Nausea and vomiting have been associated with early pregnancy since the beginning of time. Nausea and vomiting in pregnancy were documented in a physicians papyrus dated 2000 B.C. and discussed in Soranus Gynaecology, a medical text from the second century A.D.1 Although nausea and vomiting are common complaints of pregnancy (50% to 90% of all pregnant women have "morning sickness"), much remains unknown about its pathophysiology, and there is no universally agreed upon therapy. Even today, women are told to grin and bear it until symptoms abate after the first trimester.
An extreme form of pregnancy-induced vomiting is hyperemesis gravidarum (HG). HG is defined as intractable vomiting in pregnancy that causes dehydration, electrolyte disturbances, nutritional deficiencies and weight loss. Before intravenous fluid therapy was available, hyperemesis was a major cause of maternal mortality. In fact, therapeutic abortion was introduced in 1813 as a treatment of hyperemesis gravidarum.1 Charlotte Bronte, the famous 19th century author of Jane Eyre, died of hyperemesis in 1855 in her fourth month of pregnancy.2
Vomiting has been reported in 56% of all pregnant women.2 Hyperemesis gravidarum, however, occurs in 3.5 per 1,000 pregnancies.1,3,4 It is more common in urban than in rural populations and it is very rare in native American, Eskimo, African and Asian communities.1 In affected populations, it is more common in women who weigh more than 170 pounds, are nonsmokers, have twin pregnancies, trophoblastic disease and are less than 20 years old.1 Hyperemesis is most common in first pregnancies and tends to recur in subsequent pregnancies. Epidemiological studies indicate that women with nausea and vomiting in pregnancy have a statistically significant decreased risk of miscarriage in the first 20 weeks.2
Fortunately, most studies do not indicate any deleterious effects of HG on the fetus unless weight gain continues to be poor during the second half of pregnancy.5 Poor maternal weight gain increases the risk of delivery of a low-birth-weight infant and associated neurodevelopmental sequelae.6
The cause of hyperemesis gravidarum is unknown, and there are many theories as to its etiology. HG has been described as the "disease of theories" and it is probably multifactorial in origin. Pathophysiologic theories include endocrine, psychological or autonomic nervous system dysfunction, gastric dysrhythmia and nutritional deficiencies.2
Throughout history, psychogenic reasons have been thought to contribute to the pathogenesis of HG. Hyperemesis was thought to represent a somatic expression of psychological conflict or psychiatric illness exacerbated by pregnancy.1 A number of studies support this theory, including one indicating that up to 70% of hyperemetic women respond to some degree to placebo.1 In addition, hyperemesis only occurs in human pregnancies, is treatable by hypnosis, and decreases in incidence during wartime and in times of famine.1 Contrary to this theory, many studies have found no difference in the incidence of psychological disorders in women with and without HG.
Gestational hormones have long been thought to contribute to hyperemesis. Some researchers have found serum hCG (human chorionic gonadotropin) levels to be higher in hyperemetic women than in normal pregnant controls.1 High levels of estrogen may play a role, since women who suffer nausea and vomiting while taking oral contraceptives are more likely to have hyperemesis when pregnant. Progesterone may contribute, since it prolongs gastric emptying time and decreases smooth muscle motility. However, no difference in estrogen or progesterone levels has been found in women with HG.
Transient elevations in serum thyroxine levels, especially the serum level of free thyroxine, have been documented in as many as 70% of pregnancies complicated by HG.7 This hyperthyroid state is usually without significant symptoms and resolves spontaneously without treatment at the same time that HG resolves. The pathogenesis of the hyperthyroid is not completely understood, although some authors believe that hCG or some variant of hCG is responsible. HG usually occurs in the first trimester when levels of hCG are high and peak levels of hCG have an inherent thyroid-stimulating effect equivalent to 310 µU/mL.7
A recent theory of the pathogenesis of HG cites a dysfunctional gastric pacesetter as the core of the problem.3,8-11 This dysfunction is characterized by reversed gastroduodenal peristaltic waves, resulting in regurgitation of duodenal content into the stomach and subsequent nausea and vomiting. The retrograde gastric contractions result in reflux of gastric contents into the esophagus even in the absence of food. This is more pronounced during the liquid phase than the solid phase of gastric emptying, and is thought to be due to maladaptation of the GI tract to gestational hormones. When adaptation occurs by the end of the first trimester, gastric motility resumes and hyperemesis ceases.
Another theory attributes HG to deficiencies of nutrients such as pyridoxine and zinc deficiencies. However, controlled studies have failed to show a difference in the levels of these nutrients in women with hyperemesis.2 Sahakian et al. demonstrated in a double-blind, placebo-controlled study of 59 hyperemetic pregnant women that pyridoxine 25 mg every 8 hours significantly reduced vomiting.12
Hyperemesis gravidarum begins between the fourth and sixth week of pregnancy, often before the woman realizes she is pregnant. Symptoms usually improve by the 15th to 20th week of gestation, although some women continue to have frequent relapses throughout pregnancy. Most affected women have numerous episodes of vomiting throughout the day without symptom-free periods. This leads to weight loss, dehydration, electrolyte disturbances, ketosis and acetonuria requiring hospitalization. If these derangements are not treated promptly, they can lead to renal and hepatic damage. Other complications include Mallory-Weiss tears (linear mucosal tears at the cardio-esophageal junction with hematemesis), acid aspiration and metabolic alkalosis. These women present to their physicians with weight loss of five to 20 pounds; however, since many are overweight to begin with, they may not appear malnourished on visual inspection. Ptyalism, or excessive salivation, often accompanies HG, and some women require an emesis basin to expectorate into.
Laboratory findings at the time of presentation include increased ketones and increased urine specific gravity associated with an increased blood urea nitrogen. There is often an increased hematocrit, indicating a contracted blood volume. Electrolyte derangements include decreased sodium, potassium, chloride and magnesium levels. In some patients tests reveal an increase in liver function, such as in aspartate aminotransferase, alanine aminotransferase or bilirubin activity.2 If the woman has been unable to eat for a prolonged period of time, vitamin deficiencies can occur and Wernickes encephalopathy (an inflammatory, hemorrhagic form of encephalopathy caused by thiamine deficiency and usually associated with chronic alcoholism) has been reported.13
Other causes of nausea and vomiting unrelated to pregnancy, such as gastroenteritis, must be ruled out before the diagnosis of hyperemesis can be made. Conditions such as diabetic gastroparesis, peptic ulcer disease, gall bladder disease, hepatitis, renal dysfunction, hypercalcemia and hyperparathyroidism can all be associated with nausea and vomiting.
Reassurance and counseling are important components of therapy. Because patients are often frustrated and depressed, it is helpful for them to gain a better understanding of their disease and realize that they are not the only the only ones to suffer from HG. A patient education pamphlet designed to answer patients questions is recommended (FIGURE 1). A diary of times of the day when vomiting occurs and common triggers, such as foods, odors and activities or events, is beneficial. Patients who are particularly withdrawn or difficult to communicate with may require a psychologic consultation. Some authors have advocated treatment of hyperemesis with behavioral therapy, using stimulus control and imaging procedures.14
For the woman who is dehydrated, correction of fluid and electrolyte disturbances is imperative. Intravenous fluids such as 5% dextrose in lactated Ringers solution will provide electrolytes. Multivitamins should be added once a day if available. If IV multivitamins are not available, thiamine and pyridoxine should be added. Many patients stop vomiting once they are rehydrated.
When vomiting has ceased for 24 hours, a diet of dry foods should be introduced. Small, frequent meals to avoid being either too hungry or too full may be effective. High carbohydrate foods, such as baked potatoes, crackers, breads, rice and pasta are best tolerated.4 These foods also contain essential B vitamins. Fatty foods should be avoided, as they slow gastric emptying. The patient should wait about 30 minutes after eating before drinking fluids. Avoidance of offensive food odors helps some women. Consulting a dietitian is helpful in difficult cases to ensure proper nutritional intake.
It may be helpful to omit prenatal vitamins with iron because they can cause considerable nausea. Some women can tolerate childrens chewable vitamins better.
For patients who fail to respond to fluids alone, pharmacologic therapy may be necessary. Many physicians are reluctant to use drugs in the first trimester since organogenesis occurs during this period. However, several medications have been prescribed for HG for decades and are known to be safe in pregnancy (TABLE 1).
Certain antihistamines, such as meclizine (Antivert), dimenhydrinate (Dramamine) and its active moiety diphenhydramine (Benadryl) are very effective antiemetics. (Benadryl is our first choice because it can be administered by IV, PO and IM.) These agents antagonize the central action of acetylcholine and decrease stimuli going to the chemoreceptor trigger zone (CTZ) and vomiting center. Bendectin, the only drug ever approved by the U.S. Food and Drug Administration for morning sickness, originally contained the antihistamine doxylamine, the antispasmodic dicyclomine and vitamin B6.
Dicyclomine was removed from the formulation in 1976, when clinical trials failed to show evidence of its benefits as a synergistic antiemetic. Millions of women were prescribed this agent and found that it relieved nausea and vomiting. However, in 1979 The National Enquirer broke a scandalous and untrue front page news story about a drug company "cover-up" of several thousand tragically deformed infants in the U.S.15 As a result of this story, lawyers began advertising in womens magazines advising mothers of malformed children to sue Merrell Dow if they had taken Bendectin during their pregnancy. The March of Dimes received thousands of phone calls from women who had been advised to have abortions as a result of having taken Bendectin. In 1983 Merrell Dow removed Bendectin from the market when legal costs surpassed the income from the drug. Evidence is clearer today than in 1983 that Bendectin does not cause birth defects. In 1983, the editor of the medical journal Teratology called Bendectin the "most famous tortogen/litogen and best- studied human non-teratogen."16
A recently published meta-analysis of epidemiological studies found that Bendectin exposure does not increase the prevalence of congenital malformations.17 This tragedy demonstrates how false and distorted information published in tabloid journals can affect medical practice, and these events will probably prevent any drug from being approved by the FDA for nausea and vomiting in pregnancy.
Doxylamine, one of the components of Bendectin, is available today in the over-the-counter sleep aid Unisom tablets. It does carry a warning on the package against its use in pregnancy and this should be explained to the pregnant women if it is recommended as an antiemetic.
Antispasmodics such as scopolamine are often used to treat motion sickness. However, these agents slow gastric emptying and prolong GI transit time. Since slowed gastric emptying is part of the etiology of HG, these agents are inappropriate for its treatment.
The phenothiazines promethazine (Phenergan) and prochlorperazine (Compazine) are effective antiemetics that work by central antagonism of dopamine in the CTZ. There is no increased risk of congenital malformations from phenothi-azines.18 Young women are particularly susceptible to dystonic reactions and extrapyramidal side effects caused by these agents; however, antihistamines such as diphenhydramine prophylax against these side effects. Therefore, it is important to include an antihistamine with the phenothiazines not only for their antiemetic action, but to prevent adverse drug reactions.
Metoclopramide (Reglan) increases lower esophageal sphincter pressure, decreasing gastroesophageal reflux. It accelerates gastric emptying and acts directly on the CTZ. Like the phenothiazines, metoclopramide can cause dystonic reactions and extrapyramidal side effects; thus, it also requires concomitant administration of an antihistamine to counteract these effects. Cisapride (Propulsid) is another prokinetic agent which stimulates motor activity in the GI tract by enhancing the release of acetylcholine from the enteric nervous system. However, a lack of published studies on its use in pregnancy and the fact that it is assigned a pregnancy category C (see sidebar), make its use in HG controversial. Cisapride may have an advantage over metoclopramide in that it does not cause dystonic or extrapyramidal reactions.
Histamine-2 blockers such as ranitidine (Zantac) are also useful in gastroesophageal reflux that occurs in HG. They reduce gastric acid production by competitive blockade of H2 receptors in gastric parietal cells. These agents provide symptomatic relief of heartburn. Ranitidine, cimetidine (Tagamet), and famotidine (Pepcid) all are assigned pregnancy category B (see sidebar).
Ondansetron (Zofran) is a selective serotonin-3 receptor antagonist usually reserved for the most highly emetogenic cancer chemotherapy due to its extremely high cost. Published literature for its use in pregnancy is limited to a few case reports and one small study; therefore, it is considered a last-line agent in HG when all other pharmacologic therapy has failed.19-22 Sullivan et al. randomized 30 women with HG to receive either IV ondan-setron or promethazine in a double-blind design. They found no difference in relief of nausea, weight gain or length of hospitalization between the two drugs.22
For patients with severe malnutrition in whom pharmacologic therapy has failed, tube feedings or total parenteral nutrition (TPN) may be necessary.23,24 Many authors have advocated iso-osmolar tube feeding by a small-bore nasogastric feeding tube.25-27 Risks with this method include tube displacement and pulmonary aspiration and experience at our institution shows poor patient acceptance. Many physicians initiate home TPN without having first attempted an adequate trial of anti-emetic medications and fail to realize the serious risks involved. Life-threatening complications of catheter insertion such as pericardial tamponade, pneumothorax and subclavian artery laceration have been described in the nutrition and surgical literature.15 Metabolic and infectious complications of parenteral nutrition are not uncommon. Pregnant women receiving home TPN have been admitted to our hospital in Gram-negative septic shock, with severe cellulitis requiring surgery, and hyperglycemia. TPN should be reserved for hyperemetic women only when medical management has failed an adequate trial and the patient is severely malnourished.
Despite counseling and discussion of the risks and benefits of HG and pharmacologic therapy, some women refuse to take any drugs during pregnancy. For these women interventions may include dietary and lifestyle changes (TABLE 2). Alternative therapies include ginger and herbal teas. Fisher-Rasmussen et al. conducted a double-blind placebo controlled crossover study of powdered ginger root for HG.28 Capsules containing 250 mg ginger or placebo were given to 30 women four times a day for four days. After a two-day washout the groups switched for four more days. The powdered ginger root provided more significant relief of vomiting than placebo.28 Ginger ale may be a more acceptable and readily available form of ginger and the carbonation may be helpful. Herbal teas that have been promoted for nausea and vomiting include raspberry leaf tea and chamomile tea.4
Treatment of HG in Chinese literature focuses on acupuncture. Effective points of stimulation are the P6 or Neiguan point and the ears. The Neiguan point is 5 cm from the distal wrist crease, between the tendons of the flexi carpi radialis and palmaris longus.29,30 For women who refuse to take medications during pregnancy, acupressure bracelets may be an alternative.15 These are sold in boating stores and by auto clubs to treat motion sickness. Their mechanism of action is based on acupuncture principles. They consist of an elastic band with a large plastic button on one side that is placed over the above mentioned tendons.
Pharmacist-initiated HG Protocol
Since the Ohio State University Medical Center is a large teaching hospital and a regional prenatal referral center, only patients with refractory HG are admitted to the facility. Many of these patients have lost a significant amount of weight and have been on home IV or TPN therapy. The obstetric residents would consult the TPN pharmacist to initiate parenteral nutrition if the patient had been on home TPN or was severely malnourished. When the obstetricians encountered limited success with drug therapy and adverse drug reactions (ADRs) in many patients, they began to consult the specialty practice pharmacist for women and infants. The TPN pharmacist and the women and infants pharmacist worked on a plan together, allowing a 24-hour trial of IV fluids and medications before TPN was instituted. If the patient stopped vomiting, a diet of dry foods was started on the second day. The obstetricians, impressed with the success of the medical therapy prescribed by the pharmacist and the ability to avoid use of TPN, invited the specialty practice pharmacist to speak at OB Grand Rounds on the treatment of HG.
The pharmacist emphasized using drugs that worked by mechanisms that would relieve the pathology of HG and discouraged those that would aggravate the condition, such as scopolamine. Medications with established safety in pregnancy that work synergistically and avoided side effects were stressed. The risks of TPN therapy were reviewed, and TPN was advised to be used only as a last resort, when pharmacologic therapy fails. As a result of the Grand Rounds presentation, a cost-effective protocol was developed for treating patients with HG, which included counseling and written patient education (FIGURE 1). Drug therapy included diphenhydramine 25 mg, promethazine 12.5 mg, metoclopramide 10 mg, and ranitidine 50 mg, all administered IV push every 6 hours. Pyridoxine 50 mg was added to the maintenance IV fluid once daily. As soon as the patient stopped vomiting, all medications were switched to the oral route and a diet of dry foods was ordered, generally beginning with a baked potato. When caloric intake was sufficient, the patient was discharged on the same four-drug regimen.
The first 12 patients entered on the pharmacist-initiated HG protocol were compared to the previous 17 patients treated by the obstetric service. The average weight loss was 17.8 pounds in the protocol group and 15.8 pounds in the standard therapy group. Three patients in the protocol group and only one patient in the standard group were hyperthyroid. The protocol proved very effective, since eight of 12 patients stopped vomiting within 24 hours of entering the hospital. The remaining four patients stopped vomiting by 48 hours. Only four of 17 patients in the standard group stopped vomiting within 24 hours.
While the only complaint in the HG protocol group was drowsiness, there were three adverse drug reactions in the other group. Most adverse reactions were due to concomitant administration of more than one phenothiazine, producing dystonic reactions. Six patients in the protocol group were on home TPN at the time they entered the hospital. TPN was able to be discontinued in all of the protocol patients within one day of admission, whereas the five patients in the standard group were kept on prolonged TPN therapy.
The pharmacist-initiated HG protocol provided rational, effective therapy that generally relieved vomiting within 24 hours. This combination of medications prevented adverse drug reactions. The medications were administered by IV push, the least expensive method. Home medications were always given orally, or by rectal suppository when vomiting prevented use of the oral route, rather than by expensive and risky methods, such as the metoclopramide continuous IV pump.
Although nausea and vomiting are common complaints during pregnancy, hyperemesis gravidarum is more rare, occurring in 3.5 per 1,000 pregnancies. The pathophysiology of HG is not completely understood, but likely involves a maladaptation of the gastrointestinal tract to gestational hormones, causing gastric reflux and retrograde peristalsis. There may also be a psychogenic component.
Treatment should involve counseling, administration of intravenous fluids and electrolyte replacement. When patients fail to respond to this regimen, pharmacologic therapy should be tried before beginning TPN. A combination of an antihistamine, a phenothiazine, an H2-blocker and metoclopramide is effective and reduces the chance of dystonic reactions. A diet of small, frequent meals consisting of low-fat, high carbohydrate foods is best tolerated by these patients. Women who refuse medications may benefit from herbal teas, ginger or acupressure bracelets. HG generally resolves by 16 to 20 weeks gestation.
|FDA-Assigned Pregnancy Categories*|
* Excerpted from Drug Information for the Health Care Professional, USP-DI, Volume 1A, 11th ed., 1991.