Introduction

Unplanned and undesired pregnancies constitute an important problem affecting women’s health worldwide. It was estimated that 43.8 million induced abortions (i.e., 28 per 1000 women of reproductive age between 15 and 44) occurred globally in 2008, of which about 49% were unsafe . This tragedy can be avoidable through proper contraceptive practice. Despite great efforts being spent in developing and propagating the use of a number of effective contraceptive methods in the recent decades, none of the currently available methods is 100% effective. Emergency contraception (EC), also called “postcoital contraception,” is an important back-up after unprotected sexual intercourse (UPSI) or contraceptive failure for the purpose of preventing unintended pregnancies. EC may be indicated after an unexpected sexual act without contraception, after an episode of suspected failure of one’s regular contraceptive method such as missed pills, condom slippage or breakage, or dislodgement of an intrauterine contraceptive device. It may also be indicated after an incident of rape.

Despite the important role played by EC in the above scenarios, there are still myths and misconceptions that prevent proper access to and use of EC. The objective of this review is to provide an updated overview of the clinical efficacy and mechanisms of action of the current methods of EC, particularly the more recent advances, a highlight of some of the important practical points and a discussion on the potential means to improve access and delivery of EC.

Historical EC methods

Various special manoeuvres have been described in historical records to be used for EC, such as jumping, sneezing and vaginal douching with various substances such as lemon juice, Coca-Cola and disinfectants . The concept of hormonal EC can probably be dated back to the 1920s when animal studies showed that postcoital administration of oestrogen could prevent pregnancy. In the early 1960s, the use of high doses of oestrogens such as ethinyl-estradiol (EE) 5 mg, conjugated oestrogen 30 mg daily or diethylstilbestrol 25–50 mg daily over 5 days were reported, which represented the first hormonal EC methods used in women. These regimens were effective, with pregnancy rates of only 0.6–1.6%, but were soon discontinued because of the high incidence of side effects, mainly nausea (54–70%) and vomiting (24–33%), associated with the high oestrogen dose, as well as teratogenicity associated with diethylstilbestrol .

Historical EC methods

Various special manoeuvres have been described in historical records to be used for EC, such as jumping, sneezing and vaginal douching with various substances such as lemon juice, Coca-Cola and disinfectants . The concept of hormonal EC can probably be dated back to the 1920s when animal studies showed that postcoital administration of oestrogen could prevent pregnancy. In the early 1960s, the use of high doses of oestrogens such as ethinyl-estradiol (EE) 5 mg, conjugated oestrogen 30 mg daily or diethylstilbestrol 25–50 mg daily over 5 days were reported, which represented the first hormonal EC methods used in women. These regimens were effective, with pregnancy rates of only 0.6–1.6%, but were soon discontinued because of the high incidence of side effects, mainly nausea (54–70%) and vomiting (24–33%), associated with the high oestrogen dose, as well as teratogenicity associated with diethylstilbestrol .

The Yuzpe regimen

In 1974, the Yuzpe method was introduced by Dr. Albert Yuzpe from Canada, which consisted of two doses of EE 100 μg in combination with dl-norgestrel 1 mg administered within 72 h of UPSI, and repeated 12 h later . This regimen, or an equivalent consisting of EE 100 μg in combination with levonorgestrel (LNG) 0.5 mg, soon became the most popular method of hormonal EC worldwide until the recent decade. In many clinical settings, the Yuzpe regimen can be provided through off-label use of several tablets of commonly used combined oral contraceptive pills making up the equivalent dose. Nausea (54%) and vomiting (16%) remained the main side effects, resulting from the high dose of oestrogen . The failure rate of the Yuzpe regimen was reported to increase with increasing interval between UPSI and treatment , with an overall failure rate of 3.2% when administered within 72 h of UPSI . There is no absolute contraindication for its use in any medical condition, except for an established pregnancy, as the oestrogen-related risks associated with a single course of treatment is likely to be negligible.

The LNG-only regimen

The first randomised controlled trial (RCT) on the progestogen-only EC regimen, conducted in Hong Kong, was published in 1993 . This comprised of two doses of LNG 0.75 mg administered 12 h apart, with the first dose administered within 48 h of UPSI. This first study revealed an apparently higher efficacy with the LNG-only regimen compared to the Yuzpe regimen, although not reaching statistical significance. Nonetheless, the incidence of side effects was significantly reduced with the LNG-only regimen . The clinical efficacy of the LNG-only regimen was subsequently confirmed to be significantly higher than that of the Yuzpe regimen (1.1% vs. 3.2%) in a multinational trial coordinated by the World Health Organisation (WHO), which recruited women with coitus-treatment interval up to 72 h . Combined data from a recent meta-analysis showed that the failure rate of the LNG-only regimen was significantly lower than that of the Yuzpe regimen (relative risk (RR) 0.54, 95% confidence interval (CI) 0.36–0.80) , and the former has since become the first-line method for hormonal EC . Meta-analysis of three trials with a total of 3343 subjects revealed that LNG 1.5 mg administered as a single dose within 72 h of UPSI was equally efficacious as the split-dose regimen (RR 0.84; 95% CI 0.53–1.33) without significant increase in the side-effect profile except for higher incidences of headache and heavy menses in the single-dose group . As this could simplify the use and avoid non-compliance, this has become the currently recommended regimen . It has been demonstrated that the failure rate of LNG as EC increased with increasing interval between UPSI and treatment , especially when comparing use within and beyond 48 h, or within and beyond 72 h of UPSI, respectively . Although the WHO actually endorsed the use of LNG as EC up to 120 h after UPSI, the recent meta-analysis showed that administration of LNG after 72 h of UPSI gave a significantly higher failure rate compared to use within 72 h , and it should be noted that the use of LNG as EC beyond 72 h after UPSI is outside the product licence.

The main mechanism of action of LNG as EC is by delaying or inhibiting follicular growth or rupture. It has been shown that administration of LNG after onset of the LH surge was not effective in inhibiting ovulation. This may explain why LNG is only effective as EC when administered before, but not after, ovulation has occurred, a phenomenon shown in some studies . This was further supported by other in vitro studies which suggested that LNG at pharmacological concentrations had no significant effect on post-ovulatory events such as sperm function , fertilisation or endometrial receptivity . Although one study suggested that LNG might have some inhibitory effect on muscular contractility of the fallopian tube , the clinical significance of this finding is uncertain.

Mifepristone

Mifepristone is a progesterone receptor modulator with wide application in gynaecological practice. It has been studied as an EC. The first studies were reported in 1992 , where mifepristone 600 mg was used within 72 h of UPSI. A subsequent study by the WHO confirmed that lower doses of mifepristone at 50 mg and 10 mg were equally effective as EC . The latest meta-analysis showed that a mid-dose of mifepristone at 25–50 mg, or a low-dose mifepristone (<25 mg) taken within 120 h of UPSI were the most effective hormonal EC method and were significantly more effective than LNG . The significance, however, was just marginal if only high-quality studies were included. Low-dose mifepristone was less effective compared to mid-dose, although again the difference was not significant if only high-quality studies were taken into account . When compared with the Yuzpe regimen, the use of mifepristone is associated with a lower incidence of nausea and vomiting, but some users may experience delay of the subsequent menstruation secondary to delayed follicular development and ovulation, which may sometimes be a cause for concern . However, mifepristone is only available in some countries, and mifepristone tablets at doses of 10 or 25 mg are only marketed in China and Russia, hence limiting its use in many localities.

The apparently superior efficacy of mifepristone over LNG may arise from its additional post-ovulatory mechanism of action in inhibiting endometrial receptivity and tubal contractility as indicated in some in vitro studies . Other in vitro studies suggested no effect of mifepristone at concentrations relevant for EC on human sperm function and fertilisation .

Ulipristal acetate

Ulipristal acetate (UPA), another selective progesterone receptor modulator, has been studied for EC in the recent decade. Since 2009, the oral tablet containing UPA 30 mg has been marketed, under the trade name ella ^® or ellaOne ^® , as a single dose to be taken within 120 h of UPSI. UPA was shown in two large RCTs to be as effective as LNG for EC when administered within 72 h of UPSI. The study by Glasier et al. also included 100 subjects taking UPA between 72 and 120 h after UPSI in which no pregnancy was observed . Meta-analysis on the combined data from results of these two RCTs indicated significantly lower failure rate of UPA compared to LNG (1.36% (22/1617) vs. 2.15% (35/1625), p = 0.046) . A further open-label study on 1241 subjects in which UPA was administered between 48 and 120 h after UPSI revealed a failure rate of 2.1% . It appeared that the efficacy of UPA does not decrease with coitus-treatment interval up to 120 h after UPSI , which is now the recommended time frame for EC in the product licence.

The main mechanism of action of UPA is to delay or inhibit ovulation, and such effect of UPA remains evident even after the onset but before the peak of LH surge . Follicular rupture can be delayed by 4–10 days after intake of UPA . It was revealed in one study that follicular rupture did not occur for at least 5 days after UPA intake in 60% of cases, even when the follicle was 18 mm or above in mean diameter (100% when given before the LH surge, 79% after the onset but before the LH peak, 8% after the LH peak) . There was also in vitro data showing that UPA could alter endometrial histology which might suggest an anti-implantation effect, although such inference is indirect and the clinical relevance is yet to be proven. There have also been preliminary in vitro data indicating that UPA at pharmacological concentrations could inhibit human sperm hyperactivation , as well as ciliary beating and muscular contraction in the human fallopian tube . These may contribute to additional post-ovulatory mechanisms of action, and, together with the wider window of its pre-ovulatory effect, might explain the superior efficacy of UPA over LNG.

Copper intrauterine contraceptive device

The use of the copper intrauterine contraceptive device (Cu-IUCD) for EC was first reported in 1976 by Lippes et al. , and is currently the most effective method of EC to be used within 120 days after UPSI , with a failure rate of 0.09% according to a recent meta-analysis. It has also been suggested that the use can be extended beyond 5 days of UPSI if it is still within 5 days post ovulation in an individual where the day of ovulation can be reasonably estimated . Another advantage of the Cu-IUCD is that it can be retained in place as ongoing regular contraception. With these advantages, the Cu-IUCD should be offered as an option to patients requiring EC whenever clinically feasible.

Cu-IUCD releases copper ions into the uterine and tubal fluid , which may exert contraceptive action through both pre-fertilisation and post-fertilisation mechanisms. Pre-fertilisation effects include oocyte toxicity, impairment of sperm functions such as motility, viability, acrosome reaction and fertilising capacity, and influence on tubal motility and the fertilisation process . Post-fertilisation effects are mainly contributed by the local inflammatory reaction in the endometrium which is unfavourable for implantation . It should be noted that no research evidence for the use of progestogen-releasing IUCD as EC is yet available .

Factors affecting efficacy of hormonal EC

Data from meta-analyses suggested that the failure rate with LNG and UPA was higher in obese women compared to women with normal body mass index . The effect of obesity on hormonal EC failure was particularly more pronounced with LNG (odds ratio 4.41; 95% CI 2.05–9.44) than with UPA (odds ratio 2.62; 95% CI 0.89–7.00) comparing subjects with body mass index ≥ 30 kg/m ² versus <25 kg/m ² . Hence, the Cu-IUCD may preferably be offered to obese women who require EC.

A significantly higher failure rate was also observed among users with further acts of UPSI in the same cycle after the index EC use, as well as those who had UPSI within the fertile window . The failure rate of LNG but not UPA also increases with increase in coitus-treatment interval as described above.

Safety issues, contraindications and drug interactions

Hormonal methods of EC are generally considered very safe. Minor side effects such as nausea, vomiting and headache may be encountered, more associated with the use of the Yuzpe regimen. Vomiting may reduce the efficacy of EC. A recent meta-analysis found two trials which suggested that in users of the Yuzpe regimen, nausea could be reduced by pre-treatment with meclizine and metoclopramide, but only meclizine pre-treatment reduced vomiting. There is no reported study on the prevention of nausea and vomiting in users of LNG and UPA, but the incidence is generally much lower compared to those using Yuzpe regimen . As suggested by the package inserts, users of LNG and UPA for EC who vomit with 2 h and 3 h of drug intake respectively can consider repeating the dose.

Generally there is no contraindication to the use of hormonal EC apart from an established pregnancy. The usual medical contraindications applying to the regular forms of hormonal contraception do not apply to hormonal EC because of the very short exposure. Physical examinations are not mandatory before administration of hormonal EC. The absolute contraindications to the use of Cu-IUCD in regular contraception, such as current genital tract infection and cavity-distorting uterine lesions, also apply to its use for EC . High risk of sexually transmitted infections, such as in rape victims, is a relative contraindication for EC-IUCD, but antibiotic prophylaxis can be considered if there is no better alternative.

It must be clearly emphasised that EC is not an abortifacient and is effective only before pregnancy is established, even though some of the EC methods may have a post-fertilisation mechanism of action. According to a Judicial Review in the United Kingdom in 2002, establishment of pregnancy was defined at the time of embryo implantation, not at fertilisation. Nonetheless, contraceptive users may have different personal or religious views and acceptance over this, and their informed choice should be respected after proper factual explanations. Users who do not accept an EC with possible post-fertilisation actions can be offered the LNG method.

Hormonal EC is considered contraindicated in a woman who is already pregnant mainly because it is not going to work, but not because of any adverse effect on pregnancy. In other words, inadvertent use of EC in a woman who is subsequently found pregnant should not be an indication for termination of pregnancy. The current evidence suggested no adverse effect of LNG and mifepristone exposure during early pregnancy on the pregnancy and foetal outcomes . To date, there have been limited data on the effect of UPA exposure in pregnancy. A recent publication on the post-marketing surveillance of UPA reported no case of birth defect out of 20 live births with exposure to UPA during pregnancy. However, the number may be still too small to draw a conclusion on the safety of UPA in pregnancy .

A recent meta-analysis suggested that LNG and mifepristone did not increase the risk of ectopic pregnancy, probably because of an overall reduction of pregnancy risk after the use of EC . The current evidence provided reassurance that a history of ectopic pregnancy should not be a contraindication to EC use.

Use of LNG in lactating women is probably safe without the need to withhold breast-feeding . There are no reported data on the effect of UPA in breast-feeding women. UPA is a lipophilic compound and may theoretically be excreted in breast milk and hence a risk to the breast-fed baby cannot be excluded. The manufacturer recommends no breast-feeding for 36 h after intake of the UPA pill.

In women who are currently on enzyme-inducing medications within the past 28 days, there are some concerns over the use of LNG or UPA which are metabolised by the CYP3A4 microsomal system, although the actual clinical effects are unclear. Medications which reduce gastric acidity may also affect UPA absorption. In these circumstances, the Cu-IUCD is preferred, or a single 3-mg dose of LNG (two tablets) may be considered as recommended by the UK guideline, although the latter is an off-label use and not evidence-based . There is generally no contraindication for use of Cu-IUCD as EC with concomitant use of other medications .