Materials and Methods
This trial was conducted at the Center for Reproductive Medicine and Surgery at the University of Pennsylvania Medical Center, Philadelphia, and the University of Pittsburgh/Magee-Womens Hospital, Pittsburgh, between August 1997 and September 1998. The study was approved by the institutional review boards at both sites and at Family Health International. All participants gave written informed consent before enrollment.
Women between 18 and 45 years old who had no contraindications to the use of combined oral contraceptive pills, meclizine, or lactose; who had no acute or chronic intestinal disorders; who had negative urine pregnancy test results; and who were protected from pregnancy by nonhormonal means were admitted into the study. After an interview to collect baseline information, each participant was assigned randomly to one of three study groups by being given a sequentially numbered, sealed, opaque, padded envelope containing study drugs. The envelopes had been prepared at Family Health International according to a computer-generated randomization list stratified by center with randomly ordered block sizes of three, six, and nine. The envelopes contained drugs as follows: emergency contraceptive pills (to be taken in two doses 12 hours apart) and meclizine (to be taken in one dose 1 hour before the first emergency contraceptive pill dose) (group M); emergency contraceptive pills and placebo (to be taken on the same schedule as meclizine) (group P); and emergency contraceptive pills only (group A).
In addition to the drugs, each envelope contained three questionnaires, instructions on study procedures, and a pen. Participants were instructed to complete the questionnaires 12, 24, and 48 hours after taking the first dose of emergency contraceptive pills. The questionnaires included questions about the time of study drug ingestion, other medications used, nausea, vomiting, other problems, and eating and sleep patterns. Participants returned the completed questionnaires to the study center in person or by mail and then received compensation of $50.
For the emergency contraceptive pill regimen, we cut packs of oral contraceptives (OCs), each containing 50 μg of ethinyl estradiol and 0.5 mg of norgestrel (Ovral; Wyeth-Ayerst Laboratories, Philadelphia, PA), into sets of four pills each, taking care not to damage the blister bubbles in which the pills were packaged originally. We put each set of four pills into a resealable plastic bag labeled with instructions to take each dose of two pills 12 hours apart.
For group M, we removed generic 25-mg meclizine tablets from the blister packs in which they were packaged by the manufacturer and put two pills into resealable plastic bags labeled “Study Drug” along with instructions to take two pills 1 hour before the first dose of emergency contraception.
For group P, we put two placebo pills containing lactose, stearic acid, and magnesium stearate (Placebo Tablets; Consolidated Midland Corp., Brewster, NY) into bags labeled as for the meclizine.
The envelopes given to the study participants all appeared and felt identical from the outside, except for the preprinted study numbers. However, funding limitations precluded us from acquiring identical placebo and meclizine tablets. The meclizine tablets were yellow and were marked with a number. They were thus easily distinguishable from the placebo tablets, which were white and unmarked. However, because we expected that participants would see only one type of tablet, we surmised that they would likely not know which they had received. This supposition was supported by results of a test in which we showed ten staff members at Family Health International a meclizine tablet and ten other staff members a placebo tablet. In each group, five staff members guessed that the tablet was meclizine and five guessed that it was placebo. To assess concealment of allocation to treatment group, we asked research staff at each site to record their guesses regarding each participant's group assignment when they dispensed the envelopes. To evaluate masking in the study population, we asked each participant who received either meclizine or placebo to record on each questionnaire her guess regarding which tablets she had received.
We calculated the sample size for this study to test the null hypothesis of no difference in the incidence of nausea between groups M and P. We made the following assumptions: a 50% incidence of nausea in group P; a 30% incidence of nausea in group M; a 10% drop-out rate (randomized but not returning any questionnaire); α = .05, two-tail test; and β = .20. Using these assumptions, we estimated that 115 women were needed per group. To enable testing of the second hypothesis of no difference in nausea between groups M and A, we determined that group A should include 115 women as well. Thus, we estimated that a total of 345 women were needed for the entire study. Because the drop-out rate was lower than expected, we decided to terminate recruitment after only 343 participants had been enrolled.
For all analyses, we used SAS statistical software (SAS Institute, Cary, NC). We considered a P value of .05 or less to be significant, and we made no adjustments for multiple comparisons. We analyzed all women according to their group assignment, regardless of compliance. Analysts were masked to treatment assignment until the primary analysis was completed. In the two primary analyses, we compared incidences of nausea ever reported between groups M and P and between groups M and A, using a Mantel-Haenszel χ2 test controlling for center. In these analyses, we counted missing responses as negative. We also performed sensitivity analyses in which we counted missing responses as positive or we excluded women with any missing responses and no recorded nausea. The results of the sensitivity analyses were almost identical to those of the primary analyses and are not reported here. We made similar comparisons between groups M and P and between groups M and A to examine differences in incidence of vomiting and incidence of other side effects. Severity of nausea between groups was compared using the mean score extension of the Mantel-Haenszel test.17 In exploratory analyses, we examined the effect of other factors on the incidence of nausea and potential interactions between treatment and other factors using logistic regression. We converted adjusted odds ratios from regression models to approximate relative risks using the method of Zhang and Yu.18 We used χ2 tests to evaluate associations between nausea after the first dose and nausea after the second dose of emergency contraceptive pills, between time of food intake and nausea, and between guesses regarding treatment group and actual treatment assignment.
Altogether, 343 women were enrolled in the study, including 173 in Philadelphia and 170 in Pittsburgh. Of these, 18 were lost to follow-up (never returned any questionnaires) and one was found to be pregnant after she was randomized but before she took any study medications (and thus she never provided follow-up information). The analysis population thus included 324 women, including 161 from Philadelphia and 163 from Pittsburgh (Table 2). These women included six who did not fulfill all admission criteria: two (one in group M, one in group P) had been pregnant less than 3 months earlier and four (two each in groups P and A) were smokers more than 34 years of age. One woman assigned to the meclizine pretreatment group failed to take the meclizine, although she did take the emergency contraceptive pills. In each group, more than 78% of women took all study medications at the time intervals specified by the protocol. Only two women (one each in groups M and A) did not return all three questionnaires.
Baseline characteristics of the women in the three groups were similar (Table 3), except that the proportion of women who previously had used OCs and the proportion who had nausea or vomiting while taking these pills were lower in the placebo group than in the other two groups.
The incidence of nausea was significantly lower among women who received meclizine than among women who received placebo or no pretreatment (Table 4). About two thirds of women in the two control groups reported nausea during the 2-day study period, whereas less than half of the women given meclizine did so. The severity of nausea and the proportion of women who vomited were also both significantly lower in group M than in groups P and A. These conclusions did not change when we excluded the 61 women who took other medications that are known to affect nausea. However, among women with nausea, the median time to onset of nausea was similar in the three groups (12, 12, and 10 hours in groups M, P, and A, respectively). In each group, less than 8% of cases of nausea started more than 24 hours after ingestion of the emergency contraceptive pills. Vomiting within the first 2 hours after ingestion was uncommon; it occurred in one woman after the first dose (group P) and in six women after the second dose (three each in groups M and P). All women who vomited also reported nausea.
The effect of meclizine on the incidence of nausea (compared both with placebo and with no pretreatment) was consistent across subgroups of the study population defined by center, age, race, years of education, weight, smoking status, and history of nausea or vomiting with OCs. We did note a statistically significant interaction between treatment and time of ingestion of emergency contraceptive pills in the comparison between groups M and P: the incidence of nausea was lower in group M than in group P except when the emergency contraceptive pills were taken between 5 and 9 PM. A similar effect was noted in the comparison of groups M and A, although the interaction was not statistically significant. Because we did not adjust for the multiple comparisons made in this study, and because time of emergency contraceptive pill ingestion was a postrandomization variable, this finding must be interpreted with caution.
To examine the effects of baseline factors on the incidence of nausea, we performed an exploratory analysis that included all three treatment groups. We first ruled out significant interactions between treatment group and baseline factors in a logistic regression model that included all baseline factors and the interaction terms. Then we used a backward stepwise approach to identify significant main effects, forcing center and treatment group to remain in the final model. Using this approach, we found that risk of nausea was significantly associated with both age and smoking, as well as treatment group. Women who were older than the median age (27.7 years) were about two-thirds as likely to have nausea as younger women (relative risk [RR] 0.69, 95% confidence interval [CI] 0.50, 0.89). Smokers were slightly less likely to have nausea than nonsmokers (RR 0.89, 95% CI 0.75, 1.00). Center, race, educational level, weight, and history of nausea or vomiting associated with OC use were not significantly related to the risk of nausea.
In all three treatment groups, nausea after the first dose of emergency contraceptive pills was associated with nausea after the second dose (P value from analyses adjusting for treatment group = .001). In the three groups combined, 62% of women who did not have nausea after the first dose also did not have it after taking the second dose. We could find no evidence that taking the emergency contraceptive pills with food affected the chance of nausea. For example, women who ate within 2 hours before or after the first dose of emergency contraceptive were not significantly more or less likely to report nausea than those who did not.
In all three groups, drowsiness, other drowsiness-related effects, headache, and abdominal discomfort were the most commonly reported side effects (Table 5). Drowsiness was reported by about twice as many of the women who received meclizine as of the women who received placebo or no pretreatment (P < .01 for both comparisons). Other drowsiness-related effects also were reported by significantly more women in group M than in group A. These findings did not change when we excluded women who had reported taking any other medication known to cause drowsiness at any time during the study. We did not collect information on severity of drowsiness. The incidence of other side effects did not differ significantly between groups.
Concealment of allocation to treatment group appeared to be well maintained in the study, given the lack of correlation between the enrolling nurse's guess at admission about each participant's treatment group and the actual group assignment (P = .93). Women in group M were more likely than women in group P to guess that they had received meclizine: on their final questionnaires, 74% of women in group M guessed that they had received meclizine compared with only 57% of women in group P (P = .01). However, among women who reported no nausea, percentages who guessed meclizine were similar (82% in group M, 76% in group P; P = .48).
In our study, meclizine pretreatment significantly reduced the incidence of nausea and vomiting associated with the Yuzpe regimen of emergency contraception. Protection was greater for vomiting than for nausea: meclizine pretreatment averted about 25% of cases of nausea, but it prevented 74% of cases of vomiting compared with placebo and 64% compared with no pretreatment.
The incidence of nausea in both control groups in our study (64%) was higher than that reported in most previous trials of emergency contraception, and pre-treatment with meclizine reduced this incidence only to the average level previously reported among women who received no antiemetic treatment (Table 1).3,4,6,9–15 This difference may be attributable to differences in the purpose of the studies and the methods for ascertainment of nausea. In our trial, participants knew that nausea (rather than contraceptive efficacy) was the primary focus of the investigation, and we specifically asked each participant three times in a 48-hour period whether she had become nauseated. Therefore, it is not surprising that we found more nausea overall than was found in previous studies in which women may have been asked only a single general question about side effects.
Drowsiness is a known side effect of meclizine. In our study, 30% of women treated with meclizine complained of this problem, about twice as many as in the other two groups. However, many women likely will accept some drowsiness in order to reduce their chance of having nausea. Women who choose to take meclizine should be cautioned against driving or engaging in other activities that require alertness. In this study, we used the highest recommended dose of meclizine in order to maximize its potential effectiveness. A lower dose or a divided dose might be as effective while causing less drowsiness.
None of the participants in our trial were at risk for pregnancy; they took the emergency contraceptive pills only because of the study. We designed the study this way to facilitate recruitment and follow-up of participants. However, we can think of no strong reason that our results would not apply to women taking emergency contraceptive pills for clinical indications.
This trial provided no evidence to support the common recommendation that emergency contraceptive pills be taken with food to reduce the chance of nausea and vomiting. This finding should be interpreted with caution, however, because participants' decisions about whether and when to eat could have been influenced by both the presence of nausea and knowledge of treatment assignment.
We did not observe a placebo effect in this study: the incidence and severity of nausea were similar in groups P and A. This finding could be due to a true absence of placebo effect or to a failure of masking in the placebo group. We know that at least one nurse who admitted participants was aware that the meclizine and placebo pills were different colors. She and all other staff working on the study stated that they never revealed this information to participants. Nevertheless, when questioned about reasons for their treatment-group guesses, four participants said they based their guesses on the appearance or taste of the pills, and all were correct. In addition, we found a significant statistical association between actual group assignment and participants' guesses. However, given that this association was not observed among women who did not report nausea, we believe that it is more likely a consequence of recognition of the drug's effectiveness in preventing nausea than a result of unmasking. According to several recent reviews,19–21 a true placebo effect rarely has been documented in properly designed clinical trials, so its absence in our study may not be unusual.
Two clinical trials, whose findings were published in 1993 and 1998, showed that a regimen containing only levonorgestrel appears to be more effective for postcoital emergency contraception than the Yuzpe regimen.12,15 Although direct comparative data are not available, the incidence of nausea associated with the levonorgestrel regimen reported in these two studies (about 20%) was substantially lower than the 47% incidence we observed among women treated with the Yuzpe regimen and meclizine. (The incidence of vomiting, however, was about 5% in all three studies.) We expect that eventually the levonorgestrel regimen will replace the Yuzpe regimen as the standard emergency contraceptive treatment and that women who use levonorgestrel will not need antiemetic pretreatment because nausea is so infrequent. A levonorgestrel-only emergency contraceptive product recently was approved by the United States Food and Drug Association. However, it is not yet marketed widely and may not be accessible to all women who need emergency contraception in this country or abroad. For convenience and cost reasons, some clinicians and patients may continue to make up the Yuzpe regimen from regular birth control pills. Therefore, we believe that our findings will remain relevant to clinical practice in many countries for years to come.
1. Trussell J, Stewart F, Guest F, Hatcher RA. Emergency contraceptive pills: A simple proposal to reduce unintended pregnancies. Fam Plann Perspect 1992;24:269–73.
2. Yuzpe AA, Thurlow HJ, Ramzy I, Leyshon JI. Post coital contraception: A pilot study. J Reprod Med 1974;13:53–8.
3. Yuzpe AA, Lancee WJ. Ethinylestradiol and dl-norgestrel as a postcoital contraceptive. Fertil Steril 1977;28:932–6.
4. Yuzpe AA, Smith RP, Rademaker AW. A multicenter clinical investigation employing ethinyl estradiol combined with dl-norgestrel as a postcoital contraceptive agent. Fertil Steril 1982;37:508–13.
5. Bagshaw SN, Edwards D, Tucker AK. Ethinyl oestradiol and D-norgestrel is an effective emergency postcoital contraceptive: A report of its use in 1,200 patients in a family planning clinic. Aust N Z J Obstet Gynaecol 1988;28:137–40.
6. Tully B. Postcoital contraception: A study. Br J Fam Plann 1983;8:119–24.
7. Van Santen MR, Haspels AA. Interception II: Postcoital low-dose estrogens and norgestrel combination in 633 women. Contraception 1985;31:275–93.
8. Percival-Smith RK, Abercrombie B. Postcoital contraception with dl-norgestrel/ethinyl estradiol combination: Six years experience in a student medical clinic. Contraception 1987;36:287–93.
9. Zuliani G, Colombo UF, Molla R. Hormonal postcoital contraception with an ethinylestradiol-norgestrel combination and two danazol regimens. Eur J Obstet Gynecol Reprod Biol 1990;37:253–60.
10. Glasier A, Thong KJ, Dewar M, Mackie M, Baird DT. Mifepristone (RU 486) compared with high-dose estrogen and progestogen for emergency postcoital contraception. N Engl J Med 1992;327:1041–4.
11. Webb AM, Russell J, Elstein M. Comparison of Yuzpe regimen, danazol, and mifepristone (RU486) in oral postcoital contraception. BMJ 1992;305:927–31.
12. Ho PC, Kwan MS. A prospective randomized comparison of levonorgestrel with the Yuzpe regimen in post-coital contraception. Hum Reprod 1993;8:389–92.
13. Evans JK, Holmes A, Browning M, Forster GE. Emergency hormonal contraception usage in genitourinary medicine clinic attenders. Genitourin Med 1996;72:217–9.
14. Sanchez-Borrego R, Balasch J. Ethinyl oestradiol plus dl-norgestrel or levonorgestrel in the Yuzpe method for post-coital contraception: Results of an observational study. Hum Reprod 1996;11:2449–53.
15. Task Force on Postovulatory Methods of Fertility Regulation. Randomised controlled trial of levonorgestrel versus the Yuzpe regimen of combined oral contraceptives for emergency contraception. Lancet 1998;352:428–33.
16. Van Look PFA, Stewart F. Emergency contraception. In: Hatcher RA, Trussell J, Stewart F, Cates W, Stewart GK, Guest F, et al. Contraceptive technology. 17th ed. New York: Ardent Media, 1998:277–96.
17. Fleiss JL. Statistical methods for rates and proportions. 2nd ed. New York: Wiley, 1981.
18. Zhang J, Yu KF. What's the relative risk? A method of correcting the odds ratio in cohort studies of common outcomes. JAMA 1998;280:1690–1.
19. Ernst E, Resch KL. Concept of true and perceived placebo effects. BMJ 1995;311:551–3.
20. Kienle GS, Kiene H. Placebo effect and placebo concept: A critical methodological and conceptual analysis of reports on the magnitude of the placebo effect. Altern Ther Health Med 1996;2:39–54.
21. Kienle GS, Kiene H. The powerful placebo effect: Fact or fiction? J Clin Epidemiol 1997;50:1311–8.