There is very little evidence on which to base treatment decisions for nausea and vomiting of pregnancy (NVP). As a result, I believe that we practice "observation-based medicine," rather than evidence-based medicine. Additional treatment influences are that NVP is self-limited and there can be a placebo effect, as in the general population. Simply put, there are three therapeutic issues:
When to treat?
What to use?
How to measure success?
For our recent review on the treatment of NVP, we did an extensive literature search of MEDLINE, the Cochrane Database of Systematic Reviews, bibliographies and texts, as well as contacting drug manufacturers regarding their pregnancy outcome data.1,2 The key search words included nausea, vomiting, pregnancy, pregnancy complications, maternal and perinatal pregnancy outcome measures of importance, and the effectiveness of therapy for NVP. We focused on observational, controlled studies for adverse foetal effects, specifically, the incidence of major malformations, because treatment of NVP usually involves administration of medication during the first trimester. For the effectiveness of maternal therapy, we sought randomized controlled trials of drug treatment versus placebo or no therapy, or another drug therapy. We were limited to physical outcome measures as no trials evaluated the impact of treatment on other aspects of women's lives.
Some women prefer more "natural," non-pharmacological therapies for NVP, such as dietary and lifestyle changes, pyridoxine, ginger, and/or stimulation of the P6 Neiguan point (e.g., Seabands). These are thought not to be dangerous to the foetus on theoretical grounds. However, effectiveness must be ensured, given that ineffective treatment, when it results in a delay in administration of effective therapy, may be dangerous for the mother. That is, these therapies may be considered harmful if they are advocated as first line therapies but are not effective. There is no scientific evidence that small, frequent, bland meals taken separately from fluid are helpful to women with NVP. Most data on dietary management of NVP is anecdotal. Three randomized controlled trials of pyridoxine versus placebo have shown no decrease in treatment failure, defined as cessation of vomiting (RR = 0.97, 95% CI 0.78-1.20).2 However, nausea was decreased in severity. It is not known if adding additional pyridoxine to recommended doses of Diclectin (doxylamine 10 mg + pyridoxine 10 mg) will result in additional benefit. Stimulation of the P6 Neiguan point, three fingers above the wrist on the palmar aspect of the forearm, has been shown to alleviate NVP by at least 50%.3 (P6 stimulation for 5 minutes four times daily, or as continuously as possible, may be administered by acupuncture, acupressure, manual pressure, Seabands, or possibly by a small TENS unit, Relief Band.) However, the small sample sizes and the failure of most trials to blind outcome assessment complicate interpretation of results. One small randomized controlled trial has also shown that ginger (250 mg four times daily) may be effective for NVP.4 There is a trial ongoing in western Canada that is comparing Diclectin, ginger and placebo. There is no evidence that hypnosis or psychotherapy is useful as primary therapy for NVP. Interesting anecdotal data exist on the use of electrical stimulation of the vestibular system which is thought to decrease afferent impulses to the vomiting centre.
Pharmacologic approaches for the treatment of NVP have been based on the pathophysiology of nausea and vomiting, and on treatments found to be successful for non-pregnant subjects. Potential treatments include antihistamines, anticholinergics, dopamine antagonists, 5-HT3 antagonists, corticosteroids, cisapride, and cannabinoids.
In Canada, Diclectin, the only drug currently labelled safe and effective for use for NVP, is a delayed-release formulation composed of an antihistamine (doxylamine succinate 10 mg) and pyridoxine (10 mg). Meta-analysis of controlled studies, with more than 170,000 first trimester exposures to doxylamine/pyridoxine/dicyclomine, revealed no increased risk of any major or specific malformation (RR = 0.98, 95% CI 0.93-1.03).5 Bendectin (doxylamine/pyridoxine/dicyclomine) has been compared with placebo in four published randomized controlled trials. Overall, Bendectin was more effective than placebo (RR = 0.53, 95% CI 0.41-0.68), but trial results among the four studies were more different than could be expected by chance alone. The trial with the most favourable results used doses of up to four tablets per day; the others use a maximum of two to three tablets per day. The recommended dosage of Diclectin is two tablets at bedtime for early morning symptoms; one tablet (morning) and one tablet (mid-afternoon) may be added, as needed. In countries where Diclectin is not available, doxylamine and pyridoxine tablets may be taken together; the onset of action will not be delayed. H1 receptor antagonists other than doxylamine, such as dimenhydrinate (Gravol, Dramamine), are not associated with an increased teratogenic risk. A meta-analysis of over 200,000 first trimester exposures (including those to doxylamine/pyridoxine/dicyclomine) actually revealed a slightly lower risk of major/minor malformations (pooled RR = 0.76, 95% CI 0.60-0.94).6 Presumably, the decreased risk reflects a maternal factor, rather than the drug therapy. This is not an unreasonable hypothesis, given that we know that NVP is associated with a decreased risk of miscarriage.7 Randomized controlled trials have shown H1 receptor antagonists to be better than placebo in decreasing the treatment failure rate for NVP.
For rapid relief of nausea and vomiting, we at the NVP Clinic at the Hospital for Sick Children (HSC) in Toronto, recommend Diclectin be supplemented with dimenhydrinate (50-100 mg orally every four hours, or 50-100 mg rectally as a suppository every 6-8 hours, to a maximum of 200 mg/day). If taking Diclectin and dimenhydrinate does not relieve NVP, more Diclectin may be needed for effectiveness. Mass-based dosing of Diclectin is currently being investigated at the HSC.
Safety data on metoclopramide exposure during the first trimester are limited to reassuring animal studies and a few case reports. There are no data on effectiveness. However, the drug has been widely used for NVP, being the drug of choice in many European countries. It is probably best to avoid first trimester exposure if possible.
Case reports on the use of phenothiazines during pregnancy have implied teratogenicity. However, when pooled, eight prospective and retrospective controlled studies showed no increased relative risk of teratogenicity (RR = 1.00, 95% CI 0.86-1.17).2 All trials have shown them to be better than placebo for (usually severe) NVP (RR = 0.21, 95% CI 0.17-0.28); it is only the magnitude of effect that is in question.
A recent meta-analysis of six cohort and one case-control study found a trend towards an increased risk of major malformations following first trimester exposure to corticosteroids (RR = 1.23, 95% CI 0.96-1.58).8 The risk of oral clefting was significantly increased (RR = 1.59, 95% CI 1.16-2.18), although individual trials were not consistent. Anecdotal data and one small trial of prednisone versus promethazine have described steroids to be effective for severe NVP. Steroids are therefore not recommended before 10 weeks' gestation unless NVP is very severe.
No teratogenicity has been associated with 18 first trimester exposures to ondansetron. There are no data on its effectiveness for NVP. One small trial of 30 women with severe NVP found that ondansetron was no more effective than promethazine.9
As for other antiemetic agents, there is no human data on the use of cannabinoids in pregnancy, and cisapride is novel with limited safety data. No increase in malformations was seen in a multicentre prospective study of 27 first trimester exposures to cisapride, but there are no data on effectiveness for NVP.10
At the NVP Clinic, HSC, for outpatients who are still symptomatic on maximum doses of Diclectin ± dimenhydrinate therapy, we suggest: (i) P6 acupressure/acupuncture, (ii) ginger, (iii) phenothiazines, and/or (iv) after 10 weeks' gestation, metoclopramide (Maxeran, Reglan, 5-10 mg orally every six hours, as needed). For patients with unacceptable oral intake, we administer parenteral vitamins (including thiamine), Ringer's lactate, and prochlorperazine (Stemetil, 5-10 mg intravenously every 6-8 hours, as needed). As soon as oral intake is tolerated, we recommend starting Diclectin and oral vitamins; management would then be as outlined above for outpatients. However, if oral intake is not tolerated, then we recommend ADDING metoclopramide (5-10 mg intravenously every six hours, as needed). If this is ineffective, corticosteroids or ondansetron (Zofran) should be considered.
Ensuring adequate intake of B vitamins throughout the course of NVP is critical, as deficiencies of these vitamins can develop over weeks, and cases of Wernicke's encephalopathy are still reported in association with NVP. Whether to initiate enteral or parenteral feeding in the patient with poor intake is less clear, given potential side effects with both approaches. Factors to consider include duration of NVP, magnitude of weight loss, response of symptoms to aggressive management as listed above, and ability to prevent urinary ketosis with intravenous sources of carbohydrates.
In summary, there are a number of safe and effective therapeutic options for the treatment of NVP. Furthermore, when to treat a patient with NVP is probably as important an issue as what to use. Evidence suggests that quality of life may be impaired before severe physical symptoms appear, and that some women choose to terminate pregnancy because of NVP.11 Consideration should be given to offering therapy to women with mild to moderate symptoms, be they physical or psychosocial in nature. It remains to be shown that this approach can decrease women's suffering and/or progression of the severity of NVP.
RELATED PUBLICATION BY THE SPEAKER
Magee LA, Redman CW. An N-of-1 trial for treatment of hyperemesis gravidarum. Br J Obstet Gynaecol 1996;103:478-80.
Back to home