1 Department of Obstetrics & Gynecology, University of Hong Kong, Queen Mary Hospital, 102 Pokfulam Road, Hong Kong SAR, 2 MRCOG, Family Planning Association of Hong Kong, Hong Kong SAR, 3 Family Planning Association, Institute of Shenzhen, 4 Shanghai Institute of Family Planning Technical Institution, 5 Jiangsu Family Planning Institute, 6 Beijing Chaoyang Hospital, affiliated to Capital University of Medical Science and 7 Department of Obstetrics & Gynecology, University of Hong Kong, Hong Kong SAR, China
8 To whom correspondence should be addressed. Email: cora{at}hkucc.hku.hk
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key words: emergency contraception/levonorgestrel/pregnancy/randomized trial/unprotected intercourse
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
Materials and methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Randomization and follow-up visits
The pharmacy department in Queen Mary Hospital generated the randomization sequence by computer program. The drug package was done by the pharmacy department according to the randomization list. Both the clinicians and the participants were unaware of the drug assignment. The pharmacy kept the randomization list and it was revealed only at the final analysis. The levonorgestrel and the placebo was supplied by the World Health Organization. The placebo was identical in colour, shape and size to the levonorgestrel.
Consenting women were given a labelled treatment bag, which contained three tablets [group 1: one tablet of 0.75 mg levonorgestrel (first dose), one identical placebo tablet (second dose), and one tablet of 0.75 mg levonorgestrel (third dose); and for group 2: one tablet of 0.75 mg levonorgestrel, one tablet of 0.75 mg levonorgestrel and one placebo tablet respectively]. All women took the first dose of levonorgestrel on admission to the study. They were asked to take the second and the third tablet 12 and 24 h later. Women were advised not to have further acts of coitus before the return of menstruation. A diary was given to record the side-effects, spotting or bleeding. A follow-up visit was arranged 1 week after the date of the expected menses. The case was completed if menses had returned by the time of the first scheduled visit. Pregnancy tests were performed for those remaining amenorrhoeic. If the pregnancy test was positive, we referred her for counselling about the plan of the unintended pregnancy. If the pregnancy test was negative, a second follow-up in 4 weeks time was arranged. A procedure similar to that of the first visit was repeated. If the pregnancy test remained negative, treatment was regarded as successful.
Assessment of outcome
The primary outcome measure was the pregnancy rate in the two treatment arms. The efficacy was measured in two ways: (i) calculating the crude and centre-adjusted pregnancy rates; (ii) the estimated reduction in the number of expected pregnancies, or prevented fraction (one minus observed pregnancies to expected pregnancies). We used the data pooled by Trussell et al. (1998) to calculate the expected number of pregnancies. It was estimated by multiplying the number of women having unprotected coitus on each day of the cycle by the conception probability on that cycle day. The date of ovulation was estimated by subtracting 14 days from the expected date of the next menstrual period. The estimates include only clinical pregnancies.
Statistical analysis
The primary outcome measure was the pregnancy rate. Secondary outcome measures included the delay in onset of next menses and the incidence of side-effects. The proposed sample size was calculated on the basis of method failures, obtained from our previous study. The failure rate of double dose (12 h regimen) levonorgestrel was 2.4% (Trussell et al., 1998). A sample of 2052 (i.e. 1026 in each group) would permit the detection of a failure rate of the 24 h regimen that is twice that of the 12 h regimen, with the assumption of 10% lost to follow-up, a type I error of 5%, and a power of 80%.
The analyses were performed according to the intention-to-treat principle. Continuous variables were compared with the use of the Wilcoxon rank-sum test, and categorical variables were compared with the use of the 2 or Fisher's exact test (the latter when there was an expected value of less than five for any cell).
In addition, we examined the interaction effect between further acts of intercourse and treatment on the pregnancy outcome by using the BreslowDay test. The same method was used to investigate the interaction between the use of condom for those who have further acts of intercourse and pregnancy rate. We also examined if the treatment would cause any delay in the return of menstruation. The efficacy of treatment in both groups for those who presented late (>72 h) was also investigated.
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
|
|
|
Comparing women who completed follow-up with those who were lost to follow-up, the latter group were significantly younger (mean age 27.0 versus 22.0 years, P<0.05), less likely to have a previous pregnancy (58.6 versus 37.0%, P<0.01) and less likely to use any form of contraceptive methods before (87.8 versus 73.6%, P<0.01).
The effect of further acts of intercourse on the pregnancy rate was analysed (Table IV). A total of 1566 women had no further acts of intercourse after the act that prompted treatment. The pregnancy rate was similar between the two groups (2.0% in 24 h group and 1.3% in 12 h group). However, there was a significant interaction effect between further acts of intercourse and treatment on the pregnancy outcome (P=0.045 by BreslowDay test). For those with further acts of intercourse after treatment, the efficacy was unaltered in the 24 h treatment group (1.8 versus 2.0%, P>0.05, 95% CI 0.292.65); whereas the pregnancy rate of the 12 h treatment group was significantly higher (4.5 versus 1.3%, P<0.01, 95% CI 1.498.81).
|
Complications and side-effects
The prevalence of side-effects for the two groups is shown in Table V. In general, the side-effects were mild and acceptable. Nausea and vomiting occurred in <10% of women. The occurrence of diarrhoea and breast tenderness was significantly more common in the 12 h group (P<0.05). The prevalence of other side-effects was otherwise similar in the treatment groups. Most women, 436 of 1020 (42.7%) in 24 h group and 437 of 998 (43.7%) in 12 h group, had their menses returned at about the expected time (within 2 days). Forty-one out of 1044 in the 24 h group (4.0%) and 51/1027 in the 12 h group (5.0%) had a delay of >1 week. The timing of menses did not differ between the two regimens (Figure 2).
|
|
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The World Health Organization randomized trial compared the efficacy of low dose mifepristone, single dose levonorgestrel (1.5 mg) and two doses levonorgestrel (0.75 mg 12 h apart) for emergency contraception within 120 h of unprotected intercourse (Von Hertzen et al., 2002). It was a double-blind, randomized, multicentre trial with 4071 women included. The pregnancy rate was 1.5% (20/1356) in those assigned single dose levonorgestrel, and 1.8% (24/1356) in women assigned two dose levonorgestrel. The pregnancy rate among Chinese women (1.8%) was higher than that of non-Chinese women (1.3%), though the difference is not statistically significant. In our study, all the study subjects were Chinese and the pregnancy rates (1.9% in the 24 h regimen and 2.0% in the 12 h regimen) were similar to that of Chinese women in the World Health Organization study. Our results also show that the pregnancy rate was higher when the coitustreatment interval was >72 h, though the difference was not statistically significant. These are also consistent with the results in the World Health Organization study (Von Hertzen et al., 2002
).
Previous studies reported that further acts of intercourse (with and without contraception) between treatment and expected menstruation resulted in higher pregnancy rates (Task Force on Post-ovulatory Methods of Fertility Regulation 1998; Von Hertzen et al., 2002
). In contrast, our results showed that while further acts of intercourse significantly increased the pregnancy rate in subjects who underwent the 12 h treatment, they did not in those who underwent the 24 h treatment. The non-significance in the 24 h group may be due to the reduced sample size in the subgroup analysis. Therefore, we still would recommend that women should be advised to abstain from further acts of intercourse during the treatment cycle.
Hapangama et al. (2001) showed that the administration of levonorgestrel on or before the day of the first significant rise in urinary LH led to a significant delay in the LH peak, anovulation or reduction in the total luteal phase LH concentrations and the length of the luteal phase. Durand et al. (2001
) also found that pre-ovulatory administration of levonorgestrel might disrupt the normal development and/or the hormonal activity of the growing follicle, but post-ovulatory administration did not impair corpus luteum function or endometrial morphology. We postulate that the 24 h regimen may cause a more prolonged effect so that further acts of intercourse will not result in a higher pregnancy rate. Further studies are needed to confirm our finding and to find out the reasons for this phenomenon. We also showed that breast tenderness and diarrhoea occur more frequently in the 12 h group when compared with the 24 h group. However, the incidence of these side-effects is low even in the 12 h regimen. The difference in incidence between the two groups is probably not clinically significant. A previous pharmacokinetic study showed that the area under the curve in the first 24 h in the 12 h regimen was significantly higher than that in the 24 h regimen although the total area under the curve was similar (Johansson et al., 2002
). The peak observed after the second dose (32.8 nmol/l) in the 12 h regimen was also higher than that after the second dose in the 24 h regimen (30.0 nmol/l). This may explain the higher incidence of side-effects in the 12 h regimen. It may be worthwhile to compare the single dose (1.5 mg) versus 24 h double dose (0.75 mg) regimen in order to find out the optimal regimen for emergency contraception.
There were few data on the risk of ectopic pregnancy following levonorgestrel treatment for emergency contraception. We reported the occurrence of 40 pregnancies in this study. None of these pregnancies was ectopic. Although case reports on ectopic pregnancy following emergency levonorgestrel treatment have been reported previously (Sheffer-Mimouni et al., 2003), there is no evidence that levonorgestrel treatment is associated with an increased risk of ectopic pregnancy.
We conclude that the 24 h double dose levonorgestrel regimen is as effective as the 12 h regimen for emergency contraception up to 120 h after unprotected intercourse. The second dose of levonorgestrel can be given in a flexible schedule (1224 h) to make it more convenient for the women. Both regimens are associated with only minimal side-effects. Further research to investigate more effective methods of emergency contraception is warranted.
![]() |
Acknowledgements |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Hapangama D, Glasier AF and Baird D (2001) The effects of peri-ovulatory administration of levonorgestrel on the menstrual cycle. Contraception 63, 123129.[CrossRef][ISI][Medline]
Henshaw SK (1998) Unintended pregnancy in the United States. Fam Plann Perspect 30, 2429,46.[ISI][Medline]
Ho PC and Kwan MS (1993) A prospective randomized comparison of levonorgestrel with the Yuzpe regimen in post-coital contraception. Hum Reprod 8, 389392.[Abstract]
Johansson E, Brache V, Alvarez F, Faundes A, Cochon L, Ranta S, Lovern M and Kumar N (2002) Pharmacokinetic study of different dosing regimens of levonorgestrel for emergency contraception in healthy women. Hum Reprod 17, 14721476.
Kook K, Gabelnick H and Duncan G (2002) Pharmacokinetics of levonorgestrel 075 mg tablets. Contraception 66, 7376.[CrossRef][ISI][Medline]
Sheffer-Mimouni G, Pauzner D, Maslovitch S, Lessing JB and Gamzu R (2003) Ectopic pregnancies following emergency levonorgestrel contraception. Contraception 67, 267269.[CrossRef][ISI][Medline]
Task Force on Postovulatory Methods of Fertility Regulation (1998) Randomised controlled trial of levonorgestrel versus the Yuzpe regimen of combined oral contraceptives for emergency contraception. Lancet 352, 428433.[CrossRef][ISI][Medline]
Trussell J, Stewart F, Guest F and Hatcher RA (1992) Emergency contraceptive pills: a simple proposal to reduce unintended pregnancies. Fam Plann Perspect 24(6), 269273.[ISI][Medline]
Trussell J, Rodriguez G and Ellertson C (1998) New estimates of the effectiveness of the Yuzpe regimen of emergency contraception. Contraception, 363369.
Von Hertzen H, Piaggio G, Ding J, Chen J, Song S, Gemzell-Danielson K et al. (2002) Low dose mifepristone and two regimens of levonorgestrel for emergency contraception: a WHO multicentre randomized trial. Lancet 360, 18031810.[CrossRef][ISI][Medline]
Submitted on June 1, 2004; accepted on October 2, 2004.
|