are old. Obstetric practice has changed during the past 40 years. The main advantage of this data base is that the presence of thousands of actual BP and albumin measurements enabled us to base our diagnoses on the raw data rather than relying on the physicians’ diagnoses or hospital discharge summaries. It also enabled us to evaluate the performance of various definitions of hypertension and avoid the difficulties of changes in diagnostic labels over the years since the study information was collected. In addition, with advances in obstetric and neonatal care, obstetricians may be more willing to treat and end pregnancies complicated by hypertension today than 40 years ago. Aggressive intervention may curtail the natural course of hypertension and attenuate clinical manifestation. The older data thus might be beneficial by presenting more of the natural history of this syndrome. Diastolic BP might have been based on either Korotkoff 4 or 5 in the project. Korotkoff 5 is now considered better in determining diastolic BP and tends to be used more often. However, this is inconsistently applied in practice even today. Automated oscillometric BP recorder might further complicate the situation.
Lacking a criterion standard, we composed an index as the yardstick, which included severe maternal and perinatal adverse outcomes. This practical approach has some deficiencies. Because factors other than hypertension might cause adverse maternal and perinatal outcomes, and some adverse outcomes occur in women whose BPs do not reach that threshold, it is less likely for any definition of hypertension to have a high sensitivity based on this method. Likewise, mild hypertension and preeclampsia might cause less severe morbidity, ie, women without severe outcomes may still have had hypertension, which is likely to compromise the specificity. Therefore, comparison among various definitions can only be made on a relative scale within the same study. Because only patients with at least three prenatal visits were included, the incidence of adverse outcomes and the positive predictive value might have been reduced.
Overdiagnosis will increase sensitivity at a cost of including more false positive subjects (reducing specificity) and, subsequently, overtreating patients in whom the maternal and perinatal outcomes will be normal. To prevent maternal and perinatal morbidities in a clinical situation, overdiagnosis to a certain extent might be justified. For clinical and epidemiologic research, over-diagnosis is misclassification, which might reduce or inflate a treatment effect or an association. The ultimate solution for some of these problems may rely on identification of a good biomarker. The marker should be sensitive and specific to the pathophysiology of this disorder and might be used as a sole criterion or in conjunction with BP and proteinuria.
|
Footnotes |
|---|
Received July 17, 2000. Received in revised form September 25, 2000. Accepted October 12, 2000.
|
References |
|---|
|
Top
AbstractMaterials and MethodsResultsDiscussionReferences |
|---|
2. Davey DA, MacGillivrary I. The classification and definition of the hypertensive disorders of pregnancy. Am J Obstet Gynecol 1988; 158:892–8.[Medline]
3. National High Blood Pressure Education Program Working Group. Report on high blood pressure in pregnancy. Am J Obstet Gynecol 1990;163:1689–1712.
4. Redman CW, Jefferies M. Revised definition of preeclampsia. Lancet 1988;i:809–12.
5. World Health Organization. The hypertensive disorders of pregnancy. Technical Report Series 758. Geneva: World Health Organization, 1987.
6. Niswander KR, Gordon M, eds. The Collaborative Perinatal Study of the National Institute of Neurological Diseases and Stroke: The women and their pregnancies. Philadelphia: WB Saunders, 1972.
7. Friedman EA, Neff RK. Pregnancy hypertension: A systematic evaluation of clinical diagnostic criteria. Littleton, MA: PSG Publishing Co., 1977.
8. Zweig MH, Campbell G. Receiver-operating characteristic (ROC) plots: A fundamental evaluation tool in clinical medicine. Clin Chem 1993;39:561–77.
9. Myrianthopoulos NC, French KS. An application of the U.S. Bureau of the Census socioeconomic index to a large, diversified patient population. Soc Sci Med 1968;2:283–99.
10. Levine RJ, Esterlitz JR, Raymond EG, DerSimonian R, Hauth JC, Curet B, et al. Trial of calcium for preeclampsia prevention (CPEP): Rationale, design and methods. Controlled Clin Trials 1996;17:442–69.[Medline]
11. Sibai BM, Caritis SN, Thom E, Klebanoff MA, McNellis D, Rocco L, et al. Prevention of preeclampsia with low-dose aspirin in healthy, nulliparous pregnant women. N Engl J Med 1993;329:1213–8.
12. Caritis S, Sibai B, Hauth J, Lindheimer MD, Klebanoff M, Thom E, et al. Low-dose aspirin to prevent preeclampsia in women at high risk. N Engl J Med 1998;338:701–5.
13. Zhang J, Harville E. Birth-weight-for-gestational-age patterns by race, sex, and parity in the United States population: The 5th percentile. Paediatr Perinatal Epidemiol 1998;12:352–4.[Medline]
This article has been cited by other articles:
|
|
 |
|
  R. B. Ness, J. Zhang, D. Bass, and M. A. Klebanoff Interactions between Smoking and Weight in Pregnancies Complicated by Preeclampsia and Small-for-Gestational-Age Birth Am. J. Epidemiol., August 15, 2008; 168(4): 427 - 433. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. W. Powers, J. M. Catov, L. M. Bodnar, M. J. Gallaher, K. Y. Lain, and J. M. Roberts Evidence of Endothelial Dysfunction in Preeclampsia and Risk of Adverse Pregnancy Outcome Reproductive Sciences, April 1, 2008; 15(4): 374 - 381. [Abstract] [PDF]
|
|
|
|
 |
|
 F. B. Pipkin and on behalf of The Genetics of Preeclampsia Consorti Smoking in Moderate/Severe Preeclampsia Worsens Pregnancy Outcome, but Smoking Cessation Limits the Damage Hypertension, April 1, 2008; 51(4): 1042 - 1046. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. M. Roberts, G. Pearson, J. Cutler, and M. Lindheimer Summary of the NHLBI Working Group on Research on Hypertension During Pregnancy Hypertension, March 1, 2003; 41(3): 437 - 445. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
