Expanding contraceptive options: A scoping review of medically approved contraceptive methods that are not in the WHO Medical Eligibility Criteria

Abstract

Objectives

The World Health Organization’s (WHO) Medical Eligibility Criteria for contraceptive use (MEC) provides recommendations on the safety and efficacy of a range of contraceptive methods for users with various medical conditions. However, there are methods with medical regulatory agency approval that are not included in WHO guidance. Our objectives were to evaluate the evidence on methods that have regulatory approval for use in any member state, but which are not included in the MEC 2015, and to determine whether the evidence should be systematically reviewed to inform possible inclusion in a new WHO guideline.

Study design

We conducted a scoping review of experimental studies found through two bibliographic databases (PubMed, Embase) from database inception to December 2023. Two reviewers screened abstracts and full texts to determine eligibility based on a priori inclusion and exclusion criteria; one additional reviewer resolved conflicts. Reviewers extracted data on method safety, efficacy, and acceptability using a standardized tool.

Results

We identified three methods that met our review criteria: ormeloxifene, quinestrol-containing contraception, and mifepristone for emergency contraception (EC). Our initial search strategies yielded 386 total results. Five publications related to ormeloxifene, five to quinestrol-containing contraception, and 10 to mifepristone for EC were included. The data on these methods was highly variable.

Conclusion

Rigorous systematic reviews of the evidence on safety, efficacy and dosage of ormeloxifene, quinestrol-containing contraception, and mifepristone for EC are needed to inform WHO recommendations on these methods. As these methods are already being used, international guidance for practitioners is essential.

1. Introduction

The World Health Organization’s (WHO) Medical Eligibility Criteria for contraceptive use (MEC) provides recommendations on the safety and efficacy of contraceptive methods based on patients’ medical conditions and medically relevant characteristics [1]. Globally, there are 1.9 billion women of reproductive age, of whom at least 164 million wish to prevent or delay childbearing and have unmet needs for contraception [2]. For many family planning programs, particularly those in low- and middle-income countries (LMICs), the MEC and its derivative tools are very influential in guiding contraceptive decision-making.

For a contraceptive method to be recommended for inclusion in the MEC and Selected Practice Recommendations (SPR), it must first go through a series of extensive developmental and approval processes, including concept creation based on reproductive physiology, and subsequent phase I through III clinical trials. In addition, it must be used in multiple WHO member countries, have approval from a recognized stringent regulatory authority, and have the potential for high impact. While this robust process offers reassurance of a method’s safety and efficacy, there are contraceptive methods that are not in the MEC, and yet are being utilized in some WHO member states. These methods are not included in the MEC and SPR because their use is limited to specific regions and/or they have not yet passed through the aforementioned stringent regulatory processes. At present, no WHO guidelines exist on these methods, and recommendations from WHO for or against their use may be warranted in order to offer guidance on safety and efficacy [3], [4].

In this scoping review we aimed to evaluate the range of evidence on methods that have (1) regulatory approval for use in any member state, and (2) are not included in MEC 2015. We sought to identify gaps in the literature, and to determine whether the evidence on methods identified should be recommended for systematic review to inform whether these methods can be included in WHO guidelines.

2. Material and methods

2.1. Contraceptive method identification

In order to ascertain which non-MEC methods are in use, we first developed and administered a member state survey (Supplementary Material). This survey was developed in English and translated into French and Spanish, and disseminated amongst partners in all 194 member states as well as to various partner groups including the International Federation of Gynecology and Obstetrics, International Planned Parenthood Foundation, Implementing Best Practices Network, and International Consortium of Midwives, using Research Electronic Database Capture (REDCap, Vanderbilt University, Version 14). The international partner survey yielded responses from 958 individuals in 141 member states across six continents, who suggested inclusion of ormeloxifene (also called centchroman) and Andro-switch, the latter of which was ultimately confirmed as not being registered for use in any WHO member state and therefore not included.

We furthermore received suggestion of methods to include from subject matter experts, based on patient care experience as well as member state inquiries to the WHO, which was the source of inclusion of the quinestrol-containing monthly pill. Mifepristone as emergency contraception (EC) was included as a result of the literature search process. We used the results of the survey, expert recommendation, member state inquiries, and the literature search to inform the list of contraceptive methods included in our scoping review search strategy. We did not search for methods through member state regulatory bodies as many countries do not have stable or searchable regulatory systems [5].

2.2. Protocol and registration

We publicly registered our protocol for this scoping review with Open Science Framework (https://osf.io/h9nj3).

2.3. Inclusion criteria

We included in this scoping review published peer-reviewed research from PubMed and EMBASE database inception including quantitative, qualitative, and mixed-methods studies investigating the efficacy, safety, and/or acceptability of the included contraceptive methods that were not included in the WHO MEC but were registered for use in at least one WHO member country. Studies were only included if they assessed a registered contraceptive dosage for the indication of pregnancy prevention. Our population of interest was reproductive-aged women, and outcomes of interest included efficacy, safety, and acceptability of the contraceptive method. Country registration was assessed using expert knowledge from international partners, publicly posted information on internet sites regarding country registration, and an international survey administered to WHO partners using REDCap. This review included studies written in non-English languages if a published English translation was available; no translation was undertaken by the study team.

2.4. Exclusion criteria

The review excluded studies concluded using animal models, studies focused exclusively on barrier method use, studies focused exclusively on the study of pharmacokinetics, studies about non-contraceptive uses and benefits of contraceptive methods (for example, acne management, treatment of abnormal uterine bleeding, management of mastalgia), and case reports.

2.5. Search strategy

We developed a list of key words, text words, Medical Subject Headings (MeSH) terms, and substance names with the assistance of a medical librarian to develop a comprehensive search strategy for selected methods in PubMed and Embase. Search terms included terms for the substance itself, and terms related to “contraception.” Our scoping approach was in accordance with the Arksey and O’Malley framework [6].

2.6. Study selection

All identified citations from each method’s search were uploaded into a systematic review manager (Covidence.org). Two independent reviewers (SM, SS) screened all titles and abstracts. The full texts of studies deemed potentially relevant were independently reviewed in detail (SS, SM). Reviewers recorded reasons for exclusion of full texts. A third reviewer (AK) resolved any conflicts in assessment at either of these two stages. The results of the search and the process of inclusion and exclusion are presented in a Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) flow diagram (Fig. 1, Fig. 2, Fig. 3).

Fig. 1.

PRISMA diagram for ormeloxifene scoping review. PRISMA = preferred reporting items for systematic reviews and meta-analyses.

Fig. 2.

PRISMA diagram for quinestrol scoping review. PRISMA = preferred reporting items for systematic reviews and meta-analyses.

Fig. 3.

PRISMA diagram for mifepristone scoping review. PRISMA = preferred reporting items for systematic reviews and meta-analyses.

2.7. Data extraction

Two independent reviewers (SS, SM) used a standardized form to extract data from papers included in the scoping review. We extracted study location, funding source, study question, years of data collection and publication, study design, intervention, control (if applicable), primary and secondary outcomes, inclusion and exclusion criteria, sample size, and results of efficacy, safety, and acceptability outcomes.

3. Results

3.1. Search outcomes

We identified three methods that met our review criteria: ormeloxifene, quinestrol-containing contraception, and mifepristone for emergency contraception. Our initial search strategies yielded 386 total results. Five publications related to ormeloxifene, five publications related to quinestrol-containing contraception, and 10 publications related to mifepristone for EC were included.

3.2. Ormeloxifene

Of the five studies that met inclusion criteria, one was a randomized control trial [7], while the remaining four were non-randomized experimental studies [8], [9], [10], [11]. Four studies assessed efficacy of ormeloxifene as a weekly contraceptive pill, four assessed its safety profile, and three included a measure of method acceptability. In addition, one study assessed regimen feasibility via proportion of doses missed at various timepoints. All studies were performed in India, and were published between 1977 and 2023. Two studies were funded by the Indian Ministry of Health (Table 1).

Table 1.

Summary of results of studies assessing the efficacy, safety, acceptability, and feasibility of ormeloxifene as a contraceptive

Study	Design	Purpose	Exposure	Outcomes	Efficacy	Safety	Acceptability	Feasibility
Chandra 1977 [7] India, single site n = 23	Randomized control trial	Assess maximum tolerated doses of ormeloxifene, assess for toxic effects	Single dose study - four tablets containing 5, 20, or 80 mg. Multi-dose study: 60 or 120 mg per day. Placebo-controlled.	Safety	Not reported	Safety tolerated to single dose of 320 mg, daily doses of 60 and 120 mg x 30 d.	Not reported	Not reported
Nityanand 1988 [8] India, multisite n = 648 Funded by Indian Ministry of Health	Non-randomized experimental	Assess contraceptive efficacy of ormeloxifene	30 mg weekly, and on first day of every subsequent menstrual period. No control.	Primary: contraceptive efficacy. Secondary: safety, acceptability	27 methods (4.2%) and 66 users (10.2%) failures. Efficacy 95.8% PI 3.7	Cycle delay (50%), scant periods (12%), GI effects (7%).	100% (3 mo), 88.3% (6 mo), 75% (12 mo) continuation	Not reported
Nityanand 1994 [9] India, multisite n = 377 Funded by Indian Ministry of Health	Non-randomized experimental	Assess contraceptive efficacy of ormeloxifene	30 mg biweekly for 3 mo, followed by weekly administration. No control.	Primary: contraceptive efficacy. Secondary: safety, acceptability	Six methods (1.6%) and 24 users (6.4%) failures. Efficacy 98.4% PI 1.83	Cycle delay (3.7%). No derangement of liver/renal labs, ovarian morphology.	7.6% drop-out over 13 mo	Not reported
Agarwal 2023 [10], India, multisite, n = 145	Non-randomized experimental study	Evaluate acceptability, safety, continuation rates of ormeloxifene as a post-abortal contraceptive pill	30 mg biweekly for 3 mo, followed by weekly administration. No control.	Primary: contraceptive efficacy. Secondary: safety, acceptability, feasibility	One method (0.69%) and One user failure. Efficacy 99.3% PI 1.67	Cycle prolongation/ delay (26%), amenorrhea (4%). GI effects (2%).	100% (3 mo), 95% (6 mo), 91% (12 mo) continuation. 88% user satisfaction and recommendation at the end of one year.	4%, 6.8%, and 19.3% of subjects reported missing medication doses at 3, 6, and 12 mo
Nair 2016 [11] India, single site n = 153	Non-randomized experimental	Assess contraceptive efficacy of ormeloxifene, & incidence of side effects	30 mg biweekly for 3 mo with barrier method for first month, followed by weekly administration. No control.	Primary: contraceptive efficacy. Secondary: safety, acceptability	Four methods (2.6%) and seven users (4.6%) failures. Efficacy 97.4% PI 2.0	Cycle delay (38%), cycle irregularity (12%), amenorrhea (7%). GI effects (<1%). No derangement of ovarian morphology.	100% (3 mo), 97% (6 mo), 85% (12 mo) continuation	Not reported

PI, Pearl Indices.

3.2.1. Efficacy of ormeloxifene as a weekly contraceptive

Four studies (non-randomized experimental studies) that were included in the scoping review examined the efficacy of ormeloxifene for contraception when administered as a weekly pill [7], [8], [9], [11]. Efficacy was calculated using the number of pregnancies resulting from method failure, with Pearl Indices (PI) calculated for all four. Pregnancies were otherwise characterized as resulting from user failure, denoting pregnancies which were conceived as a result of any non-adherence to the prescribed treatment schedule which was reported by the subject. The studies that assessed contraceptive effectiveness of ormeloxifene varied in regimen. A single study [8] included a dosing schedule of 30 mg weekly with an additional dose to be taken on the first day of every subsequent menstrual cycle, irrespective of the timing of the weekly tablet. Duration of use ranged from 1 to 53 months, with 75% of subjects using the method for at least 6–12 months. There was a total of 27 method-failure pregnancies, the majority (70.4%) occurring during the first 5 months of method use.

The dosing schedule for the remaining four studies dictated administration of 30 mg ormeloxifene twice a week for 12 weeks followed by 30 mg weekly. These studies followed subjects for a duration ranging from 1 to 27 months, and reported method failure pregnancy rates ranging from 0.69 to 2.6% with Pearl Indices from 1.67 to 2/100 WY.

3.2.2. Safety of ormeloxifene use

Four of the included studies (one randomized controlled trial and three non-randomized trials) assessed ormeloxifene’s safety profile when used as a weekly contraceptive. Reported side effects included menstrual disturbances, and general medical complaints (including headache, nausea, fatigue).

The most frequently reported side effect was menstrual disturbances, including cycle prolongation/delayed cycles (reported by four studies: 50% [8],38% [11], 26% [10], 3.7% [9] of participants), cycle irregularity (one study: 12% of participants) [11], scant periods lasting 1–2 days (one study: 12% of participants) [8], and amenorrhea (two studies: 4%, 7% of participants) [10], [11]. In all four studies that reported menstrual disturbances, this complaint was a chief reason for method discontinuation. In three of these studies [8], [9], [11], menstrual changes were noted at monthly in person follow-up interviews. In the fourth, they were noted on follow-up questionnaires administered to participants [10].

In the included randomized controlled trial, the prevalence of systemic complaints was similar between treatment and placebo groups, suggesting that symptoms reported may be more reflective of baseline prevalence than actual medication side effects [7]. In the three non-randomized experimental studies, gastrointestinal effects were reported by 7%, 2%, and 0.6% of subjects [8], [10], [11], postulated to be reflective of normal population incidence.

Laboratory and imaging studies were also performed in a subset of subjects in three studies. In one study [8], 16/648 and 122/648 subjects had platelet aggregation and lipid studies performed respectively, without any abnormalities noted. Another included assessment of liver and kidney function testing in 10% of subjects, all of which were within normal limits [9]. Two studies performed assessments of ovarian morphology in 18% and 17% of subjects respectively [9], [11], and found no significant abnormalities.

3.2.3. Acceptability of ormeloxifene

Five studies included information on the acceptability of ormeloxifene amongst subjects, all indicating high acceptability of the method for weekly contraception. Continuation rates at 3, 6, and 12 months were included in three studies [8], [10], [11] (Table 1). Another study reported drop out of 7.6% of subjects over a 13-month study period [9]. In a study by Agarwal and Dewan [10], 88% of subjects at the end of 1 year of use expressed satisfaction with the method and reported that they would recommend it to others. In all studies which reported on acceptability, menstrual disturbance was a leading cause of subject discontinuation, with desire for return to fertility as a secondary driver.

3.2.4. Feasibility of ormeloxifene

One study reported on the feasibility of ormeloxifene as a weekly contraceptive pill, describing rates of missed medication doses as 4%, 6.8%, and 19.3% at 3, 6, and 12 months respectively [10].

3.3. Quinestrol monthly pill

Five studies were identified that met inclusion criteria. Of these, all were non-randomized experimental studies, and all evaluated outcomes of efficacy, safety, and acceptability. The studies were performed between 1970 and 1974 (Table 2).

Table 2.

Summary of results of studies assessing the efficacy, safety, and acceptability of quinestrol/quingestanol acetate-containing pills as a contraceptive

Study	Design	Purpose	Exposure	Outcomes	Efficacy	Safety	Acceptability
Guiloff 1970 [12] Chile, single site n = 719	Non-randomized experimental	Assess efficacy, safety, and acceptability of Q/q as a monthly contraceptive pill	Q 2 mg on cycle day 1, Q 2 mg/q 5 mg on day 22, and then every 4 wk. No control.	Primary: contraceptive efficacy. Secondary: safety, acceptability	13 methods (2.6%) and 6 users (0.8%) failures. Efficacy 97.4% PI 2.1 (after first cycle)	No abnormality of endometrial biopsies, ovarian biopsies. Spotting (7−16%), lactational suppression (78%), nausea/vomiting (12.1% vs 3.2% at 1 and 6 mo), headache (14.5% vs 7.6% at 1 and 6 mo), mucorrhea (15% vs 0.4% at 1 and 6 mo).	69.4% discontinuation at 12 mo, driven by nausea, headache, and menstrual effects
Leguia 1973 [13] Peru, single site n = 1081	Non-randomized experimental	Assess efficacy, safety, and acceptability of Q/q as a monthly contraceptive pill	Q 2 mg/q 2.5 mg on cycle day 1, and then every 4 wk.a,bNo control.	Primary: contraceptive efficacy. Secondary: safety, acceptability	31 methods (2.8%), 10 users (0.9%) failures. Efficacy 97.1% PI 2.9	Spotting (11%), nausea/vomiting (15% vs 6% at 1 and 3 mo), thrombophlebitis (0.18%), jaundice (0.09%).	47.3% discontinuation at 12 mo
Hefnawi 1972 [14] Egypt, single site n = 600	Non-randomized experimental	Assess efficacy, safety, and acceptability of Q/q as a monthly contraceptive pill	Regimen A: Q 2 mg/unstated q dose on cycle day 25, and then every 4 wk Regimen B: Q 2 mg/unstated q dose on cycle day 1, second dose after 3 wk, then every 4 wk.aNo control.	Primary: contraceptive efficacy. Secondary: safety, acceptability	Eight methods (1.3%) and zero user failures. Efficacy 98.7% PI 2.8 (after first cycle)	No abnormality of pap tests, endometrial biopsies. Spotting (13.8% vs 0% at 1 and 12 mo), lactational suppression (56.6%), nausea/vomiting (30.4% vs 1.5% at 1 and 3 mo), headache (14.8% vs 2.1% at 1 and 6 mo), fatigue (11.6% vs 4% at 1 and 6 mo), mucorrhea (3% across all cycles).	80% discontinuation at 12 mo, 31.2% for nausea/vomiting, 5.4% for headache, 15.6% for menstrual disturbances.
Rubio 1972 [15] Country unspecified n = 256	Non-randomized experimental	Assess efficacy, safety, and acceptability of Q/q as a monthly contraceptive pill	Q 2 mg on cycle day 2, Q 2 mg/q 2.5 mg on day 22, and then every 4 wk.aNo control.	Primary: contraceptive efficacy. Secondary: safety, acceptability	Four methods (1.5%) and zero user failures. Efficacy 98.4% PI 0.83	No abnormality of endometrial biopsies, o, pap tests. Spotting (27%), nausea/vomiting (28% vs 7% at 1 and 3 mo), headache (0−5% over 24 cycles), mucorrhea (26 vs 2% at 1 and 12 mo), mastalgia (11% vs 0% at 1 and 6 mo).	5.5% discontinuation at 36 mo
Tejuja 1974 [16] India, multisite n = 666	Non-randomized experimental	Assess efficacy, safety, and acceptability of Q/q as a monthly contraceptive pill	Q 2 mg/q 2.5 mg on cycle day 1, 2, and 21, 10 every 4 wk.aNo control.	Primary: contraceptive efficacy. Secondary: safety, acceptability	27 total pregnancies. Efficacy 95.6% PI 1.9 at 3 mo, 5.9 at 1 y.	Spotting (12.5%).	85% discontinuation at 12 mo, 42.2% for drug-related reasons.

PI, Pearl Index; Q, quinestrol; q, quingestanol acetate.

^aAdditional contraceptive method use required during cycle 1.

^bq 300 mcg daily added to the regimen for the first month in 60% of subjects.

3.3.1. Efficacy of quinestrol/quingestanol as a monthly contraceptive pill

All five studies meeting inclusion criteria examined the efficacy of quinestrol/quingestanol combination pills for contraception when administered as a monthly pill. Efficacy was calculated using PI, with four out of five studies including details regarding the incidence of pregnancies due to method vs user failure (Table 2).

Various regimens were described in the studies we included in this review. Of note, all studies but one included a requirement that subjects use an additional form of contraception (sometimes assigned vs left to subject discretion) in the first month of study due to high risk of pregnancy in cycle 1 [12]. In addition to a secondary form of contraception, Leguia et al [13] added to their experimental regimen a daily dose of 300 mcg quingestanol during the first cycle. This change was made partway through the study due to four cycle 1 pregnancies, and was ultimately applied to 60% of the study sample.

The calculated effectiveness of these regimens was high; excluding cycle 1 pregnancies, there were a total of 83 pregnancies due to method failure over 33,050 menstrual cycles, for a cumulative effectiveness of 97.5%.

3.3.2. Safety of quinestrol/quingestanol use

All five studies assessed similar safety outcomes, including endometrial biopsy (EMB) assessments, changes to menstrual bleeding profile, effect on lactation for postpartum patients, and systemic effects such as GI symptoms (such as nausea/vomiting), headache, mastalgia, and mucorrhea.

Endometrial biopsy (EMB) was performed in a subset of patients in three studies [12], [14], [15]. Out of 897 EMBs performed across these studies on patients who had been exposed to the medication for at least 6 months, none demonstrated abnormality. Additional ovarian biopsies were performed on nine patients at time of tubal ligation [12], and found evidence of neither ovulation nor abnormality on histology. Two studies also performed cervical histological assessment in a fraction of patients, without any abnormalities noted [14], [15].

Lactational effects were reported on in two studies, both of which described a possible suppressive effect of the medication on milk production. In one study, of the 43% of subjects who were lactating at the start of treatment, 78% reported a decrease in breastmilk production by end of cycle 1 [12]. Similarly, of the 106 subjects who were lactating at enrollment in Hefnawi et al [14], 56.6% reported lactation stopped entirely after the first pill while an additional 23.4% of subjects had a decrease in milk supply requiring supplementation. In this study, all subjects stopped breastfeeding by the fourth cycle of treatment.

Systemic symptoms were also common. Nausea/vomiting was described in all five studies and noted to be prevalent early in treatment (12.1% at 1 month [12], 30.4% [14], 15% [13], 28% [15], 15.2% [16]) with progressive improvement with continued use (3.2% at 6 months [12], 1.5% [14], 6% [13], 7% [15]). Headache was also frequently reported and, in many studies, followed a similar pattern at 1 (14.5% [12], 14.8% [14]) and six cycles (7.6% [12], 2.1% [14]) of use. Conversely, Rubio et al [15] describe a variable and relatively low headache frequency (0–5% across 24 cycles). Of all reported systemic side effects, mucorrhea was the most commonly reported, also tended to diminish with duration of use [12], [14], [15].

Other less commonly reported-on side effects include fatigue (11.6% C1 to 4% C6) [14], mastalgia (11% C1 to 0% C6) [15], thrombophlebitis (0.18%) [13], and jaundice (0.09%) [13].

3.3.3. Acceptability of quinestrol/quingetanol use

All five studies described the medication’s bleeding profile across multiple cycles of use. In the studies which reported on cycle duration, cycles were a regular 25–33 days in length for a majority of participants [12], [14], [15]. Spotting and breakthrough bleeding were reported in 11–27% of participants [12], [13], [14], [15], [16] but improved with duration of use (0–4% at 12 months) [12], [13], [14], [15]. Rates of missed periods also varied across studies and with length of medication use, though it did not decline as predictably as inter-menstrual bleeding (4.1–8.6% cycle 1–12 [12], 0–9.9% cycle 1–12 [14], 1% of all cycles [15], 5.2% of all cycles [16]).

All of the included studies assessed acceptability of medication use through discussion of discontinuation reasons and rates. Across the studies, discontinuation rates at 12 months were variable but relatively high (1.9−85%) [12], [13], [14], [15], [16]. Key drivers of medication-related discontinuation were nausea/vomiting [12], [14], headache [12], [14], and menstrual disturbances [12], [14]. In Tejuja [16], Leguia [13], and Rubio [15], which did not give additional detail on reasons for discontinuation, 42.2% of subjects at 12 months and 5.5% and 21% at 36 months had left the study for medication-related reasons.

3.4. Mifepristone for emergency contraception

In our scoping review, 10 studies met inclusion criteria. Seven (70%) were randomized control trials, and the other three (30%) were non-randomized experimental studies. These studies were performed between 1992 and 2015. Evaluated outcomes included efficacy, safety, menstrual side effects, and user satisfaction (Table 3).

Table 3.

Summary of results of studies assessing the efficacy, safety, and acceptability of mifepristone as an emergency contraceptive

Study	Design	Purpose	Exposure	Outcomes	Crude pregnancy rate	Prevented pregnancy rate	Safety	Acceptability
Webb 1992 [17] UK, single site n = 195 Funded by WHO	Randomized control trial	Compare efficacy and acceptability of mifepristone vs Yuzpe and danazol as EC	Arm 1: Yuzpe method Arm 2: danazol 600 mg repeated after 12 h Arm 3: mifepristone 600 mg single dose	Efficacy, safety, menstrual effects	0%	100%	Nausea 18.2%, vomiting 1.5%, breast tenderness 9.1%. Menstrual delay by >1 wk 26%	Not reported
Wu 2010 [18] China, multisite n = 499	Randomized control trial	Compare efficacy of mifepristone vs gestrinone as EC	10 mg mifepristone within 72 h of UPIC	Efficacy, safety, menstrual effects	1.8%	76%	Nausea 10.2%, vomiting 0.2%, abdominal pain 1.6%, diarrhea 0.2%, fatigue 3.6%, dizziness 2.6%, breast tenderness 0.8%, intermenstrual bleeding 14.5%.	Not reported
Carbonell 2015 [19] Cuba, single site n = 2418	Randomized control trial	Compare efficacy, safety of two mifepristone doses for EC	5 vs 10 mg mifepristone within 144 h of UPIC	Efficacy, safety, menstrual effects	1.2% (5 mg), 0.7% (10 mg)	88% (5 mg), 93% (10 mg)	Nausea 3−3.2%, abdominal pain 2.9– 3.6%, fatigue 9%, dizziness 3.5−4%, breast tenderness 1.5−1.7%, intermenstrual bleeding 4−4.8%. Menstrual delay by >1 wk 4.9−11%. Heavier subsequent period 1.2−1.5%.	Not reported
Xiao 2003 [20] China, multisite n = 4917 Funded by the Rockefeller Foundation	Single-arm clinical trial	Confirm the effectiveness of mifepristone as EC	10 mg mifepristone within 120 h of UPIC	Efficacy, safety, menstrual effects	1.4%	82.2%	Nausea/vomiting 9.1%, abdominal pain 2.9%, weakness 3.7%, breast tenderness 1.3%. Menstrual delay by >1 wk 6.5%. Heavier subsequent period 10.3%.	Not reported
Xiao 2002 [21] China, multisite n = 3030 Funded by the Rockefeller Foundation	Randomized control trial	Compare efficacy of two doses of mifepristone as EC	Arm 1: 10 mg mifepristone vs Arm 2: 25 mg mifepristone administered within 120 h of UPIC	Efficacy, safety, menstrual effects	1.1% (10 mg), 1.1% (25 mg)	85.3% (10 mg), 86.5% (25 mg)	Nausea 9.8–11.3%, vomiting 0.5–0.8%, abdominal pain 4.3– 4.6%, diarrhea 0.5%, fatigue 6.1−6.9%, breast tenderness 1.3%. Menstrual delay by >1 wk 9.2−9.9.	Not reported
Von Hertzen 2002 [22] Multicountry, multisite n = 4136 Funded by WHO	Randomized control trial	Compare efficacy of mifepristone vs LNG as EC	10 mg mifepristone within 120 h of UPIC	Efficacy, safety, menstrual effects	1.5%	81%	Nausea 1.4%, vomiting 1%, abdominal pain 14%, diarrhea 5%, fatigue 15%, dizziness 9%, breast tenderness 8%, intermenstrual bleeding 19%. Menstrual delay by >1 wk 9%.	Not reported
Hamoda 2004 [23] Scotland, single site n = 2056	Randomized control trial	Compare efficacy, acceptability, and safety of mifepristone vs LNG for EC	10 mg mifepristone within 120 h of UPIC	Efficacy, safety, menstrual effects, acceptability	1.3%	77%	Nausea 26%, vomiting 2.2%, abdominal pain 33%, dizziness 14%, lethargy 25%, breast tenderness 24%. Menstrual delay by >1 wk 16%. Heavier subsequent period 10%.	93.6% satisfaction, 96% would recommend the method.
Esteve 2007 [24] Cuba, single site n = 635	Single-arm clinical trial	Assess effectiveness, safety of mifepristone as EC	10 mg mifepristone within 120 h of UPIC	Efficacy, safety, menstrual effects	1.1%	88%	Nausea 4.9%, vomiting 0.6%, fatigue 10.7%, dizziness 6.1%. Heavier subsequent period 6%.	Not reported
Taneepanichskul 2009 [25] Thailand, single site n = 120	Single-arm clinical trial	Evaluate use of mifepristone for EC	10 mg mifepristone within 120 h of UPIC	Efficacy, safety, menstrual effects, acceptability	0%	100%	Nausea/vomiting: 1.7%, intermenstrual bleeding 6.7%,	Not reported
Ashok 2004 [26] Scotland, single site n = 500	Randomized control trial	Compare side effects, acceptance, satisfaction with mifepristone vs Yuzpe for EC	100 mg mifepristone within 72 h of UPIC	Safety, menstrual effects, satisfaction	Not reported	Not reported	Nausea 15.3%, vomiting 8.25%, abdominal pain 32.7%, fatigue 48.9%, lethargy 28.3%, breast tenderness 24.6%. Menstrual delay by >1 wk 24.5%.	93.9% satisfaction, 94% would recommend the method. 2/316 (0.6%) dissatisfaction.

EC, emergency contraception; LNG, levonorgestrel; UPIC, unprotected intercourse; WHO, World Health Organization.

3.4.1. Efficacy of mifepristone as emergency contraception

Nine of the studies (90%) assessed efficacy of mifepristone as emergency contraception at various doses and timings from a single episode of unprotected intercourse. Six were randomized control trials, and three were non-randomized experimental studies. Efficacy was assessed with crude pregnancy rate per episode of unprotected sex, as well as prevented pregnancy rate based on the calculated expected number of pregnancies per episode. Two studies [17], [18] assessed the efficacy of 600 mg and 10 mg doses respectively taken within 72 hours of unprotected intercourse, and noted 0% and 1.8% crude pregnancy rate [17], [18].

One study assessed the efficacy of mifepristone in 5 vs 10 mg doses taken within 144 hours of unprotected intercourse [19]. This study described high rates of efficacy for both doses, but found a significantly increased risk of pregnancy with the 5 mg regimen for patients with weight >75 kg compared to those with weight <75 kg. Conversely, the 10 mg regimen did not demonstrate a weight-based difference. The remainder assessed efficacy when mifepristone was taken within 120 hours of unprotected intercourse. Six studies [20], [21], [22], [23], [24], [25] assessed 10 mg doses, with the addition of a 5 mg [19] or 25 mg dose [21] comparison group. Efficacy was consistently high, with a prevented pregnancy rate ranging from 76–100%.

3.4.2. Safety of mifepristone as EC

All 10 studies assessed safety outcomes, 40% of which compared mifepristone as EC to published EC regimens (Yuzpe or levonorgestrel EC). Mifepristone compared favorably, with reported incidence of side effects similar to or lower than those of Yuzpe or levonorgestrel. When different dosages of mifepristone were compared within the same study, incidence of side effects was not consistently dose-dependent. Across all 10 studies, side effects were described as mild to moderate in severity.

3.4.2.1. Gastrointestinal effects

Side effects such as nausea, vomiting, abdominal pain, and diarrhea were reported in all 10 studies at variable frequencies. Nausea and vomiting were reported in aggregate in two studies [20], [25] and separately in the remainder, with rates of nausea (1.4–26% [17], [18], [19], [21], [23], [24], [26]) consistently higher than vomiting (0.5–8.25% [17], [18], [21], [22], [23], [24], [26]). Abdominal pain was reported in seven studies (1.6–33% [18], [19], [20], [21], [22], [23], [26]). Diarrhea was reported in three studies (0.2–5% [18], [21], [22]).

3.4.2.2. Systemic side effects

Although reported as mild, systemic side effects such as fatigue (3.6–48.9% [18], [19], [21], [22], [24], [26]), weakness (3.72% [20]), dizziness (2.6–14% [18], [19], [22], [23], [24]), and lethargy (25–28.3% [23], [26]) were described with variable frequency across studies. Breast tenderness was reported in eight studies (1.3–24.6% [17], [18], [19], [20], [21], [22], [23], [26]). Similar to other reported side effects, headache frequency was variable (0.8–25.9% [18], [19], [21], [22], [23], [25], [26]) and intensity was consistently reported as mild.

3.4.2.3. Menstrual effects

Intermenstrual vaginal bleeding was described in four studies (4–19% [18], [19], [22], [25]). Some studies also reported on the timing of onset of the subsequent menstrual period, noting that while the majority of subjects reported menstrual onset within 2–3 days of the expected date (58–75% [18], [20], [21], [23]), a delay of a week or more was common (4.9–26% [17], [19], [20], [21], [22], [23], [26]). The volume of menstrual bleeding was described in four studies, with the majority of subjects reporting normal menstrual flow (81% [20], [23]), and a small proportion reporting greater than normal bleeding (1.2–10.3% [19], [20], [23], [24]).

3.4.3. Acceptability of mifepristone as EC

Two studies assessed measures of method acceptability. In their 2004 study, Ashok et al [26] surveyed mifepristone EC users and found that 93.9% expressed satisfaction with the method and 94% would recommend it to others. Out of 316 respondents, two reported dissatisfaction with mifepristone as EC – one due to delay in return of menses, and the other due to time inconvenience of administration. Hamoda et al [23] reported similar findings, with 93.6% satisfaction rate, and 96% of respondents saying that they would recommend mifepristone as EC to others.

4. Discussion

This scoping review included a total of 20 studies on three different contraceptive options which, while used in WHO member countries, are not listed in the MEC 2015. Our aim for this scoping review was to assess the extant literature on ormeloxifene, quinestrol/quingestanol, and mifepristone as EC in order to inform future guideline development work by the WHO. As we conducted a separate scoping review for each method, we discussed each separately.

4.1. Ormeloxifene

Ormeloxifene, alternatively called “centchroman,” is a nonsteroidal selective estrogen receptor modulator used as a once-weekly contraceptive pill. It has weak estrogenic and potent anti-estrogenic effects, and has an anti-implantation mechanism of action. It is typically administered as a 30 mg tablet given twice weekly for 3 months, followed by weekly dosing of a single 30 mg tablet [27]. Given its weekly schedule, it has potential for widespread acceptability. It is presently registered for use in India, where it is marketed under several names including Saheli and Chhaya.

Our scoping review identified five studies on ormeloxifene, which collectively enrolled a total of 1346 participants and assessed the efficacy, safety, acceptability, and feasibility of ormeloxifene when used as a contraceptive. Evidence indicates that 30 mg given weekly, after an initial bi-weekly induction period, is 96–99% efficacious when assessing method failure pregnancies alone. One study, which employed a more complex dosing regimen, noted a pregnancy rate of 10%, compared to 1–6% in the studies which described 3 months of biweekly followed by future weekly administration of 30 mg. This suggests that increased complexity of dosing regimen may be associated with increased user error.

The majority of studies also assessed safety, via reporting of side effect profiles. Use of the medication was described to be overall safe, with limited systemic symptoms (such as headache, nausea, and fatigue) that were described as mild and self-limited. These side effects are similar to those experienced by users of traditional combination oral contraceptive pills. Menstrual disturbance was the most frequently reported side effect, with prolonged and delayed cycles reported by as many as 50% of subjects. There was, however, significant variation between studies in the extent and duration of cycle disturbance. Given that menstrual irregularity was found to be a key driver for method discontinuation, further research is required in order to understand the extent of menstrual effects of ormeloxifene, and how this compares to other conventional methods of contraception.

Acceptability of ormeloxifene was variably reported, but found to be high overall at one year or more of use. Three studies which assessed continuation at 3, 6, and 12 months found attrition from 100% continuation at 3 months to 75–91% by 12 months. The highest continuation was reported by Nityanand et al [9], with 92.4% continuation at 13 months. It is unclear what the drivers of discontinuation were at each timepoint. However, menstrual disturbance and desire for return to fertility were frequently cited as reasons.

Given the high efficacy and continuation rates reported in the studies we identified, ormeloxifene has high potential for more widespread use. A systematic review on this method to aid in future WHO guidance on emerging methods is warranted, as are additional clinical trials in settings outside of South Asia.

4.2. Quinestrol/Quingestanol

Quinestrol is a long-acting estrogen developed in the 1960s in China, to be used in combination with a short-acting progestogen as a monthly contraceptive pill [28]. Quinestrol is available in contraceptive pills at various doses, and in combination with different progestogens (including norgestrel, levonorgestrel, and quingestanol). These combined monthly pills work via a mechanism of ovulation suppression; quinestrol becomes stored in body fat and is subsequently slowly released and metabolized to ethinyl estradiol, while the progestogen component produces a progesterone peak, resulting in monthly withdrawal bleeding [29].

Quinestrol-containing pills are registered for use in China. The Quin-Quin (quinestrol-quingestanol) regimen involves taking 2 mg of quinestrol on cycle day 1 and day 22, and 5 mg of quingestanol acetate on day 22. This is followed by 2 mg of quinestrol and 5 mg of quingestanol acetate every 4 weeks. Quinestrol was unique amongst the methods in this scoping study, as it was included on the basis of member state inquiries to the WHO. This was due to concerns regarding the method’s safety, in light of high levels of estrogen, and reports from Kenya and Zambia of complications of use including blood pressure elevation, venous thromboembolism, fetal malformation, and uterine pre-cancers.

We identified five studies of the Quin-Quin regimen that enrolled a total of 3322 participants. All of the included studies were non-randomized experimental studies and reported on efficacy, safety, and acceptability. All five studies demonstrated high levels of efficacy and safety. No increased risk of endometrial abnormalities was noted. Systemic side effects (nausea/vomiting, headache, mucorrhea) were very common but were described as mild and diminished in frequency with continued use. Bleeding complaints were frequently reported, but the majority of users described regular cycles. None of the studies reported on the more concerning side effects brought up in member state inquiries. Quinestrol/quingestanol monthly contraceptive pills were found to have low acceptability rates, with rates of discontinuation at 12 months of use as high as 85%. Discontinuation was largely driven by side effect profile, suggesting that although side effects were described by investigators as mild, they were not perceived this way by users. Given that quinestrol dosing in the monthly pills results in exposure to 4–5 times the ethinyl estradiol exposure of standard combined oral contraceptives, akin to levels seen with the mestranol-containing oral contraceptive pills first developed in the 1960s, it is reasonable to be concerned about the potential detrimental effects of these medications. Notably, the studies included in this scoping review were performed from 1970 to 1974. In consultation with WHO Guideline Development Group membership, it is likely that additional references may be found in Chinese publication databases. Additional systematic review of quinestrol-containing contraception in a broad range of international databases is warranted, with particular attention to safety outcomes.

4.3. Mifepristone as EC

Mifepristone is a norethindrone derivative, which binds strongly to progesterone and glucocorticoid receptors with an antagonistic function at each [30]. Approved widely as part of a regimen of medication abortion, mifepristone has also been noted to have potential mechanisms for emergency contraception (EC) via ovulation delay, endometrial alteration, and prevention of implantation [31]. Mifepristone has been studied in varied doses, ranging from 10 to 600 mg oral tablets. In studies of mifepristone as EC, it has compared favorably and even shown superior efficacy to Yuzpe (estrogen/progestin) and levonorgestrel regimens [23], [32]. We included 10 studies on mifepristone as EC, including a total of 18,496 participants. Relative to other studies included in this scoping study, the studies on mifepristone as EC were more recently conducted and have larger sample sizes, enabling greater reliability. Seventy percent were randomized control trials. Efficacy was reported in 90% of studies, safety in 100%, and acceptability in 20% of studies.

Efficacy of mifepristone as emergency contraception was high overall. Across variable dosages and timings from last unprotected intercourse, the crude pregnancy rate ranged from 0 to 1.8%. This is as compared to 0.9–1.8% within 120 hours for ulipristal acetate 30 mg or within 96 hours for levonorgestrel 1.5 mg. Higher doses, of 10 mg mifepristone or more, were associated with greater rate of pregnancy prevention.

Safety outcomes were overall favorable, with mifepristone’s side effect profile being comparable to or better than published EC regimens in the randomized trials. As with other methods, side effects were reported at highly variable rates, likely contributed to by differences in dose and timing of regimens. The only side effect which was cited as possibly contributing to a lack of satisfaction with this method was delayed return of menses. While the majority of subjects reported menstrual onset within 2–3 days of their expected date (58–75%), a delay of a week or more was common (up to 26%) and could contribute to user frustration. Even so, mifepristone as EC was highly acceptable to patients, with 93–94% expressing satisfaction and 94–95% reporting that they would recommend its use to others.

Mifepristone is already used as an emergency contraceptive in several countries including Moldova, Russia, Ukraine, Armenia, China, and Vietnam [33], and has rigorous safety data in its 200 mg use in medication abortion. A Cochrane review of mifepristone for EC was published in 2019 [34], but our scoping review identified four additional studies not included in the Cochrane review. An updated systematic review on mifepristone for EC would allow the most robust data for inclusion in WHO guidance.

4.4. Strengths and limitations

This scoping review offers novel contributions to the literature by aggregating evidence regarding the use in WHO member countries of contraceptive methods not included in the MEC. Because we did not restrict our literature review by year of publication, we were able to conduct an extensive and thorough review of the extant literature on the included methods, and offer a broad summary of the available data. In addition, the multiple modalities used for method identification –survey responses from over 100 member countries, stakeholder inquiry, as well as the results of the literature review itself – provided confidence in the breadth of the methods considered for inclusion.

Limitations of this work include the exclusion of literature not already available in English. Most notably, this affected our ability to review the literature on Quinestrol-containing methods, as many of the published studies on this method are available exclusively in Chinese. Future work should look to commission translation of these studies into English, or include investigators who are literate in the papers’ original languages so that the studies can be considered for inclusion. Additionally, the nature of this study as a scoping review inherently limits it to evidence synthesis and mapping. Future work in the form of a systematic review will be necessary in order to comment on the quality of the gathered evidence.

Our scoping review provides a broad review of literature on three contraceptive methods which, while not yet included in the MEC, are presently in use in WHO member countries. For all three methods we found data regarding efficacy, safety, and acceptability. Future directions include rigorous systematic reviews of the literature for each of these individual methods, in order to provide data supporting international guidance surrounding these methods.

CRediT authorship contribution statement

James Kiarie: Project administration, Conceptualization. Asantesana Kamuyango: Project administration, Methodology. Li Jiang: Methodology, Formal analysis. Nancy Kidula: Supervision, Project administration. Sarita Sonalkar: Writing – review & editing, Supervision, Project administration, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Sameera Mokkarala: Writing – original draft, Validation, Methodology, Investigation, Formal analysis, Conceptualization.

Appendix A. Supplementary material

Supplementary material

.

Supplementary material

.