Two mifepristone doses and two intervals of misoprostol administration for termination of early pregnancy: a randomised factorial controlled equivalence trial


Dr H von Hertzen, UNDP/UNFPA/WHO/World Bank Special Programme of Research, Development and Research Training in Human Reproduction, Department of Reproductive Health and Research, WHO, 1211 Geneva 27, Switzerland. Email


Objective  To compare the efficacy of 100 mg and 200 mg of mifepristone and 24- and 48-hour intervals to administration of 800μg vaginal misoprostol for termination of early pregnancy.

Design  Placebo-controlled, randomized, equivalence trial, stratified by centre.

Setting  13 departments of obstetrics and gynecology in nine countries.

Population  2181 women with 63 days or less gestation requesting medical abortion.

Methods  Two-sided 95% CI for the risk differences of failure to complete abortion were calculated and compared with 5% equivalence margin between two doses of mifepristone and two intervals to misoprostol administration. Proportions of women with adverse effects were compared between the regimens using standard testes for proportions.

Outcome measures  Rates of complete abortion without surgical intervention and adverse effects associated with the regimens.

Results  Efficacy outcome was analysed for 2126 women (97.5%) excluding 55 lost to follow up. Both mifepristone doses were found to be similar in efficacy. The rate of complete abortion was 92.0% for women assigned 100 mg of mifepristone and 93.2% for women assigned 200 mg of mifepristone (difference 1.2%, 95% CI: –1.0 to 3.5). Equivalence was also evident for the two intervals of administration: the rate of complete abortion was 93.5% for 24-hour interval and 91.7% for the 48-hour interval (difference −1.8%, 95% CI: –4.0 to 0.5). Interaction between doses and interval to misoprostol administration was not significant (P = 0.92). Adverse effects related to treatments did not differ between the groups.

Conclusions  Both the 100 and 200 mg doses of mifepristone and the 24- and 48-hour intervals have a similar efficacy to achieve complete abortion in early pregnancy when mifepristone is followed by 800 micrograms of vaginally administered misoprostol.


The antiprogestin mifepristone followed by a suitable prostaglandin analogue is registered for termination of early pregnancy in more than 35 countries. In Europe and the USA, the approved regimen consists of 600 mg of mifepristone, followed 36–48 hours later by a prostaglandin analogue, most often misoprostol.

The data from several multicentre trials conducted over the past 15 years demonstrate that a single dose of 200 mg of mifepristone has a similar efficacy to achieve complete abortion as the 600 mg dose.1–3 In addition, administration of multiple small doses (five doses of 25 mg given at 12-hour intervals), which is routine in China, was not significantly different from a single dose of 600 mg.4 On the other hand, the efficacy decreases when the dose of mifepristone is lowered to 50 mg.5 In a study on pharmacokinetics, oral administration of 100 mg or higher single doses of mifepristone resulted in similar serum concentrations.6 Similar serum levels can be explained by saturation of the binding capacity of the carrier protein as mifepristone binds to the serum transport protein alfa-1-acid glycoprotein, which limits its availability to tissues. No previous studies, however, compared the 100 mg dose with the commonly used dose of 200 mg of mifepristone.

Mifepristone pre-treatment leads to the softening and dilation of the uterine cervix and it increases the sensitivity of the uterine muscle to prostaglandins. It has been demonstrated that the maximal sensitivity is achieved when mifepristone is administered 36–48 hours before the prostaglandin analogue,7 and this is also the time interval to achieve maximal softening and dilation of the uterine cervix.8 Consequently, the 36- to 48-hour interval was adopted as a standard for the medical abortion regimen. Also, the data collected on 16 000 women who had been treated with 600 mg of mifepristone followed by a prostaglandin analogue 24–72 hours later suggested that the efficacy was highest if the interval was 48 hours.9

Different time intervals between mifepristone and misoprostol have also been tested10 or both drugs have been administered simultaneously,11 repeating the dose of misoprostol if abortion was not complete. When misoprostol was not repeated, complete abortion rates seem to be significantly lower with shorter intervals.12 No previous studies had been undertaken with a 100 mg dose of mifepristone when the interval to misoprostol is 24 hours.

Based on the pharmacokinetics of mifepristone and the published reports, we aimed to demonstrate that 100 mg has a similar efficacy to 200 mg and that the 24- and 48-hour intervals between mifepristone and misoprostol result in an equivalent efficacy when using 800 micrograms of misoprostol administered vaginally (the null hypothesis being that there is a difference in efficacy greater than a prestated margin of 5%). A 24-hour interval would be more practical than 48-hours and, if found effective, the 100 mg dose could be cheaper than 200 mg, let alone the registered dose of 600 mg. This is particularly important as the relatively high price of mifepristone has limited its use in some countries despite its major advantages as a pre-treatment.

We launched this randomised equivalence trial to test our hypotheses for pregnancy termination among women with gestational age of 63 days or less. Our main outcome was complete abortion. We also studied adverse effects associated with the use of these four regimens and the timing of expulsion. In addition, women were asked questions about their perceptions of the treatments.



The trial was undertaken in 13 departments of obstetrics and gynaecology of teaching hospitals in Beijing and Shanghai, People’s Republic of China; Hong Kong, Special Administrative Region of China, Szeged, Hungary; New Delhi, India; Ulaanbaatar, Mongolia; Targu Mures, Romania; Ljubljana, Slovenia; Johannesburg, South Africa; Hanoi (two hospitals) and Ho Chi Minh City, Viet Nam; and Novi Sad, Serbia. Institutional review boards at each of the participating hospitals and the WHO Secretariat Committee on Research on Human Subjects gave ethics approval.

Women requesting early termination of pregnancy were given information about the study, screened for eligibility by clinic personnel if willing to participate and included if they were healthy, older than the age of legal consent, had haemoglobin level ≥100 g/l, intrauterine pregnancy with duration ≤63 days verified by ultrasound, agreed to return for follow-up visit(s) and agreeable, in principle, to surgical termination of pregnancy should the treatment fail.

We excluded from this study women who had any indication of serious past or present illness, those allergic to mifepristone or misoprostol, chronic adrenal failure, severe asthma uncontrolled by corticosteroid therapy, inherited porphyria, heavy smokers (>20 cigarettes a day) with another risk factor of cardiovascular disease, a history or evidence of mitral stenosis, glaucoma, diastolic blood pressure >90 mmHg, systolic blood pressure <90 mmHg, history or evidence of tromboembolism, evidence of liver disease, presence of an IUD in utero, breastfeeding or haemolytic disorders. All participants gave written informed consent before entering the study. Relevant medical, gynaecological and obstetric histories were recorded and bacteriological tests and Rh-typing performed according to the routine of the centre.

Randomisation and interventions

A computer-generated randomisation sequence was produced by WHO staff in Geneva to assign participants within each centre to one of the four dose-interval combinations: 100 mg of mifepristone and 800 micrograms of misoprostol 24 hours later, 100 mg of mifepristone and 800 micrograms of misoprostol 48 hours later, 200 mg of mifepristone and 800 micrograms of misoprostol 24 hours later and 200 mg of mifepristone and 800 micrograms of misoprostol 48 hours later. Mifepristone was administered orally and misoprostol vaginally.

Each centre received assignments by randomly permuted blocks with a fixed block size of 8. Randomisation was also stratified by the gestational age (up to 49, 50–56 and 57–63 days). Allocation was concealed by the use of sealed, opaque, sequentially numbered envelopes, which were filled and labelled in accordance with the list of randomisation for each centre by Magistra, Geneva, Switzerland. The 100 mg dose of mifepristone consisted of one tablet of 100 mg and one placebo tablet and the 200 mg dose consisted of two tablets of 100 mg of mifepristone (Hualian Pharmaceuticals Ltd, Shanghai, China). Placebo tablets were of similar shape and colour as mifepristone tablets. The interval to misoprostol (Cytotec; Monsanto, Paris, France) administration was not blinded.

To establish the equivalent efficacy of the two doses of mifepristone (100 and 200 mg) within each interval of misoprostol administration or of the two intervals between mifepristone and misoprostol (24 and 48 hours) within each dose of mifepristone, we required the 95% CI for the difference in complete abortion rates to be within the margin of equivalence of 5% with a probability of 80%. This margin was chosen using clinical judgement. If complete abortion rates by the two doses (or with the two intervals) are both equal to 96%, about 323 subjects will be required in each group, that is a total of about 1300 subjects for the whole study. Allowing for 10% of lost to follow up, at least 1450 subjects would have to be recruited. In practice, 14 centres wished to participate, each of them willing to recruit 156 women, that is a total of 2184 women for the whole study.

The primary outcome measure was efficacy of the treatment in inducing complete abortion. Complete abortion was defined as passing of the products of conception without surgical intervention during the follow-up period; incomplete abortion as expulsion of fetus but some products of conception remaining in uterus needing evacuation; missed abortion as gestational sac in uterus without cardiac activity on ultrasound examination, needing emptying of the uterus; continuing pregnancy as growing gestational sac with fetal heart activity. If the woman discontinued the treatment and requested vacuum aspiration before misoprostol administration or soon after it before the outcome could be known or she was lost to follow up, the outcome was regarded as undetermined.

The efficacy was assessed at the follow-up visits 2 and 6 weeks after the start of treatment.

Signs and symptoms were recorded before mifepristone, before misoprostol and at 1, 2 and 3 hours after misoprostol administration at the clinic as well as daily until the first follow-up visit by women. Signs and symptoms were classified as pregnancy related, treatment related and those related to the abortion process itself.


The first dose of treatment (two tablets of 100 mg of mifepristone or one 100 mg of mifepristone tablet and one placebo tablet) was administered at the clinic. Women returned either 24 or 48 hours later for vaginal administration of four tablets of 200 micrograms of misoprostol each, after which they stayed at the clinic for 3 hours under supervision, during which period signs and symptoms were recorded hourly. When leaving the clinic, the women were asked to keep a diary about adverse effects and bleeding until the follow-up visit.

The initial clinical judgement about the outcome was made at the first follow-up visit 2 weeks after the treatment. If the findings at pelvic examination were suggestive of a continuing pregnancy, an ultrasound examination was performed. In case of live pregnancy, vacuum aspiration was performed. If clinical findings were compatible with incomplete abortion, no further action was taken unless judged necessary by the investigator. For these women, the final outcome of therapy was made at the second follow-up visit on day 43 of the study.

Statistical analysis

Data were analysed with SAS® System statistical software (version 9.1.3) centrally at WHO. This trial was compliant with CONSORT guidelines for reporting equivalence trials.13 An independent data monitoring committee reviewed the data from two interim analyses performed after 509 and 1050 subjects were included in the study. According to the protocol, a decision to discontinue a treatment regimen due to low efficacy was based on the comparison with the regimen of 200 mg of mifepristone followed by 800 micrograms of misoprostol vaginally 48 hours later.

For the main analysis, women with unknown outcomes due to lost to follow up were excluded. Additionally, a per-protocol analysis excluding noncompliant cases was performed. For each treatment regimen, percentages and two-sided 95% CI were computed for complete abortion and failure to complete abortion, which included incomplete abortion, missed abortion, continuing live pregnancy and undetermined outcomes.

Interaction between mifepristone dose and interval of misoprostol administration was assessed adding an interaction term to a logistic regression model including dose and interval effects. Equivalence of mifepristone doses was assessed within each interval of misoprostol administration and also combining both intervals. Two-sided 95% CI for the risk differences of failure to complete abortion were calculated by standard methods and compared with the equivalence margin (5%). Similarly, equivalence of intervals of misoprostol administration was assessed within each mifepristone dose and combining the doses. Analyses adjusting by centre and baseline characteristics were carried out using a logistic regression model. A stratified analysis by gestational age was carried out together with comparisons of adverse effect occurrence among the four regimens.

The role of funding source

The UNDP/UNFPA/WHO/World Bank Special Programme of Research, Development and Research Training in Human Reproduction funded this study including misoprostol tablets. Mifepristone and placebo tablets were donated by Hualian Pharmaceuticals Ltd. Neither the donors and sponsors of the Programme nor Hualian Pharmaceuticals Ltd had a role in the study design, data collection, data analysis, data interpretation, writing of the report or the decision to submit the paper for publication. The corresponding author had full access to all final data in the study and had the final responsibility for the decision to submit for publication.


A total of 2181 women were enrolled in the study. Recruitment started in March 2003 and finished in May 2005. Baseline characteristics were similar among the four treatment groups (Table 1). The mean age was about 26 years old. For 43% of the women, this was their first pregnancy and 47% of them were using some form of contraception at the time the pregnancy started.

Table 1.  Baseline characteristics by treatment group
 Mifepristone 100 mg + misoprostol 24 hours (n= 545)Mifepristone 100 mg + misoprostol 48 hours (n= 547)Mifepristone 200 mg + misoprostol 24 hours (n= 544)Mifepristone 200 mg + misoprostol 48 hours (n= 545)
  • Data are n (%) or mean (SD), unless otherwise indicated.

  • *

    Gestational age assessed by ultrasound.

Demographic and physical
Age (years)26.2 (5.7)26.2 (5.7)26.4 (5.7)26.1 (5.4)
Weight (kg)55.3 (10.9)54.7 (10.3)54.4 (10.1)55.0 (11.5)
Haemoglobin (g/L)122.1 (10.6)122.1 (10.0)122.6 (10.7)122.3 (10.5)
Ethnic group
Chinese156 (29)156 (29)155 (28)155 (28)
Non-Chinese Asian or black214 (39)216 (39)213 (39)213 (39)
White175 (32)175 (32)176 (33)177 (33)
Obstetric and gynaecological history
Nulliparity329 (60)315 (58)311 (57)324 (59)
Previous abortion187 (34)205 (37)224 (41)216 (40)
Gestational age* (days)
<49228 (42)241 (44)244 (45)245 (45)
50–56164 (30)165 (30)147 (27)164 (30)
57–63152 (28)141 (26)151 (28)136 (25)
>631 (<1)0 (0)2 (<1)0 (0)
Median (interquartile range)51 (44–57)51 (44–57)51 (44–57)51 (44–56)

All 2181 women received mifepristone (Figure 1), after which 5 women opted for surgical termination and 1 woman did not return, so that 2175 women were administered misoprostol. A total of 2126 women completed the study and the 55 women considered lost to follow up were distributed evenly across three of the four treatment regimens. A total of 47 women discontinued their participation before the outcome could be assessed. The women with undetermined outcome (n= 102) were on average 25 years old and 34% of them were primigravidas.

Figure 1.

Flow chart.

The efficacy outcome was analysed for 2126 women including the 47 women with undetermined outcome who were considered as failures to achieve complete abortion. The numbers and percentages of women in each category are shown in Table 2. The interaction of mifepristone dose by interval of misoprostol administration was not significant for failure to complete abortion (P= 0.92). Both mifepristone doses produced equivalent rates of failure to achieve complete abortion within each interval of misoprostol administration and overall for the two intervals combined (Figure 2). Compared with the 200 mg of mifepristone, the rates of failure to complete abortion were slightly higher in the 100 mg groups but the CI for risk difference fell entirely within the equivalence limits (difference 1.2%, 95% CI: −1.0 to 3.5).

Table 2.  Efficacy outcome by treatment group
OutcomeMifepristone 100 mg + misoprostol 24 hours (n= 529)Mifepristone 100 mg + misoprostol 48 hours (n= 534)Mifepristone 200 mg + misoprostol 24 hours (n= 531)Mifepristone 200 mg + misoprostol 48 hours (n= 532)Total
  1. Data are n (%, 95% CI) in different outcome categories.

  2. Excluding lost to follow up (subjects analysed: 2126, subjects excluded: 55).

Complete abortion492 (93.0, 90.5–95.0)486 (91.0, 88.3–93.3)499 (94.0, 91.6–95.8)492 (92.5, 89.9–94.6)1969 (93)
Failure to complete abortion
Total37 (7.0, 5.0–9.5)48 (9.0, 6.7–11.7)32 (6.0, 4.2–8.4)40 (7.5, 5.4–10.1)157 (7)
Incomplete abortion17 (3.2, 1.9–5.1)21 (3.9, 2.5–6.0)17 (3.2, 1.9–5.1)27 (5.1, 3.4–7.3)82 (4)
Delayed abortion3 (0.6, 0.1–1.7)3 (0.6, 0.1–1.6)1 (0.2, 0.0–1.0)0 (0.0)7 (<1)
Continuing pregnancy6 (1.1, 0.4–2.5)7 (1.3, 0.5–2.7)4 (0.7, 0.2–1.9)4 (0.7, 0.2–1.9)21 (1)
Undetermined11 (2.1, 1.0–3.7)17 (3.2, 1.9–5.1)10 (1.9, 0.9–3.4)9 (1.7, 0.8–3.2)47 (2)
Figure 2.

Risk differences of failure to achieve complete abortion.

When the intervals of misoprostol administration were compared, combining mifepristone doses, both intervals were equivalent within a 5% margin. The 24-hour interval tended to have lower failure rates than the 48-hour interval (difference −1.8, 95% CI: −4.0 to 0.5). Adjustment for centre and baseline characteristics (age, weight, haemoglobin level, parity, previous abortion and gestational age) produced almost the same results (data not shown).

Noncompliance with treatment was 5.3%. Almost all cases of noncompliance were due to misoprostol administration more than 2 hours outside the assigned time interval. The conclusions do not change when noncompliant cases are excluded (per-protocol analysis) or when all randomised subjects (n= 2181) are included in the analysis considering lost to follow up as failures.

The stratified analysis of the rate of failure to complete abortion by gestational age shows that both doses and both administration intervals are equivalent when the gestational age is 49 days or less (Table 3). The results are inconclusive when the gestational age is 50 days or more: the trial was not powered to show equivalence in subgroups.

Table 3.  Failure to achieve complete abortion by treatment group and gestational age
 n/N (%)n/N (%)Risk difference (%, 95% CI)
  1. Excluding lost to follow up and gestational age more than 63 days (subjects analysed: 2123, subjects excluded: 58 [55 lost to follow up, 3 gestational age >63 days]).

Comparison: mifepristone 100 versus 200 mg100 mg200 mg 
Overall85/1062 (8.0)72/1061 (6.8)1.2 (−1.0–3.5)
49 days or less30/459 (6.5)25/482 (5.2)1.4 (−1.7–4.4)
50–56 days27/321 (8.4)28/303 (9.2)−0.8 (−5.3–3.6)
57–63 days28/282 (9.9)19/276 (6.9)3.1 (−1.6–7.6)
Comparison: misoprostol 24 versus 48 hours24 hours48 hours 
Overall69/1057 (6.5)88/1066 (8.3)−1.8 (−4.0–0.5)
49 days or less30/465 (6.5)25/476 (5.3)1.2 (−1.8–4.2)
50–56 days20/300 (6.7)35/324 (10.8)−4.1 (−8.5–0.3)
57–63 days19/292 (6.5)28/266 (10.5)−4.0 (−8.7–0.6)

Among 2051 women with complete and incomplete abortion timing of expulsion was assessed for 1848 women (90%). The median time to expulsion was 28 hours for regimens with 24-hour interval to misoprostol administration and between 51 and 52 hours in regimens with 48-hour intervals. Therefore, it would appear that expulsion occurred in all groups approximately 4 hours after misoprostol administration. No differences were observed in time from misoprostol administration to expulsion between the four regimens.

The percentage of women who reported pregnancy-related symptoms (nausea and vomiting) and lower abdominal pain after mifepristone administration and before misoprostol was higher in the 48 hours of misoprostol administration interval than in the 24-hour interval (Table 4). After misoprostol administration, the four groups did not differ significantly in terms of pregnancy-related symptoms.

Table 4.  Symptoms and adverse effects
 Mifepristone 100 mg + misoprostol 24 hoursMifepristone 100 mg + misoprostol 48 hoursMifepristone 200 mg + misoprostol 24 hoursMifepristone 200 mg + misoprostol 48 hoursP value
n (%)n (%)n (%)n (%)
Between mifepristone and misoprostol235 (43.1)289 (52.9)237 (43.6)298 (55.2)< 0.0001
Within 3 hours after misoprostol200 (36.7)194 (35.5)204 (37.5)212 (39.3)0.6203
After misoprostol up to first follow-up visit63 (12.1)58 (11.1)49 (9.3)48 (9.2)0.3519
Between mifepristone and misoprostol87 (16.0)126 (23.1)74 (13.6)129 (23.9)< 0.0001
Within 3 hours after misoprostol83 (15.2)86 (15.8)73 (13.4)86 (16.0)0.6380
After misoprostol up to first follow-up visit24 (4.6)29 (5.5)15 (2.9)18 (3.5)0.1279
Lower abdominal pain
Between mifepristone and misoprostol159 (29.2)212 (38.8)163 (30.0)197 (36.5)0.0008
Within 3 hours after misoprostol511 (93.8)513 (94.0)521 (95.8)510 (94.4)0.4626
After misoprostol up to first follow-up visit273 (52.3)256 (48.8)274 (52.0)234 (44.9)0.0590
Between mifepristone and misoprostol16 (2.9)24 (4.4)19 (3.5)21 (3.9)0.6233
Within 3 hours after misoprostol74 (13.6)67 (12.3)89 (16.4)61 (11.3)0.0805
After misoprostol up to first follow-up visit43 (8.2)41 (7.8)45 (8.5)26 (5.0)0.1080
Between mifepristone and misoprostol73 (13.4)80 (14.7)69 (12.7)98 (18.2)0.0546
Within 3 hours after misoprostol67 (12.3)68 (12.5)51 (9.4)79 (14.7)0.0678
After misoprostol up to first follow-up visit86 (16.5)93 (17.7)95 (18.0)92 (17.7)0.9175
Within 3 hours after misoprostol25 (4.6)32 (5.9)34 (6.3)30 (5.6)0.6620
After misoprostol up to first follow-up visit11 (2.1)16 (3.1)19 (3.6)11 (2.1)0.3556
Within 3 hours after misoprostol115 (21.1)119 (21.8)125 (23.0)103 (19.1)0.4662

Lower abdominal pain increased markedly after misoprostol administration. About 20% of the women reported pain at admission, 72% 1 hour after misoprostol, 86% 2 hours after and 88% 3 hours after, however, no differences were observed between treatment groups. At the first follow-up visit about 50% of women reported having had some lower abdominal pain since the last visit. No differences were observed between the groups for diarrhoea, chills/shivering and headache.

About 61% of the women had experienced less pain and 55% fewer adverse effects than they had expected. When women were asked after abortion, where they would prefer to have medical abortion if they need another one, about 77% said that they would prefer to have it at the clinic rather than at home.

There were 282 unscheduled visits during the study, 86 of them (31%) due to lower abdominal pain as the main reason, while 80 visits (28%) were due to excessive vaginal bleeding.

At admission to the study, 80 women (3.7%) had signs of abnormal vaginal discharge. Tests for gynaecological infections were performed for 168 women (7.7%) and found positive in 16 of them, most having candida or nonspecific vaginal infection. A total of 176 (8.1%) women received antibiotic treatment between admission and first follow-up visit, 149 of them at one site, where women were given antibiotics routinely after misoprostol visit, and the rest to treat suspected or verified vaginal infection; 43 (2.0%) women received antibiotics between the two follow-up visits and two of them were diagnosed with pelvic inflammatory disease with retained tissue. Eleven cases of heavy bleeding were reported as serious adverse events; two women required blood transfusion and one woman was given plasma expanders. In all, six women were hospitalised for treatment.


Our findings show a similar efficacy for achieving complete abortion in early pregnancy within a 5% margin, for 100 and 200 mg doses of mifepristone when followed by 800 micrograms of vaginal misoprostol and for 24- and 48-hour intervals between these drugs when the gestational length is 63 days or less.

Despite similar pharmacokinetics, no studies have compared the efficacy of 100 and 200 mg doses in medical abortion regimen. The 100 mg dose of mifepristone was used in a study14 that compared oral and vaginal misoprostol. All 40 women who were randomised to receive 800 micrograms of misoprostol vaginally 48 hours after mifepristone had complete abortion, while complete abortion rate was 85% among women who received 400 micrograms of misoprostol orally and pregnancy continued in four (10%) of these women. This study suggests that 100 mg dose may not be enough if the dose of misoprostol is insufficient and/or effective uterine contractions fail to be produced by the route it is administered.15 When mifepristone dose was lowered to 50 mg in a regimen with 1 mg of gemeprost, complete abortion rate was 89.8%, while it was 92.9% in the 200 mg group.5 The rate of continuing pregnancies was also higher (2.2 versus 0.6%) in the 50 mg group.

Based on available evidence about the time interval to maximal effect of mifepristone,7 one would have expected a better efficacy after 48-hour interval compared with 24 hours. The effect on uterine contractility is a balance between the degree of sensitivity and the dose of prostaglandin. If we shorten the interval, we are likely to need a higher dose of prostaglandin and a route of administration that leads to effective contractility, especially when the gestational age increases. We used one single dose of 800 micrograms of misoprostol vaginally. This high dose of misoprostol may have masked the effect of the interval. With lower vaginal doses or with oral administration of misoprostol or in later pregnancies, we might not have obtained similar results. This was demonstrated in a second-trimester abortion study16 in which the time until abortion occurred was significantly longer (P < 0.0001) in the 1-day interval group compared with 2-day interval between 200 mg of mifepristone and the start of administration of vaginal misoprostol (400 micrograms at 3-hour intervals). Consequently, the total dose of misoprostol given to women was higher in the 1-day interval group.

A previous study that compared 1-, 2- and 3-day intervals between 200 mg of mifepristone and 800 micrograms of vaginal misoprostol also reported similar rates of complete abortion for both 1- and 2-day intervals.10 However, a second dose of misoprostol was administered if abortion was not complete, and it is not mentioned whether the need for additional dose(s) was similar in all groups. Shorter than 24-hour intervals between mifepristone and misoprostol have also been tested or both drugs have been given at the same time. It seems common for these studies that additional doses of misoprostol are given to women who either do not abort or do not have complete abortion. As abortion can also be induced by multiple doses of misoprostol without mifepristone,17 one may need to be cautious when drawing conclusions about the efficacy of simultaneous administration when additional misoprostol doses are given. One could question whether these studies really compare different intervals if insufficient success has to be compensated by additional misoprostol doses to improve the results. When misoprostol is administered at the same time with mifepristone and the dose needs to be repeated even in early first trimester, the benefits of mifepristone are partly lost and it is not clear whether women abort with the combined regimen or multiple doses of misoprostol.

The rates of complete abortion in all four arms of our study were lower than in our previous studies.18 This is likely to be due to the fact that several of the participating centres do not provide medical abortion routinely because mifepristone has not been licensed in their country. Experienced providers report higher complete abortion rates than those with less experience.19 The comparison, however, has internal validity as randomisation was within centres. Rates of continuing pregnancies were low in all arms. Whether pregnancies continue or not depends on the efficacy of the regimen and the experience of providers have little, if any, influence on it.

A higher percentage of women in the 48-hour interval groups reported pregnancy-related symptoms such as nausea and vomiting after mifepristone administration than women in the 24-hour groups. This may be related to the fact that as the time interval is longer, there is more chance to experience these symptoms. This was also the case for lower abdominal pain. As mifepristone sensitises the uterus also for endogenous prostaglandins, uterine contractility is likely to increase with time and can be experienced as painful. As the 24-hour interval was associated with lower rates of these symptoms, it is reasonable to recommend it, given its advantages and its trend to a better efficacy.

The rate of suspected or verified infections in this study was low. There has been discussion about whether routine prophylactic antibiotic treatment is necessary when using medical abortion. The low rate of gynaecological infections after treatment in this study does not support this practise.

This equivalence trial was designed with a prestated 5% margin for the difference in percentage of women achieving complete abortion. This margin was determined using clinical assessment and it implies that a smaller difference is not clinically relevant. Using a smaller margin would have implied conducting a larger trial.

This trial has strong internal validity with a clear research question and clearly defined primary and secondary outcome measures. The sample size was calculated according to the prestated hypothesis, randomisation and concealment of the random allocation were used and both as randomized and per-protocol analyses were performed (as recommended for equivalence trials).

This trial also has external validity as it enrolled women of several different populations and included centres with different levels of experience with medical abortion. Thus, the results are likely to apply to wider populations.

In conclusion, our results show that for the termination of early pregnancy (≤63 days’ of gestation), mifepristone dose could be lowered to 100 mg and the administration interval shortened to 24 hours when using vaginal misoprostol of 800 micrograms.

Contributions to authorship

H.v.H., in collaboration with the members of the Steering Committee, was responsible for the conception of the trial and selection of centres. H.v.H.and G.P. prepared the protocol. O.S.T., A.H.F., S.C.W., L.K., S.M., R.E., M.H., A.P.-D., A.K., K.D., N.D.A., N.V.T., and H.T.D.T. contributed to the final trial protocol and implemented the trial in their respective countries. H.v.H., G.P. and L.M. supervised the trial. G.P. and D.W. were responsible for the statistical analysis and N.T.M.H. and A.P. for the data management. H.v.H., G.P. and D.W. wrote the paper with inputs from the investigators.

WHO Research Group for the trial

The full list can be found in the online supporting information.

Country collaborators

The full list can be found in the online supporting information.

Conflict of interest statement

Overall responsibility for this paper lies with Dr H.v.H. of the UNDP/UNFPA/WHO/World Bank Special Programme of Research, Development and Research Training in Human Reproduction, WHO. Neither she nor the Special Programme or any of its cosponsors have a conflict of interest.

Ethics approval

Institutional review boards at each of the participating hospitals and the WHO Secretariat Committee on Research on Human Subjects gave ethics approval.


The study was funded by the UNDP/UNFPA/WHO/World Bank Special Programme of Research, Development and Research Training in Human Reproduction. We thank the members of the Data and Safety Monitoring Board for their contribution to interim analyses and their valuable comments on the manuscript. Grateful thanks are also due to P Van Look for his critical review of the manuscript.