- Email: [email protected]
A prospective study of blood pressure in pregnancy: Prediction of preeclampsia
A prospective study of blood pressure in pregnancy: Prediction of preeclampsia J. M. Moutquin, M.D., C. Rainville, R.N., L. Giroux, Ph.D., P. Raynauld, M.D., G. Amyot, M.D., R. Bilodeau, M.D., and N. Pelland, R.N. Montreal, Quebec, Canada A prospective study of blood pressure recording was conducted in 1000 patients, at each antenatal visit, with the use of an automatic random-zero sphygmomanometer. In 46 patients, among 808 primigravid women, who developed preeclampsia, the diastolic and mean blood pressures were significantly elevated compared to values at the first antenatal visit (p < 0.01, 9 to 12 weeks). This difference was sustained throughout pregnancy until delivery by at least 10 mm Hg as compared to pressures in the normotensive group. Sensitivity for predicting preeclampsia early in pregnancy with an elevated blood pressure measurement (130 to 1.35/80 to 85 mm Hg) ranged from 16% to 57% while specificity ranged from 75% to 98%. The results substantiate an early vasospasm (9 to 12 weeks) in those women destined to develop preeclampsia. (AM J 0BSTET GYNECOL 1985;151 :191-6.)
Key words: Pregnancy-induced hypertension, chronic hypertension, prediction of preeclampsia Several blood pressure studies during the antenatal period have showed a decrease in blood pressure between 16 and 20 weeks. 1 Gallery et aP demonstrated that pregnancy-induced hypertension could be predicted by a significant increase in blood pressure as early as 17 to 20 weeks. This significant difference was observed with a standardized random-zero sphygmomanometer but was not apparent with the usual sphygmomanometer. 2 However, the sample size of that study was small and there was no discrimination between primigravid women, who are more prone to preeclampsia, and multiparous women, in which chronic hypertension is more prevalent. Recently, another group3 showed that a mean arterial blood pressure of 90 mm Hg or more between 18 and 26 weeks was predictive of preeclampsia. In a preliminary study• on 366 pregnant women, we observed a similar decrease of blood pressure from 13 to 21 weeks, in systolic and diastolic blood pressure in women who remained normotensive throughout pregnancy, together with a significant rise in systolic and diastolic pressure as early as 13 to 16 weeks in 14 primigravid women who later developed preeclampsia. More observations were needed to confirm these find-
From the Department of Obstetrics and Gynecology, Notre Dame Hospital, University of Montreal. Supported by the Medical Research Council of Canada. Presented at the Fortieth Annual Meeting of The Society of Obstetricians.and Gynaecologists ofCanada, Montreal, Quebec, Canada, june 19-23, 1984. Reprint requests:]. M. Moutquin, M.D., Department of Obstetrics and Gynecology, Notre Dame Hospital, 1560 East Sherbrooke St., Montreal, Quebec, Canada H2L 4K8.
ings, not only to show the constancy of the differences but also to establish whether preeclampsia could be predicted accurately with a marker, such as one blood pressure recording, early in pregnancy. Similar studies were also carried out with the mean blood pressure. Material and methods Population. Blood pressure was recorded in 1000 consecutive pregnancies, at each antenatal visit until delivery and at the postpartum visit, between November, 1979, and April, 1984. These women were cared for by a group of three private obstetricians and no intervention was suggested for any blood pressure reading. Seventeen patients were excluded (nine twin pregnancies and eight women who were not delivered at the institution). Thus the results included 983 women (808 primigravid, 175 multiparous) with a singleton pregnancy who were delivered at Notre-Dame Hospital. More than 7800 blood pressure recordings (in duplicate) were analyzed. Mean maternal age was 25.5 ± 4.4 years ( ± SD). Mean gestational age at delivery was 39.4 ± I. 7 weeks ( ± SD). The cesarean section rate was 14.8%. Methods. Arterial blood pressure was taken from the right arm with the patient in a sitting position, after a minimum of 5 minutes of rest. This measure was taken twice at an interval of 2 minutes. If a discrepancy occurred between the two readings, then blood pressure was taken again until similar readings were observed (difference of <5 mm Hg). Then the last two recordings were kept for analysis. A nurse was responsible for the procedure and for daily calibration of the sphygmomanometer. 191
192
Moutquin et al.
January 15, 1985 Am J Obstet Gynecol
160
'0 ., +I
Cll 140
J:
E
.s w
120
"' .,.,w :;:)
...."'
100
.....
<
....w"'
"'<
Fig. 1. Averaged systolic and diastolic arterial pressures ( ± I SD) in primigravid women who remained normotensive throughout pregnancy (Q----{)) and in those who developed preeclampsia (e-----e) by periods of 4 weeks.
Table I. Incidence of hypertensive complications at delivery in primigravid and multiparous women Primigravid women No.
No. of patients Gestational hypertension Preeclampsia-ecla mpsia Mild, moderate Severe Eclampsia Hepatic* Total Chronic hypertension Without preeclampsia With preeclampsia Total
808 IS
I
l
Total
I
%
No.
1.9
I75 5
983 20
2.0
I 2 0 0 3
43 3 I 2 49
5.0
42 I I 2 46
5.2
3I
3.8
6
37
Multiparous women
2I
%
I2.0
2
4.5
23
No.
52 8 60
%
5.3 6.I
*Hepatocellular damage evidenced by altered biochemical tests, thrombocytopenia , and jaundice.
The apparatus was an automatic random-zero sphygmomanometer with digital display (Dinamap 845*) which is based on the oscillometric principle (same principle as the usual sphygmomanom eter). One advantage of this apparatus is 1ts proficiency to decrease acoustic and physical artifacts. The diastolic reading corresponded to phase IV of Korotkoff sounds. The accuracy of blood pressure measurement was validated by the obtaining of similar values (in the range of 160 mm Hg systolic to 50 mm Hg diastolic) of intra-arterial radial pressure during surgical procedures. Mean arterial pressure was calculated according to the formula: Sys*Bionetics, California.
tolic blood pressure plus twice diastolic blood pressure/3.3 Data were stored at the calculus center of the University of Montreal and analyzed with the Statistical Package for the Social Sciences. Comparisons between differences were analyzed with the unpaired t test, 5 while sensitivity, specificity, and predictive values were assessed by the method of Vecchio. 6
Results The perinatal mortality rate was 4.111000 births. There were two stillbirths and two neonatal deaths (one due to congenital cardiopathy) thus providing a corrected perinatal mortality rate of 3.1/l 000 births.
Blood pressure in pregnancy
Volume 151 Number 2
193
160
14
•
c.,
*
*
*
*
•
*
~120
J:
E
E ......
Lr_
w100 Ill:
.,.,w ::I
Ill: A. 80
_.._I----tr._____r___._I-----I•
*
'(
*
*
• Chronic hypertension o Normotension
1
* P< ·001
Y~--~~--~~~--~~-9-12
13-16
17-20
21-24
25-28
29-32
33-36
37-40
WEEKS
Fig. 2. Averaged systolic and diastolic arterial pressures ( ± 1 SD) in multiparous women who remained normotensive throughout pregnancy (o---o) and in those who developed chronic hypertension (---) by periods of 4 weeks.
120
.,w +I
"'E
J:
HO
!, w
Ill:
.,.,::I w
100
.....
90
Ill: A. <(
;;;:
w ....
Ill: <(
80
z <(
w ~
70
chronic hypertension • preeclampsia * P< ·001 • normotension •• P< ·05 L~~~~--~~--~~~0
9-12
13-16
17-20
21-24
25-28
2!'-32
33-36
37-40
WEEKS
Fig. 3. Averaged mean arterial blood pressures ( ± 1 SE) in women who remained normotensive throughout pregnancy (o---o), in women who developed preeclampsia (.---....),and in women who developed hypertension (---) by periods of 4 weeks.
Hypertensive complications are illustrated in Table I. Gestational hypertension was diagnosed when hy-
pertension alone (:;;;.140/90 mm Hg) occurred after 20 weeks and disappeared at the postpartum visit. The diagnosis of preeclampsia was ascertained by the presence of hypertension :;;;.}40/90 mm Hg on two occasions 6 hours apart, edema (> 1 kg/wk) or proteinuria (AI-
bustix 1 +) on two occasions 6 hours apart after 20 weeks, elevated serum urate levels (4.6 mg/dl or more), and a blood pressure of ,..120/80 mm Hg at 6 weeks post partum. Chronic hypertension was defined by a prepregnancy personal history of hypertension or persistence of hypertension 6 weeks post partum. Averaged systolic and diastolic blood pressures
194 Moutquin et al.
January 15, 1985 Am J Obstet Gynecol
Table II. Sensitivity, specificity, and positive and negative predictive values (percentages) of recordings of 130 and 135 mm Hg systolic blood pressure as predictors of preeclampsia Weeks' gestation 130 mmHg 9-12 13-16 17-20 21-24 25-28 135mmHg 9-12 13-16 17-20 21-24 25-28
Sensitivity
Specificity
Positive predictive value
Negative predictive value
46.2 57.6 43.3 41.2 42.1
75.4 81.2 85.3 86.2 82.7
12.5 16.0 15.1 14.9 13.1
94.9 96.9 96.1 96.2 95.8
30.8 42.4 18.9 29.4 28.9
87.1 90.2 93.5 92.8 91.4
15.4 21.2 14.9 19.2 17.2
94.3 96.2 95.0 95.7 95.4
Table III. Sensitivity, specificity, and positive and negative predictive values (percentages) of recordings of 80 and 85 mm Hg diastolic blood pressure as predictors of preeclampsia Weeks' gestation 80mmHg 9-12 13-16 17-20 21-24 25-28 85mmHg 9-12 13-16 17-20 21-24 25-28
Sensitivity
Specificity
Positive predictive value
46.2 45.5 32.4 38.2 44.7
84.2 89.8 91.0 93.4 91.7
18.2 21.7 17.9 25.5 25.0
95.4 96.4 95.7 96.3 96.4
30.8 30.3 16.2 20.2 18.4
93.0 96.6 97.4 97.9 98.0
25.0 35.7 27.3 36.3 36.8
94.6 95.7 95.1 95.5 95.1
( ± SD) by 4-week periods in the group of normotensive
nulliparous women were compared to the blood pressures in the group of nulliparous women who later developed preeclampsia (Fig. 1). From 13 weeks onward, a significant difference in systolic pressure was observed between the two groups; the increase ranged from 7 to 22 mm Hg in the group destined to develop preeclampsia (p < 0.001, Fig. 1). From the first antenatal visit (9 to 12 weeks), the diastolic pressure was consistently more elevated in patients eventually developing preeclampsia; elevations ranged from 7 mni Hg (p < 0.01, at 9 to 12 weeks) to 10 to 15 mm Hg until delivery (p < 0.001, Fig. 1). In the multiparous women, averaged systolic and diastolic blood pressures ( ± SD) showed, as expected, a significant increase of about 20 mm Hg systolic and 15 to 17 mm Hg diastolic throughout pregnancy in women with chronic hypertension compared to normotensive multiparous women (p < 0.001, Fig. 2). Comparisons of mean arterial pressures ( ± SE) between patients eventually developing preeclampsia, patients with chronic hypertension, and normotensive pa-
Negative predictive value
tients showed significant increases in both women with eventual preeclampsia and those with chronic hypertension compared to the normotensive group (p < 0.05, 9 to 12 weeks, preeclampsia; p < 0.001, all other periods of pregnancy; Fig. 3). The difference in excess of mean blood pressure was 10 mm Hg in patients with eventual preeclampsia and 16 to 20 mm Hg in patients with chronic hypertension as compared to the normotensive patients at any period of pregnancy (Fig. 3). Although significant differences in blood pressure were dearly established in the group of women with eventual preeclampsia, such a prediction needed to be substantiated in individual patients. Clinical markers of easy access to identify early in pregnancy those patients who will develop preeclampsia are yet to be found. Single values of blood pressure (systolic, diastolic, mean) were tested as potential predictors of preeclampsia in the first five periods of observations (9 to 12, 13 to 16, 17 to 20, 21 to 24, and 25 to 28 weeks); these values were examined with the method of Vecchio.• With a systolic value of 130 mm Hg, the sensitivity
Blood pressure in pregnancy 195
Volume 151 Number 2
Table IV. Sensitivity, specificity, and positive and negative predictive values (percentages) of mean blood pressures of 85 and 90 mm Hg as predictors of preeclampsia Weeks' gestation
85mmHg 9-12 13-16 17-20 21-24 25-28 90mmHg 9-12 13-16 17-20 21-24 25-28
Sensitivity
Specificity
Positive predictive value
84.6 81.8 86.5 85.3 86.8
38.6 43.2 50.4 49.7 48.9
9.5 8.2 9.5 9.0 9.5
97.1 97.5 98.4 98.3 98.4
61.5 69.7 67.6 55.9 71.1
62.0 66.4 75.4 78.1 73.9
11.0 11.4 14.2 13.0 14.4
95.5 97.3 97.5 96.8 97.6
was around 50% and specificity was about 80%. At a value of 135 mm Hg, the number of eventually preeclamptic patients who exhibited such systolic pressure was small (18% to 42%) while as few as 6.5% to 12.9% of women who remained normotensive had such systolic pressure (Table II). The positive predictive value for 130 and 135 mm Hg was rather low, as expected," because only 5% of the population studied had the disease; conversely, the negative predictive value was 94% to 96% (Table II) because the fraction of the population not having preeclampsia was large.• When diastolic values were examined as predictors of preeclampsia, sensitivity was between 38% and 46% at 80 mm Hg and specificity ranged from 84% to 98% at 80 and 85 mm Hg (Table III). Here again positive and negative predictive values yielded percentages similar to those with systolic pressure (Table III). The mean arterial pressure showed an improved sensitivity (82% to 87%) with a value of only 85 mm Hg but specificity was <50% in most instances (Table IV). Specificity increased somewhat at a value of 90 mm Hg but the sensitivity was then decreased because several eventually preeclamptic women did not have a mean blood pressure of ;;;.go mm Hg before 29 weeks (Table IV). Comment A cohort of private patients with non-high-risk pregnancies yielded 5% preeclampsia and 6% chronic hypertension. This incidence of chronic hypertension is a reflection of a rather older population (up to 40 years of age). However, the frequency of superimposed preeclampsia was only 16% in primigravid women and 13% in the to~al sample, which is less than the reported frequency of 20% to 40% for this complication. The fall in systolic and diastolic blood pressures between 13 and 25 weeks in both primigravid and multiparous women confirmed results in previous reports, 1• 2 but a larger sample size permitted us to extend this fall during the entire second trimester rather than
Negative predictive value
between 16 and 20 weeks. 2 Such a fall, coincident with increasing cardiac output and blood volume, illustrates a lowering of systemic vascular resistance in normotensive women. This study also confirms the increase of blood pressure in those women destined to develop preeclampsia. However, recording from the first antenatal visit has demonstrated for the first time a significant increase in diastolic pressure as early as 9 to 12 weeks rather than at 17 to 20 weeks. 2 This latter observation suggests a potential to identify from the first antenatal visit, in the first trimester, those women destined to develop preeclampsia with a mean diastolic blood pressure value of as low as 77 mm Hg or with a mean arterial blood pressure of93 mm Hg. In addition, this finding may be of some help for the clinician so that a particular pregnancy may be observed more closely and other tests may be carried out to confirm future preeclampsia.' More important, the increase in diastolic blood pressure suggests, for the first time, an early vasospasm at 9 to 12 weeks. Such patients did not have chronic hypertension because the blood pressure was < 140/90 mm Hg before 20 weeks, the well-defined criteria of preeclampsia were present, the blood pressure returned to normal at 6 weeks postpartum, and also because the reported incidence of preeclampsia of 5% in this study was not different from the incidence of preeclampsia in our institution. Such an early vasospasm favors early vascular pathophysiologic changes because of the presence of the conceptus in the first trimester. Another objective was to identify predictors of preeclampsia from a cohort of pregnant women. Single values of systolic, diastolic, and mean arterial pressures were tested to ascertain their potential as markers for future preeclampsia. The sensitivity and specificity for diastolic values of 80 and 85 mm Hg were not different from those of previous preliminary reports. 4 The rather poor sensitivity suggests that the systemic manifestation of vaso-
Moutquin et al.
January 15, 1985
Am J Obstet Gynecol
spasm does not appear at the same period in every woman eventually developing preeclampsia and that its expression is not necessarily evident at rest, in a sitting position. The specificity, however, increases with increasing values of both systolic and diastolic blood pressures. Percentages of specificity are similar to those previously reported! Predictive positive values were higher in a preliminary report• than in the present study; this can be attributed to the larger sample size.6 With the mean arterial pressure, the sensitivity and specificity approached those reported by others. 3 However, the predictive positive value was less in the present study because the sample size was larger than the 85 patients reported by Oney and Kaulhausen. 3 From this information, clinical evaluation of blood pressure with an automatic random-zero sphygmomanometer will identify a subgroup of patients at potentially high risk of developing preeclampsia. Concurrently, a diastolic blood pressure <85 mm Hg as early as 9 to 12 weeks has a 95% chance to predict normotension throughout· pregnancy. Only 2% of women with that diastolic blood pressure in the second trimester will develop preeclampsia; this enables the
identification of the low-risk group in a cohort of pregnant women. REFERENCES I. Wilson M, Morgani AA, Zervoudakis I, et al. Blood
2.
3.
4. 5. 6. 7.
pressure, the renin-aldosterone system and sex steroids throughout normal pregnancy. Am1 Med 1980;68:97. Gallery EDM, Hunyor SN, Ross M, Gyory AA. Predicting the development of pregnancy-associated hypertension. The place of standardised blood-pressure measurement. Lancet 1977;1:1273. Oney T, Kaulhausen H. The value of the mean arterial blood pressure in the second trimester (MAP-2 value) as a predictor of pregnancy-induced hypertension and preeclampsia. A preliminary report. Clin Exp Hypertens [B) 1983;2:211. Moutquin 1M, Bilodeau R, Raynauld P, et al. Etude prospective de Ia tension arterielle au cours de Ia grossesse. 1 Gynecol Obstet Bioi Reprod 1982;11:833. Swinscow TDV. Statistics at square one. 6th ed. London: British Medical Association, 1980. Vecchio T1. Predictive value of a single diagnostic test in unselected populations. N Eng! 1 Med 1966;274: 1171. Thurnau GR, Dyer A, Depp OR, Martin AO. The development of a profile scoring system for early identification and severity assessment of pregnancy-induced hypertension. AM 1 0BSTET GYNECOL 1983; 146:406.
Minibolus diazoxide in the management of severe hypertension in pregnancy Denis K. L. Dudley, M.B., B.S.
Ottawa, Ontario, Canada Severe hypertension poses immediate and potentially lethal maternal risks. Prompt reduction in blood pressure should be an urgent priority, with concomitant anticonvulsant therapy. Previous reports confirm intravenous diazoxide to be an effective agent. Concern exists over adverse reactions associated with bolus injection of the standard 300 mg dose. Many physicians are unaware that diazoxide may be given in increments. This study describes the use of diazoxide in 34 patients with severe preeclampsia. They received minibolus injections of 30 mg every 1 to 2 minutes. The maximum dose required was 150 mg and no maternal hypotension occurred. Maternal side effects were minimal and no cases of fetal distress occurred; a perinatal mortality was zero. It appears that minibolus diazoxide rapidly lowers maternal blood pressure without the adverse effects described with a standard bolus dose. Its advantages are contrasted with the limitations of conventional hydralazine therapy. (AM J OssTET GvNECOL 1985; 151 :196-200.)
Key words: Minibolus diazoxide, severe hypertension From the Department of Obstetrics and Gynecology, Ottawa Civic Hospital, and the University of Ottawa. Presented at the Fortieth Annual Meeting of The Society of Obstetricians and Gynaecologists ofCa'T!a(]n, Montreal, Quebec, Canada, june 19-23, 1984. Reprint requests: Denis K. L. Dudley, Department of Obstetrics and Gynecology, Ottawa Civic Hospital, Ottawa, Ontario, Ca'T!a(]n K1 Y 4£9.
196
Severe preeclampsia or pregnancy-induced hypertension should be considered as a form of malignant (accelerated) hypertension. In previously normotensive patients rapidly progressive hypertension can cause vascular damage and hypertensive encephalopathy at lower blood pressures than in chronic hypertensive pa-
Key words: Pregnancy-induced hypertension, chronic hypertension, prediction of preeclampsia Several blood pressure studies during the antenatal period have showed a decrease in blood pressure between 16 and 20 weeks. 1 Gallery et aP demonstrated that pregnancy-induced hypertension could be predicted by a significant increase in blood pressure as early as 17 to 20 weeks. This significant difference was observed with a standardized random-zero sphygmomanometer but was not apparent with the usual sphygmomanometer. 2 However, the sample size of that study was small and there was no discrimination between primigravid women, who are more prone to preeclampsia, and multiparous women, in which chronic hypertension is more prevalent. Recently, another group3 showed that a mean arterial blood pressure of 90 mm Hg or more between 18 and 26 weeks was predictive of preeclampsia. In a preliminary study• on 366 pregnant women, we observed a similar decrease of blood pressure from 13 to 21 weeks, in systolic and diastolic blood pressure in women who remained normotensive throughout pregnancy, together with a significant rise in systolic and diastolic pressure as early as 13 to 16 weeks in 14 primigravid women who later developed preeclampsia. More observations were needed to confirm these find-
From the Department of Obstetrics and Gynecology, Notre Dame Hospital, University of Montreal. Supported by the Medical Research Council of Canada. Presented at the Fortieth Annual Meeting of The Society of Obstetricians.and Gynaecologists ofCanada, Montreal, Quebec, Canada, june 19-23, 1984. Reprint requests:]. M. Moutquin, M.D., Department of Obstetrics and Gynecology, Notre Dame Hospital, 1560 East Sherbrooke St., Montreal, Quebec, Canada H2L 4K8.
ings, not only to show the constancy of the differences but also to establish whether preeclampsia could be predicted accurately with a marker, such as one blood pressure recording, early in pregnancy. Similar studies were also carried out with the mean blood pressure. Material and methods Population. Blood pressure was recorded in 1000 consecutive pregnancies, at each antenatal visit until delivery and at the postpartum visit, between November, 1979, and April, 1984. These women were cared for by a group of three private obstetricians and no intervention was suggested for any blood pressure reading. Seventeen patients were excluded (nine twin pregnancies and eight women who were not delivered at the institution). Thus the results included 983 women (808 primigravid, 175 multiparous) with a singleton pregnancy who were delivered at Notre-Dame Hospital. More than 7800 blood pressure recordings (in duplicate) were analyzed. Mean maternal age was 25.5 ± 4.4 years ( ± SD). Mean gestational age at delivery was 39.4 ± I. 7 weeks ( ± SD). The cesarean section rate was 14.8%. Methods. Arterial blood pressure was taken from the right arm with the patient in a sitting position, after a minimum of 5 minutes of rest. This measure was taken twice at an interval of 2 minutes. If a discrepancy occurred between the two readings, then blood pressure was taken again until similar readings were observed (difference of <5 mm Hg). Then the last two recordings were kept for analysis. A nurse was responsible for the procedure and for daily calibration of the sphygmomanometer. 191
192
Moutquin et al.
January 15, 1985 Am J Obstet Gynecol
160
'0 ., +I
Cll 140
J:
E
.s w
120
"' .,.,w :;:)
...."'
100
.....
<
....w"'
"'<
Fig. 1. Averaged systolic and diastolic arterial pressures ( ± I SD) in primigravid women who remained normotensive throughout pregnancy (Q----{)) and in those who developed preeclampsia (e-----e) by periods of 4 weeks.
Table I. Incidence of hypertensive complications at delivery in primigravid and multiparous women Primigravid women No.
No. of patients Gestational hypertension Preeclampsia-ecla mpsia Mild, moderate Severe Eclampsia Hepatic* Total Chronic hypertension Without preeclampsia With preeclampsia Total
808 IS
I
l
Total
I
%
No.
1.9
I75 5
983 20
2.0
I 2 0 0 3
43 3 I 2 49
5.0
42 I I 2 46
5.2
3I
3.8
6
37
Multiparous women
2I
%
I2.0
2
4.5
23
No.
52 8 60
%
5.3 6.I
*Hepatocellular damage evidenced by altered biochemical tests, thrombocytopenia , and jaundice.
The apparatus was an automatic random-zero sphygmomanometer with digital display (Dinamap 845*) which is based on the oscillometric principle (same principle as the usual sphygmomanom eter). One advantage of this apparatus is 1ts proficiency to decrease acoustic and physical artifacts. The diastolic reading corresponded to phase IV of Korotkoff sounds. The accuracy of blood pressure measurement was validated by the obtaining of similar values (in the range of 160 mm Hg systolic to 50 mm Hg diastolic) of intra-arterial radial pressure during surgical procedures. Mean arterial pressure was calculated according to the formula: Sys*Bionetics, California.
tolic blood pressure plus twice diastolic blood pressure/3.3 Data were stored at the calculus center of the University of Montreal and analyzed with the Statistical Package for the Social Sciences. Comparisons between differences were analyzed with the unpaired t test, 5 while sensitivity, specificity, and predictive values were assessed by the method of Vecchio. 6
Results The perinatal mortality rate was 4.111000 births. There were two stillbirths and two neonatal deaths (one due to congenital cardiopathy) thus providing a corrected perinatal mortality rate of 3.1/l 000 births.
Blood pressure in pregnancy
Volume 151 Number 2
193
160
14
•
c.,
*
*
*
*
•
*
~120
J:
E
E ......
Lr_
w100 Ill:
.,.,w ::I
Ill: A. 80
_.._I----tr._____r___._I-----I•
*
'(
*
*
• Chronic hypertension o Normotension
1
* P< ·001
Y~--~~--~~~--~~-9-12
13-16
17-20
21-24
25-28
29-32
33-36
37-40
WEEKS
Fig. 2. Averaged systolic and diastolic arterial pressures ( ± 1 SD) in multiparous women who remained normotensive throughout pregnancy (o---o) and in those who developed chronic hypertension (---) by periods of 4 weeks.
120
.,w +I
"'E
J:
HO
!, w
Ill:
.,.,::I w
100
.....
90
Ill: A. <(
;;;:
w ....
Ill: <(
80
z <(
w ~
70
chronic hypertension • preeclampsia * P< ·001 • normotension •• P< ·05 L~~~~--~~--~~~0
9-12
13-16
17-20
21-24
25-28
2!'-32
33-36
37-40
WEEKS
Fig. 3. Averaged mean arterial blood pressures ( ± 1 SE) in women who remained normotensive throughout pregnancy (o---o), in women who developed preeclampsia (.---....),and in women who developed hypertension (---) by periods of 4 weeks.
Hypertensive complications are illustrated in Table I. Gestational hypertension was diagnosed when hy-
pertension alone (:;;;.140/90 mm Hg) occurred after 20 weeks and disappeared at the postpartum visit. The diagnosis of preeclampsia was ascertained by the presence of hypertension :;;;.}40/90 mm Hg on two occasions 6 hours apart, edema (> 1 kg/wk) or proteinuria (AI-
bustix 1 +) on two occasions 6 hours apart after 20 weeks, elevated serum urate levels (4.6 mg/dl or more), and a blood pressure of ,..120/80 mm Hg at 6 weeks post partum. Chronic hypertension was defined by a prepregnancy personal history of hypertension or persistence of hypertension 6 weeks post partum. Averaged systolic and diastolic blood pressures
194 Moutquin et al.
January 15, 1985 Am J Obstet Gynecol
Table II. Sensitivity, specificity, and positive and negative predictive values (percentages) of recordings of 130 and 135 mm Hg systolic blood pressure as predictors of preeclampsia Weeks' gestation 130 mmHg 9-12 13-16 17-20 21-24 25-28 135mmHg 9-12 13-16 17-20 21-24 25-28
Sensitivity
Specificity
Positive predictive value
Negative predictive value
46.2 57.6 43.3 41.2 42.1
75.4 81.2 85.3 86.2 82.7
12.5 16.0 15.1 14.9 13.1
94.9 96.9 96.1 96.2 95.8
30.8 42.4 18.9 29.4 28.9
87.1 90.2 93.5 92.8 91.4
15.4 21.2 14.9 19.2 17.2
94.3 96.2 95.0 95.7 95.4
Table III. Sensitivity, specificity, and positive and negative predictive values (percentages) of recordings of 80 and 85 mm Hg diastolic blood pressure as predictors of preeclampsia Weeks' gestation 80mmHg 9-12 13-16 17-20 21-24 25-28 85mmHg 9-12 13-16 17-20 21-24 25-28
Sensitivity
Specificity
Positive predictive value
46.2 45.5 32.4 38.2 44.7
84.2 89.8 91.0 93.4 91.7
18.2 21.7 17.9 25.5 25.0
95.4 96.4 95.7 96.3 96.4
30.8 30.3 16.2 20.2 18.4
93.0 96.6 97.4 97.9 98.0
25.0 35.7 27.3 36.3 36.8
94.6 95.7 95.1 95.5 95.1
( ± SD) by 4-week periods in the group of normotensive
nulliparous women were compared to the blood pressures in the group of nulliparous women who later developed preeclampsia (Fig. 1). From 13 weeks onward, a significant difference in systolic pressure was observed between the two groups; the increase ranged from 7 to 22 mm Hg in the group destined to develop preeclampsia (p < 0.001, Fig. 1). From the first antenatal visit (9 to 12 weeks), the diastolic pressure was consistently more elevated in patients eventually developing preeclampsia; elevations ranged from 7 mni Hg (p < 0.01, at 9 to 12 weeks) to 10 to 15 mm Hg until delivery (p < 0.001, Fig. 1). In the multiparous women, averaged systolic and diastolic blood pressures ( ± SD) showed, as expected, a significant increase of about 20 mm Hg systolic and 15 to 17 mm Hg diastolic throughout pregnancy in women with chronic hypertension compared to normotensive multiparous women (p < 0.001, Fig. 2). Comparisons of mean arterial pressures ( ± SE) between patients eventually developing preeclampsia, patients with chronic hypertension, and normotensive pa-
Negative predictive value
tients showed significant increases in both women with eventual preeclampsia and those with chronic hypertension compared to the normotensive group (p < 0.05, 9 to 12 weeks, preeclampsia; p < 0.001, all other periods of pregnancy; Fig. 3). The difference in excess of mean blood pressure was 10 mm Hg in patients with eventual preeclampsia and 16 to 20 mm Hg in patients with chronic hypertension as compared to the normotensive patients at any period of pregnancy (Fig. 3). Although significant differences in blood pressure were dearly established in the group of women with eventual preeclampsia, such a prediction needed to be substantiated in individual patients. Clinical markers of easy access to identify early in pregnancy those patients who will develop preeclampsia are yet to be found. Single values of blood pressure (systolic, diastolic, mean) were tested as potential predictors of preeclampsia in the first five periods of observations (9 to 12, 13 to 16, 17 to 20, 21 to 24, and 25 to 28 weeks); these values were examined with the method of Vecchio.• With a systolic value of 130 mm Hg, the sensitivity
Blood pressure in pregnancy 195
Volume 151 Number 2
Table IV. Sensitivity, specificity, and positive and negative predictive values (percentages) of mean blood pressures of 85 and 90 mm Hg as predictors of preeclampsia Weeks' gestation
85mmHg 9-12 13-16 17-20 21-24 25-28 90mmHg 9-12 13-16 17-20 21-24 25-28
Sensitivity
Specificity
Positive predictive value
84.6 81.8 86.5 85.3 86.8
38.6 43.2 50.4 49.7 48.9
9.5 8.2 9.5 9.0 9.5
97.1 97.5 98.4 98.3 98.4
61.5 69.7 67.6 55.9 71.1
62.0 66.4 75.4 78.1 73.9
11.0 11.4 14.2 13.0 14.4
95.5 97.3 97.5 96.8 97.6
was around 50% and specificity was about 80%. At a value of 135 mm Hg, the number of eventually preeclamptic patients who exhibited such systolic pressure was small (18% to 42%) while as few as 6.5% to 12.9% of women who remained normotensive had such systolic pressure (Table II). The positive predictive value for 130 and 135 mm Hg was rather low, as expected," because only 5% of the population studied had the disease; conversely, the negative predictive value was 94% to 96% (Table II) because the fraction of the population not having preeclampsia was large.• When diastolic values were examined as predictors of preeclampsia, sensitivity was between 38% and 46% at 80 mm Hg and specificity ranged from 84% to 98% at 80 and 85 mm Hg (Table III). Here again positive and negative predictive values yielded percentages similar to those with systolic pressure (Table III). The mean arterial pressure showed an improved sensitivity (82% to 87%) with a value of only 85 mm Hg but specificity was <50% in most instances (Table IV). Specificity increased somewhat at a value of 90 mm Hg but the sensitivity was then decreased because several eventually preeclamptic women did not have a mean blood pressure of ;;;.go mm Hg before 29 weeks (Table IV). Comment A cohort of private patients with non-high-risk pregnancies yielded 5% preeclampsia and 6% chronic hypertension. This incidence of chronic hypertension is a reflection of a rather older population (up to 40 years of age). However, the frequency of superimposed preeclampsia was only 16% in primigravid women and 13% in the to~al sample, which is less than the reported frequency of 20% to 40% for this complication. The fall in systolic and diastolic blood pressures between 13 and 25 weeks in both primigravid and multiparous women confirmed results in previous reports, 1• 2 but a larger sample size permitted us to extend this fall during the entire second trimester rather than
Negative predictive value
between 16 and 20 weeks. 2 Such a fall, coincident with increasing cardiac output and blood volume, illustrates a lowering of systemic vascular resistance in normotensive women. This study also confirms the increase of blood pressure in those women destined to develop preeclampsia. However, recording from the first antenatal visit has demonstrated for the first time a significant increase in diastolic pressure as early as 9 to 12 weeks rather than at 17 to 20 weeks. 2 This latter observation suggests a potential to identify from the first antenatal visit, in the first trimester, those women destined to develop preeclampsia with a mean diastolic blood pressure value of as low as 77 mm Hg or with a mean arterial blood pressure of93 mm Hg. In addition, this finding may be of some help for the clinician so that a particular pregnancy may be observed more closely and other tests may be carried out to confirm future preeclampsia.' More important, the increase in diastolic blood pressure suggests, for the first time, an early vasospasm at 9 to 12 weeks. Such patients did not have chronic hypertension because the blood pressure was < 140/90 mm Hg before 20 weeks, the well-defined criteria of preeclampsia were present, the blood pressure returned to normal at 6 weeks postpartum, and also because the reported incidence of preeclampsia of 5% in this study was not different from the incidence of preeclampsia in our institution. Such an early vasospasm favors early vascular pathophysiologic changes because of the presence of the conceptus in the first trimester. Another objective was to identify predictors of preeclampsia from a cohort of pregnant women. Single values of systolic, diastolic, and mean arterial pressures were tested to ascertain their potential as markers for future preeclampsia. The sensitivity and specificity for diastolic values of 80 and 85 mm Hg were not different from those of previous preliminary reports. 4 The rather poor sensitivity suggests that the systemic manifestation of vaso-
Moutquin et al.
January 15, 1985
Am J Obstet Gynecol
spasm does not appear at the same period in every woman eventually developing preeclampsia and that its expression is not necessarily evident at rest, in a sitting position. The specificity, however, increases with increasing values of both systolic and diastolic blood pressures. Percentages of specificity are similar to those previously reported! Predictive positive values were higher in a preliminary report• than in the present study; this can be attributed to the larger sample size.6 With the mean arterial pressure, the sensitivity and specificity approached those reported by others. 3 However, the predictive positive value was less in the present study because the sample size was larger than the 85 patients reported by Oney and Kaulhausen. 3 From this information, clinical evaluation of blood pressure with an automatic random-zero sphygmomanometer will identify a subgroup of patients at potentially high risk of developing preeclampsia. Concurrently, a diastolic blood pressure <85 mm Hg as early as 9 to 12 weeks has a 95% chance to predict normotension throughout· pregnancy. Only 2% of women with that diastolic blood pressure in the second trimester will develop preeclampsia; this enables the
identification of the low-risk group in a cohort of pregnant women. REFERENCES I. Wilson M, Morgani AA, Zervoudakis I, et al. Blood
2.
3.
4. 5. 6. 7.
pressure, the renin-aldosterone system and sex steroids throughout normal pregnancy. Am1 Med 1980;68:97. Gallery EDM, Hunyor SN, Ross M, Gyory AA. Predicting the development of pregnancy-associated hypertension. The place of standardised blood-pressure measurement. Lancet 1977;1:1273. Oney T, Kaulhausen H. The value of the mean arterial blood pressure in the second trimester (MAP-2 value) as a predictor of pregnancy-induced hypertension and preeclampsia. A preliminary report. Clin Exp Hypertens [B) 1983;2:211. Moutquin 1M, Bilodeau R, Raynauld P, et al. Etude prospective de Ia tension arterielle au cours de Ia grossesse. 1 Gynecol Obstet Bioi Reprod 1982;11:833. Swinscow TDV. Statistics at square one. 6th ed. London: British Medical Association, 1980. Vecchio T1. Predictive value of a single diagnostic test in unselected populations. N Eng! 1 Med 1966;274: 1171. Thurnau GR, Dyer A, Depp OR, Martin AO. The development of a profile scoring system for early identification and severity assessment of pregnancy-induced hypertension. AM 1 0BSTET GYNECOL 1983; 146:406.
Minibolus diazoxide in the management of severe hypertension in pregnancy Denis K. L. Dudley, M.B., B.S.
Ottawa, Ontario, Canada Severe hypertension poses immediate and potentially lethal maternal risks. Prompt reduction in blood pressure should be an urgent priority, with concomitant anticonvulsant therapy. Previous reports confirm intravenous diazoxide to be an effective agent. Concern exists over adverse reactions associated with bolus injection of the standard 300 mg dose. Many physicians are unaware that diazoxide may be given in increments. This study describes the use of diazoxide in 34 patients with severe preeclampsia. They received minibolus injections of 30 mg every 1 to 2 minutes. The maximum dose required was 150 mg and no maternal hypotension occurred. Maternal side effects were minimal and no cases of fetal distress occurred; a perinatal mortality was zero. It appears that minibolus diazoxide rapidly lowers maternal blood pressure without the adverse effects described with a standard bolus dose. Its advantages are contrasted with the limitations of conventional hydralazine therapy. (AM J OssTET GvNECOL 1985; 151 :196-200.)
Key words: Minibolus diazoxide, severe hypertension From the Department of Obstetrics and Gynecology, Ottawa Civic Hospital, and the University of Ottawa. Presented at the Fortieth Annual Meeting of The Society of Obstetricians and Gynaecologists ofCa'T!a(]n, Montreal, Quebec, Canada, june 19-23, 1984. Reprint requests: Denis K. L. Dudley, Department of Obstetrics and Gynecology, Ottawa Civic Hospital, Ottawa, Ontario, Ca'T!a(]n K1 Y 4£9.
196
Severe preeclampsia or pregnancy-induced hypertension should be considered as a form of malignant (accelerated) hypertension. In previously normotensive patients rapidly progressive hypertension can cause vascular damage and hypertensive encephalopathy at lower blood pressures than in chronic hypertensive pa-