Effect of MgSO4 Prophylaxis and Preeclampsia

8/11/2019 Effect of MgSO4 Prophylaxis and Preeclampsia

1/5

1475

Effect of magnesium prophylaxis and preeclampsia on the

duration of labor

Sara E. Szal, MD,aMary S. Croughan-Minihane, PhD,a, b and Sarah J. Kilpatrick, MD, PhDa

San Francisco, California

OBJECTIVE: Our goals were to compare duration of labor at term for (1) women with preeclampsia versus

normotensive nulliparous women and (2) nulliparous women with preeclampsia who received magnesium for

seizure prophylaxis versus those who did not.

STUDY DESIGN: We performed a retrospective cohort study of all nulliparous, term vaginal deliveries from

1989 through 1995 at University of California, San Francisco.The perinatal database and medical records

were reviewed for information on duration of labor, maternal and labor characteristics, and neonatal out-

comes. The 2 odds ratio, and Student t tests were used to compare categoric and continuous variables be-

tween women with preeclampsia and control women and between women with preeclampsia who did and

those who did not receive magnesium. Logistic regression was used to evaluate variables predictive of labor

duration.

RESULTS: Our study subjects were 4083 normotensive nulliparous women and 154 women with preeclamp-

sia. A sample size calculation revealed that 1764 normotensive control subjects were needed to show a 10%difference in labor duration with 80% power and alpha of 0.05. Among women with preeclampsia, 93 (60%)

were treated with magnesium and 61 (40%) were not. More women with preeclampsia than normotensive

women had induction of labor and received epidural anesthesia, prostaglandin gel, and oxytocin (P< .003).

Total labor duration did not differ between women with preeclampsia and normotensive women (P= .15) or

between women with preeclampsia who received magnesium and those who did not (P= .09). In compari-

son with normotensive women, those with preeclampsia had a higher rate of postpartum hemorrhage (31%

vs 22%, P= .005), and the rate was even higher among preeclamptic women treated with magnesium ver-

sus those who received no magnesium (34% vs 26%, P= .002). Logistic regression, with prolonged first

stage of labor (>12 hours) used as the outcome variable, indicated that epidural anesthesia (odds ratio 2.3,

95% confidence interval 1.9-2.6), oxytocin (odds ratio 1.8, 95% confidence interval 1.6-2.2), and persistent

occipitoposterior presentation (odds ratio 1.6, 95% confidence interval 1.1-2.4) were associated with pro-

longed labor, whereas preeclampsia (odds ratio 0.9, 95% confidence interval 0.7-1.1) and treatment with

magnesium were not (odds ratio 1.1, 95% confidence interval 0.9-1.4). Induction (odds ratio 0.5, 95% confi-

dence interval 0.4-0.6) and birth weight


2/5

vere preeclampsia and found the relative risk of eclamp-

sia to be 0.09 (95% confidence interval 0.1-0.69) among

those treated with magnesium. The rate of cesarean de-

livery was high in both groups (50%), but labor duration

was not compared. Another study evaluated women with

mild preeclampsia.8 They found no difference in labor

duration or rate of cesarean delivery.

A clinical observation that women with preeclampsia

have rapid labors has not been supported in the litera-

ture. We sought to test the hypothesis that women with

preeclampsia have faster labors and that magnesium

treatment increases labor duration. Our goals were (1) to

compare duration of labor for women with preeclampsia

at term versus normotensive women and (2) to compare

duration of labor for women with preeclampsia at termwho were treated with magnesium versus those who were

not.

Methods

We performed a retrospective cohort study of nulli-

parous, term vaginal deliveries from 1989 through 1995,

using the University of California, San Francisco,

Perinatal Database. The database contains >200 items of

obstetric data on every delivery at Moffitt Hospital since

1975. Physicians and midwives record intrapartum data

after each delivery. Antenatal, postnatal, and neonatal

data are abstracted from maternal and neonatal charts by

professional chart abstraction. Internal validity and logicchecks ensure optimal accuracy.

We identified all women with a diagnosis of hyperten-

sion at 36 weeks from 1989 through 1995. Hypertension

was defined as blood pressure 140/90 mm Hg during

labor. Magnesium data were first recorded in 1989. One

of us (S.E.S.) performed a chart review of all patients

with hypertension to identify women meeting the criteria

for preeclampsia, which included 2 blood pressure read-

ings at least 6 hours apart of 140/90 mm Hg and pro-

teinuria (2+ on clean-catch specimen or 1+ on

catheterized specimen). Criteria for severe disease in-

cluded one of the following: blood pressure 160 mm Hg

systolic or 110 mm Hg diastolic; proteinuria (>5 g/24 h

or persistent 3+-4+ on dipstick); elevated serum creati-

nine concentration (>1.0 mg/dL); eclampsia; pulmonary

edema; oliguria (2.5-fold higher than upper limit of

normal); thrombocytopenia (6 hours apart but no proteinuria.

Predictor variables included preeclampsia and magne-

sium treatment, and outcome variables included dura-

tion of labor (stages 1, 2, and 3 and total); rate of pro-

longed first stage of labor (>12 hours, chosen for clinical

relevance; and >16.6 hours, which is at the 95th per-

centile in our patients10); demographic data (maternal

age and race); obstetric outcomes (gestational age, in-

duction, use of prostaglandin, oxytocin, epidural anes-

thesia, persistent occipitoposterior position, and esti-

mated blood loss); and neonatal outcomes (birth weight


3/5

delivery was 115. A chart review of hypertensive women

delivered vaginally revealed that 154 met the criteria for

preeclampsia. Among the women with preeclampsia, 93

(60%) were treated with magnesium for seizure prophy-

laxis, whereas the remaining 61 (40%) were not.

Preeclampsia was diagnosed more frequently in

women 16.6 hours10). The women with preeclampsia who were

treated with magnesium did have a higher rate of post-

partum hemorrhage (Table IV).

Logistic regression, with >12 hours for the first stage of

labor used as the outcome variable, demonstrated that

preeclampsia and use of magnesium were not associated

with prolonged labor (Table V). Univariate and multi-

variate results are shown for all significant obstetric pre-dictors. Epidural medication, oxytocin, and persistent oc-

cipitoposterior position were associated with stage 1

labor of >12 hours. Induction and birth weight 16.6 hours)10 showed

that the only significant predictor was epidural anesthe-

Volume 180, Number 6, Part 1 Szal, Croughan-Minihane, and Kilpatrick 1477Am J Obstet Gynecol

Table III. Demographic data for women with

preeclampsia treated and not treated with magnesium

Treated with Not treated withmagnesium magnesium Statistical

Category (n = 93) (n = 61) significance*

Maternal age 25 7.1 26 6.4 P= .8020 y 25 (27%) 13 (21%) P= .43>35 y 12 (13%) 6 (10%) P= .56

EthnicityWhite 31 (36%) 24 (42%) P= .42Black 21 (24%) 16 (28%) P= .008

Hispanic 16 (18%) 7 (12%)P

= .047Asian 7 (8%) 4 (7%) P= .77Filipina 8 (9%) 3 (5%) P= .39Other 4 (5%) 3 (5%) P= .86

Data expressed as mean SD or frequency and percentage.*Significance determined by 2 or ttest.

Table II. Obstetric outcomes for all women with preeclampsia versus normotensive control subjects

Women with Control subjects Statistical Category preeclampsia (n = 154) (n = 4083) significance*

Gestational age (wk) 39.0 1.4 39.5 1.4 P= .000Induction 78 (50%) 571 (14%) P= .000Use of prostaglandin 57 (37%) 309 (8%) P= .000

Use of epidural anesthesia 113 (73%) 2488 (61%) P= .003Use of oxytocin 126 (81%) 1733 (42%) P= .000Duration of labor (h)

Stage 1 9.1 5.9 10.0 6.0 P= .07Stage 2 1.6 1.1 1.7 1.4 P= .31Stage 3 0.2 0.2 0.2 0.2 P= .25TOTAL 10.9 6.2 11.9 6.4 P= .06

Stage l labor >12 h 39 (25%) 1252 (31%) P= .15Stage 1 labor >16.6 h 16 (10%) 508 (13%) P= .44Persistent occipitoposterior position 6 (4%) 118 (3%) P= .47Estimated blood loss >500 mL 48 (31%) 881 (22%) P= .005Birth weight


4/5

sia, with an odds ratio of 3.12 (95% confidence interval

1.02-9.56, data not shown).

Preeclampsia became eclampsia in 3 women, 1 before,

1 during, and 1 after birth. None were receiving magne-

sium before seizure. More women who received magne-

sium met criteria for severe preeclampsia (39% vs 21% in

the nonmagnesium group, P= .02). No statistically signif-

icant difference in labor duration was found betweenthose with severe disease and those with mild disease or

between those who received magnesium and those who

did not (data not shown). However, the data were sug-

gestive of a trend toward faster labor among women with

severe preeclampsia not treated with magnesium (total

labor duration 9.6 hours, n = 13) than among women

with severe preeclampsia treated with magnesium (12.3

hours, n = 34), women with mild preeclampsia not

treated with magnesium (10.0 hours, n = 48), and women

with mild preeclampsia treated with magnesium (10.9

hours, n = 53).

Comment

Our data demonstrated that labor duration in nulli-

parous women with preeclampsia at term did not differ

significantly from labor duration in normotensive

women. In addition, use of magnesium prophylaxis wasnot associated with a change in labor duration. Multiple

predictors of prolonged labor were identified, including

epidural use, oxytocin use, and persistent occipitoposte-

rior position. Induction and birth weight 12 hours

Unadjusted 95% Confidence Adjusted 95% ConfidencePredictor odds ratio interval odds ratio interval

Maternal age 1.03 0.92-1.16 0.97 0.85-1.11Induction 0.74* 0.61-0.89 0.49* 0.39-0.63Prostaglandin gel 0.76* 0.59-0.97 0.82 0.61-1.10Oxytocin 1.86* 1.62-2.13 1.83* 1.57-2.15Epidural anesthesia 2.61* 2.25-3.03 2.25* 1.92-2.63Persistent occipitoposterior position 2.26* 1.60-3.23 1.63* 1.13-2.36Preeclampsia 0.87 0.73-1.04 0.87 0.71-1.07

Severe disease 0.88 0.59-1.30 0.94 0.61-1.45Magnesium 1.12 0.86-1.38 1.13 0.89-1.43Birth weight 12 h 27 (29%) 12 (20%) P= .32Stage 1 labor >16.6 h 12 (13%) 4 (7%) P= .70Persistent occipitoposterior position 5 (5%) 1 (2%) P= .27Estimated blood loss >500 mL 32 (34%) 16 (26%) P= .002Birth weight


5/5

controlled for parity by evaluating only nulliparous

women, whereas other investigations of labor duration in

preeclampsia have included both nulliparous and multi-

parous women.8, 11 Initial cervical dilatation is more diffi-

cult to assess, although we uniformly recorded in the

database the first stage of labor as the time from onset of

painful, regular contractions until complete dilation.However, advanced cervical dilation at admission may

have been associated with a decision against use of mag-

nesium; this potential confounder may be evaluated only

by a prospective study. Finally, severity of disease may have

affected our results because significantly more of the

women with preeclampsia who were given magnesium had

severe preeclampsia, in comparison with the patients not

given magnesium. However, severe preeclampsia was not

associated with faster labor in the multivariate analysis.

One recent study,12which was a secondary analysis of a

randomized study of women with pregnancy-induced hy-

pertension who were treated with phenytoin versus mag-

nesium, showed no difference in labor outcomes, such as

time from admission to delivery, rate of prolonged sec-

ond stage, or rate of cesarean delivery. Our study differs

from the previous one in that we had different treatment

groups, different outcomes (we looked at duration of

labor, whereas the authors of the previous study looked

at the outcomes listed here), and, finally, different ran-

domization methods (ie, it is not clear whether the au-

thors initial randomized sample was, in fact, a random

sample, because they excluded, post hoc, 58% of the pa-

tients for epidural use, multiparity, severe preeclampsia,

and birth weight 2500 g, women with severe

preeclampsia, and those receiving epidural anesthesia,

the authors excluded 58% (1233/2138) of their initial

randomized sample.

Our logistic regression identified several predictors

that were independently associated with prolonged

labor: oxytocin, epidural anesthesia, and persistent oc-

cipitoposterior position. As predicted, smaller babies

were associated with faster labor.

Our logistic regression showed an interesting associa-tion between induction and decreased labor duration.

There are several possible explanations. First, more of

the patients with induced labor may have undergone ce-

sarean delivery, thereby enriching the vaginally delivered

group with those with fast labors. Second, more women

undergoing induction may have had more aggressive

management of labor.

The only difference found between women with

preeclampsia and control subjects that appeared to be as-

sociated with magnesium treatment was an increased risk

of postpartum hemorrhage. We did not evaluate rate of

transfusion, which may be a more clinically significantoutcome. However, these findings are in agreement with

those of other investigators.8, 11

Our study appears to be the first to address the dura-

tion of labor between women with preeclampsia treated

and those not treated with magnesium at term, with the

elimination of multiple confounders, including parity

and severity of disease. Our study had sufficient power

(80%) that we can state that the difference in duration of

labor was not >10%. These data suggest that the clinical

concern that magnesium slows labor is unfounded and

should not influence the obstetricians decision for mag-

nesium prophylaxis.

REFERENCES

1. Lazard EM. A preliminary report on the intravenous use of mag-nesium sulphate in puerperal eclampsia. Am J Obstet Gynecol1925;9:178-88.

2. Sibai BM. Eclampsia. VI. Maternal-perinatal outcome in 254consecutive cases. Am J Obstet Gynecol 1990;163:1049-55.

3. Burrows RF, Burrows EA. The feasibility of a control populationfor a randomized control trial of seizure prophylaxis in the hy-pertensive disorders of pregnancy. Am J Obstet Gynecol1995;173:929-35.

4. Coetzee EJ, Dommisse J, Anthony J. A randomised, controlledtrial of intravenous magnesium sulphate versus placebo in themanagement of women with severe pre-eclampsia. Br J ObstetGynaecol 1998;105:300-3.

5. Sibai BM, Ramadan MK, Usta I, Salama M, Mercer BM, FriemanSA. Maternal morbidity and mortality in 442 pregnancies withhemolysis, elevated liver enzymes, and low platelets (HELLPsyndrome). Am J Obstet Gynecol 1993;169:1000-6.

6. Sibai BM, Abdella TN, Spinnato JA, Anderson GD. Eclampsia. V.The incidence of nonpreventable eclampsia. Am J ObstetGynecol 1986;154:581-6.

7. Fuentes A, Rojas A, Porter KB, Saviello G, OBrien WF. The ef-fect of magnesium sulfate on bleeding time in pregnancy. Am JObstet Gynecol 1995;173:1246-9.

8. Witland AG, Friedman SA, Sibai BM. The effect of magnesiumsulfate therapy on the duration of labor in women with mildpre-eclampsia at term: a randomized, double-blind, placebo-controlled trial. Am J Obstet Gynecol 1997;176:623-7.

9. American College of Obstetricians and Gynecologists.Hypertension in pregnancy. Washington: The College; 1996.p. 1-8. Technical Bulletin No.: 219.

10. Kilpatrick SJ, Laros RK. Characteristics of normal labor. ObstetGynecol 1989;74:85-7.11. Friedman SA, Lim K-H, Baker CA, Repke JT. Phenytoin versus

magnesium sulfate in preeclampsia: a pilot study. Am J Perinatol1993;10:233-8.

12. Leveno KJ, Alexander JM, McIntire DD, Lucas MJ. Does magne-sium sulfate given for the prevention of eclampsia affect the out-come of labor? Am J Obstet Gynecol 1998;178:707-12.

Volume 180, Number 6, Part 1 Szal, Croughan-Minihane, and Kilpatrick 1479Am J Obstet Gynecol

Documents

Effect of MgSO4 Prophylaxis and Preeclampsia