Nausea and vomiting are the most common symptoms experienced in early pregnancy and affect 50% to 80% of women.1 One to 3% of these women experience hyperemesis gravidarum, a severe form of vomiting due to pregnancy characterized by weight loss, electrolyte abnormalities, dehydration, and ketonuria.1 Severe nausea and vomiting is the third leading cause for hospitalization during pregnancy,2 with a financial burden on the American health system estimated at $130 million per year.1 It is estimated that 206 hours are lost from paid work for each woman who has nausea and vomiting during early pregnancy.3 Complications, albeit uncommon, associated with hyperemesis gravidarum include Wernicke,5 encephalopathy,4 rhabdomyolysis,5 coagulopathy,6 and low birth weight infants.7
In 1959, Geiger and et al,8 in a double-blind, placebo-controlled trial, showed that Bendectin aided in the treatment of nausea and vomiting of pregnancy. Bendectin became the first pharmaceutical option designated specifically for controlling symptoms of nausea and vomiting due to pregnancy, but it was removed from the market in 1983 because of claims of teratogenicity, which were subsequently proven to be unsubstantiated.9 Since then, there have been few randomized trials to assess different treatment modalities. Corticosteroids have been proposed to modify the chemoreceptor trigger zone in the brain10 that is responsible for nausea and vomiting, and on this basis, corticosteroids have been used successfully for many years to treat cancer chemotherapy–induced emesis.11,12 Our goal was to estimate the effect of corticosteroids in reducing the number of women requiring rehospitalization for hyperemesis gravidarum as has been described in one randomized trial.13
MATERIALS AND METHODS
From July 6, 1998, to August 22, 2001, all women presenting to Parkland Memorial Hospital, Dallas, complaining of nausea and vomiting during the first half of pregnancy (less than 20 weeks' gestation) were asked to participate in a study of the efficacy of intravenous methylprednisolone and oral prednisone. Women considered eligible for this study were those who previously had not responded to outpatient therapy and who demonstrated 3+ or 4+ dipstick urinary ketones as evidence of severe dehydration. Outpatient therapy for hyperemesis is standardized at our hospital and consists of promethazine 25 mg every 6 hours as needed. Before enrollment into the study, an ultrasound was performed to exclude molar pregnancy, to confirm a live fetus, and to establish gestational age. The study protocol was approved by the Institutional Review Board of the University of Texas Southwestern Medical Center, and informed written consent was obtained from all the women. Staffing of the obstetric service is provided by faculty from the Department of Obstetrics and Gynecology at the University of Texas Southwestern Medical School.
The women enrolled in the study were all provided the prevailing treatment for persistent hyperemesis gravidarum at our hospital. This treatment included admission and intravenous hydration with crystalloid until ketonuria cleared. The first liter of crystalloid included thiamine 100 mg. Conventional treatment also included promethazine 25 mg and metoclopramide 10 mg intravenously every 6 hours for 24 hours, followed by the same regimen administered as needed orally until discharge from the hospital. The women were also randomly administered, in a double-blind fashion, methyl-prednisolone 125 mg intravenously or placebo. This was followed by a tapering regimen of oral prednisone or an identical-appearing placebo (40 mg for 1 day, 20 mg for 3 days, 10 mg for 3 days, and 5 mg for 7 days). These study drugs were dispensed by the Investigational Drug Service of Parkland Hospital. Randomization was performed by computer-generated blocks of 20.
Women with persistent vomiting on day 2 of hospitalization and randomized to methylprednisolone received an additional 80-mg dose, and similar women in the placebo arm received an identical-appearing placebo drug. The decision to give an additional dose of study drug was made exclusively by the principal investigator (NY). Both groups of women were provided crackers and juice upon request and were advanced to a regular diet as tolerated, at which time discharge was permitted. Each woman was counseled by a nutritionist before discharge. At discharge, all women also received promethazine 25 mg every 6 hours as needed and metoclopramide 10 mg every 6 hours as needed in addition to their oral study drug taper. This treatment approach was also used in women requiring readmission for hyperemesis gravidarum. Study drug assignment during subsequent admissions was identical to that used for the initial admission; there were no crossovers.
All women enrolled into this trial routinely underwent a battery of laboratory tests to include assessments of thyroid function, liver function, pancreas (amylase and lipase), electrolytes, and human chorionic gonadotropin.
A minimum sample size of 70 women was estimated to be necessary, assuming that corticosteroid therapy would reduce the number of women requiring rehospitalization from 30% to 5%. This assumption was based on a published report that suggested corticosteroids virtually eliminated the need for rehospitalization.13 This sample size provided a power of 80% for a two-sided test with a significance of error rate .05.
Statistical analysis included χ2, Student t test, and Wilcoxon signed-rank test. P < .05 was considered statistically significant. Analysis was performed by SAS 8.2 (SAS Institute, Cary, NC).
A total of 126 women were enrolled into this trial, and 110 women (87%) were subsequently delivered at Parkland Hospital (Figure 1). The remainder were lost to follow-up. The 16 excluded women were not significantly different from the remainder of the cohort with respect to maternal characteristics, laboratory tests, and course before randomization. The results were analyzed on an intent-to-treat basis using the 56 women with follow-up and randomized to corticosteroid therapy compared with the 54 women administered placebo. The characteristics of the women in these two study groups were similar (Table 1). The mean gestational age at randomization was 11.0 ± 2.7 weeks in women administered corticosteroids compared with 10.8 ± 2.7 (P = .69) in women who received placebo. Shown in Table 2 are laboratory results at randomization for women according to study drug assignment. There were no significant differences. The disease course before and after randomization was not different between the two study groups (Table 3). Of the 19 women in each group who required readmission, 11 in the placebo group and 8 in the corticosteroid group were rehospitalized within 2 weeks of their first admission (P = .33).
Shown in Table 4 are pregnancy complications in women randomized to corticosteroids compared with placebo. There were no significant differences between the study groups. Similarly, neonatal outcomes were not different (Table 5). There was one stillborn fetus during the trial; this occurred in a woman who received corticosteroids. This patient had 6 admissions for hyperemesis gravidarum, the last at 27 weeks' gestation, and the fetal death occurred at 29 weeks. The infant weighed 1090 g and was delivered with a nuchal umbilical cord and no other findings. One woman in the placebo group delivered an infant with microcephaly. This infant weighed 1144 g and was delivered at 27 weeks due to spontaneous labor.
The addition of parenteral and oral corticosteroids to the treatment of pregnant women with hyperemesis gravidarum had no appreciable effect—neither beneficial nor deleterious—on the course of hyperemesis or pregnancy outcome. Specifically, the primary outcome in our study, rehospitalization for hyperemesis, occurred in 34% of women who received corticosteroids compared with 35% in women who received placebo. We are disappointed with this result because we had hoped that corticosteroids would prove useful in the management of a common and very difficult to cure complication of pregnancy. It has been argued, however, that no single therapy can be shown to be beneficial because hyperemesis gravidarum is an obstetrical syndrome with several etiologic factors.14 This may account for much of the frustration in the treatment of hyperemesis gravidarum. It might also be argued that our sample size was too small because it was based on an overly optimistic improvement in the primary outcome from 30% to 5%. We cannot disagree with this argument; however, we point out that the lack of a trend in the proportion of women rehospitalized in this study makes it unlikely that a larger sample size would have revealed a significant benefit for corticosteroid therapy.
Corticosteroids are a logical choice for treatment of nausea and vomiting due to pregnancy because they have been used effectively as an antiemetic in oncology patients.11,12 Indeed, this use prompted Nelson-Piercy et al10 to use corticosteroids in four women with refractory hyperemesis gravidarum. These investigators followed this experience with a randomized, placebo-controlled trial of prednisolone in 24 women and found no significant improvement in the frequency of vomiting.15
Safari et al,13 in the only other randomized trial of corticosteroids for hyperemesis gravidarum, assessed the use of oral methylprednisolone versus promethazine in a total of 40 women and concluded that methylprednisolone was superior. However, there was no significant difference in the number of women in each group who stopped vomiting within 2 days of initiating therapy. Fewer women who received methylprednisolone, however, required readmission within 2 weeks (five women who received promethazine versus none who received methylprednisolone). This is contrary to our findings. The study by Safari et al differs from ours in several ways: their study groups each received one drug oral therapy (methylprednisolone or promethazine), whereas our patients received two medications (metoclopramide and promethazine) in addition to the corticosteroid or placebo regimen. Also, our regimen was administered intravenously until an oral regimen was tolerated. Study drug assignment during subsequent admissions was identical to that used for the initial admission, whereas Safari et al allowed crossover. The two studies were similar in that they were conducted in a randomized, double-blind fashion with a 2-week steroid taper.
In a recent evidence-based review of pharmacologic therapy for hyperemesis gravidarum, Magee et al16 concluded that the pooled results from the literature failed to show that corticosteroids reduced the number of readmissions for hyperemesis gravidarum. Our results support this conclusion.
1. Miller F. Nausea and vomiting in pregnancy: The problem of perception—Is it really a disease? Am J Obstet Gynecol 2002;186:S182–3.
2. Bennett TA, Kotelchuck M, Cox CE, Tucker MJ, Nadeau DA. Pregnancy-associated hospitalizations in the United States in 1991 and 1992: A comprehensive view of maternal morbidity. Am J Obstet Gynecol 1998;178:346–54.
3. Mazzotta P, Maltepe C, Navioz Y, Magee LA, Koren G. Attitudes, management and consequences of nausea and vomiting of pregnancy in the United States and Canada. Int J Gynecol Obstet 2000;70:359–65.
4. Lavin PJM, Smith D, Kori SH, Ellenberger C. Wernicke's encephalopathy: A predictable complication of hyperemesis gravidarum. Obstet Gynecol 1983;62:13S–15.
5. Fukada Y, Ohta S, Mizuno K, Hoshi K. Rhabdomyolysis secondary to hyperemesis gravidarum. Acta Obstet Gynecol Scand 1999;78:71–3.
6. Robinson JN, Banerjee R, Thiet MP. Coagulopathy secondary to vitamin K deficiency in hyperemesis gravidarum. Obstet Gynecol 1998;92:673–5.
7. Chin RKH, Lao TT. Low birth weight and hyperemesis gravidarum. Eur J Obstet Gynecol Reprod Biol 1988;28:179–83.
8. Geiger CJ, Fahrenbuch DM, Healy FJ. Bendectin in the treatment of nausea and vomiting of pregnancy. Obstet Gynecol 1959;14:688–90.
9. Koren G, Levichek Z. The teratogenicity of drugs for nausea and vomiting of pregnancy: Perceived versus true risk. Am J Obstet Gynecol 2002;186:S248–52.
10. Nelson-Piercy C, de Swiet M. Corticosteroids for the treatment of hyperemesis gravidarum. Br J Obstet Gynaecol 1994;101:1013–5.
11. Italian Group for Antiemetic Research. Dexamethasone, granisetron, or both for the prevention of nausea and vomiting during chemotherapy for cancer. N Engl J Med 1995;332:1–5.
12. Italian Group for Antiemetic Research. Dexamethasone alone or in combination with ondansetron for the prevention of delayed nausea and vomiting induced by chemotherapy. N Engl J Med 2000;342:1554–9.
13. Safari HR, Fassett MJ, Souter IC, Alsulyman OM, Goodwin TM. The efficacy of methylprednisolone in the treatment of hyperemesis gravidarum: A randomized, double-blind, controlled study. Am J Obstet Gynecol 1998;179:921–4.
14. Goodwin TM. Nausea and vomiting of pregnancy: An obstetric syndrome. Am J Obstet Gynecol 2002;186:S184–9.
15. Nelson-Piercy C, Fayers P, de Swiet M. Randomised, double-blind, placebo-controlled trial of corticosteroids for the treatment of hyperemesis gravidarum. Br J Obstet Gynaecol 2001;108:9–15.
© 2003 by The American College of Obstetricians and Gynecologists.
16. Magee LA, Mazzotta P, Koren G. Evidence-based view of safety and effectiveness of pharmacologic therapy for nausea and vomiting of pregnancy (NVP). Am J Obstet Gynecol 2002;186:S256–61.