Original ...... Article Comparison of Three Outpatient Regimens in the Management of Nausea and Vomiting in Pregnancy

Fadi A. Bsat, MD of pregnancies.3 As nausea and vomiting during pregnancy are Despina E. Hoffman, BA difficult to treat, many therapies have been utilized.4 David E. Seubert, MD Prochlorperazine, promethazine, metoclopramide, and pyridoxine 5 (vitamin B6 or B6) are often used as first-line therapy. Newer regimens using ondansetron or steroids are also utilized, either as first-line therapy or more likely after initial therapy failure. The Objective: majority of studies tend to focus on the management of hyperemesis gravidarum with little investigation of nausea and This study compares pyridoxine–metoclopramide combination therapy to vomiting in pregnancy.6 Since nausea and vomiting can adversely prochlorperazine and promethazine monotherapies in the outpatient affect the quality of the pregnant woman’s life,7 we investigated treatment of nausea and vomiting in pregnancy. therapies that may abate these symptoms. Study Design: The purpose of this study was to evaluate the effectiveness of In total, 174 first trimester, singleton pregnancies were evaluated for three commonly prescribed pharmaceutical regimens in the nausea and vomiting. Patients were prospectively randomized into three outpatient management of nausea and vomiting in pregnancy: a treatment groups: pyridoxine–metoclopramide, prochlorperazine, or B6–metoclopramide combination, prochlorperazine, and promethazine. Prior to, and on the third day, patients recorded their promethazine. We assessed therapeutic successes subjectively and subjective responses to the given treatment and their number of emesis objectively by the degree of symptom resolution, the occurrence of episodes. The three treatment groups were compared for therapy response. side effects, and the rate of hospitalization.

Results: There were no differences in the number of emesis episodes prior to treatment. Both subjective and objective responses to treatment differed MATERIALS AND METHODS among the three groups when comparing the combination therapy to the We prospectively evaluated 174 patients with singleton gestations monotherapies (p<0.05). presenting in the first trimester to their obstetrical provider with Conclusion: nausea and/or vomiting between January 1994 and December Combination therapy with pyridoxine and metoclopramide appears to be 1996. With ethical board approval, informed consent was obtained superior to either monotherapy in the treatment of nausea and vomiting from the patients enrolled in the study. The patients were recruited in pregnancy. from both private and clinic general obstetric population. All Journal of Perinatology (2003) 23, 531–535. doi:10.1038/sj.jp.7210986 participants underwent general physical and obstetrical evaluations. An ultrasound examination confirmed gestational age, viability, and plurality of the pregnancy. Laboratory studies obtained included complete blood count, urinalysis with culture, free T4 and thyroid-stimulating hormone, and hepatitis B surface INTRODUCTION antigen testing. All patients were advised to discontinue their iron Nausea and vomiting occur in about 80% of pregnancies.1,2 or vitamin supplementation until their symptoms improved. Hyperemesis gravidarum is a more severe form affecting up to 1% Patients excluded from the study prior to randomization included women with a medical condition manifesting as nausea or vomiting, women necessitating hospital admission upon initial assessment, women hospitalized or treated for hyperemesis Department of Obstetrics & Gynecology (D.E.H.), Division of Maternal-Fetal Medicine Baystate gravidarum during their current pregnancy, and patients lost to Medical Center, Springfield, MA, USA; and Tufts University School of Medicine (F.A.B., D.E.S.), Boston, MA, USA. follow-up. Clinical thyroid disease was an exclusion factor.

Address correspondence and reprint requests to Fadi B., MD, Baystate Medical Center, However, subclinical patients with only mild dysfunction (low TSH, Department Obstetrics and Gynecology, 759 Chestnut Street, Springfield, MA 01199-0001, USA. normal free T4) and no prior thyroid disease were included.

Journal of Perinatology 2003; 23:531–535 r 2003 Nature Publishing Group All rights reserved. 0743-8346/03 $25 www.nature.com/jp 531 Bsat et al. Outpatient treatments for Nausea and Vomiting in Pregnancy

Women with both abnormal TSH and abnormal free T4 were respectively). One patient withdrew from the study due to acute excluded. dystonic reactions, thought to be secondary to metoclopramide on Patients were divided into three groups based on a computer- day 3 of treatment. The symptoms resolved spontaneously 6 to 8 generated randomization. The patients in Group A received one hours after the last dose. This patient declined further antiemetic 50 mg intramuscular injection of pyridoxine, with metoclopramide treatment and proceeded to have an uneventful pregnancy. We 10 mg orally every 6 hours as needed. The patients in Group B completed data analysis for 54 patients in group A, 50 patients in received prochlorperazine, as needed, 25 mg rectal suppositories group B, and 52 patients in group C. There were no statistically every 12 hours, or 10 mg tablets orally every 6 hours as needed. significant differences among the groups when compared for The patients in Group C received promethazine 25 mg orally every maternal age, gestational age at enrollment, parity, and 6 hours as needed. subsequent hospitalization for nausea or vomiting (Table 1). Drug Symptoms were subjectively assessed and recorded by the usage and compliance was comparable between all the three patients on the third day of treatment. Subjective scores of 1–5 groups. Patients in group A did not express difficulty or concern were noted by the patient as follows: much worse (1), worse (2), regarding the use of pyridoxine as an intramuscular injection same (3), better (4), and much better (5). The responses were then rather than as an oral form. divided into two subgroups: those that answered 1–3 (same-worse) While there were no statistically significant differences between and those that answered 4–5 (better). In addition, patients also the groups in the number of emesis episodes before treatment, reported their medication compliance. group A had significantly lower number of emesis after treatment To obtain quantitative data on treatment effectiveness, all than either group B or C (Table 2). B6–metoclopramide patients recorded the number of emesis episodes the day before and combination was better than prochlorperazine (relative risk on the third day of treatment. Dry heaves were counted as nausea, (RR) ¼ 0.59; 95% confidence interval (CI) ¼ 0.39 to 0.88) or but not vomiting episodes. Worsening of symptoms was evaluated promethazine (RR ¼ 0.62; 95% CI ¼ 0.42 to 0.91) in improving and patient admission for hydration or inpatient management was the subjective symptoms of these patients. Figure 1 depicts the considered on an individual basis. Hospitalization for the specific subjective responses of the patients following the different management of nausea or vomiting was noted. regimens. Statistical analysis employed w2, analysis of variance, and the The Kruskal–Wallis test demonstrated that the mean number of Kruskal–Wallis tests. Statistical significance was defined as emesis after treatment for the patients in the pyridoxine– p<0.05. Based on our preliminary data, a power analysis metoclopramide group was significantly lower, that is, better than calculation indicated that we needed at least 46 patients in each the other two groups (Table 2). The Kruskal–Wallis test was group to reach statistical significance when comparing groups A also used to evaluate the three groups with respect to their and B, or A and C (a ¼ 0.05; b ¼ 0.20). subjective responses to the different antiemetic regimens used. The groups did differ (p ¼ 0.012), with an apparent advantage to the pyridoxine–metoclopramide regimen compared to the other two agents. RESULTS At enrollment, 14 patients had borderline elevation in free T4, Of 174 pregnant women screened for eligibility, 169 were enrolled. but normal thyroid-stimulating hormone levels. Thyroid function Five were excluded before randomization for medical reasons. In testing was repeated in the second trimester and was normal in all. all, 12 patients were lost to follow-up, similarly distributed among Subsequent pregnancy courses were similar among the three the groups (three, five, and four from groups A, B, and C, groups. No neonatal anomalies were seen, except for one

Table 1 Group Characteristics

B6–metoclopramide Prochlorperazine Promethazine p n 54 50 52 Age (years)* 25.1±6.8 25.9±5.6 27.5±6.4 NSw Gestational age (weeks)* 8.5±2.0 7.9±1.8 8.6±2.0 NSw Nulliparous (%) 37 (69%) 36 (72%) 35 (67%) NSz Hospitalization (%) 3 (5.6%) 3 (6.0%) 6 (11.5%) NSz

NS=not significant. *Mean±standard deviation. wAnalysis of variance. zw2 analysis.

532 Journal of Perinatology 2003; 23:531–535 Outpatient treatments for Nausea and Vomiting in Pregnancy Bsat et al.

Table 2 Patient Response to the Different Antiemetic Regimens

B6–metoclopramide Prochlorperazine Promethazine p n 54 50 52 Emesis before* 2.3±1.1 2.3±1.1 2.4±1.2 NSw Emesis after* 0.6±0.8 1.1±0.8 0.8±0.8 0.014w

NS=not significant *Mean±standard deviation. wKruskal–Wallis test.

Having small meals at more frequent intervals may help as well. Estrogens in the combined birth control pill have been shown to induce nausea or vomiting in a dose-related fashion.11 Nausea or vomiting of pregnancy is also more common in the morning hours of the day,9 hence ‘‘morning sickness’’. Social and psychological factors can also contribute to the severity of the nausea.12 Some have even described an evolutionary advantage to nausea and vomiting, protecting the fetus from certain harmful substances in food.13 Pyridoxine was first used to manage nausea and vomiting in pregnancy by Willis et al.14 in 1942. Pyridoxine requirements are increased in pregnancy, but low serum concentrations are not usually seen in normal pregnancies before the second and third trimesters of pregnancy. Pyridoxine deficiency without clinical symptoms is common in pregnancy.15 Pyridoxine and doxylamine Figure 1. Subjective patient responses to the three antiemetic were included in the formulation of Bendectins, which was regimens. *p<0.05 when group A compared to either group B or approved by the Food and Drug Administration for the use of group C. nausea in pregnancy, before subsequently being pulled from the market. When pyridoxine alone is not successful, a combination regimen have been suggested.16 ventricular septal defect in the prochlorperazine group. None of the We have been using 50 mg single pyridoxine injection with patients had any relevant allergic reaction history. apparent success in the treatment of nausea of pregnancy. The effect noted was often an all-or-none effect. Although we could not duplicate their results, others have described oral pyridoxine as effective therapy when given orally.5 Our past experience as well as SIGNIFICANCE others did not mirror that result.17 Therefore, oral pyridoxine was Nausea and vomiting are common complaints during the first not used. Patient satisfaction may be increased when a trimester of pregnancy. Symptoms are usually worse in the combination of parenteral and oral medications is given. This has morning, but may continue throughout the day. The genesis of prompted us in the past to combine intramuscular pyridoxine pregnancy-induced nausea and pregnancy is not clear. Chorionic injection and oral metoclopramide. The reason for combining two gonadotropin has been implicated on the basis that levels are high agents and comparing that treatment with two other single agents at the same time when nausea and vomiting are more common.8,9 was to specifically mirror local common practices. We acknowledge Moreover, nausea and vomiting are prominent features in that these regimens may not necessarily be the first choice conditions where chorionic gonadotropin is significantly elevated, regimens in other communities. We also understand that such such as in women with hydatidiform mole. Others found no grouping may raise questions of fairness in combining two agents relationship between chorionic gonadotropin and nausea of that may also work independently, and compare that with pregnancy.3,10 A progesterone effect has been implicated by some, monotherapy. A placebo effect in the route of administration may especially in relation to relaxation of the esophageal-gastric also play a role in the patient’s response, which is not shared by sphincter. Agents that enhance sphincter tone and promote the groups taking prochlorperazine or promethazine. We did not gastrointestinal motility, such as metoclopramide, may help intend to compare route of administration (intramuscular vs oral prevent reflux, which may be manifested as nausea or vomiting. vs rectal) or placebo effect, which may have led to a more attractive

Journal of Perinatology 2003; 23:531–535 533 Bsat et al. Outpatient treatments for Nausea and Vomiting in Pregnancy

but different study. We intended to mirror established local double-blind fashion, and possibly to establish a placebo group. practices, hence the corresponding grouping of the patients in this The intramuscular injection of pyridoxine may have been study. evaluated on its own, or may have been given to all the three Metoclopramide counteracts some of the physiologic changes of groups. However, the groups were established as such to mirror the gastrointestinal tract in pregnancy that may lead to nausea or outpatient treatment regimens already in practice in our local vomiting. It helps strengthen the basal sphincter tone of the lower community. This study shows favorable results in the use of a esophageal sphincter, which is usually decreased in pregnancy.18 pyridoxine–metoclopramide regimen in the treatment of nausea In the small intestines, propulsive motility is decreased, and transit and vomiting of pregnancy, possibly because it addresses the time is increased. This may lead to bloating and bowel distention, physiologic changes of pregnancy that may have led to these which may be manifested as nausea or malaise. Some have symptoms. attributed these changes to effects of progesterone or chorionic gonadotropin.19 From a pharmacological basis, metoclopramide appears to address these specific issues by promoting upper gastrointestinal tract motility, relaxing the pyloric sphincter and References duodenal bulb, and increasing the lower esophageal sphincter tone. 1. Jewell D, Young G. Interventions for nausea and vomiting in early The subjective response of the patients to the different regimens pregnancy. Cochrane Database Syst Rev 2002;CD000145. is a very important measure of the treatment success, since a major 2. Klebanoff MA, Koslowe PA, Kaslow R, et al. Epidemiology of vomiting in indication of the treatment is to alleviate the subjective symptoms. early pregnancy. Obstet Gynecol 1985;66:612–6. The pyridoxine–metoclopramide group appeared to have an 3. Depue RH, Bernstein L, Ross RT, et al. Hyperemesis gravidarum in relation advantage when the subjective responses of the patients are to estradiol levels, pregnancy outcome, and other maternal factors: a evaluated. As shown on Table 2, the number of emesis after the seroepidemiologic study. Am J Obstet Gynecol 1987;156:1137–41. treatments appeared to differ as well. 4. Nageotte MP, Briggs GG, Towers CV, et al. Droperidol and diphenhydramine in the management of hyperemesis gravidarum. Am J Obstet Gynecol Prochlorperazine is a neuroleptic phenothiazine that is believed 1996;174:1801–5; discussion 1805–6. to work centrally in relieving nausea or pregnancy. Promethazine 5. Sahakian V, Rouse D, Sipes S, et al. Vitamin B6 is effective therapy for is a phenothiazine antihistamine derivative, with H1 receptor nausea and vomiting of pregnancy: a randomized, double-blind placebo- blocking properties. They both have sedative effects. controlled study. Obstet Gynecol 1991;78:33–6. Metoclopramide and prochlorperazine increase the risk of tardive 6. Briggs GG. A guideline for treating hyperemesis gravidarum. Contemporary dyskinesia, although this is more common in the elderly, and OB/GYN 1997;42:70–9. thought to be dose-dependent. 7. Attard CL, Kohli MA, Coleman S, et al. The burden of illness of severe Metoclopramide and promethazine have been evaluated for the nausea and vomiting of pregnancy in the United States. Am J Obstet treatment of emesis during labor. They were equally effective, and Gynecol 2002;185:S220–7. both better than placebo, in reducing the incidence of nausea and 8. Prenatal care. In: Cunningham FG, Gant NF, Leveno KJ, Gilstrap III LC, th vomiting after the administration of pethidine.20 However, Hankins GDV, Clark SL, editors. Williams Obstetrics, 20 edition. Stamford: promethazine was found to be associated with a more persistent Appleton & Lange; 1997. p. 227–50. 9. Gadsby R, Barnie-Adshead AM, Jagger C. A prospective study of nausea and sedative effect (66%), compared to metoclopramide (41%). vomiting during pregnancy. Br J Gen Pract 1993;43:245–8. Sedation may be considered a serious side effect, especially in active 10. Soules MR, Hughes Jr CL, Garcia JA, et al. Nausea and vomiting of women, and the availability of a less-sedating medication with pregnancy: role of human chorionic gonadotropin and 17-hydroxyproges- similar or better effectiveness and safety profile should be terone. Obstet Gynecol 1980;55:696–700. encouraged. 11. Goldzieher JW, Moses LE, Averkin E, et al. A placebo-controlled double-blind A thorough comparison of the serious side effects of the different crossover investigation of the side effects attributed to oral contraceptives. agents was not practically possible because of the small sample Fertil Steril 1971;22:609–23. population and the low risk of serious side effects. Among our 12. Deuchar N. Nausea and vomiting in pregnancy: a review of the problem recruited patients, only one who received metoclopramide had with particular regard to psychological and social aspects. Br J Obstet acute dystonic reactions that resolved spontaneously. In its product Gynaecol 1995;102:6–8. information (A.H. Robbins, July 2002), metoclopramide may be 13. Sherman PW, Flaxman SM. Nausea and vomiting of pregnancy in an associated with this complication in one in 500 patients treated. evolutionary perspective. Am J Obstet Gynecol 2002;185:S190–7. 14. Willis RS, Morris AT, Newsom AA, et al. Clinical observations in treatment of However, evidence-based view regarding the utilized medications in nausea and vomiting in pregnancy with vitamin B1 and B6. A preliminary the therapy of nausea and vomiting of pregnancy concluded that report. Am J Obstet Gynecol 1942;44:265–71. 21 they are safe and effective. 15. Heller S, Salkeld RM, Korner WF. Vitamin B6 status in pregnancy. Am J Clin Nausea remains a difficult symptom to quantify, and we admit Nutr 1973;26:1339–48. that assigning a score to the different responses was arbitrary but 16. Niebyl JR, Goodwin TM. Overview of nausea and vomiting of pregnancy with logical. It would have been ideal to randomize the patients in a an emphasis on vitamins and ginger. Am J Obstet Gynecol 2002;185:S253–5.

534 Journal of Perinatology 2003; 23:531–535 Outpatient treatments for Nausea and Vomiting in Pregnancy Bsat et al.

17. Vutyavanich T, Wongtra-ngan S, Ruangsri R. Pyridoxine for nausea and 20. Vella L, Francis D, Houlton P, et al. Comparison of the antiemetics vomiting of pregnancy: a randomized, double-blind, placebo-controlled metoclopramide and promethazine in labour. Br Med J (Clin Res Ed) trial. Am J Obstet Gynecol 1995;173:881–4. 1985;290:1173–5. 18. Van Thiel DH, Gavaler JS, Joshi SN, et al. Heartburn of pregnancy. 21. Magee LA, Mazzotta P, Koren G. Evidence-based view of safety and Gastroenterology 1977;72:666–8. effectiveness of pharmacologic therapy for nausea and vomiting of 19. Kumar D. In vitro inhibitory effect of progesterone on extrauterine human pregnancy (NVP). Am J Obstet Gynecol 2002;185:S256–61. smooth muscle. Am J Obstet Gynecol 1962;84:1300.

Journal of Perinatology 2003; 23:531–535 535