Early medical abortion using low-dose mifepristone followed by buccal misoprostol: a large Australian observational study

The antiprogesterone mifepristone combined with a prostaglandin analogue (misoprostol or gemeprost) is effective for terminating early pregnancy and has a favourable safety profile.1 Early medical abortion (EMA) regimens using 200 mg oral mifepristone and misoprostol have been endorsed by the Royal College of Obstetricians and Gynaecologists (RCOG) as an effective and appropriate method of terminating early pregnancy.2 Randomised controlled trials have shown that mifepristone 200 mg followed 24–48 hours later by misoprostol is as effective as surgical abortion in gestations up to 63 days, with complete abortion generally occurring in 93%–98% of cases, incomplete abortion in 1.1%–4.2%, and ongoing pregnancy in 0.2%–2.7%.3-10

Since its first registration in France in 1988, mifepristone has been registered in around 50 countries and is on the World Health Organization list of essential medicines.11 Mifepristone is currently an unapproved medicine in Australia, but can be accessed under the Authorised Prescriber provisions of the Therapeutic Goods Act 1989.12 Results for Authorised Prescriber use of mifepristone in Australia12-14 have been similar to those reported for other countries.1

The not-for-profit, non-governmental organisation Marie Stopes International Australia (MSIA) started Authorised Prescriber use of mifepristone–buccal misoprostol for EMA in Australia in 2009. This article describes outcomes from the use of this regimen in 13 345 EMAs.

This is the first large-scale Australian study of mifepristone for EMA up to 63 days of gestation. Clinic-administered mifepristone (200 mg) in combination with home-administered buccal misoprostol (800 μg) was associated with a low failure rate, low rates of haemorrhage and infection, and was well tolerated.

The low failure rate (3.5%) with mifepristone in our study was similar to that reported for mifepristone in randomised controlled trials (failure rates, 2%–7%)3-10 and in other Australian Authorised Prescriber Schemes (failure rates, 2.9%–5.6%).13,14 Notably, the rate of surgical intervention for incomplete abortion declined from 4.3% during the first 3-month period of our study to 1.6%–2.6% during subsequent 3-month periods. We propose that as clinic staff gained experience with the EMA method, they felt more comfortable with expectant management rather than proceeding directly to surgical intervention.

In our study, there were four cases of known infection reported (0.03%), including one death from sepsis. Infection following medical termination of pregnancy is reported to occur in 0.02%–0.92% of cases.15,16 Death resulting from infection and fatal toxic shock after medical abortion with mifepristone is a known, but very rare, risk (1.1/100 000).15 Although the RCOG2 recommends universal prophylactic antibiotics effective against C. trachomatis and anaerobes to reduce the risk of infection after medical abortion, the WHO17 does not. At MSIA, all women are screened for C. trachomatis, unless they decline, and for other STIs based on a risk assessment. MSIA has upgraded its warnings to women regarding serious infections and also now provides an opt-in SMS service that reminds women 3–5 days after their visit about the after-care phone number and signs and symptoms of complications.

The regimen we used is similar to those recommended by the WHO17 and the RCOG,2 with buccal administration of misoprostol. Vaginal misoprostol has been associated with deaths from toxic shock syndrome in the United States18 and some women may prefer buccal, rather than vaginal, administration. In addition, oral misoprostol is not sufficiently effective beyond 49 days of gestation.1,2,17 Randomised controlled trials have shown mifepristone with buccal misoprostol to have efficacy similar to or greater than mifepristone with vaginal7 or oral19 misoprostol for terminating pregnancies up to 56 and 63 days, respectively. Our study provides further support for using 200 mg of mifepristone, which is lower than the 600 mg dose approved in the US and Europe.1 Randomised controlled trials directly comparing the two mifepristone doses have shown similar efficacy for the 200 mg and 600 mg doses.3,6 A Cochrane meta-analysis also found no difference in failure rates between the 200 mg and 600 mg doses.20

In our study, all women were informed of the need for mandatory follow-up to confirm pregnancy termination; contact with most women was achieved. Those who did not attend the clinic were telephoned and, if telephone contact was not possible, a registered letter was sent (the letter included a pregnancy test kit to facilitate home-testing). Despite these efforts, follow-up information was missing for around 17% of women. The loss to follow-up in our real-world setting of mifepristone use was higher than that observed in rigorous clinical trial settings (1.7%–8.6%).3-5,8,9,21 As might be expected in clinical practice, not all patients may return for follow-up, even if instructed to attend a follow-up visit to assess treatment outcome. Given the emotional issues surrounding abortion, women who felt the abortion was complete and had no physical complications may have been reluctant to return to the clinic. While this is of concern, a recent study using early telephone follow-up after medical abortion to determine the need for further evaluation found that women who initially had not required evaluation returned to the clinic of their own accord if they experienced problems, such as bleeding or continuing pregnancy.21 This finding suggests that women who experience adverse effects or feel that the abortion is not complete are likely to seek appropriate medical care. Indeed, in our study, about one-third of women did call the MSIA after-care service about issues such as pain or bleeding. While many women required reassurance only, some callers were referred to the clinic or a hospital for further assessment. In addition, it is possible that some women received follow-up care elsewhere. Not all women with complications, however, will decide to seek follow-up care and, in very rare cases (eg, one out of 13 345 cases in our study), this can have serious consequences. Understandably, such cases can raise concerns about the safety of home-based versus clinic-based administration of misoprostol. Although misoprostol could be administered in a clinic under medical supervision, home-based administration of misoprostol has been associated with a low complication rate, is preferred by women, and is common practice in France and the US.1,22 Further, clinic-based administration of misoprostol would not necessarily enhance outcomes as medical abortion complications can occur hours or days after misoprostol is administered.

Overall, there was a high level of satisfaction with the mifepristone–buccal misoprostol regimen used, with most women who completed the study questionnaire reporting the bleeding, pain/cramps, and overall experience to be as they had expected or better than expected. Satisfaction with the EMA method used in our study was similar to that reported in other studies.1,23,24 Notably, there has been a significant uptake of medical abortion in our clinics, with about a third of women requesting an abortion up to 63 days of gestation opting for the medical method (data not shown).

Our mifepristone–buccal misoprostol EMA method could be used outside a hospital setting. The diagnosis and initiation of treatment (mifepristone) can be conducted in a clinic and the second part of the treatment (misoprostol) can be carried out by the woman at home, with 24-hour access to after-care services to manage any issues. This arrangement, along with the 24–48-hour window for misoprostol administration, allows the abortion process to take place in the privacy of the woman’s home and at the time of her choosing, factors known to be of importance to women undergoing an abortion.1,23,24 The clinic- and home-based nature of the mifepristone–buccal misoprostol regimen makes it feasible for use in regional centres of Australia. These centres may have smaller hospitals than urban areas, but have the facilities and staff to manage spontaneous miscarriage.24 The MSIA clinics are located in major metropolitan areas and women are provided with discharge letters. Any woman contacting the after-care service can be referred back to the MSIA clinic or to her local doctor or, if necessary, a nearby accident and emergency facility. In clinical practice, proximity to emergency services should be evaluated when considering the use of the mifepristone–buccal misoprostol regimen. The potential need for access to appropriate 24-hour emergency help could limit the use of this regimen in more remote areas of Australia.

The observational nature of our study allowed assessment of the mifepristone–buccal misoprostol regimen in a real-world setting. However, as discussed, one of the study limitations was the relatively high rate of loss to follow-up. As a result, it was not possible to accurately determine the complete abortion rate; instead the known failure rate was reported. In addition, the use of multiple databases for data collection led to some small discrepancies in the datasets used for the analyses (eg, different total numbers of EMAs), because of differences in how the data were collected.

In conclusion, our Australian study of 13 345 EMAs has shown that mifepristone followed by buccal misoprostol has a low failure rate, including a low ongoing pregnancy rate. The regimen, which involved self-administration of misoprostol and completion of the abortion process at home, had a favourable safety profile and was well accepted by the women in this study. While the potential risk of serious infection should be kept in mind and monitored, these results indicate that the mifepristone–buccal misoprostol regimen is an effective option for Australian women seeking an abortion up to 63 days of gestation.

5 Results of patient survey on the early medical abortion method, April 2010 – August 2011

Women were asked (i) to describe the severity of their bleeding (A; n = 6330) and pain/cramps (B; n = 6381); (ii) whether the bleeding (C; n = 6233), pain/cramps (D; n = 6340), and overall experience (E; n = 6265) were better than expected, as expected, or worse than expected; (iii) whether they would choose the method again (F; n = 6335); and (iv) whether they would recommend the method to a friend (G; n = 5196).