Cyclical progestogens for heavy menstrual bleeding

Abstract

Background

Heavy menstrual bleeding (HMB) is a menstrual blood loss perceived by women as excessive that affects the health of women of reproductive age, interfering with their physical, emotional, social and material quality of life. Whilst abnormal menstrual bleeding may be associated with underlying pathology, in the present context, HMB is defined as excessive menstrual bleeding in the absence of other systemic or gynaecological disease. The first‐line therapy is usually medical, avoiding possibly unnecessary surgery. Of the wide variety of medications used to reduce HMB, oral progestogens were originally the most commonly prescribed agents. This review assesses the effectiveness of two different types and regimens of oral progestogens in reducing ovulatory HMB.

This is the update of a Cochrane review last updated in 2007, and originally named "Effectiveness of cyclical progestagen therapy in reducing heavy menstrual bleeding" (1998).

Objectives

To determine the effectiveness, safety and tolerability of oral progestogen therapy taken either during the luteal phase (short cycle) or for a longer course of 21 days per cycle (long cycle), in achieving a reduction in menstrual blood loss in women of reproductive age with HMB.

Search methods

In January 2019 we searched Cochrane Gynaecology and Fertility's specialized register, CENTRAL, MEDLINE, Embase, CINAHL and PsycInfo. We also searched trials registers, other sources of unpublished or grey literature and reference lists of retrieved trials. We also checked citation lists of review articles to identify trials.

Selection criteria

Randomized controlled trials (RCTs) comparing different treatments for HMB that included cyclical oral progestogens were eligible.

Data collection and analysis

Two review authors independently selected trials for inclusion, assessed trials for risk of bias and extracted data. We contacted trial authors for clarification of methods or additional data when necessary. We only assessed adverse events if they were separately measured in the included trials. We compared cyclical oral progestogen in different regimens and placebo or other treatments. Our primary outcomes were menstrual blood loss and satisfaction with treatment; the secondary outcomes were number of days of bleeding, quality of life, compliance and acceptability of treatment, adverse events and costs.

Main results

This review identified 15 randomized controlled trials (RCTs) with 1071 women in total. Most of the women knew which treatment they were receiving, which may have influenced their judgements about menstrual blood loss and satisfaction. Other aspects of trial quality varied among trials.

We did not identify any RCTs comparing progestogen treatment with placebo. We assessed comparisons between oral progestogens and other medical therapies separately according to different regimens.

Short‐cycle progestogen therapy during the luteal phase (medroxyprogesterone acetate or norethisterone for 7 to 10 days, from day 15 to 19) was inferior to other medical therapy, including tranexamic acid, danazol and the progestogen‐releasing intrauterine system (Pg‐IUS (off the market since 2001)), releasing 60 mcg of progesterone daily, with respect to reduction of menstrual blood loss (mean difference (MD) 37.29, 95% confidence interval (CI) 17.67 to 56.91; I² = 50%; 6 trials, 145 women, low‐quality evidence). The rate of satisfaction and the quality of life with treatment was similar in both groups. The number of bleeding days was greater on the short cycle progestogen group compared to other medical treatments. Adverse events (such as gastrointestinal symptoms and weight gain) were more likely with danazol when compared with progestogen treatment. We note that danazol is no longer in general use for treating HMB.

Long‐cycle progestogen therapy (medroxyprogesterone acetate or norethisterone), from day 5 to day 26 of the menstrual cycle, is also inferior compared to the levonorgestrel‐releasing intrauterine system (LNG‐IUS), tranexamic acid and ormeloxifene, but may be similar to the combined vaginal ring with respect to reduction of menstrual blood loss (MD 16.88, 95% CI 10.93 to 22.84; I² = 87%; 4 trials, 355 women, very low‐quality evidence). There was no clear evidence of a difference between progestogen therapy long cycle and other medical therapy in terms of headache (OR 1.45, 95% CI 0.40 to 5.31; I² = 0%; 2 trials, 189 women; low‐quality evidence). Breakthrough bleeding or spotting was more likely in women with the LNG‐IUS (OR 0.18, 95% CI 0.06 to 0.55; I² = 0%; 3 trials, 220 women; low‐quality evidence). No trials reported on days of bleeding or quality of life for this comparison.

The evidence supporting these findings was limited by low or very low gradings of quality; thus, we are uncertain about the findings and there is a potential that they may change if we identify other trials.

Authors' conclusions

Low‐ or very low‐quality evidence suggests that short‐course progestogen was inferior to other medical therapy, including tranexamic acid, danazol and the Pg‐IUS with respect to reduction of menstrual blood loss. Long cycle progestogen therapy (medroxyprogesterone acetate or norethisterone) was also inferior to the LNG‐IUS, tranexamic acid and ormeloxifene, but may be similar to the combined vaginal ring with respect to reduction of menstrual blood loss.

Plain language summary

Are cyclical progestogens an effective and safe treatment for heavy menstrual bleeding compared to other medical treatments?

Background

Heavy menstrual bleeding (HMB) is menstrual bleeding (periods) that interferes with a woman's quality of life, either physical, emotional, social or material, independently of the actual amount of blood loss. Most women with HMB do not have any associated physical cause, such as fibroids, so getting help that does not involve surgery is an attractive alternative. A cyclical progestogen is a hormone tablet that can be taken by mouth for either 10 days or 3 to 4 weeks per month for the treatment of HMB (short or long course cyclical progestogen).

Trial characteristics

This review identified 15 randomized controlled trials (clinical studies where people are randomly put into one of two or more treatment groups) with 1071 women in total comparing oral progestogens to other medical treatment for HMB (other oral treatments, intrauterine device and vaginal ring). Our primary outcomes were menstrual blood loss and satisfaction with treatment; the secondary outcomes were number of days of bleeding, quality of life, compliance and acceptability of treatment, adverse events and costs. Evidence is current to January 2019.

Key results

This review of trials found that progestogen hormone tablets taken by mouth for 10 days per month (short course) were less effective at reducing menstrual blood loss when compared to other medical treatments. We are uncertain whether they improved satisfaction or quality of life in women with HMB, or were associated with any difference in adverse effects when compared to other medical treatments. Even though it was less effective at reducing menstrual blood loss, satisfaction with treatment was similar to other medical treatments such as tranexamic acid and Pg‐IUS.

We found that progestogen hormone tablets, taken by mouth for three to four weeks from day 5 to 26 of the menstrual cycle (long course), reduced menstrual blood loss but this treatment may be less effective than tranexamic acid, combined hormonal contraceptives and the levonorgestrel‐releasing intrauterine device. No studies in this comparison reported on quality of life. Satisfaction with treatment was similar to women using the combined vaginal ring, but there were no data to compare satisfaction between long cycle and LNG‐IUS or tranexamic acid. There was no evidence of a difference in the occurrence of headache, but long course oral progestogens were associated with a significantly lower incidence of breakthrough bleeding compared with other medical treatment.

Quality of the evidence

The quality of the evidence that compared oral progestogens (short and long course) to other medical treatments for HMB was either low or very low which means that we are very uncertain of the findings of the review. The main limitations were risk of bias (women and researchers were aware of the treatment they were receiving which was likely to interfere with the responses, and there was a high number of dropouts from studies) and inconsistency (the results varied among studies).

Background

Description of the condition

Heavy menstrual bleeding (HMB), also known as menorrhagia, is an important cause of morbidity in otherwise healthy women. The definition of HMB was considered for a long time as a menstrual blood loss over 80 mL per menstrual cycle; this objective cut‐off was broadly utilized in clinical trials, but such measurement (involving the collection of sanitary pads and tampons and extracting haemoglobin) has proven to be impractical outside research settings. Furthermore, this objectively measured cut‐off of 'heavy' menstrual bleeding may not reflect the woman's experience, nor the impact of the HMB on her quality of life (Warner 2004). The Pictorial Blood Assessment Chart (PBAC; Higham 1990), is a semi‐quantitative tool to measure HMB. Currently, the International Federation of Gynaecology and Obstetrics (FIGO) defines HMB focused on the woman's perspective, as "an excessive menstrual blood loss that interferes with the woman's physical, emotional, social, and material quality of life, and can occur alone or in combination with other symptoms" such as dysmenorrhoea, headache or fatigue (Munro 2012).

In this context, HMB is not related to pregnancy, systemic or gynaecological disease.

The prevalence of HMB ranges from 9% to 51%, according to the assessment method, varying from 9% to 15% if measured objectively, to between 20% and 51%, if assessed based on the woman's perception (Fraser 2015; NICE 2007; NICE 2018). HMB represents a significant clinical challenge and is responsible for up to one‐third of all gynaecological consultations (Liu 2007; Miller 2015).

Description of the intervention

Surgery (hysterectomy) was the most common treatment for women with HMB in the UK in the 1980s, with 60% of all secondary care referrals for HMB resulting in hysterectomy (Van der Meij 2016). Currently, hysterectomy is reserved for women for whom other treatment options have failed or are contra‐indicated.

A wide variety of medications may be used to reduce HMB. In the UK, norethisterone (norethindrone in the USA) was the most frequently prescribed drug during the 1980s and 1990s, when progestogens constituted 55% of total prescriptions for HMB (Coulter 1995).

Prolonged use of high‐dose progestogens is sometimes associated with side effects, including fatigue, mood changes, weight gain, nausea, bloating, oedema, headaches, depression, loss of libido, irregular bleeding, and atherogenic changes in the lipid profile.

How the intervention might work

In the anovulatory woman, progestogens help to co‐ordinate regular uterine shedding when given as a late luteal replacement treatment, on days 19 to 26 of the cycle (Conyngham 1965). However, the use of progestogens as luteal phase supplementation in the ovulatory woman with HMB is more questionable. Consequently, an increase in the duration and dosage of progestogen therapy was investigated in women with ovulatory HMB. Progesterone is the dominant hormone during the secretory phase of the menstrual cycle and is a potent anti‐inflammatory agent. Progesterone exposure limits endometrial inflammation (Maybin 2016), and long cycle use of progestogens may reduce menstrual bleeding.

Why it is important to do this review

The levonorgestrel intrauterine system (LNG‐IUS) has solid evidence of efficacy in managing HMB (Gupta 2013; Qiu 2014). Gupta published a randomized controlled trial (RCT) comparing LNG‐IUS with other medical therapies for HMB (Gupta 2013), and Qiu published a systematic analysis including Gupta's results (Qiu 2014). Both authors concluded that LNG‐IUS is more effective for the treatment of HMB than oral medical treatment. However, for women who wish to avoid an intrauterine device, or for whom such a device is contra‐indicated, other medical treatments may have a role. The role and efficacy of oral progestogens in treating HMB remains controversial and the objective of this review is to address this issue.

This is the update of a Cochrane review last updated in 2007, originally named "Effectiveness of cyclical progestagen therapy in reducing heavy menstrual bleeding" (Lethaby 1998b).

Objectives

To determine the effectiveness, safety and tolerability of oral progestogen therapy taken either during the luteal phase (short cycle) or for a longer course of 21 days per cycle (long cycle), in achieving a reduction in menstrual blood loss in women of reproductive age with HMB.

Methods

Criteria for considering studies for this review

Types of studies

1. All RCTs comparing oral progestogen therapies, taken either during the luteal phase (short cycle) or for a longer course of 21 days per cycle (long cycle), when used to reduce HMB versus placebo.

2. All RCTs comparing oral progestogen therapies, taken either during the luteal phase (short cycle) or for a longer course of 21 days per cycle (long cycle), versus other medical therapies when used to reduce HMB.

Types of participants

1. Women of reproductive age

2. Women with regular heavy periods measured either subjectively by the woman, objectively by the alkaline haematin method (more than 80 mL per cycle) or semi‐objectively by the pictorial blood assessment chart (Hallberg 1964; Higham 1990; Newton 1977).

3. Women attending primary care, family planning or specialist clinics

Exclusion criteria

1. Post‐menopausal bleeding (more than one year following the last period)

2. Irregular menses and intermenstrual bleeding

3. Pathological causes of HMB

4. Iatrogenic causes of HMB

Types of interventions

1. Oral progestogen therapy (norethisterone, dydrogesterone, medroxyprogesterone acetate) versus placebo

2. Oral progestogen therapy versus any other medical therapy: antifibrinolytic agents, danazol, non‐steroidal anti‐inflammatory drugs (NSAIDs), gonadotrophin‐releasing hormone analogues, combined hormonal contraceptives (oral or vaginal ring), selective oestrogen receptor modulators (SERMs), progestogen‐releasing intrauterine systems (Pg‐IUS) and levonorgestrel‐releasing intrauterine device (LNG‐IUS).

We considered variable doses and lengths of administration of treatments. Oral progestogens can be given as luteal phase replacement treatment (for 7 to 10 days from day 15 or day 19 of the menstrual cycle) or as a longer course of 21 out of 28 days (days 5 to 26 of the cycle).

Types of outcome measures

Where available, we recorded the following outcomes.

Primary outcomes

1. Menstrual blood loss

   1. Objective assessment of blood loss, either during the intervention or for up to one year of follow‐up after the intervention, measured by the alkaline haematin method (Hallberg 1964), or the pictorial blood assessment chart (Higham 1990)

   2. Subjective (woman's perception) assessment of blood loss, also measured either during the intervention or for up to one year of follow‐up after the intervention.

2. Satisfaction: overall satisfaction with treatment.

Secondary outcomes

1. Number of days bleeding during the intervention menstrual cycle

2. Quality of life: women's perceived change in quality of life, provided this was recorded in a reproducible and validated format

3. Compliance with treatment

4. Acceptability of treatment

5. Adverse events, of any degree, reported either spontaneously by the woman or elicited from specific questioning

6. Resource use and cost

Search methods for identification of studies

We searched for all published and unpublished RCTs of cyclical progestogens for the treatment of HMB, without language restriction and in consultation with the Cochrane Gynaecology and Fertility (CGF) Information Specialist.

Electronic searches

For the update of this review in 2019, we searched the Cochrane Gynaecology and Fertility Register of controlled trials (searched January 2019; Procite platform), the Cochrane Central Register of Studies (CRSO), MEDLINE (searched from 1946 to January 2019) via Ovid, Embase (searched from 1980 to January 2019) via Ovid, PsycINFO (searched from 1806 to January 2019) via Ovid, and CINAHL (searched from 1961 to January 2019) via Ebsco. See Appendix 1; Appendix 2; Appendix 3; Appendix 4; Appendix 5; Appendix 6.

We also searched citation lists of included trials and relevant review articles. For most of the included trials, we contacted the first author to clarify issues relating to data extraction.

Searching other resources

Trials registers for ongoing and registered trials:

1. clinicaltrials.gov (a service of the US National Institutes of Health) (January 2019)

2. apps.who.int/trialsearch/Default.aspx (World Health Organization International Trials Registry Platform search portal). (January 2019)

Data collection and analysis

We conducted data collection and analysis in accordance with the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011a).

Selection of studies

For the 2019 update, Marian Showell (CGF Information Specialist) conducted the initial search. MB and CL conducted an initial screen of titles and abstracts using Covidence, and then we retrieved the full texts of all potentially eligible trials. Two review authors (MB and AL) independently examined these full‐text articles for compliance with the inclusion criteria and selected eligible trials. We resolved disagreements by discussion. If any reports required translation, we described the process used for data collection. We documented the selection process with a PRISMA flow chart (Moher 2009; Figure 1).

1.

Trial flow diagram

Data extraction and management

In the 2019 update, two review authors (MB and AL) independently performed the data extraction from three trials (Ashraf 2017; Kiseli 2016; Shravage 2011), using forms designed according to Cochrane guidelines (Higgins 2011b). The remaining five new trials (Goshtasebi 2013; Hashim 2012; Kaunitz 2010; Kriplani 2006; Zhang 2008), were already included in other Cochrane Reviews of treatments for HMB. We obtained data from those reviews.

In the previous versions of the review two authors (AL and GI) performed the screening, data extraction and risk of bias assessment.

We analysed included trials for the following criteria and methodological details.

Trial characteristics

1. Method of randomization

2. Presence or absence of blinding to treatment allocation

3. Quality of allocation concealment

4. Number of women randomized, excluded or lost to follow‐up

5. Whether an intention‐to‐treat analysis was done

6. Whether a power calculation was done

7. Duration, timing and location of the trial

8. Source of funding

9. Conflict of interest

Characteristics of the women in the trial

1. Age and any other recorded characteristics of women in the trial

2. Methods used to define HMB

3. Other inclusion criteria

4. Exclusion criteria

Interventions used

1. Types of medical therapy used

2. Dose, duration and timing of administration of medical therapy

Outcomes

1. Methods used to measure menstrual blood loss at or after intervention

2. Methods used to evaluate patient satisfaction, symptoms and change in quality of life

Two review authors (MB and AL) independently assessed the quality of the included trials. Where necessary, we sought additional information on trial methodology or actual trial data from the principal author of any trial that appeared to meet the eligibility criteria. In cases where trials presented results in graphs and did not give any actual data, we extracted the data from the graphs.

Assessment of risk of bias in included studies

Two review authors (MB and AL) independently assessed the included trials for risk of bias using the Cochrane 'Risk of bias' assessment tool (Higgins 2011c; Higgins 2017). Individual domains included sequence generation, allocation concealment, blinding and incomplete outcome data; we scored each domain separately as either low, unclear or high risk of bias for three new trials (Ashraf 2017; Kiseli 2016; Shravage 2011). For the remaining five new trials (Goshtasebi 2013; Hashim 2012; Kaunitz 2010; Kriplani 2006; Zhang 2008) that were previously included in other Cochrane Reviews of treatments for HMB, we extracted risk of bias from the reviews.

This assessment is presented in 'Risk of bias' tables in the Characteristics of included studies and summary figures (Figure 2; Figure 3).

2.

'Risk of bias' graph: review authors' judgements about each 'Risk of bias' item presented as percentages across all included trials

3.

'Risk of bias' summary: review authors' judgements about each 'Risk of bias' item for each included trial

In the previous version of the review two authors (from AL, GI or IC) assessed risk of bias.

Measures of treatment effect

For dichotomous data (for example proportion of women who found the treatment unacceptable), we expressed results for each trial as an odds ratio (OR) with 95% confidence intervals (CI) and combined them for meta‐analysis with Review Manager 5 (RevMan 5) software using the Peto‐modified Mantel‐Haenszel method (Review Manager 2014).

We encountered difficulties with the reporting of continuous outcomes (for example menstrual blood loss after treatment). Meta‐analysis with RevMan 5 software uses a mean difference (MD) to combine outcomes and requires data to be presented as absolute values of post‐treatment means with their standard deviations (SD). For some outcomes, particularly menstrual blood loss, the data are skewed and trial authors correctly presented their data as medians with a range. Where possible, we obtained original data from the principal trial authors but post‐treatment means and standard deviations were not always either available or calculable. Where only medians and ranges were available, we regarded the median as being identical to the mean and we calculated a crude estimate of the SD from the range ((range x 0.95)/4). This method is not ideal for skewed data and is likely to result in an over‐estimation of the SD but we had planned a priori to perform sensitivity analysis with and without these trials included in the meta‐analysis. Without exception, the distribution of data on menstrual blood loss are usually skewed and the ideal and most appropriate statistical analysis is nonparametric rather than by inclusion in a meta‐analysis assessing a mean difference.

Unit of analysis issues

We undertook all analyses per woman randomized. Participants were individually randomized to the groups and a single measurement for each outcome from each participant was collected and analysed.

Dealing with missing data

We analysed the data on an intention‐to‐treat basis as far as possible (i.e. including all randomized participants in analysis, in the groups to which they were randomized). We attempted to obtain missing data from the original trial authors (Kiseli 2016; Shravage 2011). Where these were unobtainable, we undertook imputation of individual values for primary outcomes only. For other outcomes, we analysed only the available data (Shravage 2011).

Assessment of heterogeneity

We considered whether the clinical and methodological characteristics of the included trials were sufficiently similar for meta‐analysis to provide a clinically meaningful summary. We assessed statistical heterogeneity by the measure of the I² statistic. If the I² statistic measurement was greater than 50%, we took this as an indication of substantial heterogeneity (Deeks 2017).

Assessment of reporting biases

To minimize the potential impact of the difficulty of detecting and correcting for publication bias and other reporting biases, we ensured a comprehensive search for eligible trials, and were alert for duplication of data. If there were 10 or more trials in an analysis, we planned to use a funnel plot to explore the possibility of small‐trial effects (a tendency for estimates of the intervention effect to be more beneficial in smaller trials; Sterne 2017).

Data synthesis

If the trials were sufficiently similar, we combined the data using a fixed‐effect model in the following comparisons:

1. Progestogen versus placebo

2. Progestogen therapy (luteal phase only) versus NSAIDs

3. Progestogen therapy (luteal phase only) versus danazol

4. Progestogen therapy (luteal phase only) versus tranexamic acid

5. Progestogen therapy (norethisterone luteal phase only) versus progestogen‐releasing IUS: Pg‐IUS and LNG‐IUS

6. Progestogen therapy (medroxyprogesterone acetate luteal phase only) versus LNG‐IUS

7. Progestogen therapy (3‐4 weeks) versus LNG‐IUS

8. Progestogen therapy (3‐4 weeks) versus tranexamic acid

9. Progestogen therapy (3‐4 weeks) versus combined hormonal vaginal ring

10. Progestogen therapy (3‐4 weeks) versus ormeloxifene

11. Overall comparison: progestogen therapy (luteal phase only) versus other medical therapy

12. Overall comparison: progestogen therapy long cycle (3‐4 weeks) versus other medical therapy

Subgroup analysis and investigation of heterogeneity

We assessed differences between the action of luteal phase progestogen and long‐cycle (i.e. 3 to 4 weeks) progestogen. There was no substantial heterogeneity.

Sensitivity analysis

We planned sensitivity analysis to compare potential differences in participants, interventions, outcomes and whether allocation concealment was adequate in the included trials. We could not undertake any sensitivity analysis because of the small numbers of included trials for each outcome.

Overall quality of the body of evidence

We generated 'Summary of findings' tables using GRADEpro GDT software (GRADEpro GDT), to evaluate the overall quality of the body of evidence for the two overall comparisons: luteal phase and long‐cycle progestogen versus other medical therapy, for the following outcomes: menstrual blood loss; satisfaction; menstruation days; quality of life; and adverse effects. Two review authors (MBR and AL) made independent judgments on the overall quality of trials for each of these outcomes, according to the GRADE criteria (trial limitations (i.e. risk of bias), consistency of effect, imprecision, indirectness and publication bias). For each GRADE criterion, if there were concerns about quality, the assessment could be downgraded either one or two levels. Overall quality for each outcome could be categorized as either high, moderate, low or very low, according to these assessments.

Results

Description of studies

Results of the search

We included seven RCTs of oral progestogen therapy for regular HMB in the 2007 version of the review.

We considered a further 11 potential trials eligible for the 2019 update and retrieved the full texts of the papers (when available) for closer inspection. We included eight new trials (Ashraf 2017; Goshtasebi 2013; Hashim 2012; Kaunitz 2010; Kiseli 2016; Kriplani 2006; Shravage 2011; Zhang 2008). There were no trials awaiting classification; we excluded three trials (Gupta 2013; Karakus 2009; Kucuk 2008), and there was one ongoing trial (NCT02943655), comparing the combined contraceptive pill, medroxyprogesterone acetate and mefenamic acid.

Thus, we included a total of 15 RCTs of oral progestogen therapy for regular HMB in the review. Summary details of the trials are given in the Characteristics of included studies.

Included studies

Noresthisterone and medroxyprogesterone acetate were the only two types of oral progestogen included in trials for the treatment of HMB.

Design and setting

All 15 included RCTs have a parallel‐group design.

Eleven trials were single‐centre trials; the country settings included Egypt, India, Iran, Ireland, Turkey, and UK. Three trials recruited participants from various sites across individual countries: Pakistan (Ashraf 2017), Iran (Goshtasebi 2013), and China (Zhang 2008), and one trial was multicentre based in Brazil, Canada and USA (Kaunitz 2010).

Time frame

Six trials reported their recruiting time frame, three between 2002 and 2008 (Kaunitz 2010; Kriplani 2006; Zhang 2008), and five between 2008 and 2014 (Ashraf 2017; Goshtasebi 2013; Hashim 2012; Kiseli 2016; Shravage 2011). The remaining trials did not report their recruiting time and were all published prior to 1999 (Bonduelle 1991; Buyru 1995; Cameron 1987; Cameron 1990; Higham 1993; Irvine 1998; Preston 1995).

Participants

The 15 included trials contained 1071 participants with age ranging mostly between 25 and 40 years.

All the trials but one (Zhang 2008), stated that inclusion criteria required a normal uterus and no associated pathology.

Three trials mentioned weight as a characteristic of women: one included only women with a body mass index (BMI) between 19 and 29 (Goshtasebi 2013), one excluded women with a BMI over 35 (Hashim 2012), and one excluded women who weighed over 110 kg (Higham 1993).

Interventions

There were only two different oral progestogens included in the trials: norethisterone and medroxyprogesterone acetate, used either as luteal phase (for 10 to 14 days from day 14 of the menstrual cycle) or long cycle (3 to 4 weeks from day 2 to 5 of the cycle).

Nine RCTs (Bonduelle 1991; Buyru 1995; Cameron 1987; Cameron 1990; Higham 1993; Kaunitz 2010; Kiseli 2016; Preston 1995; Zhang 2008), with a total of 355 women, assessed the effects of oral progestogen therapy administered during the luteal phase of the menstrual cycle (starting between days 15 or 19 and finishing between days 23 and 26 of the menstrual cycle). Three of these compared norethisterone with danazol (Bonduelle 1991; Buyru 1995; Higham 1993); one with mefenamic acid (Cameron 1990); three with tranexamic acid (Kiseli 2016; Preston 1995; Zhang 2008); and one trial with small numbers (Cameron 1987) compared norethisterone with danazol, mefenamic acid and a Pg‐IUS (Progestasert). One of the danazol trials compared norethisterone with two separate danazol groups (Higham 1993): a continuous regimen of 200 mg a day for three cycles and a reducing regimen of 200 mg a day for the first cycle, 100 mg a day during the next cycle and 50 mg a day during the last cycle. One trial (Kaunitz 2010), compared medroxyprogesterone acetate during the luteal phase to the LNG‐IUS. One trial compared short cycle norethisterone with tranexamic acid and LNG‐IUS (Kiseli 2016).

Six RCTs (Ashraf 2017; Goshtasebi 2013; Hashim 2012; Irvine 1998; Kriplani 2006; Shravage 2011), with a total of 487 women, assessed the effects of oral progestogen therapy administered from day three to five of the menstrual cycle for 21 to 25 days. Three used norethisterone 5 mg three times a day from day 5 to 26 (Ashraf 2017; Hashim 2012; Irvine 1998), and three used medroxyprogesterone acetate 10 mg to 20 mg a day for 21 days a month (Goshtasebi 2013; Kriplani 2006; Shravage 2011). Two compared a regimen of three to four weeks of oral progestogen to the LNG‐IUS (Ashraf 2017; Irvine 1998), two to tranexamic acid (Goshtasebi 2013; Kriplani 2006), one to the combined vaginal ring (Hashim 2012), and one to ormeloxifene (Shravage 2011).

Comparisons

None of the included trials compared cyclical progestogen to placebo.

Twelve trials assessed luteal phase progestogen.

1. Two compared luteal norethisterone to NSAIDs (Cameron 1987; Cameron 1990)

2. Four compared luteal norethisterone to danazol (Bonduelle 1991; Buyru 1995; Cameron 1987; Higham 1993)

3. Three compared luteal norethisterone to tranexamic acid (Kiseli 2016; Preston 1995; Zhang 2008)

4. Two compared luteal norethisterone to progestogen‐releasing intrauterine devices, one releasing 65 mcg progestogen daily (Pg‐IUS; Cameron 1987), and one releasing 20 mcg of levonorgestrel daily (LNG‐IUS; Kiseli 2016).

5. One compared luteal medroxyprogesterone acetate to Pg‐IUS (Kaunitz 2010)

Six trials compared long cycle oral progestogens.

1. Two trials compared long cycle norethisterone to LNG‐IUS (Ashraf 2017; Irvine 1998), both using the intrauterine system releasing 20 mcg levonorgestrel daily

2. Two trials compared long cycle medroxyprogesterone acetate to tranexamic acid (Goshtasebi 2013; Kriplani 2006)

3. One trial compared long cycle norethisterone to the combined vaginal ring (Hashim 2012)

4. One trial compared long cycle medroxyprogesterone acetate to ormeloxifene (Shravage 2011)

Two of the trials were multi‐arm.

1. One with four arms (Cameron 1987), compared luteal norethisterone, danazol, NSAIDs and Pg‐IUS

2. One with three arms (Kiseli 2016), compared long cycle norethisterone, tranexamic acid and LNG‐IUS

Primary outcomes

Menstrual blood loss

Thirteen trials measured bleeding as an outcome.

Only 13 trials reported bleeding. The criteria for HMB varied among trials: blood loss over 80 mL per cycle in four trials (Cameron 1990; Irvine 1998; Kaunitz 2010; Preston 1995); and blood loss over 50 mL per cycle in one trial (Cameron 1987). Five trials used the Pictorial Blood Assessment Chart (PBAC): two used a cut off over 100 for two cycles(Ashraf 2017; Kiseli 2016), and one for two cycles over 185 (Hashim 2012). Two used only one cycle: one trial used PBAC over 100 (Kriplani 2006) and one used PBAC over 130 (Zhang 2008). One trial used a subjective assessment of menstrual blood loss, using a modified PBAC (Goshtasebi 2013). One trial used the clinical history of excessive menstrual loss proving socially and domestically disruptive associated with anaemia (Bonduelle 1991), and one trial, the use of more than three pads a day (Buyru 1995). Five trials assessed menstrual blood loss using the alkaline haematin method (Cameron 1987; Cameron 1990; Higham 1993; Irvine 1998; Preston 1995): one recorded amenorrhoea (Irvine 1998); and one the woman's perception of improvement of bleeding (Higham 1993). Four trials reported haemoglobin (Goshtasebi 2013; Hashim 2012; Irvine 1998; Kriplani 2006), and one of them also reported ferritin levels (Goshtasebi 2013).

Satisfaction

Three trials reported overall satisfaction (Hashim 2012; Irvine 1998; Kriplani 2006). This was added as an outcome at the 2019 update.

Secondary outcomes

1. Eight trials reported the number of days of menstrual bleeding (Bonduelle 1991; Buyru 1995; Cameron 1987; Cameron 1990; Higham 1993; Kriplani 2006; Preston 1995; Zhang 2008)

2. Quality of life was reported in six trials, using different approaches: two did not specify which questionnaire of quality of life was used (Preston 1995; Zhang 2008), one used the HRQoL‐4 (Hashim 2012), one the World Health Organization QoL‐Short Form (Turkish version) (Kiseli 2016), one the SF‐36 QoL ( Goshtasebi 2013); and one asked women how their periods interfered with their quality of life both before and after treatment (Irvine 1998).

3. Patient compliance with the treatment was reported by two trials (Cameron 1990; Irvine 1998).

4. Two trials reported the acceptability of treatment (Higham 1993; Irvine 1998).

5. The prevalence of side effects of any degree, reported either spontaneously by the woman or elicited from specific questioning, was reported in nine trials (Bonduelle 1991; Buyru 1995; Cameron 1990; Goshtasebi 2013; Higham 1993; Irvine 1998; Kaunitz 2010; Kriplani 2006; Preston 1995).

6. Resource use and cost: no trial reported this outcome.

Excluded studies

We excluded four trials.

• Azizkhani 2018: the comparison was an alternative treatment (traditional dry cupping).

• Gupta 2013: the randomization was LNG‐IUS versus any other medical treatment, the choice of which medical treatment was according to the woman's preference. The comparison was a variety of different types of other medical treatment and there were no data specifically for cyclical progestogen (even though it was one of the comparators).

• Karakus 2009: the population was women with irregular bleeding.

• Kucuk 2008: this is a quasi‐randomized trial.

Risk of bias in included studies

Allocation

The random sequence generation received a low risk score for the majority of the trials (Ashraf 2017; Cameron 1987; Cameron 1990; Goshtasebi 2013; Hashim 2012; Irvine 1998; Kaunitz 2010; Kiseli 2016; Kriplani 2006; Preston 1995; Shravage 2011; Zhang 2008). Three received an unclear risk because they did not clearly state their randomization technique (Bonduelle 1991; Buyru 1995), and one had high risk of bias because they used a sequential order for randomization (Higham 1993).

Eight trials had an unclear risk of bias for allocation (Ashraf 2017; Bonduelle 1991; Buyru 1995; Cameron 1987; Goshtasebi 2013; Kiseli 2016; Kriplani 2006; Zhang 2008), mostly because they did not clearly state their allocation method. Seven trials received a low risk score for allocation based on adequate concealment prior to randomization (Cameron 1990; Hashim 2012; Higham 1993; Irvine 1998; Kaunitz 2010; Preston 1995; Shravage 2011). However, one of these trials (Preston 1995), had a very high dropout rate (more than 50%) after randomization. In this trial, 103 women were randomized to two parallel groups, either norethisterone or tranexamic acid, and proceeded to take placebo medication for two menstrual cycles. Fifty‐seven of the participants (55%) were then excluded from the treatment phase because the measured menstrual blood loss was less than 80 mL per cycle, there was poor compliance or lack of ovulation.

Blinding

Performance bias

Only two trials were double‐blinded and scored as at low risk of performance bias (Preston 1995; Shravage 2011), for objective and subjective outcomes. The remaining trials (Ashraf 2017; Bonduelle 1991; Buyru 1995; Cameron 1987; Cameron 1990; Goshtasebi 2013; Hashim 2012; Higham 1993; Irvine 1998; Kaunitz 2010; Kiseli 2016; Kriplani 2006; Zhang 2008) scored at high risk for performance bias of subjective outcomes, as the lack of blinding was likely to influence the outcomes. For the objective outcome (alkaline haematin) six trials scored at low risk (Cameron 1987; Cameron 1990; Higham 1993; Irvine 1998; Preston 1995; Shravage 2011), as the lack of blinding was unlikely to influence the outcome. We were obliged to score seven trials as unclear for performance bias of objective outcomes, as they did not report alkaline haematin (Ashraf 2017; Goshtasebi 2013; Hashim 2012; Kaunitz 2010; Kiseli 2016; Kriplani 2006; Zhang 2008).

Detection bias

Two of the trials scored at low risk for detection bias of objective and subjective outcomes (Preston 1995; Shravage 2011), both were double‐blind, placebo‐controlled trials. Of the remaining trials, for subjective outcomes two scored as unclear, one because it was single‐blinded and placebo controlled (likely to have blinded the participants; Higham 1993), and the other because it was unclear whether or not personnel were blinded (Goshtasebi 2013). Eleven trials scored at high risk for detection bias (Ashraf 2017; Bonduelle 1991; Buyru 1995; Cameron 1987; Cameron 1990; Hashim 2012; Irvine 1998; Kaunitz 2010; Kiseli 2016; Kriplani 2006; Zhang 2008), as the lack of blinding was likely to influence the outcomes. For the objective outcome, we scored the detection bias of six trials as low (Cameron 1987; Cameron 1990; Higham 1993; Irvine 1998; Preston 1995; Shravage 2011), because the blinding (presence or absence) was unlikely to influence the outcome. The remaining nine trials did not report alkaline haematin so we were obliged to score them as unclear (Ashraf 2017; Bonduelle 1991; Buyru 1995; Goshtasebi 2013; Hashim 2012; Kaunitz 2010; Kiseli 2016; Kriplani 2006; Zhang 2008).

Incomplete outcome data

Four trials were scored as at high risk of attrition bias; one for 20% lost after randomization and further 26% during treatment (Bonduelle 1991), the other three for substantial dropouts or loss to follow up (Cameron 1987; Kiseli 2016; Kriplani 2006).

Four trials scored as at unclear risk of attrition bias; one because no withdrawals or exclusions were reported (Buyru 1995), one for significant loss to follow‐up, but with similar numbers and reasons for each group (Goshtasebi 2013) and one because authors stated that they performed both ITT and per protocol analyses but it appears that this was only for two outcomes (Irvine 1998) and one trial excluded 55% after randomization (Preston 1995).

The remaining seven trials scored at low risk of attrition bias (Ashraf 2017; Cameron 1990; Hashim 2012; Higham 1993; Kaunitz 2010; Shravage 2011; Zhang 2008).

Selective reporting

We scored eight trials as at unclear risk of selective reporting. Most of these took place before 1996 and did not have a registered protocol (Bonduelle 1991; Buyru 1995; Cameron 1987; Cameron 1990; Higham 1993; Kiseli 2016; Preston 1995; Shravage 2011). We scored five trials as at low risk of reporting bias, because they reported all their previously stated outcomes in the results (Ashraf 2017; Goshtasebi 2013; Irvine 1998; Kaunitz 2010; Kriplani 2006). We scored two trials as at high risk for reporting bias (Hashim 2012; Zhang 2008); Hashim 2012 because the protocol in the trial register reported fewer outcomes than those reported in the publication, and Zhang 2008 because the trial authors only reported significant results in the abstract, which appears to favour the experimental treatment. Quality‐of‐life scores after one month were significantly improved after one cycle of treatment in the experimental group, but after two cycles of treatment and at follow‐up the scores were similar. Zhang 2008 did not report this finding, which suggests that the conclusions were influenced by lack of reporting.

The authors of two trials provided original data. In one case, we calculated means and SDs from the data provided by the trial author, since they only reported medians and ranges in the publication (Cameron 1990). Extra data from the other trial allowed us to assess additional outcomes not reported in the publication (Higham 1993).

Other potential sources of bias

Baseline

Almost all trials had similar baseline groups. We only scored one trial as unclear because it had no clear indication that the groups were similar at baseline (Shravage 2011) and one as high risk because of substantial imbalance at baseline (Cameron 1987).

Funding

Two trials were funded by University Grants (Ashraf 2017; Goshtasebi 2013), one by a Birthright Research Grant from the Royal College of Obstetricians & Gynaecologists (Cameron 1987), and one by the Indian Council of Medical Research (Kriplani 2006). Six trials received funding from drug companies (Cameron 1990; Higham 1993; Irvine 1998; Kaunitz 2010; Preston 1995; Zhang 2008). One trial denied receiving any funding (Kiseli 2016). Three trials did not state any funding source (Bonduelle 1991; Buyru 1995; Shravage 2011), and one trial that did not state funding declared that the treatment was provided by a drug company and the sanitary pads were supplied by Industry (Hashim 2012).

Conflict of interest

Only two trials clearly reported not having any conflicts of interest (Ashraf 2017; Hashim 2012). The remaining thirteen studies did not state whether they had any conflicts of interest.

The risk of bias for other potential source of bias was reported according to the 3 criteria above. One trial was scored as high risk of bias for substantial imbalance at baseline in menstrual bleeding (Cameron 1987). Four trials were scored as unclear as no details were provided for the other potential sources (Buyru 1995;Hashim 2012; Shravage 2011) and one had different age groups at baseline and was funded by a drug company (Higham 1993). The remaining ten trials scored low risk of other potential sources of bias (Ashraf 2017; Bonduelle 1991; Cameron 1990; Goshtasebi 2013; Irvine 1998; Kaunitz 2010; Kiseli 2016; Kriplani 2006; Preston 1995; Zhang 2008)

Effects of interventions

See: Table 1; Table 2

Summary of findings for the main comparison. Progestogen therapy (luteal phase only) compared to other medical therapy for heavy menstrual bleeding.

Progestogen therapy (luteal phase only) compared to other medical therapy for heavy menstrual bleeding
Patient or population: women with heavy menstrual bleeding Setting: outpatient Intervention: progestogen therapy, luteal phase only Comparison: other medical therapy
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	№ of participants (trials)	Quality of the evidence (GRADE)	Comments
	Risk with other medical treatment	Risk with progestogen therapy (luteal phase only)
Menstrual blood loss Assessed with alkaline haematin method	The mean menstrual blood loss ranged from 51 to 97	MD 37.29 higher (17.67 higher to 56.91 higher)	‐	145 (4 RCTs)	⊕⊕⊝⊝ Lowa,b
Satisfaction	738 per 1000	693 per 1000 (324 to 915)	OR 0.80 (0.17 to 3.80)	52 (1 RCT)	⊕⊝⊝⊝ Very lowc,d
Days of bleeding	The mean number of days of bleeding ranged from 2 to 7 days	MD 1.05 days more (0.69 more to 1.4 more)	‐	171 (5 RCTs)	⊕⊝⊝⊝ Very lowb,e
Quality of life Physical functioning	The mean quality of life: physical functioning was 14.12	MD 0.43 lower 5.10 lower to 4.24 higher	‐	62 (1 RCT)	⊕⊝⊝⊝ Very lowc,d
Total adverse effects	680 per 1000	591 per 1000 (433 to 731)	OR 0.68 (0.36 to 1.28)	197 (5 RCTs)	⊕⊝⊝⊝ Very lowb,c
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; MD: mean difference; OR: odds ratio; RCT: randomized controlled trial
GRADE Working Group grades of evidence High quality: further research is very unlikely to change our confidence in the estimate of effect. Moderate quality: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Low quality: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Very low quality: we are very uncertain about the estimate.

^aDowngraded one level due to moderate heterogeneity.
 ^bThere are two different luteal phase progestogens, so we downgraded the evidence for this outcome by 1 level.
 ^cDowngraded two levels for high risk of detection and attrition bias.
 ^dBased only on one trial, downgraded one level.
 ^eDowngraded two levels due to high heterogeneity (> 75%).

Summary of findings 2. Progestogen therapy long cycle (3 to 4 weeks) compared to other medical therapy for heavy menstrual bleeding.

Progestogen therapy long cycle (3 to 4 weeks) compared to other medical therapy for heavy menstrual bleeding
Patient or population: women with heavy menstrual bleeding Setting: outpatient Intervention: progestogen therapy long cycle (3 to 4 weeks) Comparison: other medical therapy
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	№ of participants (trials)	Quality of the evidence (GRADE)	Comments
	Risk with other medical therapy	Risk with progestogen long cycle (3 to 4 weeks)
Menstrual blood loss PBAC after 3 months	The mean bleeding PBAC after 3 months ranged from 92‐155	MD 16.88 more (10.93 more to 22.84 more)	‐	355 (4 RCTs)	⊕⊝⊝⊝ Very lowa
Menstrual blood loss PBAC after 6 months	The mean bleeding PBAC after 6 months was 80	MD 21.42 more (14.24 more to 28.6 more)	‐	76 (1 RCT)	⊕⊕⊝⊝ Lowa,b
Satisfaction	708 per 1000	430 per 1000 (240 to 633)	OR 0.31 (0.13 to 0.71)	95 (1 RCT)	⊕⊕⊝⊝ Lowa,b
Days of bleeding	Not reported in any trial
Quality of life	Not reported in any trial
Adverse effect Headache	43 per 1000	62 per 1000 (18 to 194)	OR 1.45 (0.40 to 5.31)	189 (2 RCTs)	⊕⊕⊝⊝ Lowc
Adverse effect Breakthrough bleeding/spotting	189 per 1000	40 per 1000 (14 to 114)	OR 0.18 (0.06 to 0.55)	220 (3 RCTs)	⊕⊕⊝⊝ Lowc
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; MD: mean difference; OR: Odds ratio; PBAC: Pictorial Blood Assessment Chart; RCT: randomized controlled trial
GRADE Working Group grades of evidence High quality: further research is very unlikely to change our confidence in the estimate of effect. Moderate quality: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Low quality: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Very low quality: we are very uncertain about the estimate.

^aHigh risk of detection bias. Unknown risk of selection bias. Downgraded one level.
 ^bBased only on one trial, downgraded one level.
 ^cAll the trials had a high risk of detection bias and unknown risk of selection bias. One had high risk of attrition bias. Downgraded two levels.

1. Oral progestogen therapy versus placebo

No included trials compared placebo to cyclical progestogen.

2. Progestogen therapy (luteal phase only) versus non‐steroidal anti‐inflammatory drugs (NSAIDs)

Two trials with a total of 177 women compared luteal phase norethisterone versus NSAIDs (Cameron 1987; Cameron 1990).

Primary outcomes

2.1 Menstrual blood loss

There was no clear evidence of difference in the menstrual blood loss measured by the alkaline haematin method between luteal phase norethisterone and NSAIDs in either trial (MD 22.97, 95% CI −0.62 to 46.57; I² = 71%; 2 trials; 48 women; Analysis 2.1; Figure 4).

2.1. Analysis.

Comparison 2 Progestogen therapy (luteal phase only) versus NSAIDs, Outcome 1 Menstrual blood loss (alkaline haematin method).

4.

Forest plot of comparison 12: Overall analysis 2: progestogen therapy long cycle (3‐4 weeks) versus other medical therapy, outcome: 12.1 Bleeding PBAC after 3 months.

2.2 Satisfaction

Neither trial reported data on the women's satisfaction with the treatment.

Secondary outcomes

2.3 Days of bleeding

The number of days of bleeding during the intervention menstrual cycle did not report a clear difference between the norethisterone and NSAIDs groups (MD 0.41, 95% CI −0.13 to 0.95; I² = 35%; 2 trials, 48 women; Analysis 2.2).

2.2. Analysis.

Comparison 2 Progestogen therapy (luteal phase only) versus NSAIDs, Outcome 2 Duration of menstruation (days).

2.4 Quality of life

No trials reported data on changes in the quality of life with treatment.

2.5 Compliance

One trial (Cameron 1990), did not report clear evidence of a difference between the luteal phase norethisterone and NSAIDs in women's compliance with treatment (Peto OR 0.88, 95% CI 0.05 to 14.78; 1 trial, 32 women; Analysis 2.3).

2.3. Analysis.

Comparison 2 Progestogen therapy (luteal phase only) versus NSAIDs, Outcome 3 Compliance with treatment.

2.6 Acceptability

Neither of the trials reported data on the acceptability of the treatment for this comparison.

2.7 Adverse events

One trial (Cameron 1990), reported adverse events, of any degree, reported either spontaneously by the woman or elicited from specific questioning. There was no clear evidence of difference between luteal phase norethisterone and NSAIDs in:

1. total adverse events (Peto OR 1.86, 95% CI 0.44 to 7.86; 1 trial, 32 women; Analysis 2.4);

2. headache (Peto OR 1.60, 95% CI 0.35 to 7.31; 1 trial, 32 women; Analysis 2.5);

3. gastrointestinal symptoms (Peto OR 0.56, 95% CI 0.05 to 5.83; 1 trial, 32 women; Analysis 2.6);

4. dysmenorrhoea (Peto OR 1.16, 95% CI 0.20 to 6.69; 1 trial, 32 women; Analysis 2.7).

2.4. Analysis.

Comparison 2 Progestogen therapy (luteal phase only) versus NSAIDs, Outcome 4 Adverse events (total).

2.5. Analysis.

Comparison 2 Progestogen therapy (luteal phase only) versus NSAIDs, Outcome 5 Adverse events ‐ headache.

2.6. Analysis.

Comparison 2 Progestogen therapy (luteal phase only) versus NSAIDs, Outcome 6 Adverse events ‐ gastrointestinal symptoms.

2.7. Analysis.

Comparison 2 Progestogen therapy (luteal phase only) versus NSAIDs, Outcome 7 Adverse events ‐ dysmenorrhoea.

2.8 Resource use and cost

The included trials did not report data on resource use or costs for this comparison.

3. Progestogen therapy (luteal phase only) versus danazol

There were four trials comparing cyclical progestogen with danazol, all published before 1996. Three trials (Bonduelle 1991; Buyru 1995; Cameron 1987), used 200 mg danazol daily and one (Higham 1993), used decreasing dosages of danazol (200 mg, 100 mg and 50 mg per cycle) with a total of 107 women.

Primary outcomes

3.1 Menstrual blood loss

There was clear evidence of a difference in menstrual blood loss, measured by the alkaline haematin method, that favoured danazol compared with luteal phase norethisterone (MD 55.63, 95% CI 14.73 to 96.54; I² = 0%; 2 trials, 51 women; Analysis 3.1 (Cameron 1987; Higham 1993)).

3.1. Analysis.

Comparison 3 Progestogen therapy (luteal phase only) versus danazol, Outcome 1 Menstrual blood loss (alkaline haematin method).

There was clear evidence of a difference that favoured danazol in relation to the proportion of women with improvement in menstrual blood loss measured subjectively in two trials (Peto OR 0.24, 95% CI 0.08 to 0.69; I² = 0%; 2 trials, 54 women; Analysis 3.2 (Bonduelle 1991; Higham 1993)).

3.2. Analysis.

Comparison 3 Progestogen therapy (luteal phase only) versus danazol, Outcome 2 Proportion of women with improvement in MBL (subjective).

3.2 Satisfaction

None of the trials reported data on the women's satisfaction with the treatment.

Secondary outcomes

3.3 Days of bleeding

All four trials reported the number of days bleeding during the intervention menstrual cycle, reporting clear evidence of a difference that favoured danazol (MD 1.60, 95% CI 1.10 to 2.11; I² = 85%; 4 trials, 107 women; Analysis 3.3).

3.3. Analysis.

Comparison 3 Progestogen therapy (luteal phase only) versus danazol, Outcome 3 Duration of menstruation (days).

3.4 Quality of life

None of the trials reported data on women's satisfaction or perceived change in quality.

3.5 Compliance

Two trials (Bonduelle 1991; Higham 1993), reported no clear evidence of a difference in women's compliance with treatment between luteal phase norethisterone and danazol (Peto OR 1.56, 95% CI 0.45 to 5.45; I² = 0%; 2 trials, 61 women; Analysis 3.4).

3.4. Analysis.

Comparison 3 Progestogen therapy (luteal phase only) versus danazol, Outcome 4 Compliance with treatment.

3.6 Acceptability

None of the trials reported data on the acceptability of the treatment for this comparison.

3.7 Adverse events

Three trials (Bonduelle 1991; Buyru 1995; Higham 1993), reported adverse events, of any degree, reported either spontaneously by the woman or elicited from specific questioning. Two trials (Buyru 1995; Higham 1993), reported no clear evidence of difference in total adverse events between luteal phase norethisterone and danazol (Peto OR 0.34, 95% CI 0.13 to 0.88; I² = 0%; 2 trials, 77 women; Analysis 3.5).

3.5. Analysis.

Comparison 3 Progestogen therapy (luteal phase only) versus danazol, Outcome 5 Adverse events (total).

There was clear evidence that favoured luteal progestogen for:

1. headache (Peto OR 0.29, 95% CI 0.09 to 0.95; I² = 0%; 2 trials, 64 women; Analysis 3.6 (Bonduelle 1991; Buyru 1995));

2. gastrointestinal symptoms (Peto OR 0.22, 95% CI 0.06 to 0.90; I² = 0%; 2 trials, 64 women; Analysis 3.7 (Bonduelle 1991; Buyru 1995));

3. weight gain (Peto OR 0.24, 95% CI 0.09 to 0.65; I² = 0%; 3 trials, 101 women; Analysis 3.8 (Bonduelle 1991; Buyru 1995; Higham 1993));

4. skin changes (Peto OR 0.18, 95% CI 0.04 to 0.75; I² = 0%; 2 trials, 64 women; Analysis 3.9 (Bonduelle 1991; Buyru 1995)).

3.6. Analysis.

Comparison 3 Progestogen therapy (luteal phase only) versus danazol, Outcome 6 Adverse events ‐ headache.

3.7. Analysis.

Comparison 3 Progestogen therapy (luteal phase only) versus danazol, Outcome 7 Adverse events ‐ gastrointestinal symptoms.

3.8. Analysis.

Comparison 3 Progestogen therapy (luteal phase only) versus danazol, Outcome 8 Adverse events ‐ weight gain.

3.9. Analysis.

Comparison 3 Progestogen therapy (luteal phase only) versus danazol, Outcome 9 Adverse events ‐ skin changes.

There was clear evidence that favoured danazol for:

1. breast changes (Peto OR 8.73, 95% CI 1.14 to 67.13; 1 trial, 40 women; Analysis 3.15 (Buyru 1995)).

3.15. Analysis.

Comparison 3 Progestogen therapy (luteal phase only) versus danazol, Outcome 15 Adverse events ‐ breast changes.

There was no clear evidence of difference in:

1. neurological changes (Peto OR 1.00, 95% CI 0.20 to 4.90; 1 trial, 24 women; Analysis 3.10 (Bonduelle 1991));

2. bloating (Peto OR 0.21, 95% CI 0.04 to 1.07; 1 trial, 24 women; Analysis 3.11 (Bonduelle 1991));

3. voice change (Peto OR 6.00, 95% CI 0.34 to 106.33; 1 trial, 24 women; Analysis 3.12 (Bonduelle 1991));

4. muscle cramps (Peto OR 1.40, 95% CI 0.28 to 7.02; 1 trial, 40 women; Analysis 3.13 (Buyru 1995));

5. intermenstrual bleeding (Peto OR 0.26, 95% CI 0.04 to 1.67; 1 trial, 40 women; Analysis 3.14 (Buyru 1995)).

3.10. Analysis.

Comparison 3 Progestogen therapy (luteal phase only) versus danazol, Outcome 10 Adverse events ‐ neurological changes.

3.11. Analysis.

Comparison 3 Progestogen therapy (luteal phase only) versus danazol, Outcome 11 Adverse events ‐ bloating.

3.12. Analysis.

Comparison 3 Progestogen therapy (luteal phase only) versus danazol, Outcome 12 Adverse events ‐ voice changes.

3.13. Analysis.

Comparison 3 Progestogen therapy (luteal phase only) versus danazol, Outcome 13 Adverse events ‐ muscle cramps.

3.14. Analysis.

Comparison 3 Progestogen therapy (luteal phase only) versus danazol, Outcome 14 Adverse events ‐ intermenstrual bleeding.

3.8 Resource use and cost

The included trials did not report data on resource use or costs for this comparison.

4. Progestogen therapy (luteal phase only) versus tranexamic acid

Three trials with a total of 216 women compared luteal norethisterone and tranexamic acid (Kiseli 2016; Preston 1995; Zhang 2008).

Primary outcomes

4.1 Menstrual blood loss

One trial (Preston 1995), showed clear evidence of a difference in menstrual blood loss measured by the alkaline haematin method favouring tranexamic acid when compared to luteal phase norethisterone (MD 111.00, 95% CI 43.54 to 178.46; 1 trial, 46 women; Analysis 4.1).

4.1. Analysis.

Comparison 4 Progestogen therapy (luteal phase only) versus tranexamic acid, Outcome 1 Menstrual blood loss (alkaline haematin method).

One trial (Zhang 2008), showed no clear evidence of a difference between luteal phase norethisterone and tranexamic acid on the PBAC at three months (MD 25.00, 95% CI −15.54 to 65.54; 1 trial, 128 women; Analysis 4.2).

4.2. Analysis.

Comparison 4 Progestogen therapy (luteal phase only) versus tranexamic acid, Outcome 2 Bleeding PBAC at 3 months.

Three trials (Kiseli 2016; Preston 1995; Zhang 2008), showed clear evidence of a difference in the improvement of bleeding favouring tranexamic acid (Peto OR 0.43, 95% CI 0.25 to 0.75;; 3 trials, 216 women; I² = 9%; Analysis 4.3).

4.3. Analysis.

Comparison 4 Progestogen therapy (luteal phase only) versus tranexamic acid, Outcome 3 Bleeding: improvement rates.

4.2 Satisfaction

One trial (Kiseli 2016), showed no clear evidence of a difference between luteal phase norethisterone and tranexamic acid for satisfaction with treatment (OR 1.33, 95% CI 0.37 to 4.85; 1 trial, 42 women; Analysis 4.4).

4.4. Analysis.

Comparison 4 Progestogen therapy (luteal phase only) versus tranexamic acid, Outcome 4 Satisfaction with treatment.

Secondary outcomes

4.3 Days of bleeding

None of the trials reported data for the number of days of bleeding.

4.4 Quality of life

There was no evidence of difference in the improvement in general health (Peto OR 0.45, 95% CI 0.14 to 1.47), dysmenorrhoea symptoms (Peto OR 0.44, 95% CI 0.12 to 1.59), ability to enjoy social activities (Peto OR 0.42, 95% CI 0.13 to 1.38) and improvement in sexual life (OR 0.29, 95% CI 0.07 to 1.29) between groups according to one trial with 44 women (Preston 1995). The same trial reported clear evidence of a difference favouring tranexamic acid for improvement of episodes of flooding and leakage (OR 0.16, 95% CI 0.04 to 0.66). See Analysis 4.5.

4.5. Analysis.

Comparison 4 Progestogen therapy (luteal phase only) versus tranexamic acid, Outcome 5 Proportion with improvement in quality of life/dysmenorrhoea.

One trial (Kiseli 2016) with 42 women showed no clear evidence of a difference between luteal phase norethisterone and tranexamic acid in quality of life measured with SF36 (Analysis 4.6) in the following areas:

4.6. Analysis.

Comparison 4 Progestogen therapy (luteal phase only) versus tranexamic acid, Outcome 6 Quality of life SF36.

1. physical functioning (MD −0.73, 95% CI −2.25 to 0.79);

2. social functioning (MD −0.37, 95% CI −2.46 to 1.72);

3. mental health (MD −0.27, 95% CI −1.73 to 1.19);

4. environmental domain (MD −1.20, 95% CI −2.57 to 0.17).

4.5 Compliance

None of the trials reported data on compliance with treatment for this comparison.

4.6 Acceptability

None of the trials reported data on the acceptability of the treatment for this comparison.

4.7 Adverse events

There was no clear evidence of a difference between luteal phase norethisterone and tranexamic acid in adverse events:

1. total adverse effects (Peto OR 0.93, 95% CI 0.33 to 2.64; I² = 15%; 2 trials, 88 women; Analysis 4.7) and gastrointestinal symptoms (Peto OR 1.69, 95% CI 0.54 to 5.34; I² = 67%; 2 trials, 88 women; Analysis 4.9) in two trials (Kiseli 2016; Preston 1995);

2. headache (Peto OR 1.90, 95% CI 0.59 to 6.15; Analysis 4.8), weight gain (Peto OR 0.15, 95% CI 0.01 to 2.54; Analysis 4.10), dysmenorrhoea (Peto OR 1.48, 95% CI 0.32 to 6.71; Analysis 4.12) in one trial with 46 women (Preston 1995);

3. bloating (Peto OR 8.61, 95% CI 0.52 to 142.87; Analysis 4.11) in one trial with 42 women (Kiseli 2016).

4.7. Analysis.

Comparison 4 Progestogen therapy (luteal phase only) versus tranexamic acid, Outcome 7 Adverse events (total).

4.9. Analysis.

Comparison 4 Progestogen therapy (luteal phase only) versus tranexamic acid, Outcome 9 Adverse events ‐ gastrointestinal symptoms.

4.8. Analysis.

Comparison 4 Progestogen therapy (luteal phase only) versus tranexamic acid, Outcome 8 Adverse events ‐ headache.

4.10. Analysis.

Comparison 4 Progestogen therapy (luteal phase only) versus tranexamic acid, Outcome 10 Adverse events ‐ weight gain.

4.12. Analysis.

Comparison 4 Progestogen therapy (luteal phase only) versus tranexamic acid, Outcome 12 Adverse events ‐ dysmenorrhoea.

4.11. Analysis.

Comparison 4 Progestogen therapy (luteal phase only) versus tranexamic acid, Outcome 11 Adverse events ‐ bloating.

4.8 Resource use and cost

The included trials did not report data on resource use or costs for this comparison.

5. Progestogen therapy (luteal phase only) versus progestogen‐releasing intrauterine system

Two trials with a total of 49 women (Cameron 1987; Kiseli 2016), compared luteal phase norethisterone versus progestogen‐releasing intrauterine devices: one trial (Cameron 1987), used a device releasing 65 μg progesterone daily (Pg‐IUS), and the other trial (Kiseli 2016), used an intrauterine system releasing 20 μg of levonorgestrel daily (LNG‐IUS).

Primary outcomes

5.1 Menstrual blood loss

One trial (Cameron 1987), showed clear evidence of a difference in menstrual blood loss between luteal phase norethisterone and Pg‐IUS, favouring the Pg‐IUS, using the alkaline haematin method (MD 51.00, 95% CI 18.38 to 83.62; 1 trial, 16 women; Analysis 5.1).

5.1. Analysis.

Comparison 5 Progestogen therapy (NE luteal phase only) versus Progestogen‐ IUS, Outcome 1 Menstrual blood loss (alkaline haematin method).

One trial (Kiseli 2016), showed clear evidence of an improvement in menstrual blood loss between luteal phase norethisterone and LNG‐IUS, favouring the LNG‐IUS, assessed by the proportion of women with a PBAC less than 100 at six‐month follow‐up (Peto OR 0.11, 95% CI 0.03 to 0.38; 1 trial, 40 women; Analysis 5.2).

5.2. Analysis.

Comparison 5 Progestogen therapy (NE luteal phase only) versus Progestogen‐ IUS, Outcome 2 Improvement in MBL PBAC < 100 at 6 months.

5.2 Satisfaction

One trial (Kiseli 2016), showed no clear evidence of difference between treatments on the satisfaction rate (Peto OR 0.43, 95% CI 0.10 to 1.87; 1 trial, 40 women; Analysis 5.3).

5.3. Analysis.

Comparison 5 Progestogen therapy (NE luteal phase only) versus Progestogen‐ IUS, Outcome 3 Satisfaction with treatment.

Secondary outcomes

5.3 Days of bleeding

One trial (Cameron 1987), showed no clear evidence of difference between treatments in the number of days of bleeding (MD 1.00, 95% CI −0.19 to 2.19; 1 trial, 16 women; Analysis 5.4).

5.4. Analysis.

Comparison 5 Progestogen therapy (NE luteal phase only) versus Progestogen‐ IUS, Outcome 4 Duration of menstruation (days).

5.4 Quality of life

One trial with 40 women (Kiseli 2016), showed no clear evidence of a difference between treatments on quality of life in the following areas (Analysis 5.5):

5.5. Analysis.

Comparison 5 Progestogen therapy (NE luteal phase only) versus Progestogen‐ IUS, Outcome 5 Quality of life.

1. physical domain (MD 0.03, 95% CI −6.21 to 6.27);

2. psychological domain (MD −0.67, 95% CI −7.49 to 6.15);

3. social domain (MD −0.14, 95% CI −9.33 to 9.05);

4. environmental domain (MD −0.20, 95% CI −6.34 to 5.94).

5.5 Compliance

Neither of the included trials reported data on compliance with treatment for this comparison.

5.6 Acceptability

Neither of the included trials reported data on the acceptability of the treatment for this comparison.

5.7 Adverse events

Neither of the included trials reported data on side effects for this comparison.

5.8 Resource use and cost

Neither of the included trials reported data on resource use or costs for this comparison.

6. Progestogen therapy (medroxyprogesterone acetate, (luteal phase only) versus progestogen‐releasing intrauterine system

One trial, with 162 women in three countries (Kaunitz 2010), compared 10 mg a day of medroxyprogesterone acetate from day 16 to 25 of the cycle versus LNG‐IUS.

Primary outcomes

6.1 Menstrual blood loss

There was clear evidence of a difference in menstrual blood loss, measured by a reduction of PBAC from baseline, favouring the LNG‐IUS (MD −49.30, 95% CI −69.91 to −28.69; 1 trial, 165 women; Analysis 6.1).

6.1. Analysis.

Comparison 6 Progestogen therapy (MPA luteal phase only) versus LNG‐IUS, Outcome 1 Menstrual blood loss: PBAC reduction from baseline.

6.2 Satisfaction

Kaunitz 2010 did not report data on satisfaction with treatment.

Secondary outcomes

6.3 Days of bleeding

Kaunitz 2010 did not report data on days of bleeding for this comparison.

6.4 Quality of life

Kaunitz 2010 did not report data on quality of life for this comparison.

6.5 Compliance

Kaunitz 2010 did not report data on compliance with treatment for this comparison.

6.6 Acceptability

Kaunitz 2010 did not report data on the acceptability of the treatment for this comparison.

6.7 Adverse events

There was clear evidence of a difference in the adverse effect 'ovarian cyst', favouring luteal progestogen (OR 0.17, 95% CI 0.04 to 0.83; Analysis 6.2). All the ovarian cysts in the trial were characterized as mild events that were thought to be related to treatment with trial medication.

6.2. Analysis.

Comparison 6 Progestogen therapy (MPA luteal phase only) versus LNG‐IUS, Outcome 2 Adverse effects.

Kaunitz 2010 showed no clear evidence of a difference in 162 women for the following adverse effects (Analysis 6.2).

1. Pelvic pain (Peto OR 0.49, 95% CI 0.10 to 2.49)

2. Headache (Peto OR 1.56, 95% CI 0.64 to 3.83)

3. Breast tenderness (Peto OR 0.72, 95% CI 0.16 to 3.28)

4. Vaginitis (Peto OR 3.00, 95% CI 0.93 to 9.70)

5. Acne (Peto OR 1.03, 95% CI 0.29 to 3.68)

6. Hypertension (Peto OR 4.06, 95% CI 0.80 to 20.65)

7. Sinusitis (Peto OR 1.73, 95% CI 0.42 to 7.14)

8. Fatigue (Peto OR 2.04, 95% CI 0.40 to 10.38)

9. Urinary tract infection (Peto OR 2.07, 95% CI 0.54 to 7.91)

10. Increased weight (Peto OR 0.81, 95% CI 0.21 to 3.10)

6.8 Resource use and cost

Kaunitz 2010 did not report data on resource use or costs for this comparison.

7. Progestogen therapy (3 to 4 weeks) versus progestogen‐releasing intrauterine system (LNG‐IUS)

Two trials, with 120 women in total (Ashraf 2017; Irvine 1998), compared norethisterone from day 5 to 26 of the cycle and the LNG‐IUS.

Primary outcomes

7.1 Menstrual blood loss

One trial (Irvine 1998), showed clear evidence of a difference between groups for menstrual blood loss favouring the LNG‐IUS, as measured by alkaline haematin; although there was a substantial reduction from baseline in both groups. We have not included these results in the graphical display because the data were not normally distributed (see Analysis 7.1).

7.1. Analysis.

Comparison 7 Progestogen therapy (3 to 4 weeks) versus LNG‐IUS, Outcome 1 Menstrual blood loss (alkaline haematin method).

Menstrual blood loss (alkaline haematin method)
Study
Irvine 1998	Median menstrual blood loss (range) after 3 months of NET: 20 mls (4‐137) Median menstrual blood loss (range) after 3 months of LNG IUS: 6 mls (0‐284) p=0.033, t=315.5, Wilcoxon rank‐sum test, n=44

There was clear evidence of a difference in the proportion of women with menstrual blood loss lower than 80 mL after treatment favouring the LNG‐IUS (Peto OR 0.23, 95% CI 0.09 to 0.57; I² = 0%; 2 trials, 120 women; Analysis 7.2).

7.2. Analysis.

Comparison 7 Progestogen therapy (3 to 4 weeks) versus LNG‐IUS, Outcome 2 Proportion of women with MBL < 80 mL after treatment.

One trial, with 76 women (Ashraf 2017), did not report a clear difference in menstrual blood loss at three months' follow‐up using the PBAC (MD 29.00, 95% CI −609.14 to 667.14; Analysis 7.3); but the same trial reported clear evidence of a difference favouring the LNG‐IUS at six months (MD 21.42, 95% CI 14.24 to 28.60; Analysis 7.4).

7.3. Analysis.

Comparison 7 Progestogen therapy (3 to 4 weeks) versus LNG‐IUS, Outcome 3 PBAC at 3 months.

7.4. Analysis.

Comparison 7 Progestogen therapy (3 to 4 weeks) versus LNG‐IUS, Outcome 4 PBAC at 6 months.

7.2 Satisfaction

Neither of the included trials reported women's satisfaction with the treatment for this comparison.

Secondary outcomes

7.3 Days of bleeding

Neither of the included trials reported days of bleeding for this comparison.

7.4 Quality of life

There was no evidence of a clear difference in improvement in quality of life between groups for dysmenorrhoea (Peto OR 2.12, 95% CI 0.42 to 10.79; 1 trial, 31 women; Analysis 7.5).

7.5. Analysis.

Comparison 7 Progestogen therapy (3 to 4 weeks) versus LNG‐IUS, Outcome 5 Proportion with improvement in quality of life‐dysmenorrhoea.

7.5 Compliance

Neither of the included trials reported compliance with treatment for this comparison.

7.6 Acceptability

There was clear evidence of a difference in treatment acceptability of the treatment favouring the LNG‐IUS (Peto OR 0.12, 95% CI 0.03 to 0.40: 1 trial, 40 women; Analysis 7.6).

7.6. Analysis.

Comparison 7 Progestogen therapy (3 to 4 weeks) versus LNG‐IUS, Outcome 6 Proportion who find treatment acceptable.

7.7 Adverse events

There was clear evidence of differences in the following side effects favouring the oral progestogens:

1. intermenstrual bleeding (Peto OR 0.23, 95% CI 0.05 to 0.99; 1 trial, 31 women; Analysis 7.7);

2. breast tenderness (Peto OR 0.11, 95% CI 0.03 to 0.46; 1 trial, 31 women; Analysis 7.8).

7.7. Analysis.

Comparison 7 Progestogen therapy (3 to 4 weeks) versus LNG‐IUS, Outcome 7 Adverse events ‐ intermenstrual bleeding.

7.8. Analysis.

Comparison 7 Progestogen therapy (3 to 4 weeks) versus LNG‐IUS, Outcome 8 Adverse events ‐ breast tenderness.

There was no clear evidence of a difference between groups for mood swings (Peto OR 0.82, 95% CI 0.19 to 3.54; 1 trial, 31 women; Analysis 7.9).

7.9. Analysis.

Comparison 7 Progestogen therapy (3 to 4 weeks) versus LNG‐IUS, Outcome 9 Adverse events ‐ mood swings.

7.8 Resource use and cost

Neither of the included trials reported data on resource use or costs for this comparison.

8. Progestogen therapy (3 to 4 weeks) versus tranexamic acid

Two trials, with 184 women in total (Goshtasebi 2013; Kriplani 2006), compared medroxyprogesterone acetate for 21 days per cycle (from day 5 to 26) with tranexamic acid.

Primary outcomes

8.1 Menstrual blood loss

There was no clear evidence of a difference between groups in the PBAC at three months (MD 8.82, 95% CI −12.09 to 29.73; I² = 0%; 2 trials, 184 women; Analysis 8.1).

8.1. Analysis.

Comparison 8 Progestogen therapy (3 to 4 weeks) versus tranexamic, Outcome 1 Bleeding PBAC at 3 months.

One trial (Kriplani 2006), showed clear evidence of an improvement in bleeding rates favouring tranexamic acid (Peto OR 0.20, 95% CI 0.07 to 0.59; 1 trial, 94 women; Analysis 8.2).

8.2. Analysis.

Comparison 8 Progestogen therapy (3 to 4 weeks) versus tranexamic, Outcome 2 Bleeding: improvement rate.

8.2 Satisfaction

The included trials did not report satisfaction with treatment.

Secondary outcomes

8.3 Days of bleeding

The included trials did not report days of bleeding.

8.4 Quality of life

One trial with 90 women (Goshtasebi 2013), reported no clear evidence of a difference in quality of life between treatments using the SF36 for the following aspects (Analysis 8.3).

8.3. Analysis.

Comparison 8 Progestogen therapy (3 to 4 weeks) versus tranexamic, Outcome 3 Quality of life SF‐36.

1. Physical functioning (MD −1.50, 95% CI −6.77 to 3.77)

2. Role physical (MD −4.00, 95% CI −13.63 to 5.63)

3. Bodily pain (MD 2.80, 95% CI −6.01 to 11.61)

4. General health (MD −5.00, 95% CI −12.49 to 2.49)

5. Vitality (MD −1.40, 95% CI −9.42 to 6.62)

6. Social functioning (MD 1.10, 95% CI −5.36 to 7.56)

7. Role emotional (MD 0.50, 95% CI −11.47 to 12.47)

8. Mental health (MD 2.90, 95% CI −4.83 to 10.63)

8.5 Compliance

The included trials did not report compliance with treatment for this comparison.

8.6 Acceptability

The included trials did not report the acceptability of the treatment for this comparison.

8.7 Adverse events

One trial, with 94 women (Kriplani 2006), showed no clear evidence of a difference between treatments for any of the following adverse effects (Analysis 8.4).

8.4. Analysis.

Comparison 8 Progestogen therapy (3 to 4 weeks) versus tranexamic, Outcome 4 Adverse effects.

1. Any adverse effects (Peto OR 0.40, 95% CI 0.16 to 1.03)

2. Gastrointestinal event (Peto OR 1.39, 95% CI 0.23 to 8.36)

3. Headache (Peto OR 1.39, 95% CI 0.23 to 8.36)

4. Allergic reaction (Peto OR 6.81, 95% CI 0.13 to 344.42)

5. Intermenstrual bleeding (Peto OR 0.11, 95% CI 0.02 to 0.68)

6. Giddiness (Peto OR 0.33, 95% CI 0.04 to 2.39)

7. Breast tenderness (Peto OR 0.12, 95% CI 0.01 to 1.97)

8. Mood changes (Peto OR 0.12, 95% CI 0.00 to 6.26)

8.8 Resource use and cost

The included trials did not report data on resource use or costs for this comparison.

9. Progestogen therapy (3 to 4 weeks) versus ormeloxifene

One trial, with 72 women (Shravage 2011), compared medroxyprogesterone acetate and ormeloxifene. Shravage 2011 presented data in a format that did not allow us to combine them. We contacted the trial authors for clarification.

At the end of the first month, no statistically significant difference was reported in terms of PBAC. At the end of the third month, there was clear evidence of a higher reduction in PBAC score favouring ormeloxifene (P = 0.014).

Mean PBAC after 3 months' treatment
Mean PBAC	Ormeloxifene twice a week	Medroxyprogesterone acetate 21 d
< 100	29	11
100‐185	7	22
> 185	1	2
Total	42	42

10. Progestogen therapy (3 to 4 weeks) versus combined vaginal ring

One trial, with 95 women (Hashim 2012), compared norethisterone for 21 days (day 5 to 26 of the cycle) and the combined vaginal ring. The combined vaginal ring is a polymeric drug delivery device designed to provide controlled release of drugs for intravaginal administration over extended periods of time. The combined vaginal ring releases 15 ug of ethinyl oestradiol and 120 ug of etonogestrel daily over a single cycle, the ring was inserted between days 1 and 5 of the menstrual cycle for 3 weeks, followed by a 1‐week, ring‐free period.

Primary outcomes

10.1 Menstrual blood loss

There was no clear evidence of difference between treatments in menstrual blood loss at the end of the trial (MD 2.10, 95% CI −8.15 to 12.35; Analysis 10.1) or in the percentage change or reduction in menstrual blood loss (Analysis 10.2).

10.1. Analysis.

Comparison 10 Progestogen therapy (3‐4 weeks) versus combined vaginal ring, Outcome 1 Menstrual blood loss: PBAC after 3 cycles.

10.2. Analysis.

Comparison 10 Progestogen therapy (3‐4 weeks) versus combined vaginal ring, Outcome 2 Percentage reduction in MBL (at end of study).

Percentage reduction in MBL (at end of study)
Study	Comparison	N	Results	Conclusion/comment
Hashim 2012	Oral Pg (NET) versus CVR	N = 95 (47 in NET group and 48 in CVR group)	Mean (SD) PBAC score reduction (no measure of variation reported) NET: 69.5 CVR: 68.6	The authors concluded that there were no differences between randomised groups

10.2 Satisfaction

There was clear evidence of a difference in satisfaction with treatment, favouring the combined vaginal ring (Peto OR 0.32, 95% CI 0.14 to 0.72; Analysis 10.3).

10.3. Analysis.

Comparison 10 Progestogen therapy (3‐4 weeks) versus combined vaginal ring, Outcome 3 Satisfaction with treatment.

Secondary outcomes

10.3 Days of bleeding

Hashim 2012 did not report any data on days of bleeding for this comparison.

10.4 Quality of life

There was no clear evidence of a difference between treatments for:

1. HRQoL‐ 4 self‐rated health (Peto OR 0.78, 95% CI 0.33 to 1.82; Analysis 10.4);

2. HRQoL‐4 impairment or lost days in terms of feeling physically unwell (MD 0.20, 95% CI −0.28 to 0.68) or mentally unwell (MD 0.40, 95% CI −0.10 to 0.90; Analysis 10.5).

10.4. Analysis.

Comparison 10 Progestogen therapy (3‐4 weeks) versus combined vaginal ring, Outcome 4 Quality of life (HRQoL‐4 self‐rated health).

10.5. Analysis.

Comparison 10 Progestogen therapy (3‐4 weeks) versus combined vaginal ring, Outcome 5 Quality of life (HRQoL‐4 impairment or lost days).

There was clear evidence of a difference favouring the combined vaginal ring in the number of lost days (days with no regular activities) (MD 0.90, 95% CI 0.38 to 1.42; Analysis 10.5).

10.5 Compliance

Hashim 2012 did not report any data on compliance with treatment for this comparison.

10.6 Acceptability

Hashim 2012 did not report any data on the acceptability of the treatment for this comparison.

10.7 Adverse events

There was no clear evidence of a difference for the following adverse effects (Analysis 10.6).

10.6. Analysis.

Comparison 10 Progestogen therapy (3‐4 weeks) versus combined vaginal ring, Outcome 6 Adverse events.

1. Nausea (Peto OR 2.02, 95% CI 0.20 to 19.90)

2. Headache (Peto OR 0.67, 95% CI 0.11 to 4.04)

3. Breast tenderness (Peto OR 1.55, 95% CI 0.26 to 9.31)

4. Breakthrough bleeding or spotting (Peto OR 3.01, 95% CI 0.71 to 12.73)

5. Vaginal discharge (leukorrhoea) (Peto OR 0.24, 95% CI 0.05 to 1.27)

6. Vaginal discomfort (Peto OR 0.14, 95% CI 0.01 to 2.20)

7. Vaginitis (Peto OR 0.29, 95% CI 0.05 to 1.75)

10.8 Resource use and cost

Hashim 2012 did not report data on resource use or costs for this comparison.

11. Overall analysis 1: progestogen therapy (luteal phase only) versus any other medical therapy (all types combined)

Primary outcomes

11.1 Menstrual blood loss

There was clear evidence of a difference between luteal progestogen and other medical therapy, favouring other medical therapy. Measuring blood loss using the alkaline haematin method, the progestogen group blood loss was on average 37 mL higher than on the other medical treatments (MD 37.29, 95% CI 17.67 to 56.91; I² = 50%; 4 trials, 145 women; low‐quality evidence; Analysis 11.1).

11.1. Analysis.

Comparison 11 Overall analysis 1: progestogen therapy (luteal phase only) versus other medical therapy, Outcome 1 Menstrual blood loss (alkaline haematin method).

11.2 Satisfaction

There was no clear evidence of a difference between treatments in terms of satisfaction (OR 0.80, 95% CI 0.17 to 3.80; I² = 0%; 1 trial, 52 women; very low‐quality evidence; Analysis 11.2).

11.2. Analysis.

Comparison 11 Overall analysis 1: progestogen therapy (luteal phase only) versus other medical therapy, Outcome 2 Satisfaction.

Secondary outcomes

11.3 Days of bleeding

There was clear evidence of a difference in the number of days of bleeding between treatments favouring other medical treatments (MD 1.05, 95% CI 0.69 to 1.40; I² = 81%; 5 trials, 171 women; very low‐quality evidence; Analysis 11.3). The high heterogeneity is probably because of the nature of the outcome. It is inherently unreliable as it is subjective and self‐reported, a day has 24 hours and it is not specific to the days of menstrual bleeding (so it could include days of intermenstrual bleeding too).

11.3. Analysis.

Comparison 11 Overall analysis 1: progestogen therapy (luteal phase only) versus other medical therapy, Outcome 3 Menstruation days.

11.4 Quality of life

There was no clear evidence of a difference between treatments in terms of physical well‐being (MD −0.43, 95% CI −5.10 to 4.24; I² = 0%; 1 trial, 62 women; very low‐quality evidence; Analysis 11.4).

11.4. Analysis.

Comparison 11 Overall analysis 1: progestogen therapy (luteal phase only) versus other medical therapy, Outcome 4 Quality of life: physical functioning.

11.5 Compliance

The included trials did not report any data on compliance with treatment for this comparison.

11.6 Acceptability

The included trials did not report any data on the acceptability of the treatment for this comparison.

11.7 Adverse events

There was no clear evidence of a difference between treatments on total adverse effects (OR 0.68, 95% CI 0.36 to 1.28; I² = 31%; 5 trials, 197 women; very low‐quality evidence; Analysis 11.5).

11.5. Analysis.

Comparison 11 Overall analysis 1: progestogen therapy (luteal phase only) versus other medical therapy, Outcome 5 Total adverse effects.

11.8 Resource use and cost

The included trials did not report data on resource use or costs for this comparison.

12. Overall analysis 2: progestogen therapy long cycle (3 to 4 weeks) versus any other medical therapy (all types combined)

Primary outcomes

12.1 Menstrual blood loss

There was clear evidence of a difference in the PBAC:

1. at three months' follow‐up between groups, favouring other medical therapy (MD 16.88, 95% CI 10.93 to 22.84; I² = 87%; 4 trials, 355 women; very low‐quality evidence; Analysis 12.1). Heterogeneity is high because the comparison is versus other medical treatments in general and the other medical treatments vary considerably.

2. at six months' follow‐up between groups, favouring other medical therapy (MD 21.42, 95% CI 14.24 to 28.60; 1 trial, 76 women; low‐quality evidence; Analysis 12.2).

12.1. Analysis.

Comparison 12 Overall analysis 2: progestogen therapy long cycle (3‐4 weeks) versus other medical therapy, Outcome 1 Bleeding PBAC after 3 months.

12.2. Analysis.

Comparison 12 Overall analysis 2: progestogen therapy long cycle (3‐4 weeks) versus other medical therapy, Outcome 2 Bleeding PBAC after 6 months.

12.2 Satisfaction

There was evidence of a difference in terms of satisfaction favouring other medical therapies (OR 0.31, 95% CI 0.13 to 0.71; 1 trial, 95 women; low‐quality evidence; Analysis 12.3).

12.3. Analysis.

Comparison 12 Overall analysis 2: progestogen therapy long cycle (3‐4 weeks) versus other medical therapy, Outcome 3 Satisfaction.

Secondary outcomes

12.3 Days of bleeding

None of the included trials reported data on the number of days of bleeding for this comparison.

12.4 Quality of life

The included trials did not report data on quality of life for this comparison.

12.5 Compliance

The included trials did not report data on compliance with treatment for this comparison.

12.6 Acceptability

There was clear evidence of difference on acceptability of treatment favours other medical treatment, specifically LNG IUS, compared to long cycle oral progesterone (OR 0.08, 95% CI 0.02 to 0.37; 40 women; one study; Analysis 12.4).

12.4. Analysis.

Comparison 12 Overall analysis 2: progestogen therapy long cycle (3‐4 weeks) versus other medical therapy, Outcome 4 Treatment acceptability.

12.7 Adverse events

1. There was no clear evidence of a difference between progestogen therapy long cycle and other medical therapy in terms of headache (OR 1.45, 95% CI 0.40 to 5.31; I² = 0%; 2 trials, 189 women; low‐quality evidence; Analysis 12.5).

2. There was clear evidence of a difference between groups in terms of breakthrough bleeding or spotting favouring the progestogen long cycle (OR 0.18, 95% CI 0.06 to 0.55; I² = 0%; 3 trials, 220 women; low‐quality evidence; Analysis 12.6).

12.5. Analysis.

Comparison 12 Overall analysis 2: progestogen therapy long cycle (3‐4 weeks) versus other medical therapy, Outcome 5 Adverse effect ‐ headache.

12.6. Analysis.

Comparison 12 Overall analysis 2: progestogen therapy long cycle (3‐4 weeks) versus other medical therapy, Outcome 6 Adverse effect ‐ breakthrough bleeding/spotting.

12.8 Resource use and cost

The included trials did not report data on resource use or costs for this comparison.

Discussion

Summary of main results

Based on the available evidence, this review was not able to meet all of its objectives in determining the effectiveness of oral progestogens in the treatment of HMB. We did not identify any RCTs comparing oral progestogens with placebo, and none of the trials under consideration assessed resource use or cost effectiveness.

Primary outcomes

Progestogen therapy (luteal phase only) compared to other medical treatment

See Table 1.

Luteal phase progestogens were less effective at reducing menstrual blood loss (as measured by the alkaline haematin method and the number of days bleeding during menstruation) when compared to other medical treatments such as tranexamic acid, danazol and the progestogen‐releasing intrauterine system. We are uncertain whether luteal phase progestogen improves satisfaction or quality of life of women with HMB, or is associated with any difference in adverse effects when compared to other medical treatments. Even though luteal phase progestogen was less effective at reducing menstrual blood loss, satisfaction with treatment was similar to other medical treatments such as tranexamic acid and Pg‐IUS.

Progestogen therapy long cycle (three to four weeks) compared to other medical therapy

See Table 2.

Long‐cycle oral progestogens (3 to 4 weeks, either norethisterone or medroxyprogesterone acetate) produced a reduction in menstrual blood loss from baseline measurements, but appeared to be less effective than the LNG‐IUS in reducing menstrual loss at three and six months' follow‐up. There was no difference in the occurrence of headache, but long‐cycle oral progestogens were associated with a significantly lower incidence of breakthrough bleeding compared with other medical treatment. Satisfaction with treatment was similar to women using the combined vaginal ring, but there were no data to compare satisfaction between long cycle and LNG‐IUS or tranexamic acid.

This review does not support the hypothesis that treatment with oral progestogen is more effective than other medical therapies in reducing menstrual blood loss. Luteal phase progestogens were less effective at reducing menstrual blood loss when compared to tranexamic acid, danazol and the Pg‐IUS, and there was no clear evidence of a difference when compared with NSAIDs. Long‐cycle oral progestogens (3 to 4 weeks) produced a significant reduction in menstrual blood loss, but appeared to be less effective than the LNG‐IUS.

Other outcomes

Oral progestogens in the luteal phase were associated with a lower incidence of adverse events than danazol, but no difference was observed when oral progestogens were compared with mefenamic acid or tranexamic acid. Women treated with the longer course of norethisterone, from day 5 to day 26 of the cycle, had a lower incidence of intermenstrual bleeding and breast tenderness than those treated with the LNG‐IUS. Without placebo trials, it is difficult to put these findings into context, since both danazol and the LNG‐IUS are known to be associated with a significant number of adverse effects. A high incidence of adverse effects can result in poor patient compliance, but we noted no difference in patient compliance when comparing luteal phase progestogen with mefenamic acid or danazol; the LNG‐IUS was associated with a higher patient acceptability than the longer regimen of norethisterone.

Only six of the 15 trials assessed quality of life. They demonstrated no significant difference in improved general health, symptoms of dysmenorrhoea or an ability to enjoy life, between progestogen‐ and tranexamic acid‐treated women, although a greater proportion of these women had improvement in their sex life, and flooding and leakage problems with tranexamic acid.

Overall completeness and applicability of evidence

This Cochrane Review included 15 trials with data that were relevant to this review question.

The review illustrates the importance of conducting objective RCTs of medical therapy for the treatment of HMB. The first subjective trial using progestogens for the treatment of HMB was published in 1960 (Bishop 1960). It was not until 27 years later that a randomized trial using objective measurements of menstrual blood loss was carried out (Cameron 1987), demonstrating no significant reduction in menstrual blood loss with norethisterone 5 mg twice daily from day 15 to 25 of the cycle (short cycle). Additional RCTs have investigated the effectiveness of a longer cycle of progestogens (from days 5 to 26 of the menstrual cycle) but have found no evidence that this treatment is better than other medical treatments, such as LNG‐IUS and tranexamic acid. Despite this, and the other papers included in this review, oral progestins remain commonly prescribed for the treatment of HMB. According to a 2009 national survey in the USA, oral progestogen is one of the first three options for treating HMB for up for up to 60% of gynaecologists (Matteson 2011), In South Korea, oral progestogen is the second most common choice for HMB, and the first option for 13% of gynaecologists (Lee 2015).

Quality of the evidence

The methodological quality of the evidence varied from low to very low. The main limitations were risk of bias (associated with high risk of detection and attrition bias) and inconsistency (due to high heterogeneity).

For the primary outcome of menstrual bleeding, different trials used different ways to assess the bleeding and to determine if there was any improvement. This is a major limitation when attempting to pool data.

For some comparisons, only one trial reported data, which contributes to the lack of power.

Potential biases in the review process

We attempted to identify and include all relevant work through the standardized method of identifying trials, but it is always possible that we may have missed some trials. Not all trial authors responded to our requests for more information, so we used imputation for some data. Some trial authors responded to the communications but did not submit the data requested.

We attempted to minimize potential biases in the review process by undertaking duplicate selection of trials, data extraction, and assessment of risk of bias. If there was disagreement, we resolved this through discussion and reaching consensus, but no intervention of a third review author was needed.

Agreements and disagreements with other studies or reviews

The results of this review are in accordance with other reviews.

The ECLIPSE trial (Gupta 2013), randomly assigned 571 women with HMB to the LNG‐IUS or to other usual medical treatment, including oral cyclical progesterone. Both groups improved from baseline and the improvement was maintained over two years, but was significantly better in the LNG‐IUS group.

According to a recent comprehensive review of the literature (Bitzer 2015), the data available assessing the use of luteal phase progestogen were inconsistent or suggested a very limited efficacy in reducing HMB, with an eventual increase after treatment. Long‐cycle progestogens (3 to 4 weeks per cycle) were consistent in reducing HMB. In the proposed algorithm of treatment in the review, luteal phase progestogen was not encouraged and long‐cycle progestogen was only suggested as treatment for women with HMB not seeking pregnancy, those who did not wish to, or could not use an IUS, or those who preferred oral hormonal treatment (Bitzer 2015).

The 2018 NICE guidelines for treating HMB (NICE 2018), suggested that the LNG‐IUS should be the first‐line treatment for HMB in women without pathology, with fibroids less than 3 cm in diameter (and not causing distortion of the uterine cavity), or with suspected or diagnosed adenomyosis. The guideline reported that the available evidence did not show clinically important differences in effectiveness and acceptability among the other pharmacological treatments, so if a woman declines LNG‐IUS, or it is not suitable, there are several options that may be considered. Cyclical progestogens are included in these options, but the guideline did not specify which type, dosage or length of treatment.

Authors' conclusions

Implications for practice.

Progestogens administered from day 15 or 19 to day 26 of the cycle (short cycle) are inferior to other treatment in the treatment of heavy menstrual bleeding (HMB) in women with ovulatory cycles, and offer no advantage over other medical therapies including danazol, tranexamic acid, nonsteroidal anti‐inflammatory drugs (NSAIDs), progestogen‐releasing intrauterine system (Pg‐IUS) and levonorgestrel‐releasing intrauterine system (LNG‐IUS).

Oral progestogen therapy for 21 days of the cycle (long cycle) is also inferior to tranexamic acid and the LNG‐IUS but may be similar to the combined vaginal ring. Women find the LNG‐IUS more acceptable than oral progestogen. This regimen of oral progestogen may have a role in the short‐term treatment of HMB.

Implications for research.

Quality‐of‐life issues merit further attention since the majority of women complaining of HMB will be offered treatment based on their symptoms alone with no objective measurement of menstrual blood loss.

Longer duration of progestogen use, which results in a significant reduction in menstrual blood loss, at present, has been compared with the LNG‐IUS, tranexamic acid and the combined vaginal ring. This merits further assessment against other forms of medical therapy such as NSAIDs and the oral contraceptive pill.

Bleeding outcomes were reported in several ways at several time points, affecting the possibility to combine them in meta‐analysis. It is imperative to create CORE outcomes (Khan 2016), for HMB, in order to be able to compare trials and make strong conclusions and recommendations.

What's new

Date	Event	Description
11 September 2019	Amended	Correction of typographical errors

History

Protocol first published: Issue 1, 1998
 Review first published: Issue 4, 1998

Date	Event	Description
25 February 2019	New citation required but conclusions have not changed	Overall satisfaction with treatment was added as a primary outcome and mortality was deleted as an outcome of this review.
25 February 2019	New search has been performed	Review updated. Eight new trials were added (Ashraf 2017; Goshtasebi 2013; Hashim 2012; Kaunitz 2010; Kiseli 2016; Kriplani 2006; Shravage 2011; Zhang 2008). Conclusions did not change.
25 September 2007	New citation required and conclusions have changed	Substantive amendment

Notes

The updated searches in 2019 were undertaken by Marian Showell, Cochrane Gynaecology and Fertility's Information Specialist.

Appendices

Appendix 1. Cochrane Gynaecology and Fertility search strategy

Searched 07 January 2019

Procite platform

Keywords CONTAINS "menorrhagia" or "heavy bleeding" or "heavy menstrual bleeding" or "heavy menstrual loss" or "dysfunctional bleeding" or "dysfunctional uterine bleeding" or "abnormal uterine bleeding" or "abnormal vaginal bleeding" or "excessive menstrual bleeding" or "excessive menstrual loss" or Title CONTAINS "menorrhagia" or "heavy bleeding" or "heavy menstrual bleeding" or "heavy menstrual loss" or "dysfunctional bleeding" or "dysfunctional uterine bleeding" or "abnormal uterine bleeding" or "abnormal vaginal bleeding" or "excessive menstrual bleeding" or "excessive menstrual loss"

AND

Keywords CONTAINS "progestagen" or "Progesterone" or "progestin" or "progestins" or "progestogen" or "progestogens" or "Gestagen" or "noresthisterone" or "Norethindrone" or "norethindrone acetate" or "Norethisterone" or "norethisterone acetate" or "Norgestimate" or "Norgestrel" or "dydrogesterone" or "Medrogestone" or "Medroxyprogesterone" or "medroxyprogesterone acetate" or Title CONTAINS "progestagen" or "Progesterone" or "progestin" or "progestins" or "progestogen" or "progestogens" or "Gestagen" or "noresthisterone" or "Norethindrone" or "norethindrone acetate" or "Norethisterone" or "norethisterone acetate" or "Norgestimate" or "Norgestrel" or "dydrogesterone" or"Medrogestone" or "Medroxyprogesterone" or "medroxyprogesterone acetate"

(119 hits)

Appendix 2. Cochrane Central Register of Studies (CENTRAL) via CRSO

Searched 07 January 2019

Web platform

#1 MESH DESCRIPTOR Menorrhagia EXPLODE ALL TREES 330

#2 menorrhag*:TI,AB,KY 718

#3 (menstrua* adj5 (bleed* or blood)):TI,AB,KY 860

#4 (heavy adj5 menstrua*):TI,AB,KY 236

#5 (dysfunctional adj3 uter*):TI,AB,KY 139

#6 hypermenorrh*:TI,AB,KY 24

#7 #1 OR #2 OR #3 OR #4 OR #5 OR #6 1435

#8 MESH DESCRIPTOR Progestins EXPLODE ALL TREES 2426

#9 (Progestin* or progestogen* or progesterone or norethisterone or norethindrone or dydrogesterone or medroxyprogesterone):TI,AB,KY 9082

#10 (progestagen* or progestational):TI,AB,KY 246

#11 #8 OR #9 OR #10 9429

#12 #7 AND #11 331

Appendix 3. MEDLINE search strategy

Searched from 1946 to 07 January 2019

Ovid platform

1 menorrhagia/ (4092)
 2 menorrhag$.tw. (3177)
 3 (menstrua$ adj5 (bleed$ or blood)).tw. (4449)
 4 (heavy adj5 menstrua$).tw. (945)
 5 (dysfunctional adj5 uter$).tw. (1009)
 6 hypermenorrh$.tw. (289)
 7 or/1‐6 (9981)
 8 exp Progestins/ (66516)
 9 (Progestin$ or progestogen$ or progesterone or norethisterone or norethindrone or dydrogesterone or medroxyprogesterone).tw. (95821)
 10 (progestagen$ or progestational).tw. (3797)
 11 or/8‐10 (119214)
 12 7 and 11 (1567)
 13 randomized controlled trial.pt. (473863)
 14 controlled clinical trial.pt. (92838)
 15 Randomized Controlled Trials/ (120184)
 16 Random allocation/ (97058)
 17 Double‐blind method/ (148903)
 18 Single‐blind method/ (26063)
 19 or/13‐18 (771199)
 20 clinical trial.pt. (513959)
 21 exp clinical trials/ (0)
 22 (clin$ adj25 trial$).ti,ab,sh. (404663)
 23 ((singl$ or doubl$ or tripl$ or trebl$) adj25 (blind$ or mask$)).ti,ab,sh. (167874)
 24 Placebos/ (34185)
 25 placebo$.ti,ab,sh. (214731)
 26 random$.ti,ab,sh. (1239751)
 27 Research design/ (99221)
 28 or/20‐27 (1846860)
 29 animal/ not (human/ and animal/) (4499580)
 30 19 or 28 (1871757)
 31 30 not 29 (1702841)
 32 12 and 31 (315)

Appendix 4. Embase search strategy

Searched from 1980 to 07 January 2019

Ovid platform

1 exp menorrhagia/ (8826)
 2 menorrhag$.tw. (5070)
 3 (menstrua$ adj5 (bleed$ or blood)).tw. (5807)
 4 (heavy adj5 menstrua$).tw. (1648)
 5 (dysfunctional adj5 uter$).tw. (1204)
 6 hypermenorrh$.tw. (389)
 7 or/1‐6 (14829)
 8 exp gestagen/ (142039)
 9 (Progestin$ or progestogen$ or progesterone or norethisterone or norethindrone or dydrogesterone or medroxyprogesterone).tw. (103908)
 10 gestagen$.tw. (1256)
 11 (progestagen$ or progestational).tw. (3034)
 12 or/8‐11 (177591)
 13 Clinical Trial/ (943095)
 14 Randomized Controlled Trial/ (525520)
 15 exp randomization/ (80582)
 16 Single Blind Procedure/ (33489)
 17 Double Blind Procedure/ (153616)
 18 Crossover Procedure/ (57605)
 19 Placebo/ (314683)
 20 Randomi?ed controlled trial$.tw. (193503)
 21 Rct.tw. (30758)
 22 random allocation.tw. (1845)
 23 randomly.tw. (393172)
 24 randomly allocated.tw. (31235)
 25 allocated randomly.tw. (2383)
 26 (allocated adj2 random).tw. (798)
 27 Single blind$.tw. (21833)
 28 Double blind$.tw. (186587)
 29 ((treble or triple) adj blind$).tw. (868)
 30 placebo$.tw. (276887)
 31 prospective study/ (492047)
 32 or/13‐31 (2181426)
 33 case study/ (58345)
 34 case report.tw. (359641)
 35 abstract report/ or letter/ (1041120)
 36 or/33‐35 (1449925)
 37 32 not 36 (2131226)
 38 (exp animal/ or animal.hw. or nonhuman/) not (exp human/ or human cell/ or (human or humans).ti.) (5594799)
 39 37 not 38 (1983619)
 40 7 and 12 and 39 (953)

Appendix 5. PsycINFO search strategy

Searched from 1806 to 07 January 2019

Ovid platform

1 exp Menstrual Disorders/ (1195)
 2 menorrhag$.tw. (83)
 3 (menstrua$ adj5 (bleed$ or blood)).tw. (242)
 4 (heavy adj5 menstrua$).tw. (29)
 5 (dysfunctional adj5 uter$).tw. (26)
 6 hypermenorrh$.tw. (2)
 7 or/1‐6 (1474)
 8 exp Progestational Hormones/ (2309)
 9 (Progestin$ or progestogen$ or progesterone or norethisterone or norethindrone or dydrogesterone or medroxyprogesterone).tw. (4800)
 10 (progestagen$ or progestational).tw. (73)
 11 or/8‐10 (4904)
 12 7 and 11 (99)

Appendix 6. CINAHL search strategy

Searched from 1961 to 07 January 2019

Ovid platform

#	Query	Results
S24	S11 AND S23	61
S23	S12 OR S13 OR S14 OR S15 OR S16 OR S17 OR S18 OR S19 OR S20 OR S21 OR S22	1,301,554
S22	TX allocat* random*	9,763
S21	(MH "Quantitative Studies")	21,772
S20	(MH "Placebos")	11,101
S19	TX placebo*	55,070
S18	TX random* allocat*	9,763
S17	(MH "Random Assignment")	53,035
S16	TX randomi* control* trial*	163,662
S15	TX ( (singl* n1 blind*) or (singl* n1 mask*) ) or TX ( (doubl* n1 blind*) or (doubl* n1 mask*) ) or TX ( (tripl* n1 blind*) or (tripl* n1 mask*) ) or TX ( (trebl* n1 blind*) or (trebl* n1 mask*) )	1,000,920
S14	TX clinic* n1 trial*	238,474
S13	PT Clinical trial	86,722
S12	(MH "Clinical Trials+")	254,738
S11	S6 AND S10	165
S10	S7 OR S8 OR S9	10,573
S9	TX(progestagen* or progestational)	2,226
S8	TX(Progestin* or progestogen* or progesterone or norethisterone or norethindrone or dydrogesterone or medroxyprogesterone)	9,546
S7	(MM "Progestational Hormones+")	2,192
S6	S1 OR S2 OR S3 OR S4 OR S5	1,606
S5	TX heavy N5 menstrua*	460
S4	TX hypermenorrh*	24
S3	TX (dysfunctional N5 uter*)	134
S2	TX menorrhag*	1,284
S1	(MM "Menorrhagia")	691

61 hits and 42 hits when date limited from 01.01.07 to 30.01.19

Data and analyses

Comparison 1. Progestogen therapy versus placebo.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Menstrual blood loss (alkaline haematin method)	0	0	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2 Proportion of women with no improvement in MBL (subjective)	0	0	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.0 [0.0, 0.0]
3 Duration of menstruation (days)	0	0	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
4 Proportion with no improvement in quality of life ‐ dysmenorrhoea	0	0	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.0 [0.0, 0.0]
5 Proportion of participant non‐compliance with treatment	0	0	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.0 [0.0, 0.0]
6 Proportion of women who find treatment unacceptable	0	0	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.0 [0.0, 0.0]
7 Adverse events (total)	0	0	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.0 [0.0, 0.0]
8 Adverse events ‐ headache	0	0	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.0 [0.0, 0.0]
9 Adverse events ‐ gastrointestinal symptoms	0	0	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.0 [0.0, 0.0]
10 Adverse events ‐ weight gain	0	0	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.0 [0.0, 0.0]
11 Adverse events ‐ skin changes	0	0	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.0 [0.0, 0.0]
12 Adverse events ‐ neurological symptoms	0	0	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.0 [0.0, 0.0]
13 Adverse events ‐ bloating	0	0	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.0 [0.0, 0.0]
14 Adverse events ‐ voice changes	0	0	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.0 [0.0, 0.0]
15 Adverse events ‐ dysmenorrhoea	0	0	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.0 [0.0, 0.0]

Comparison 2. Progestogen therapy (luteal phase only) versus NSAIDs.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Menstrual blood loss (alkaline haematin method)	2	48	Mean Difference (IV, Fixed, 95% CI)	22.97 [‐0.62, 46.57]
2 Duration of menstruation (days)	2	48	Mean Difference (IV, Fixed, 95% CI)	0.41 [‐0.13, 0.95]
3 Compliance with treatment	1	32	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.88 [0.05, 14.78]
4 Adverse events (total)	1	32	Peto Odds Ratio (Peto, Fixed, 95% CI)	1.86 [0.44, 7.86]
5 Adverse events ‐ headache	1	32	Peto Odds Ratio (Peto, Fixed, 95% CI)	1.60 [0.35, 7.31]
6 Adverse events ‐ gastrointestinal symptoms	1	32	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.56 [0.05, 5.83]
7 Adverse events ‐ dysmenorrhoea	1	32	Peto Odds Ratio (Peto, Fixed, 95% CI)	1.16 [0.20, 6.69]

Comparison 3. Progestogen therapy (luteal phase only) versus danazol.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Menstrual blood loss (alkaline haematin method)	2	51	Mean Difference (IV, Fixed, 95% CI)	55.63 [14.73, 96.54]
2 Proportion of women with improvement in MBL (subjective)	2	54	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.24 [0.08, 0.69]
3 Duration of menstruation (days)	4	107	Mean Difference (IV, Fixed, 95% CI)	1.60 [1.10, 2.11]
4 Compliance with treatment	2	61	Peto Odds Ratio (Peto, Fixed, 95% CI)	1.56 [0.45, 5.45]
5 Adverse events (total)	2	77	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.34 [0.13, 0.88]
6 Adverse events ‐ headache	2	64	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.29 [0.09, 0.95]
7 Adverse events ‐ gastrointestinal symptoms	2	64	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.22 [0.06, 0.90]
8 Adverse events ‐ weight gain	3	101	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.24 [0.09, 0.65]
9 Adverse events ‐ skin changes	2	64	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.18 [0.04, 0.75]
10 Adverse events ‐ neurological changes	1	24	Peto Odds Ratio (Peto, Fixed, 95% CI)	1.0 [0.20, 4.90]
11 Adverse events ‐ bloating	1	24	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.21 [0.04, 1.07]
12 Adverse events ‐ voice changes	1	24	Peto Odds Ratio (Peto, Fixed, 95% CI)	6.00 [0.34, 106.33]
13 Adverse events ‐ muscle cramps	1	40	Peto Odds Ratio (Peto, Fixed, 95% CI)	1.40 [0.28, 7.02]
14 Adverse events ‐ intermenstrual bleeding	1	40	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.26 [0.04, 1.67]
15 Adverse events ‐ breast changes	1	40	Peto Odds Ratio (Peto, Fixed, 95% CI)	8.73 [1.14, 67.13]

Comparison 4. Progestogen therapy (luteal phase only) versus tranexamic acid.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Menstrual blood loss (alkaline haematin method)	1	46	Mean Difference (IV, Fixed, 95% CI)	111.0 [43.54, 178.46]
2 Bleeding PBAC at 3 months	1	128	Mean Difference (IV, Fixed, 95% CI)	25.0 [‐15.54, 65.54]
3 Bleeding: improvement rates	3	216	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.43 [0.25, 0.75]
4 Satisfaction with treatment	1	42	Peto Odds Ratio (Peto, Fixed, 95% CI)	1.32 [0.37, 4.71]
5 Proportion with improvement in quality of life/dysmenorrhoea	1		Peto Odds Ratio (Peto, Fixed, 95% CI)	Subtotals only
5.1 Improvement in general health	1	44	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.45 [0.14, 1.47]
5.2 Improvement in dysmenorrhoea symptoms	1	44	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.44 [0.12, 1.59]
5.3 Improvement in ability to enjoy social activities	1	44	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.42 [0.13, 1.38]
5.4 Improvement in flooding or leakage	1	44	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.19 [0.05, 0.65]
5.5 Improvement in sex life	1	44	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.33 [0.09, 1.23]
6 Quality of life SF36	1	168	Mean Difference (IV, Fixed, 95% CI)	‐0.70 [‐1.48, 0.07]
6.1 Physical functioning	1	42	Mean Difference (IV, Fixed, 95% CI)	‐0.73 [‐2.25, 0.79]
6.2 Social functioning	1	42	Mean Difference (IV, Fixed, 95% CI)	‐0.37 [‐2.46, 1.72]
6.3 Mental health	1	42	Mean Difference (IV, Fixed, 95% CI)	‐0.27 [‐1.73, 1.19]
6.4 Environmental domain	1	42	Mean Difference (IV, Fixed, 95% CI)	‐1.20 [‐2.57, 0.17]
7 Adverse events (total)	2	88	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.93 [0.33, 2.64]
8 Adverse events ‐ headache	1	46	Peto Odds Ratio (Peto, Fixed, 95% CI)	1.90 [0.59, 6.15]
9 Adverse events ‐ gastrointestinal symptoms	2	88	Peto Odds Ratio (Peto, Fixed, 95% CI)	1.69 [0.54, 5.34]
10 Adverse events ‐ weight gain	1	46	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.15 [0.01, 2.54]
11 Adverse events ‐ bloating	1	42	Peto Odds Ratio (Peto, Fixed, 95% CI)	8.61 [0.52, 142.87]
12 Adverse events ‐ dysmenorrhoea	1	46	Peto Odds Ratio (Peto, Fixed, 95% CI)	1.48 [0.32, 6.71]

Comparison 5. Progestogen therapy (NE luteal phase only) versus Progestogen‐ IUS.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Menstrual blood loss (alkaline haematin method)	1	16	Mean Difference (IV, Fixed, 95% CI)	51.0 [18.38, 83.62]
2 Improvement in MBL PBAC < 100 at 6 months	1	40	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.11 [0.03, 0.38]
3 Satisfaction with treatment	1	40	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.43 [0.10, 1.87]
4 Duration of menstruation (days)	1	16	Mean Difference (IV, Fixed, 95% CI)	1.0 [‐0.19, 2.19]
5 Quality of life	1	200	Mean Difference (IV, Fixed, 95% CI)	‐0.21 [‐3.18, 2.76]
5.1 Physical domain	1	40	Mean Difference (IV, Fixed, 95% CI)	0.03 [‐6.21, 6.27]
5.2 Psychological domain	1	40	Mean Difference (IV, Fixed, 95% CI)	‐0.67 [‐7.49, 6.15]
5.3 Social domain	1	40	Mean Difference (IV, Fixed, 95% CI)	‐0.14 [‐9.33, 9.05]
5.4 Environmental domain	1	40	Mean Difference (IV, Fixed, 95% CI)	‐0.20 [‐6.34, 5.94]
5.5 Environmental domain‐TR	1	40	Mean Difference (IV, Fixed, 95% CI)	‐0.11 [‐6.14, 5.92]

Comparison 6. Progestogen therapy (MPA luteal phase only) versus LNG‐IUS.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Menstrual blood loss: PBAC reduction from baseline	1	165	Mean Difference (IV, Fixed, 95% CI)	‐49.3 [‐69.91, ‐28.69]
2 Adverse effects	1	1782	Peto Odds Ratio (Peto, Fixed, 95% CI)	1.21 [0.82, 1.80]
2.1 Ovarian cyst	1	162	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.23 [0.07, 0.75]
2.2 Pelvic pain	1	162	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.49 [0.10, 2.49]
2.3 Headache	1	162	Peto Odds Ratio (Peto, Fixed, 95% CI)	1.56 [0.64, 3.83]
2.4 Breast tenderness	1	162	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.72 [0.16, 3.28]
2.5 Vaginitis	1	162	Peto Odds Ratio (Peto, Fixed, 95% CI)	3.00 [0.93, 9.70]
2.6 Acne	1	162	Peto Odds Ratio (Peto, Fixed, 95% CI)	1.03 [0.29, 3.68]
2.7 Hypertention	1	162	Peto Odds Ratio (Peto, Fixed, 95% CI)	4.06 [0.80, 20.65]
2.8 Sinusitis	1	162	Peto Odds Ratio (Peto, Fixed, 95% CI)	1.73 [0.42, 7.14]
2.9 Fatigue	1	162	Peto Odds Ratio (Peto, Fixed, 95% CI)	2.04 [0.40, 10.38]
2.10 Urinary tract infection	1	162	Peto Odds Ratio (Peto, Fixed, 95% CI)	2.07 [0.54, 7.91]
2.11 Increased weight	1	162	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.81 [0.21, 3.10]

Comparison 7. Progestogen therapy (3 to 4 weeks) versus LNG‐IUS.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Menstrual blood loss (alkaline haematin method)			Other data	No numeric data
2 Proportion of women with MBL < 80 mL after treatment	2	120	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.23 [0.09, 0.57]
3 PBAC at 3 months	1	76	Mean Difference (IV, Fixed, 95% CI)	29.00 [‐609.14, 667.14]
4 PBAC at 6 months	1	76	Mean Difference (IV, Fixed, 95% CI)	21.42 [14.24, 28.60]
5 Proportion with improvement in quality of life‐dysmenorrhoea	1	31	Peto Odds Ratio (Peto, Fixed, 95% CI)	2.12 [0.42, 10.79]
6 Proportion who find treatment acceptable	1	40	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.12 [0.03, 0.40]
7 Adverse events ‐ intermenstrual bleeding	1	31	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.23 [0.05, 0.99]
8 Adverse events ‐ breast tenderness	1	31	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.11 [0.03, 0.46]
9 Adverse events ‐ mood swings	1	31	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.82 [0.19, 3.54]

Comparison 8. Progestogen therapy (3 to 4 weeks) versus tranexamic.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Bleeding PBAC at 3 months	2	184	Mean Difference (IV, Fixed, 95% CI)	8.82 [‐12.09, 29.73]
2 Bleeding: improvement rate	1	94	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.20 [0.07, 0.59]
3 Quality of life SF‐36	1	720	Mean Difference (IV, Fixed, 95% CI)	‐0.66 [‐3.33, 2.00]
3.1 Physical functioning	1	90	Mean Difference (IV, Fixed, 95% CI)	‐1.5 [‐6.77, 3.77]
3.2 Role physical	1	90	Mean Difference (IV, Fixed, 95% CI)	‐4.0 [‐13.63, 5.63]
3.3 Bodily pain	1	90	Mean Difference (IV, Fixed, 95% CI)	2.80 [‐6.01, 11.61]
3.4 General health	1	90	Mean Difference (IV, Fixed, 95% CI)	‐5.0 [‐12.49, 2.49]
3.5 Vitality	1	90	Mean Difference (IV, Fixed, 95% CI)	‐1.40 [‐9.42, 6.62]
3.6 Social functioning	1	90	Mean Difference (IV, Fixed, 95% CI)	1.10 [‐5.36, 7.56]
3.7 Role emotional	1	90	Mean Difference (IV, Fixed, 95% CI)	0.5 [‐11.47, 12.47]
3.8 Mental health	1	90	Mean Difference (IV, Fixed, 95% CI)	2.90 [‐4.83, 10.63]
4 Adverse effects	1		Peto Odds Ratio (Peto, Fixed, 95% CI)	Subtotals only
4.1 Any adverse event	1	94	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.40 [0.16, 1.03]
4.2 Gastrointestinal events	1	94	Peto Odds Ratio (Peto, Fixed, 95% CI)	1.39 [0.23, 8.36]
4.3 Headache	1	94	Peto Odds Ratio (Peto, Fixed, 95% CI)	1.39 [0.23, 8.36]
4.4 Allergic reaction	1	94	Peto Odds Ratio (Peto, Fixed, 95% CI)	6.81 [0.13, 344.42]
4.5 Intermenstrual bleeding	1	94	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.11 [0.02, 0.68]
4.6 Giddiness	1	94	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.33 [0.04, 2.39]
4.7 Breast tenderness	1	94	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.12 [0.01, 1.97]
4.8 Mood changes	1	94	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.12 [0.00, 6.26]

Comparison 10. Progestogen therapy (3‐4 weeks) versus combined vaginal ring.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Menstrual blood loss: PBAC after 3 cycles	1	95	Mean Difference (IV, Random, 95% CI)	2.10 [‐8.15, 12.35]
2 Percentage reduction in MBL (at end of study)			Other data	No numeric data
3 Satisfaction with treatment	1	95	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.32 [0.14, 0.72]
4 Quality of life (HRQoL‐4 self‐rated health)	1	95	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.78 [0.33, 1.82]
5 Quality of life (HRQoL‐4 impairment or lost days)	1		Mean Difference (IV, Random, 95% CI)	Subtotals only
5.1 Number of days feeling physically unwell	1	95	Mean Difference (IV, Random, 95% CI)	0.20 [‐0.28, 0.68]
5.2 Number of days feeling mentally unwell	1	95	Mean Difference (IV, Random, 95% CI)	0.40 [‐0.10, 0.90]
5.3 Number of lost days (with no regular activity)	1	95	Mean Difference (IV, Random, 95% CI)	0.90 [0.38, 1.42]
6 Adverse events	1		Peto Odds Ratio (Peto, Fixed, 95% CI)	Subtotals only
6.1 Nausea	1	95	Peto Odds Ratio (Peto, Fixed, 95% CI)	2.02 [0.20, 19.90]
6.2 Headache	1	95	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.67 [0.11, 4.04]
6.3 Breast tenderness	1	95	Peto Odds Ratio (Peto, Fixed, 95% CI)	1.55 [0.26, 9.31]
6.4 Breakthrough bleeding/spotting	1	95	Peto Odds Ratio (Peto, Fixed, 95% CI)	3.01 [0.71, 12.73]
6.5 Leukorrhea	1	94	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.24 [0.05, 1.27]
6.6 Vaginal discomfort	1	95	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.14 [0.01, 2.20]
6.7 Vaginitis	1	95	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.29 [0.05, 1.75]

Comparison 11. Overall analysis 1: progestogen therapy (luteal phase only) versus other medical therapy.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Menstrual blood loss (alkaline haematin method)	4	145	Mean Difference (IV, Fixed, 95% CI)	37.29 [17.67, 56.91]
1.1 Versus NSAIDs	2	43	Mean Difference (IV, Fixed, 95% CI)	16.78 [‐9.00, 42.56]
1.2 Versus danazol	2	45	Mean Difference (IV, Fixed, 95% CI)	56.25 [12.03, 100.46]
1.3 Versus tranexamic acid	1	46	Mean Difference (IV, Fixed, 95% CI)	111.0 [43.54, 178.46]
1.4 Versus Pg‐IUS	1	11	Mean Difference (IV, Fixed, 95% CI)	51.0 [‐1.51, 103.51]
2 Satisfaction	1	52	Odds Ratio (M‐H, Fixed, 95% CI)	0.80 [0.17, 3.80]
2.1 versus tranexamic acid	1	27	Odds Ratio (M‐H, Fixed, 95% CI)	0.86 [0.12, 6.26]
2.2 Versus LNG‐IUS	1	25	Odds Ratio (M‐H, Fixed, 95% CI)	0.71 [0.06, 8.70]
3 Menstruation days	5	171	Mean Difference (IV, Fixed, 95% CI)	1.05 [0.69, 1.40]
3.1 Versus NSAIDs	2	48	Mean Difference (IV, Fixed, 95% CI)	0.41 [‐0.13, 0.95]
3.2 Versus danazol	4	107	Mean Difference (IV, Fixed, 95% CI)	1.60 [1.10, 2.11]
3.3 Versus Pg‐IUS	1	16	Mean Difference (IV, Fixed, 95% CI)	1.0 [‐0.19, 2.19]
4 Quality of life: physical functioning	1	62	Mean Difference (IV, Fixed, 95% CI)	‐0.43 [‐5.10, 4.24]
4.1 versus tranexamic acid	1	32	Mean Difference (IV, Fixed, 95% CI)	‐0.73 [‐6.71, 5.25]
4.2 versus LNG‐ IUS	1	30	Mean Difference (IV, Fixed, 95% CI)	0.03 [‐7.45, 7.51]
5 Total adverse effects	5	197	Odds Ratio (M‐H, Fixed, 95% CI)	0.68 [0.36, 1.28]
5.1 Versus NSAIDs	1	32	Odds Ratio (M‐H, Fixed, 95% CI)	1.93 [0.43, 8.61]
5.2 Versus danazol	2	77	Odds Ratio (M‐H, Fixed, 95% CI)	0.32 [0.12, 0.88]
5.3 Versus tranexamic acid	2	88	Odds Ratio (M‐H, Fixed, 95% CI)	0.92 [0.32, 2.66]

Comparison 12. Overall analysis 2: progestogen therapy long cycle (3‐4 weeks) versus other medical therapy.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Bleeding PBAC after 3 months	4	355	Mean Difference (IV, Fixed, 95% CI)	16.88 [10.93, 22.84]
1.1 Versus LNG‐IUS	1	76	Mean Difference (IV, Fixed, 95% CI)	29.0 [21.20, 36.80]
1.2 Versus tranexamic acid	2	184	Mean Difference (IV, Fixed, 95% CI)	8.82 [‐12.09, 29.73]
1.3 Versus combined vaginal ring	1	95	Mean Difference (IV, Fixed, 95% CI)	‐2.10 [‐12.35, 8.15]
2 Bleeding PBAC after 6 months	1	76	Mean Difference (IV, Fixed, 95% CI)	21.42 [14.24, 28.60]
2.1 Versus LNG‐IUS	1	76	Mean Difference (IV, Fixed, 95% CI)	21.42 [14.24, 28.60]
3 Satisfaction	1	95	Odds Ratio (M‐H, Fixed, 95% CI)	0.31 [0.13, 0.71]
4 Treatment acceptability	1		Odds Ratio (M‐H, Fixed, 95% CI)	Subtotals only
4.1 Compared to LNG‐IUS	1	40	Odds Ratio (M‐H, Fixed, 95% CI)	0.08 [0.02, 0.37]
5 Adverse effect ‐ headache	2	189	Odds Ratio (M‐H, Fixed, 95% CI)	1.45 [0.40, 5.31]
5.1 Versus tranexamic acid	1	94	Odds Ratio (M‐H, Fixed, 95% CI)	1.40 [0.22, 8.80]
5.2 Versus combined vaginal ring	1	95	Odds Ratio (M‐H, Fixed, 95% CI)	1.5 [0.24, 9.41]
6 Adverse effect ‐ breakthrough bleeding/spotting	3	220	Odds Ratio (M‐H, Fixed, 95% CI)	0.18 [0.06, 0.55]
6.1 Versus LNG‐IUS	1	31	Odds Ratio (M‐H, Fixed, 95% CI)	0.18 [0.03, 1.05]
6.2 Versus tranexamic acid	1	94	Odds Ratio (M‐H, Fixed, 95% CI)	0.07 [0.00, 1.39]
6.3 Versus combined vaginal ring	1	95	Odds Ratio (M‐H, Fixed, 95% CI)	0.30 [0.06, 1.55]

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Ashraf 2017.

Methods	Parallel, prospective, randomized, multicentre, clinical trial
Participants	Setting: conducted at the gynaecology units of Bahawal Victoria Hospital (BVH), Jubilee Female Hospital, Civil Hospital and private clinics of consultant gynaecologists in Bahawalpur, Pakistan Inclusion criteria: women aged 18‐45 with dysfunctional uterine bleeding measuring PBAC score > 100 for 2 consecutive cycles uterus size < 10 cm on ultrasonography negative cervical cytology on Pap smear Exclusion criteria: contraindications for LNG‐IUS and NET use pregnancy post‐menopausal bleeding uterine neoplastic disease women with concomitant use of medications that could influence the trial objectives including sex steroids, any treatment for HMB (including tranexamic acid and NSAIDs) women who had intramural or subserous fibroids of mean diameter > 4 cm or submucous fibroids adenomyosis, or endometrial abnormalities women with coagulation disorders, liver disease or pelvic inflammatory disease Follow‐up: 6 months
Interventions	NET‐containing tablet was given orally at a dose of 5 mg 3 times/day for days 5‐26 of cycle over consecutive cycles: 40 LNG‐IUS (Mirena®): 40
Outcomes	PBAC at baseline, 3 and 6 months
Notes	Time frame: March‐August 2014
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomly allocated by lottery method into 2 equal groups; A and B
Allocation concealment (selection bias)	Unclear risk	Not stated
Blinding of participants and personnel (performance bias) PBAC, satisfaction, Adverse effect, response to treatment, bleeding	High risk	Only outcome was PBAC (scored by participants), which could be influenced by participants' knowledge of treatment
Blinding of participants and personnel (performance bias) Haematin alkaline	Unclear risk	Alkaline haematin was not measured.
Blinding of outcome assessment (detection bias) PBAC, satisfaction, Adverse effect, response to treatment	High risk	Not stated. Very unlikely for the difference of the treatments (1 oral and 1 IUS)
Blinding of outcome assessment (detection bias) Haematin Alkaline	Unclear risk	Alkaline haematin was not measured.
Incomplete outcome data (attrition bias) All outcomes	Low risk	No apparent dropouts after participants withdrawn for hysterectomy (2 in each arm)
Selective reporting (reporting bias)	Low risk	Trial reported all previously stated outcomes. Objective of trial authors was only to measure PBAC, however they acknowledged that oral NET is associated with adverse events – surprising then that they did not set out to measure these
Other bias	Low risk	Similar at baseline: yes Disclaimer: none CoI: none Source of funding: trial was funded by the departmental grant of the Post Graduate Medical College (PGMC), Islamic University of Bahawalpur March‐August 2014

Bonduelle 1991.

Methods	Single centre, parallel‐group design with no blinding Number of women randomized: n = 30 (15 in each treatment group) Withdrawals post‐randomization: n = 6 (2, selection criteria not met; 4, did not return), 5 from danazol group and 1 from NET group Loss to follow‐up: n = 8 (during treatment), 4 from each group, all because of adverse events
Participants	Patients (with mean ages 39 and 32) recruited from the Menstrual Disorders Clinic at the Royal Infirmary, Glasgow, UK Inclusion criteria: complaint of MBL requiring > 5 pads/tampons/day for > 6 days in cycle; presence of flooding or clots on any day of the cycle; presence of secondary anaemia; excessive MBL proving socially and domestically disruptive. Exclusion criteria: underlying pathology (from history, examination and D&C within the last year)
Interventions	NET 5 mg 3 times/day from day 19‐day 26 Danazol 200 mg, daily Duration: 3 cycles (1 pre‐treatment)
Outcomes	Duration of menstruation (days) Prevalence of side effects
Notes	Considered by authors to be a "pilot study". Poor‐quality trial because of high proportion of participants excluded or lost to follow‐up No source of funding given
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Randomization technique not stated
Allocation concealment (selection bias)	Unclear risk	Not stated
Blinding of participants and personnel (performance bias) PBAC, satisfaction, Adverse effect, response to treatment, bleeding	High risk	Open trial, with different dosage schedule. All outcomes likely to be influenced by lack of blinding
Blinding of participants and personnel (performance bias) Haematin alkaline	Unclear risk	Alkaline haematin was not measured.
Blinding of outcome assessment (detection bias) PBAC, satisfaction, Adverse effect, response to treatment	High risk	Open trial, all outcomes likely to be influenced by lack of blinding
Blinding of outcome assessment (detection bias) Haematin Alkaline	Unclear risk	Alkaline haematin was not measured.
Incomplete outcome data (attrition bias) All outcomes	High risk	30 women randomized, 6 lost post randomization and 8 during treatment
Selective reporting (reporting bias)	Unclear risk	No protocol available
Other bias	Low risk	Similar at baseline: yes Disclaimer: not stated CoI: not stated Source of funding: not given

Buyru 1995.

Methods	Single‐centre, parallel‐group Number of women randomized: n = 40 No power calculation reported
Participants	Patients aged 25‐50 years with a complaint of HMB recruited from Istanbul University Obstetrics and Gynaecology outpatient's department Inclusion criteria: women using > 3 pads/day during menstruation Exclusion criteria: organic pathology confirmed by US Women > 35 years also had endometrial biopsies
Interventions	NET 5 mg twice/day on days 16‐25 of the menstrual cycle Danazol 200 mg daily continuously Duration: 3 months
Outcomes	Duration of menstruation (days) Side effects
Notes	Paper written in Turkish and translated by Metin Gulmezoglu of Cochrane Pregnancy and Childbirth No source of funding given
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomization not stated
Allocation concealment (selection bias)	Unclear risk	Unclear
Blinding of participants and personnel (performance bias) PBAC, satisfaction, Adverse effect, response to treatment, bleeding	High risk	No blinding
Blinding of participants and personnel (performance bias) Haematin alkaline	Unclear risk	Alkaline haematin was not measured
Blinding of outcome assessment (detection bias) PBAC, satisfaction, Adverse effect, response to treatment	High risk	No blinding
Blinding of outcome assessment (detection bias) Haematin Alkaline	Unclear risk	Alkaline haematin was not measured.
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	No withdrawals or exclusions reported
Selective reporting (reporting bias)	Unclear risk	No protocol available
Other bias	Unclear risk	Similar at baseline: not stated Disclaimer: not stated CoI: not stated Source of funding: not given

Cameron 1987.

Methods	Single‐centre, parallel‐group design with no blinding Number of women randomized: n = 30 No power calculation. ITT analysis
Participants	Patients, aged 29‐50, recruited from Royal Infirmary, Edinburgh, UK Inclusion criteria: MBL > 50 mL/cycle No exclusion criteria stated
Interventions	NET, 5 mg twice/day, on days 15‐25 of cycle Danazol, 200 mg, daily Mefenamic acid, 500 mg 3 times/day, for first 5 days of menstruation Pg‐IUS, 65 ug progesterone daily Duration: 2 cycles
Outcomes	MBL (alkaline haematin method) Duration of menstruation (days)
Notes	Groups not comparable at baseline. Baseline MBL in danazol group significantly greater than in mefenamic acid and progestogen IUS groups Original data not available from principal trial author; MBL data reported as median and range. Median substituted for mean in meta‐analysis and standard deviation estimated from the range. Source of funding, Birthright Research Grant, RCOG
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomization computer‐generated
Allocation concealment (selection bias)	Unclear risk	Not reported
Blinding of participants and personnel (performance bias) PBAC, satisfaction, Adverse effect, response to treatment, bleeding	High risk	No blinding, lack of blinding likely to influence outcomes
Blinding of participants and personnel (performance bias) Haematin alkaline	Low risk	No blinding, lack of blinding unlikely to influence outcome
Blinding of outcome assessment (detection bias) PBAC, satisfaction, Adverse effect, response to treatment	High risk	No blinding, lack of blinding likely to influence outcomes
Blinding of outcome assessment (detection bias) Haematin Alkaline	Low risk	No blinding, lack of blinding unlikely to influence outcome
Incomplete outcome data (attrition bias) All outcomes	High risk	Substantial dropout (7/30 (23%)) ‐ no reasons given
Selective reporting (reporting bias)	Unclear risk	Adverse effects not reported
Other bias	High risk	Similar at baseline: substantial imbalance in MBL at baseline between groups ‐ as this is the primary outcome, the estimates are likely to be biased Disclaimer: not stated CoI: not stated Source of funding: Birthright Research Grant, RCOG

Cameron 1990.

Methods	Single‐centre with parallel‐group design MBL over 2 cycles assessed in 102 women with a subjective complaint of HMB. 20 women refused to collect sanitary pads, 10 women had anovulatory cycles and 40 women had MBL < 80 mL/cycle. The remaining women (n = 32) were randomized to treatment arms. No power calculation or ITT analysis No source of funding stated
Participants	Patients aged 21‐51 years recruited from the Outpatient Department, Royal Infirmary, Edinburgh, UK Inclusion criteria: MBL > 80 mL/cycle over 2 cycles measured by the alkaline haematin method Exclusions: organic disease non‐ovulatory cycles non compliance with collecting pads
Interventions	NET 5 mg twice daily on days 19‐26 of the menstrual cycle Mefenamic acid 500 mg 3 times/day on days 1‐5 of the menstrual cycle Duration over 2 cycles
Outcomes	MBL Number of days bleeding Cycle length Side effects Patient compliance
Notes	Original data provided by the lead author
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomization technique centralized and controlled by pharmacy
Allocation concealment (selection bias)	Low risk	Adequate
Blinding of participants and personnel (performance bias) PBAC, satisfaction, Adverse effect, response to treatment, bleeding	High risk	No blinding, lack of blinding likely to influence outcomes
Blinding of participants and personnel (performance bias) Haematin alkaline	Low risk	No blinding, lack of blinding unlikely to influence outcomes
Blinding of outcome assessment (detection bias) PBAC, satisfaction, Adverse effect, response to treatment	High risk	No blinding, lack of blinding likely to influence outcomes
Blinding of outcome assessment (detection bias) Haematin Alkaline	Low risk	No blinding, lack of blinding unlikely to influence outcomes
Incomplete outcome data (attrition bias) All outcomes	Low risk	No withdrawals/exclusions reported
Selective reporting (reporting bias)	Unclear risk	No protocol available
Other bias	Low risk	Similar at baseline: groups balanced on baseline Disclaimer: not stated CoI: not stated Source of funding: Not stated

Goshtasebi 2013.

Methods	RCT, with parallel‐group technique Block randomization technique Participants unable to be blinded, given the differences between the regimens
Participants	Country: Iran Number of participants: 90, 19 of whom withdrew by the end of the follow‐up period Age: 20‐45 years old Inclusion criteria: reported regular HMB; BMI 19‐29 Exclusion criteria: “organic cause of HMB”, iron‐deficiency anaemia, previous VTE, history of chronic diseases, history of diseases known to interfere with menstrual bleeding (e.g. fibroids, anticoagulant use, COCP or other hormonal drug use), IUS in situ
Interventions	TXA 500 mg 4 times daily for days 1‐5 of menses MPA 5 mg twice daily, for days 5‐26 of the menstrual cycle
Outcomes	Subjective assessment of MBL, using a modified PBAC (end scores) Serum Hb and ferritin SF‐36 for QoL (Farsi version) HMB questionnaire (Farsi version) Side effects
Notes	Funded by Tarbiat Modares University
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Block randomization
Allocation concealment (selection bias)	Unclear risk	Not reported
Blinding of participants and personnel (performance bias) PBAC, satisfaction, Adverse effect, response to treatment, bleeding	High risk	Unable to blind participants due to different regimens. Potential knowledge of treatment may have influenced the primary outcome of MBL, which was measured by PBAC
Blinding of participants and personnel (performance bias) Haematin alkaline	Unclear risk	Alkaline haematin was not measured.
Blinding of outcome assessment (detection bias) PBAC, satisfaction, Adverse effect, response to treatment	Unclear risk	It is unclear whether or not personnel were blinded
Blinding of outcome assessment (detection bias) Haematin Alkaline	Unclear risk	Alkaline haematin was not measured.
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Significant loss to follow‐up, but similar numbers and reasons for each group 71/90 randomized women (79%) included in analysis
Selective reporting (reporting bias)	Low risk	All prespecified outcomes were reported
Other bias	Low risk	Similar at baseline: groups balanced on baseline Disclaimer: not stated CoI: not stated Source of funding: funded by Tarbiat Modares University

Hashim 2012.

Methods	Multicentre (2 centres in Egypt), parallel‐group RCT undertaken from July 2008‐September 2010
Participants	Recruited from women complaining of regular heavy periods attending an outpatient clinic in Mansoura University Hospitals, Mansoura University, Egypt and a private practice setting Inclusion: HMB based on a PBAC score > 185 (mean of 2 control cycles) parous women desiring contraception and willing to use a male condom if required aged 20‐35 years in good general health and with a regular menstrual cycle with evidence of ovulation diagnosed when mid‐luteal phase serum progestogen level was ≥ 5 ng/mL normal pelvic examination with a sound measurement of the uterus of < 10 cm no pathology from pelvic US normal histology on endometrial biopsy negative cervical smear no contraindications to either treatment Exclusion: pregnancy age > 35 years obesity (BMI > 30 kg/m2 smokers current IUD users abnormal uterine bleeding not fully investigated hormone therapy or any medication that might affect MBL within the previous 3 months women who used injectable hormones for contraception during the previous 12 months use of drugs that interfere with contraceptive hormone metabolism previous endometrial resection/ablation and other pathology (e.g. participants with fibroids of any size etc) HMB of endocrine or systemic origin (e.g. thyroid disease and coagulopathies) participants unwilling to use contraception or medical management
Interventions	NET acetate (NETA) tablets (n = 47): dose of 5 mg 3 times/day from days 5‐26 of the cycle. Duration: 3 cycles CVR (NuvaRing) (n = 48): released 15 ug of ethinyl oestradiol and 120 ug of etonogestrel daily over a single cycle. Ring inserted between days 1 and 5 of the menstrual cycle for 3 weeks, followed by a 1‐week, ring‐free period
Outcomes	Primary: PBAC score at the end of treatment Secondary: Hb, adverse events, QoL (measured by HRQoL‐4), overall satisfaction
Notes	No CoI reported NuvaRing provided by Organon Egypt, and the sanitary pads were supplied by Procter & Gamble, Cairo, Egypt
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer‐generated random numeric table prepared by independent statistician
Allocation concealment (selection bias)	Low risk	Sealed opaque envelopes that were given to a third party (nurse) who assigned participants to trial arms
Blinding of participants and personnel (performance bias) PBAC, satisfaction, Adverse effect, response to treatment, bleeding	High risk	Blinding not possible; outcomes (other than Hb) likely to be influenced
Blinding of participants and personnel (performance bias) Haematin alkaline	Unclear risk	Alkaline haematin was not measured.
Blinding of outcome assessment (detection bias) PBAC, satisfaction, Adverse effect, response to treatment	High risk	Blinding not possible; outcomes (other than Hb) likely to be influenced
Blinding of outcome assessment (detection bias) Haematin Alkaline	Unclear risk	Alkaline haematin was not measured.
Incomplete outcome data (attrition bias) All outcomes	Low risk	No missing data
Selective reporting (reporting bias)	High risk	Protocol in trial register reported fewer outcomes than were reported in the publication; high risk of data mining
Other bias	Unclear risk	Similar at baseline: not stated Disclaimer: not stated CoI: not reported Source of funding: trial authors reported no funding, but one of the interventions was supplied by a pharmaceutical company (and not the other) July 2008‐September 2010

Higham 1993.

Methods	Parallel‐group design Number of women randomized: n = 57 Exclusions post randomization: n = 3 (2 from danazol‐reducing group because of pharmacy error, 1 from NET group because of menopausal symptoms) No power calculation made. ITT analysis
Participants	Patients aged 26‐48 years recruited from Royal Free Hospital and Edgware General Hospital, London Inclusion criteria: MBL > 80 mL/cycle regular cycle 21‐35 days aged 20‐50 years weight 45 kg‐110 kg endometrial sampling within previous 3 years no sensitivity to danazol or NET or ingestion 10 weeks previously informed consent Exclusion criteria: Underlying pathological conditions (from clinical assessment, pelvic US or endometrial biopsy) concomitant treatment with hormonal, anti‐prostaglandin or anti‐coagulant medication pregnancy or lactation or desire to become pregnant perimenopausal IUD wearers.
Interventions	NET 5 mg, 3 times/day on days 19‐26 of the cycle Danazol 200 mg/day Danazol 200 mg/day for 1 cycle, 100 mg/day during next cycle and 50 mg/day during last cycle Duration: 3 cycles
Outcomes	MBL (alkaline haematin method) Improvement in MBL (patient assessment) Duration of menstruation Side effects Treatment acceptability
Notes	Comparison between NET group and danazol 200 mg/day group. Comparisons between NET and danazol‐reducing group used in sensitivity analysis. A similar proportion of anovulatory participants were found in each treatment group but there was no difference in response of these women to medication compared to ovulatory women. Source of funding: Sanofi Winthrop Ltd
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk	Randomization method in sequential order and controlled by pharmacy
Allocation concealment (selection bias)	Low risk	Controlled by pharmacy
Blinding of participants and personnel (performance bias) PBAC, satisfaction, Adverse effect, response to treatment, bleeding	High risk	Stated as single blinding. Lack of blinding likely to influence outcomes
Blinding of participants and personnel (performance bias) Haematin alkaline	Low risk	Lack of blinding unlikely to influence this outcome
Blinding of outcome assessment (detection bias) PBAC, satisfaction, Adverse effect, response to treatment	Unclear risk	The assessors are blinded, but subjective outcomes are assessed by participants.
Blinding of outcome assessment (detection bias) Haematin Alkaline	Low risk	Stated as single blinding. Assessor blinded.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Loss to follow‐up during Rx: n = 10 (5 from danazol‐reducing group (3, side effects; 2, trial too demanding); 3 from danazol group (side effects) and 2 from NET group (side effects) from a total of 54 women (57‐3)
Selective reporting (reporting bias)	Unclear risk	No protocol available
Other bias	Unclear risk	Similar at baseline: at baseline participants in the danazol‐reducing group were younger than in the other 2 groups: danazol‐reducing 36 (26‐46), danazol 40.5 (30‐48) and NET 40.1 (32‐47) CoI: not stated Source of funding: Sanofi Winthrop Ltd

Irvine 1998.

Methods	Single‐centre, parallel‐group design with no blinding Power calculation for sample size and ITT analysis
Participants	Country: UK Women aged 18‐45 years all referred to specialist clinic complaining of regular HMB. 152 women were screened but 107 were excluded from eligibility (41 measured MBL < 80 mL, 62 declined to do MBL measurements, 4 declined to participate) Inclusion criteria: > 80 mL/cycle loss (as measured by alkaline haematin method parous (≥ 1 children) normal pelvic examination negative cervical cytology regular menstrual cycle good general health uterine cavity sound length < 10 cm Exclusion criteria: abnormal pelvic examination recent use of oestrogens progestogens or anticoagulants (within 3 months) injectable hormones for contraception (within 12 months)
Interventions	NET 5 mg 3 times/day from day 5‐26 of the cycle LNG‐IUS (Mirena) fitted into the uterus within 7 days of the onset of a menstrual period Duration: 3 menstrual cycles
Outcomes	MBL (alkaline haematin method) Hb and serum Ferritin at pretreatment and 3 months (or sooner if premature termination) Participant symptom/side effect questionnaire at pretreatment, 1 and 3 months Participant satisfaction categorized as liking treatment very well, well, moderately, poorly Women were asked how their periods interfered with their quality of life both before and after treatment. Proportion of women with amenorrhoea Proportion of women with specified side effects Withdrawal from treatment because of adverse events relating to treatment Acceptability of treatment (willingness to continue)
Notes	Outcomes assessed at 3 months, which is a relatively short period to assess the effectiveness of the LNG‐IUS. Source of funding not stated. Co‐investigator Prof Iain Cameron advised/email on 25 February 2019 that this trial was funded by the pharmaceutical company Leiras Oy, Turku, Finland
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "computer generated"
Allocation concealment (selection bias)	Low risk	Quote: "sealed opaque consecutively numbered envelopes"
Blinding of participants and personnel (performance bias) PBAC, satisfaction, Adverse effect, response to treatment, bleeding	High risk	Does not mention blinding. Lack of blinding likely to influence outcomes
Blinding of participants and personnel (performance bias) Haematin alkaline	Low risk	Does not mention blinding. Unlikely to influence outcome
Blinding of outcome assessment (detection bias) PBAC, satisfaction, Adverse effect, response to treatment	High risk	Does not mention blinding. Lack of blinding likely to influence outcomes
Blinding of outcome assessment (detection bias) Haematin Alkaline	Low risk	Does not mention blinding. Unlikely to influence outcome
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Trial authors stated that they performed both ITT and per protocol analyses but it appears that this was only for MBL and satisfaction. They undertook per protocol analyses for all other outcomes. Completers of the trial at 3 months were 20/22 (90.9%) in LNG‐IUS group and 16/22 (72.7%) in NET group. Side effects were collected in only 12/22 (54.5%) of NET group
Selective reporting (reporting bias)	Low risk	All prespecified outcomes reported
Other bias	Low risk	Similar at baseline: yes Disclaimer: not stated CoI: not reported Source of funding: Co‐investigator Prof Iain Cameron advised/email on 25 February 2019 that this trial was funded by the pharmaceutical company Leiras Oy, Turku, Finland

Kaunitz 2010.

Methods	Multicentre (n = 55), parallel‐group RCT ITT analysis and use of last observation carried forward for dropouts Power calculation for sample size: allowed for 20% dropouts, 69 women per treatment group to detect a significant difference in HMB between groups (83% reduction with LNG‐IUS and 50% reduction with MPA) (90% power) and 40% difference in proportions with successful treatment (99% power)
Participants	Country: USA, Canada and Brazil Women with mean age 38 or 39 years Inclusion criteria: parous women aged ≥ 18 years with idiopathic HMB (MBL ≥ 80 mL/cycle (assessed by alkaline haematin method) desiring intrauterine contraception willing to use barrier contraception Exclusion criteria: changes in menstrual irregularity hot flushes sleeping disorders changes in mood within the 3 months before the trial breastfeeding congenital or acquired uterine abnormality including fibroids if they distorted the uterine cavity or cervical canal history of organic causes of abnormal uterine bleeding use of LNG‐IUS or a copper IUS during the 30 days before the trial history of vascular or coagulation disorders concomitant use of medication or presence of an underlying disease/condition known to affect the metabolism or pharmacokinetics of the trial medication BMI > 35k g/m2
Interventions	MPA 10 mg once/day for 10 consecutive days of the cycle starting on day 16. LNG‐IUS (placed within 7 days of the onset of menstruation) (only 1 attempt at replacement could be made) 165 women randomized
Outcomes	Primary Absolute change in MBL from baseline to end of trial Proportion of women in which the treatment was successful (defined as MBL < 80 mL at end of trial and ≥ 50% reduction in HMB from baseline) Secondary Adverse events Follow‐up at 3 and 6 months
Notes	Screening phase for 2‐3 menstrual cycles to assess baseline HMB (alkaline haematin method) 2 publications: HMB outcomes assessed at 6 months and Hb and ferritin levels at 6 months. At the end of the trial, women assigned to MPA were able to choose to use a LNG‐IUS and those allocated to LNG‐IUS were allowed to continue its use. Funding: Bayer Schering Pharma AG
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	"centralised interactive voice system" with "random permuted block lengths of 4 to attain balance within the strata and by country"
Allocation concealment (selection bias)	Low risk	Centralized system of allocation
Blinding of participants and personnel (performance bias) PBAC, satisfaction, Adverse effect, response to treatment, bleeding	High risk	Not blinded for participants
Blinding of participants and personnel (performance bias) Haematin alkaline	Unclear risk	Alkaline haematin was not measured.
Blinding of outcome assessment (detection bias) PBAC, satisfaction, Adverse effect, response to treatment	High risk	Not blinded for personnel
Blinding of outcome assessment (detection bias) Haematin Alkaline	Unclear risk	Alkaline haematin was not measured.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Minimal dropouts and analysis by ITT and last observation carried forward
Selective reporting (reporting bias)	Low risk	All prespecified outcomes reported
Other bias	Low risk	Similar at baseline: groups appeared similar at baseline Disclaimer: not stated CoI: not reported Source of funding: Bayer Schering Pharma AG

Kiseli 2016.

Methods	Single‐centre RCT Central computer‐generated randomization schedule used for allocation Participants unable to be blinded, given the differences between the regimens Duration of trial: 2 control cycles pretreatment; 6 cycles on treatment
Participants	Country: Turkey Mean age: 42.1 years Number of participants: 84 women were randomized and 62 completed the trial Inclusion criteria: aged between 18 and 45 years PBAC score > 100 during 2‐month run‐in phase Exclusion criteria: abnormal pelvic US or endometrial biopsy Hb < 10 g/dL abnormal Pap smear result thyroid disease hypertension diabetes, or coronary artery disease history of previously taking medications for HMB contra‐indication to current therapy
Interventions	TXA 1 g 4 times/day from day 1 for 4 days NET 5 mg 3 times/day from day 14‐23 of the cycle LNG‐IUS (20 μg/day) inserted during the first few days of menses Treatment for 6 cycles
Outcomes	PBAC score and associated percentage reduction in blood loss (end scores) WHO QoL‐Short Form (Turkish version), in which women report limitations in physical health, psychological status, social support, and "relating to their environment"
Notes	The trial authors deny receiving any funding Authors contacted: time frame (recruitment and follow up was): May 2005 to December 2008
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer‐generated randomization scheme
Allocation concealment (selection bias)	Unclear risk	No details provided
Blinding of participants and personnel (performance bias) PBAC, satisfaction, Adverse effect, response to treatment, bleeding	High risk	Quote: Neither patients nor researchers were blinded to treatment'' Different dosage schedules make blinding impossible Potential knowledge of treatment may have influenced the primary outcome of MBL, which was measured by PBAC
Blinding of participants and personnel (performance bias) Haematin alkaline	Unclear risk	Alkaline haematin was not measured.
Blinding of outcome assessment (detection bias) PBAC, satisfaction, Adverse effect, response to treatment	High risk	Quote: ''Neither patients nor researchers were blinded to treatment'' Different dosage schedules make blinding impossible Potential knowledge of treatment may have influenced the primary outcome of MBL, which was measured by PBAC
Blinding of outcome assessment (detection bias) Haematin Alkaline	Unclear risk	Alkaline haematin was not measured.
Incomplete outcome data (attrition bias) All outcomes	High risk	Significant loss to follow‐up, but similar numbers for each group 62/84 randomized women (74%) included in analysis
Selective reporting (reporting bias)	Unclear risk	All prespecified outcomes reported
Other bias	Low risk	Similar at baseline: groups appeared comparable at baseline Disclaimer: not stated CoI: not reported Source of funding: authors deny receiving any funding Authors contacted: time frame (recruitment and follow up): May 2005 to December 2008

Kriplani 2006.

Methods	Single‐centre, prospective RCT Randomization by "computerised randomization table" Blinding not possible, given the different dosage schedules of the 2 arms of the trial
Participants	Country: India Age: 15‐50 years old Number of participants: 94 women Inclusion criteria: women presenting with HMB and PBAC score > 100 Excusion: fibroids adenomyosis endometriosis atypia on endometrial histopathology thyroid disease history of hormone therapy in previous 3 months unwilling to trial medical management
Interventions	TXA 500 mg 4 times/day from day 1 for 5 days MPA 10 mg twice/day from day 5‐25 of the cycle Treatment for 3 cycles, then participants were followed up for 3 months after treatment had been stopped
Outcomes	PBAC score and associated percentage reduction in blood loss (end scores) Recurrence of HMB Further surgery Participant satisfaction Duration of bleeding Hb level Side effects
Notes	Funded by the Indian council of Medical Research Time (recruitment and follow up was): from November 2002‐November 2004
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomization via "computerised randomization table"
Allocation concealment (selection bias)	Unclear risk	No details provided
Blinding of participants and personnel (performance bias) PBAC, satisfaction, Adverse effect, response to treatment, bleeding	High risk	No blinding. Different dosage schedules make blinding very difficult
Blinding of participants and personnel (performance bias) Haematin alkaline	Unclear risk	Alkaline haematin was not measured.
Blinding of outcome assessment (detection bias) PBAC, satisfaction, Adverse effect, response to treatment	High risk	No evidence of blinding; different dosage schedules make blinding very difficult. Potential knowledge of treatment may have influenced the primary outcome of MBL, which was measured by PBAC
Blinding of outcome assessment (detection bias) Haematin Alkaline	Unclear risk	Alkaline haematin was not measured.
Incomplete outcome data (attrition bias) All outcomes	High risk	Trial authors stated that there were minimal dropouts but at the end of treatment, only 66% of MPA group but 94% of TXA group were analyzed No reasons for withdrawal were given, but response levels were reported for all participants
Selective reporting (reporting bias)	Low risk	All prespecified outcomes were reported
Other bias	Low risk	Similar at baseline: groups appeared comparable at baseline Disclaimer: not stated CoI: not reported Source of funding: funded by the Indian council of Medical Research Time frame (recruitment and follow up): from November 2002‐November 2004

Preston 1995.

Methods	Double‐blind, placebo‐controlled RCT Participant and carer were blind to the treatment allocation Months 1 and 2 were placebo for all women In months 3 and 4, women took tablets on days 1‐4 and days 19‐26 of each cycle but only 1 of these treatments was active
Participants	Country: UK Age: > 18 years old Number of participants: 103 recruited and underwent 2 cycles of placebo to screen for eligibility, of whom 46 were randomized to the treatment phase of the trial Inclusion criteria: cycle length 28 ± 7 days average MBL over 2 cycles > 80 mL/cycle no hormone therapy within 3 months no medication that may affect MBL confirmed to be ovulating had complied with the protocol during the 2 months of placebo treatment Exclusion criteria: abnormal renal function abnormal pelvic examination abnormal cervical smear anovulatory cycles lack of compliance during the placebo cycle
Interventions	2 months of placebo (to assess eligibility) Followed by 2 months of: TXA (25 participants) 1 g taken 4 times/day on days 1‐4 NET (21 participants) 5 mg taken twice/day on days 19‐26 of cycle Duration 2 cycles
Outcomes	Objective MBL: measured via the alkaline haematin method (end scores) QoL assessed using a questionnaire (at end of cycle 2 and cycle 4) using 5‐point scale for general health, amount of flooding and leakage experienced, abdominal pain, limitation to social life, effect on sex life Diary of days bleeding, number of sanitary towels used and side effects recorded by participants
Notes	Funded by Pharmacia (a drug company)
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomization by computer‐generated numbers, which were placed in sealed envelopes
Allocation concealment (selection bias)	Low risk	Adequate
Blinding of participants and personnel (performance bias) PBAC, satisfaction, Adverse effect, response to treatment, bleeding	Low risk	Authors stated that the other group took "placebo of identical appearance to the active drug" and the treatment regimen was also the same for each group.
Blinding of participants and personnel (performance bias) Haematin alkaline	Low risk	All outcomes unlikely to be influenced as blinding assured
Blinding of outcome assessment (detection bias) PBAC, satisfaction, Adverse effect, response to treatment	Low risk	Double‐blind, placebo‐controlled RCT
Blinding of outcome assessment (detection bias) Haematin Alkaline	Low risk	All outcomes unlikely to be influenced as blinding assured
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Originally 103 women were recruited but 57 were excluded during the placebo cycles because of lack of objective HMB (> 80 mL/cycle), anovulation or lack of compliance with protocol
Selective reporting (reporting bias)	Unclear risk	No protocol available
Other bias	Low risk	Similar at baseline: groups appeared comparable at baseline Disclaimer: not stated CoI: not reported Source of funding: funded by Pharmacia (drug company)

Shravage 2011.

Methods	Single‐centre, parallel‐group Setting: teaching hospital JN Medical College, Belgaum Karnataka India.
Participants	72 women Inclusion criteria: women 35‐50. with menorrhagia (2 cycles with PBAC > 100 pre‐treatment) Exclusion criteria: pelvic pathology systemic disorders severe anaemia chronic cervicitis cervical dysplasia
Interventions	Ormeloxifene 60 mg twice/week + placebo form of capsule MPA x 21 days from day 2‐5 of the cycle 21 days of MPA from day 2‐5 + placebo form of ormeloxifene twice/week
Outcomes	Efficacy of PBAC before and after treatment Endometrial thickness before and after treatment
Notes	No source of funding stated Time frame (recruitment and follow up): September 2008 ‐ June 2009
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer‐generated with block size of 2
Allocation concealment (selection bias)	Low risk	Central control of randomization process by pharmacy
Blinding of participants and personnel (performance bias) PBAC, satisfaction, Adverse effect, response to treatment, bleeding	Low risk	Double‐blind, real placebo. All placebo used to ensure blinding and pharmacy only revealed allocation after analysis was done
Blinding of participants and personnel (performance bias) Haematin alkaline	Low risk	All outcomes unlikely to be influenced as blinding assured
Blinding of outcome assessment (detection bias) PBAC, satisfaction, Adverse effect, response to treatment	Low risk	Double blind, real placebo. All placebo used to ensure blinding and pharmacy only revealed allocation after analysis was done
Blinding of outcome assessment (detection bias) Haematin Alkaline	Low risk	All outcomes unlikely to be influenced as blinding assured
Incomplete outcome data (attrition bias) All outcomes	Low risk	96 randomized and 1 lost to follow‐up. No reasons stated
Selective reporting (reporting bias)	Unclear risk	Conclusion without the data on the paper ”the predominant side effect was"… there was no list or report of side effects and no measure on the control group
Other bias	Unclear risk	Similar at baseline: not clear indication that the groups were similar at baseline Disclaimer: not stated CoI: not reported Source of funding: not stated Time frame (recruitment and follow up): September 2008 ‐ June 2009

Zhang 2008.

Methods	Multi‐centre, open‐label RCT: method of randomization used computer software Blinding was highly unlikely, given the differences in the treatment regimens (1 group took tablets twice a day, the other 3 times a day) Given treatment for 2 cycles, then followed‐up for 1 further cycle
Participants	Country: China Age: mean age 35 years old Number of participants: 106 women Inclusion criteria: "proven ovulatory menorrhagia" attending gynaecological clinics PBAC score > 130 Exclusion criteria: heart, kidney, liver or haematological disease having had any hormonal treatments in the 3 months prior, including an IUS previous VTE
Interventions	TXA 1 g 3 times/day during days 1‐5 (69 participants) NET 5 mg twice/day on days 19‐26 (59 participants) Administered for 2 consecutive cycles
Outcomes	MBL (PBAC) Length of menstrual period 6‐item QoL questionnaire collected in the 2nd week before, during and after each treatment cycle and a 3rd (follow‐up) cycle
Notes	Funded by Daiichi Sankyo Co Ltd (a Japanese drug company)
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Used computer software to allocate participants randomly
Allocation concealment (selection bias)	Unclear risk	Not reported
Blinding of participants and personnel (performance bias) PBAC, satisfaction, Adverse effect, response to treatment, bleeding	High risk	Treatment and control groups were given their pills on different days of the menstrual cycle
Blinding of participants and personnel (performance bias) Haematin alkaline	Unclear risk	Alkaline haematin was not measured.
Blinding of outcome assessment (detection bias) PBAC, satisfaction, Adverse effect, response to treatment	High risk	Open‐label
Blinding of outcome assessment (detection bias) Haematin Alkaline	Unclear risk	Alkaline haematin was not measured.
Incomplete outcome data (attrition bias) All outcomes	Low risk	It appears that there was minimal dropout
Selective reporting (reporting bias)	High risk	The trial authors only report significant results in the abstract, which appears to favour the experimental treatment (TXA) For example, QoL scores were significantly improved after 1 cycle of treatment in the TXA group, but after 2 cycles of treatment and at follow‐up the scores were similar These latter findings were not reported, which suggests that the conclusions are influenced by lack of reporting
Other bias	Low risk	Similar at baseline: groups of participants appeared similar at baseline Disclaimer: not stated CoI: not reported Source of funding: funded by Daiichi Sankyo Co Ltd (a Japanese drug company) Time frame (recruitment and follow up): July 2004‐March 2006

BMI: body mass index; COCP: combined oral contraceptive pill; CoI: conflict of interest; CVC: combined hormone vaginal ring D&C: dilation and curettage; Hb: haemoglobin; HMB: heavy menstrual bleeding; HRQoL‐4: Health‐Related Quality of Life questionnaire 4; ITT: intention‐to‐treat; IUD: intrauterine device; IUS: intrauterine system; LNG‐IUS: levonorgestrel intrauterine system; MBL: menstrual blood loss; MPA: medroxyprogesterone acetate; NET: norethisterone; NSAIDs: nonsteroidal anti‐inflammatory drugs; PBAC: Pictorial Blood Assessment Chart; Pg‐IUS: progestogen‐releasing intrauterine system; QoL: quality of life; RCOG: Royal College of Obstetricians and Gynaecologists; SF‐36: Short‐form 36‐item health survey; TXA: tranexamic acid; US: ultrasound; VTE: venous thromboembolism; WHO: World Health Organization

Characteristics of excluded studies [ordered by study ID]

Study	Reason for exclusion
Azizkhani 2018	Comparison is alternative not medical treatment.
Gupta 2013	The comparison is using all medical treatment and there is no data specifically for cyclical progestogen (even though it is one of the comparators). No randomization between other medical treatments, women were allowed to choose their treatment when allocated to other medical treatment.
Karakus 2009	The population is women with irregular bleeding.
Kucuk 2008	Quasi‐randomized trial; it was done by a predefined application order.

Characteristics of ongoing studies [ordered by study ID]

NCT02943655.

Trial name or title	Combined oral contraceptives, progestogens, and non‐steroidal anti‐inflammatory drugs for heavy and/or prolonged menstrual bleeding without organic cause
Methods	Parallel‐group, randomized
Participants	Inclusion criteria: regular menstrual cycles with BMI 19‐29 kg/m2 HMB and/or prolonged menstrual bleeding involving at ≥ 3 consecutive menstrual cycles Exclusion criteria: postmenopausal bleeding (> 1 year since the last menstrual period) irregular menses or intermenstrual bleeding organic causes of heavy menstrual bleeding suspected or confirmed by experienced abdominal and transvaginal ultrasound after thorough general and gynecological examination. iatrogenic (treatment‐related) causes of HMB (e.g. non‐Pg‐IUS, oral contraceptives, other hormonal drug use or anticoagulant agent) iron deficiency anaemia history of chronic diseases known to interfere with menstrual bleeding or prevent the use of any of the listed drugs e.g. previous or current thromboembolic disease
Interventions	COCP second generation, 1 tablet/day MPA oral 5 mg NSAID (mefenamic acid) oral 500 mg 3 times/day
Outcomes	Amount of MBL (time frame: 3 months)
Starting date	2016
Contact information	Ahmed Abbas, Assiut University, Cairo, Egypt, 002 bmr90@hotmail.com
Notes

BMI: body mass index; COCP: combined oral contraceptive pill; HMB: heavy menstrual bleeding; MBL: menstrual blood loss; MPA: medroxyprogesterone acetate; NSAIDs: nonsteroidal anti‐inflammatory drugs; Pg‐IUS: progestogen‐releasing intrauterine system;

Differences between protocol and review

For the 2019 update:

1. we expanded the scope of the review to include other medical treatments, such as the combined vaginal ring and ormeloxifene that were not available as medical treatment for heavy menstrual bleeding when the protocol was written;

2. we added satisfaction as a primary outcome as it is now considered highly important to women and clinicians;

3. we deleted mortality as an outcome, as the risk is extremely low.

Contributions of authors

In the 2019 update, Magdalena Bofill conducted searches for relevant trials, assessed the included trials for quality, performed data extraction, entered data and updated all sections. Anne Lethaby performed data extraction, assessed the included trials for quality and commented on the draft and final version of the review. Cindy Low conducted searches for relevant trials. Iain Cameron read and commented on the final version of the review.

In the 1998 and 2008 versions Anne Lethaby registered the title, prepared the protocol, conducted searches for relevant trials, assessed the included trials for quality, performed data extraction, entered data, wrote all sections of the review except for the discussion and conclusions and incorporated suggested changes from the peer review process.
 Gill Irvine read and commented on the draft of the protocol, assessed the included trials for quality, performed data extraction and wrote the discussion and conclusions sections of the review.
 Iain Cameron read and commented on both the draft of the protocol and the final review.

Sources of support

Internal sources

• Department of Obstetrics and Gynaecology, University of Auckland, Auckland, New Zealand.

External sources

• Health Research Council, Auckland, New Zealand.

Declarations of interest

MBR: no conflict of interest to declare
 AL: no conflict of interest to declare
 CL: no conflict of interest to declare
 ITC received funding from the makers of mefenamic acid and the levonorgestrel intrauterine system to undertake investigator‐initiated clinical trials (before 1999), and to present results at international symposia (1999‐2002).

Edited (no change to conclusions)