Antifibrinolytics for heavy menstrual bleeding

Alison C Bryant‐Smith; Anne Lethaby; Cindy Farquhar; Martha Hickey

doi:10.1002/14651858.CD000249.pub2

. 2018 Apr 15;2018(4):CD000249. doi: 10.1002/14651858.CD000249.pub2

Antifibrinolytics for heavy menstrual bleeding

Alison C Bryant‐Smith ^1,^✉, Anne Lethaby ², Cindy Farquhar ², Martha Hickey ³

Editor: Cochrane Gynaecology and Fertility Group

PMCID: PMC6494516 PMID: 29656433

Abstract

Background

Heavy menstrual bleeding (HMB) is an important physical and social problem for women. Oral treatment for HMB includes antifibrinolytic drugs, which are designed to reduce bleeding by inhibiting clot‐dissolving enzymes in the endometrium.

Historically, there has been some concern that using the antifibrinolytic tranexamic acid (TXA) for HMB may increase the risk of venous thromboembolic disease. This is an umbrella term for deep venous thrombosis (blood clots in the blood vessels in the legs) and pulmonary emboli (blood clots in the blood vessels in the lungs).

Objectives

To determine the effectiveness and safety of antifibrinolytic medications as a treatment for heavy menstrual bleeding.

Search methods

We searched the Cochrane Gynaecology and Fertility (CGF) Group trials register, CENTRAL, MEDLINE, Embase, PsycINFO and two trials registers in November 2017, together with reference checking and contact with study authors and experts in the field.

Selection criteria

We included randomized controlled trials (RCTs) comparing antifibrinolytic agents versus placebo, no treatment or other medical treatment in women of reproductive age with HMB. Twelve studies utilised TXA and one utilised a prodrug of TXA (Kabi).

Data collection and analysis

We used standard methodological procedures expected by Cochrane. The primary review outcomes were menstrual blood loss (MBL), improvement in HMB, and thromboembolic events.

Main results

We included 13 RCTs (1312 participants analysed). The evidence was very low to moderate quality: the main limitations were risk of bias (associated with lack of blinding, and poor reporting of study methods), imprecision and inconsistency.

Antifibrinolytics (TXA or Kabi) versus no treatment or placebo

When compared with a placebo, antifibrinolytics were associated with reduced mean blood loss (MD −53.20 mL per cycle, 95% CI −62.70 to −43.70; I² = 8%; 4 RCTs, participants = 565; moderate‐quality evidence) and higher rates of improvement (RR 3.34, 95% CI 1.84 to 6.09; 3 RCTS, participants = 271; moderate‐quality evidence). This suggests that if 11% of women improve without treatment, 43% to 63% of women taking antifibrinolytics will do so. There was no clear evidence of a difference between the groups in adverse events (RR 1.05, 95% CI 0.93 to 1.18; 1 RCT, participants = 297; low‐quality evidence). Only one thromboembolic event occurred in the two studies that reported this outcome.

TXA versus progestogens

There was no clear evidence of a difference between the groups in mean blood loss measured using the Pictorial Blood Assessment Chart (PBAC) (MD −12.22 points per cycle, 95% CI −30.8 to 6.36; I² = 0%; 3 RCTs, participants = 312; very low quality evidence), but TXA was associated with a higher likelihood of improvement (RR 1.54, 95% CI 1.31 to 1.80; I² = 32%; 5 RCTs, participants = 422; low‐quality evidence). This suggests that if 46% of women improve with progestogens, 61% to 83% of women will do so with TXA.

Adverse events were less common in the TXA group (RR 0.66, 95% CI 0.46 to 0.94; I² = 28%; 4 RCTs, participants = 349; low‐quality evidence). No thromboembolic events were reported in any group.

TXA versus non‐steroidal anti‐inflammatory drugs (NSAIDs)

TXA was associated with reduced mean blood loss (MD −73.00 mL per cycle, 95% CI −123.35 to −22.65; 1 RCT, participants = 49; low‐quality evidence) and higher likelihood of improvement (RR 1.43, 95% CI 1.18 to 1.74; 1² = 0%; 2 RCTs, participants = 161; low‐quality evidence). This suggests that if 61% of women improve with NSAIDs, 71% to 100% of women will do so with TXA. Adverse events were uncommon and no comparative data were available. No thromboembolic events were reported.

TXA versus ethamsylate

TXA was associated with reduced mean blood loss (MD 100 mL per cycle, 95% CI −141.82 to −58.18; 1 RCT, participants = 53; low‐quality evidence), but there was insufficient evidence to determine whether the groups differed in rates of improvement (RR 1.56, 95% CI 0.95 to 2.55; 1 RCT, participants = 53; very low quality evidence) or withdrawal due to adverse events (RR 0.78, 95% CI 0.19 to 3.15; 1 RCT, participants = 53; very low quality evidence).

TXA versus herbal medicines (Safoof Habis and Punica granatum)

TXA was associated with a reduced mean PBAC score after three months' treatment (MD −23.90 pts per cycle, 95% CI −31.92 to −15.88; I² = 0%; 2 RCTs, participants = 121; low‐quality evidence). No data were available for rates of improvement. TXA was associated with a reduced mean PBAC score three months after the end of the treatment phase (MD −10.40 points per cycle, 95% CI −19.20 to −1.60; I² not applicable; 1 RCT, participants = 84; very low quality evidence). There was insufficient evidence to determine whether the groups differed in rates of adverse events (RR 2.25, 95% CI 0.74 to 6.80; 1 RCT, participants = 94; very low quality evidence). No thromboembolic events were reported.

TXA versus levonorgestrel intrauterine system (LIUS)

TXA was associated with a higher median PBAC score than LIUS (median difference 125.5 points; 1 RCT, participants = 42; very low quality evidence) and a lower likelihood of improvement (RR 0.43, 95% CI 0.24 to 0.77; 1 RCT, participants = 42; very low quality evidence). This suggests that if 85% of women improve with LIUS, 20% to 65% of women will do so with TXA. There was insufficient evidence to determine whether the groups differed in rates of adverse events (RR 0.83, 95% CI 0.25 to 2.80; 1 RCT, participants = 42; very low quality evidence). No thromboembolic events were reported.

Authors' conclusions

Antifibrinolytic treatment (such as TXA) appears effective for treating HMB compared to placebo, NSAIDs, oral luteal progestogens, ethamsylate, or herbal remedies, but may be less effective than LIUS. There were too few data for most comparisons to determine whether antifibrinolytics were associated with increased risk of adverse events, and most studies did not specifically include thromboembolism as an outcome.

Plain language summary

Antifibrinolytics (such as tranexamic acid) for treatment of heavy menstrual bleeding

Review question

Antifibrinolytic agents are designed to reduce bleeding by inhibiting endometrial clot‐dissolving enzymes (in the uterine lining); Cochrane researchers reviewed the evidence about the effect of these medications (such as tranexamic acid, TXA) versus placebo and other medical therapies in women with heavy menstrual bleeding (HMB: defined as more than 80 millilitres (> 80 mL) of blood loss per menstrual cycle).

Background

Antifibrinolytic agents (such as tranexamic acid, TXA) are commonly used to manage HMB. However, historically there has been concern that they may cause dangerous blood clots in the legs or lungs. There are a variety of other medications that can be used to treat HMB. We compared the benefits and risks of the treatments.

Study characteristics

We found 13 randomized controlled trials (RCTs) comparing an antifibrinolytic medication with a different medical therapy, in a total of 1312 women with heavy menstrual bleeding. The evidence is current to November 2017.

Key results

Antifibrinolytic medication may improve HMB in women aged 15 to 50 years old, without substantially increasing the rate of adverse events. Evidence suggests there is a 40% to 50% reduction in the amount of menstrual blood lost per menstrual cycle for participants taking TXA. Antifibrinolytic treatment was better at improving HMB loss than other medical treatments, except for the levonorgestrel intrauterine system (LIUS), a plastic device placed in the uterus which releases hormone to prevent conception.

The evidence suggests that if 10.9% of women taking placebo report an improvement in HMB, 36.3% of women taking TXA will do so.

TXA probably improves quality of life for women with HMB.

We did not find any evidence that side effects (including life‐threatening blood clots) were increased in women taking antifibrinolytic treatment compared to placebo or other treatments for HMB. Two studies measured venous thromboembolic events: unfortunately these studies did not have enough participants to distinguish a real effect of a certain size from pure luck.

Quality of the evidence

The evidence was of very low to moderate quality. The main limitations were: risk of bias, due to participants/investigators being aware of which medication they were receiving (known as lack of blinding), or the study's methods not being reported very clearly; imprecision (i.e. repeated measurements being far apart from each other), and inconsistency (i.e. as the sample size increases, the sampling distribution becomes increasingly wide around the true parameter value).

Summary of findings

Background

Description of the condition

‘Normal’ menstrual blood loss (MBL) has been defined as 30 mL to 40 mL per menstrual cycle, whilst heavy menstrual bleeding has traditionally been defined as greater than 80 mL blood loss per cycle (Duckitt 2012). Whilst this objective cut‐off has been broadly utilized in clinical trials, such measurement (involving extracting haemoglobin from sanitary wear) is impractical outside research settings. Also, this objectively measured cut‐off of 'heavy' menstrual bleeding may not reflect the woman's experience, nor the impact of heavy menstrual bleeding (HMB) on her quality of life (QoL) (Warner 2004). International guidelines (such as that produced by the United Kingdom's National Institute for Health and Care Excellence (NICE)) base the diagnosis of HMB on women’s and clinicians’ subjective perceptions of MBL, and its resultant impact (e.g. iron‐deficiency anaemia, days off work).

HMB is an important cause of ill health in women: prevalence estimates range from 4% to 51%. This wide range is due to these studies being undertaken in different countries and clinical settings (NICE 2007). It has been estimated that HMB accounts for 5% of general practitioner consultations by women aged 30 to 49 years old (Turner 2000), whilst HMB accounts for up to one‐third of all gynaecological consultations (El‐Hemaidi 2007).

The most widely used classification system for causes of abnormal uterine bleeding in reproductive‐aged women is that of the International Federation of Gynecology and Obstetrics, called the PALM‐COEIN system. As outlined in Munro 2011, the basic classification system breaks aetiologies down into: Polyp; Adenomyosis; Leiomyoma (submucosal/other); Malignancy and hyperplasia; Coagulopathy; Ovulatory dysfunction; Endometrial; Iatrogenic; and Not yet classified. In general, the PALM aetiologies are structural, whilst the COEIN categories are non‐structural.

Around 80% of women treated for heavy menstrual bleeding have no underlying uterine abnormality (i.e. would fall into the COEIN categories listed above), yet up to 60% of women referred to a gynaecologist for HMB undergo a hysterectomy within five years of referral (Edlund 2003; Qiu 2014). HMB is the primary indication for approximately 50% of all hysterectomies (emergency plus elective) in the UK (Turner 2000), and 38% of all elective hysterectomies (Butt 2012).

Hence medical therapy, with the avoidance of potential complications of surgical management, is an attractive alternative. A wide variety of medications are available to reduce heavy menstrual bleeding, but there is considerable variation in practice, and some uncertainty about the most appropriate first line therapy: a universally applicable step‐wise approach is sorely lacking (Fox 2012; Marret 2010).

Description of the intervention

Trans‐4‐aminomethylcyclohexanocarboxylic acid (or tranexamic acid and its precursors) is an antifibrinolytic medication. Tranexamic acid (TXA) has been used to treat HMB for over four decades in many European countries; in the UK, TXA is prescribed as first‐ or second‐line medical management of HMB for over 64% of women not requiring contraception (Turner 2000). In the US, TXA was not approved for the treatment of menorrhagia until 2009 (Kaunitz 2010).

How the intervention might work

Women with HMB have been found to have increased fibrinolytic activity in their menstrual fluid (Edlund 2003).

TXA exerts its antifibrinolytic effect by reversibly blocking lysine‐binding sites on plasminogen, thus preventing plasmin from interacting with lysine residues on the fibrin polymer. By preventing plasmin and lysine residues from interacting, TXA thus slows subsequent fibrin degradation, thereby slowing the dissolution of clots. Antifibrinolytic agents have, therefore, been promoted as a treatment for heavy menstrual bleeding.

Why it is important to do this review

There is growing evidence of the utility of the levonorgestrel intrauterine system (LIUS) in managing heavy menstrual bleeding. Qiu et al published a systematic analysis comparing the levonorgestrel intrauterine device to medical management of HMB, and included Gupta 2013, a randomized controlled trial (RCT) which compared LIUS to medical management of HMB (Gupta 2013; Qiu 2014). The authors concluded that LIUS is more effective for the treatment of HMB compared with oral medical treatment. However, for women for whom an intrauterine device is contra‐indicated or who wish to avoid an LIUS, antifibrinolytic treatment still plays an important role.

Although TXA has been credited with reducing MBL by up to 60% (Leminen 2012), one study based in Somerset (UK) found that less than 15% of women presenting to their general practitioner complaining of HMB were offered antifibrinolytic treatment (Grant 2000). Recent data indicate that prescribing patterns may be changing: a randomized controlled trial carried out in general practices in East Anglia (in the UK) indicated rates of antifibrinolytic use among women with heavy menstrual bleeding were 57% among those practices given a specific evidence‐based education package, and 35% among control practices (Fender 1999).

As these medications slow the breakdown of clots, there has been anecdotal concern that antifibrinolytic agents may be associated with an increased risk of thromboembolic disease (such as deep venous thrombosis). However, venous thromboembolic events have not been reported in treatment studies, and (to date) data from population‐based studies do not support an increased incidence of venous thromboembolism with antifibrinolytic use (Leminen 2012). Long‐term studies in Sweden have shown that the incidence of thrombosis in women treated with TXA is comparable to that of women not being treated with TXA (Berntorp 2001).

Hence, it is important to do this review to assess antifibrinolytic treatment's efficacy in managing heavy menstrual bleeding, which will be particularly relevant for women who are unable/unwilling to have a LIUS inserted.

Objectives

To determine the effectiveness and safety of antifibrinolytics as a treatment for heavy menstrual bleeding

Methods

Criteria for considering studies for this review

Types of studies

We accepted as eligible for inclusion published and unpublished RCTs of antifibrinolytic therapy versus placebo, no treatment or any other medical (non‐surgical) therapy when used to reduce heavy menstrual bleeding. We excluded non‐randomized studies (e.g. studies with evidence of inadequate sequence generation such as alternate days, patient numbers) as they are associated with a high risk of bias.

Cross‐over trials were only eligible for inclusion if they reported first‐phase data, in order to minimise the chance of cross‐over bias. This is a change from the original protocol criteria for inclusion, where cross‐over trials could be included regardless of whether data were provided for the first phase of the trial. Where cross‐over trials only report findings at the end of the study the likelihood of significant bias is increased, because no adjustment is made for cross‐over effects.

Types of participants

Women of reproductive age, who are having regular heavy periods (measured either objectively or subjectively), undertake at least two months' follow‐up whilst on treatment, and who are recruited from primary care, family planning, or a specialist clinic setting were eligible for inclusion.

Exclusion criteria included: post‐menopausal bleeding; irregular menses, inter‐menstrual bleeding or both; pathological causes of HMB (e.g. a coagulopathy); and iatrogenic causes of HMB (e.g. intrauterine device/system, or anti‐coagulant medication).

Types of interventions

We included trials comparing antifibrinolytic agents (e.g. tranexamic acid and its precursors) versus no treatment, placebo, or any other medical (non‐surgical) therapy. We excluded studies that used combined treatments (e.g. a LIUS with concurrent oral TXA).

Types of outcome measures

Primary outcomes

1. Menstrual blood loss (MBL), measured by either or both of the following.

a) Objective assessment of mean blood loss in mL (using alkaline haematin method or similar, Hallberg 1964), using either change scores or end scores: where studies reported both, we used end scores.

b) Subjective assessment of blood loss using continuous measures such as Pictorial Blood Assessment Chart (PBAC) scores, using either change scores or end scores: where studies reported both, we used end scores. PBAC scores range from 0 to more than 500, and rely on women scoring each tampon and sanitary towel they use during the course of a period, in terms of how heavily blood‐stained they are. Blood clots and episodes of flooding are also recorded (Higham 1990). Studies have confirmed a significant correlation between PBAC score and the alkaline haematin method, with a PBAC score of more than 100 being indicative of HMB (Zakherah 2011).

2. Improvement in HMB: binary measures (improved/not improved) as reported by the study, giving priority to subjective measures if studies reported both.

3. Thromboembolic events.

Secondary outcomes

4. Quality of life: participant's perceived change in quality of life from baseline provided this has been recorded in a reproducible and validated format (e.g. Menorrhagia Impact Questionnaire (MIQ), SF‐36, WHOQOL‐BREF).

5. Adverse events (other than thromboembolic events), including but not limited to: any adverse event; gastrointestinal side effects; abdominal discomfort; headaches; dizziness; breast tenderness; dysmenorrhoea; changes in weight; and changes in mood.

Search methods for identification of studies

We searched for all published and unpublished randomized controlled trials of antifibrinolytic therapy for the treatment of heavy menstrual bleeding, without language restriction and in consultation with the Gynaecology and Fertility Group (CGF) Information Specialist.

Electronic searches

We searched the following electronic databases for relevant trials.

1. The Cochrane Gynaecology and Fertility Group (CGF) Specialised Register of Controlled Trials (PROCITE platform) (searched 7 November 2017) (Appendix 1).

2. The Cochrane Central Register of Controlled Trials in the Cochrane Library, via the Cochrane Register of Studies Online (CRSO Web platform) (searched 7 November 2017) (Appendix 2).

3. MEDLINE (OVID platform) (searched from 1946 to 7 November 2017) (Appendix 3).

4. Embase (OVID platform) (searched from 1980 to 7 November 2017) (Appendix 4).

5. PsycINFO (OVID platform) (searched from 1806 to 7 November 2017) (Appendix 5).

The MEDLINE search was combined with the Cochrane Highly Sensitive Search Strategy for identifying randomized trials which appears in the Cochrane Handbook of Systematic Reviews of Interventions (Version 5.1.0 chapter 6, 6.4.11). The Embase, PsycINFO and CINAHL searches are combined with trial filters developed by the Scottish Intercollegiate Guidelines Network (SIGN) (www.sign.ac.uk/assets/search‐filters‐randomised‐controlled‐trials.docx).

Other electronic sources of trials included:

6. trial registers for ongoing and registered trials –

clinicaltrials.gov (a service of the US National Institutes of Health)
apps.who.int/trialsearch/Default.aspx (The World Health Organization International Trials Registry Platform search portal)

7. LILACS (Latin American and Caribbean Health Science Information database (from 1982 ongoing) and other Spanish and Portuguese language databases, found in the Virtual Health Library Regional Portal (VHL): lilacs.bvsalud.org  8. PubMed and Google Scholar (for recent trials not yet indexed in the major databases)

These databases were searched using the following subject headings and keywords: menorrhagia, dysfunctional uterine bleeding, heavy menstrual bleeding, antifibrinolytic, tranexamic acid, trans‐4‐aminomethylcyclohexanocarboxylic acid, KABI. Please see the Appendices for details.

Searching other resources

We handsearched reference lists of relevant trials and systematic reviews retrieved by the search and contact experts in the field to obtain additional data. We also handsearched relevant journals and conference abstracts that are not covered in the CGF register, in liaison with the Information Specialist. Marian Showell (Information Specialist for the Cochrane Gynaecology and Fertility Group (CGF)) performed the initial search, whilst AB‐S searched other electronic sources and resources.

Data collection and analysis

Selection of studies

The initial search was conducted by Marian Showell. After an initial screen of titles and abstracts retrieved by the search, conducted by AB‐S and AL, we retrieved the full texts of all potentially eligible studies. Two review authors (AB‐S and AL) independently examined these full text articles for compliance with the inclusion criteria and selected eligible studies. We corresponded with study investigators as required, to clarify study eligibility. Disagreements were resolved 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 (Figure 1).

Data extraction and management

Two review authors (AB‐S and AL) independently extracted data from eligible studies using a data extraction form designed and pilot‐tested by the authors. We resolved any disagreements by discussion; or if deadlock persisted, by involving a third review author as arbitrator. Data extracted included study characteristics and outcome data (see Data collection and analysis section for details). Where studies had multiple publications, we used the main trial report as the reference and derived additional details from secondary papers.

We corresponded with study investigators for further data on methods or results (or both), as required.

Assessment of risk of bias in included studies

Two review authors (AB‐S and AL) independently assessed the included studies for risk of bias using the Cochrane 'Risk of bias' assessment tool to assess: selection (random sequence generation and allocation concealment); performance (blinding of participants and personnel); detection (blinding of outcome assessors); attrition (incomplete outcome data); reporting (selective reporting); and other bias (Higgins 2011). We assigned judgement as recommended in the Cochrane Handbook for Systematic Reviews of Interventions Section 8.5 (Higgins 2011). We resolved disagreements by discussion; or if deadlock persisted, by involving a third review author as arbitrator. We described all judgements fully and present the conclusions in the 'Risk of bias' table, which we incorporated into the interpretation of review findings by means of sensitivity analyses (see below).

With respect to within‐trial selective reporting, where identified studies fail to report the primary outcome of live birth, but do report interim outcomes such as pregnancy, we assessed whether the interim values are similar to those reported in studies that also report live birth.

Measures of treatment effect

For dichotomous data (e.g. adverse event rates), we used the numbers of events in the control and intervention groups of each study to calculate Mantel‐Haenszel risk ratios (RRs). For continuous data (e.g. MBL in mL), if all studies report exactly the same outcomes we calculated mean difference (MDs) between treatment groups. If similar outcomes are reported on different scales we calculated the standardized mean difference (SMD). We reversed the direction of effect of individual studies, if required, to ensure consistency across trials. We treated ordinal data (e.g. quality of life scores) as continuous data. We presented 95% confidence intervals for all outcomes. Where data to calculate ORs or MDs are not available, we utilized the most detailed numerical data available that facilitated similar analyses of included studies (e.g. test statistics, P values). We assessed whether the estimates calculated in the review for individual studies are compatible in each case with the estimates reported in the study publications.

We included either end score or change score data for measuring MBL. Where studies report both, we used end scores. For the primary outcome, we considered whether the data underlying the published result were likely to be skewed by examining the ratio of each group mean to its standard deviation. Where end scores were reported, a ratio considerably less than two indicates positive skewness, due to the fact that 'menstrual bleeding' cannot take values less than zero. Where this was deemed to be the case, we reported the results in an additional table, as they could not be pooled in the meta‐analysis without access to the raw underlying data. Where no standard deviations were reported, we took a corresponding value reported in another similar study.

Where studies reported standard deviations that were implausibly small, we assumed that these were in fact standard errors, and converted them to standard deviations, using standard methods (Higgins 2011).

Unit of analysis issues

The primary analysis was per woman randomized. Only first‐phase data from cross‐over trials was included.

Dealing with missing data

We analyzed 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). Attempts were made to obtain missing data from the original trialists. Where these are unobtainable, we undertook imputation of individual values for primary outcomes only. For other outcomes, we analyzed only the available data. Any imputation undertaken was subjected to sensitivity analysis (see below).

If studies reported sufficient detail to calculate mean differences but no information on associated standard deviation (SD), we assumed the outcome to have a standard deviation equal to the highest SD from other studies within the same analysis.

Assessment of heterogeneity

We considered whether the clinical and methodological characteristics of the included studies were sufficiently similar for meta‐analysis to provide a clinically meaningful summary. We assessed statistical heterogeneity by the measure of the I² statistic. An I² measurement greater than 50% was taken to indicate substantial heterogeneity (Higgins 2011).

Assessment of reporting biases

In view of the difficulty of detecting and correcting for publication bias and other reporting biases, the authors aimed to minimise their potential impact by ensuring a comprehensive search for eligible studies, and by being alert for duplication of data. If there were 10 or more studies in an analysis, we planned to use a funnel plot to explore the possibility of small‐study effects (a tendency for estimates of the intervention effect to be more beneficial in smaller studies).

Data synthesis

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

Antifibrinolytics versus placebo or no treatment.
Antifibrinolytics versus any other medical (non‐surgical) treatment.

Any increase in the odds of a particular outcome, either beneficial (e.g. decreased MBL) or detrimental (e.g. adverse effects), were displayed graphically in the meta‐analyses to the right of the centre line, and a decrease in the odds of an outcome to the left of the centre line.

Subgroup analysis and investigation of heterogeneity

Where data are available, we conducted subgroup analyses to determine the separate evidence within the following subgroups.

Tranexamic acid dose (< 3 g/day versus ≥ 3 g/day)
Different methods of measuring MBL: objectively, by the alkaline haematin method and subjectively by the PBAC
Differences between the control interventions (e.g. luteal phase norethisterone (NET) or medroxyprogesterone (MPA))

If we detected substantial heterogeneity, we explored possible explanations in sensitivity analyses. We took any statistical heterogeneity into account when interpreting the results, especially if there was any variation in the direction of effect.

Sensitivity analysis

We conducted sensitivity analyses for the primary outcomes to determine whether the conclusions are robust to arbitrary decisions made regarding the eligibility and analysis. These analyses included consideration of whether the review conclusions would have differed if:

eligibility were restricted to studies without high risk of bias;
a random‐effects model had been adopted;
alternative imputation strategies had been implemented;
the summary effect measure was relative risk rather than odds ratio.

Overall quality of the body of evidence: 'Summary of findings' table

We prepared a 'Summary of findings' table using GRADEpro and Cochrane methods. This table evaluates the overall quality of the body of evidence for the main review outcomes (MBL, adverse events) for the main review comparison (antifibrinolytic agent versus placebo or other medical therapy). Additional 'Summary of findings' tables were also prepared for the main review outcomes for other important comparisons (antifibrinolytic agent versus progestogens, antifibrinolytic agent versus ethamsylate etc.). We assessed the quality of the evidence using GRADE criteria: risk of bias, consistency of effect, imprecision, indirectness and publication bias). Two review authors working independently made judgements about evidence quality (high, moderate, low or very low), with disagreements resolved by discussion. Judgements were justified, documented, and incorporated into reporting of results for each outcome.

We extracted study data, formatted our comparisons in data tables, and prepared a 'Summary of findings' table before writing the results and conclusions of our review.

Results

Description of studies

Results of the search

The 2010 version of this review included seven studies. Our new search retrieved 191 articles. Eighteen studies not included in the previous version of this review were potentially eligible and were retrieved in full text. Nine of these new studies met our inclusion criteria. We excluded three studies from the 2010 version of this review, and there are no studies awaiting classification; (see study tables: Characteristics of included studies, Characteristics of excluded studies).

See Figure 1 for the relevant PRISMA flow chart.

Included studies

Study design and setting

Thirteen studies are included in the current update of the review, twelve with a parallel group design and one cross‐over trial.

Seven of the studies were single‐centre; the country settings included Ireland, UK, India, Turkey, and Iran. The remaining studies drew subjects from various sites across individual countries (Sweden, USA, Thailand, Iran and China).

Participants

The studies (1312 participants) included 582 women in the control (non‐antifibrinolytic) groups and 778 in the intervention (i.e. tranexamic acid) groups. Their age ranged across studies from 15 to 50 years. Of note: Kriplani 2006 included women less than 18 years old without confirming that their HMB was ovulatory, whilst Bonnar 1996 and Callender 1970 only included women more than 32 years old.

Goshtasebi 2013 was the only study to use BMI as an inclusion criterion (19 to 29 kg/m²).

Several studies used serum haemoglobin as an inclusion criterion: Fathima 2012, Freeman 2011 and Lukes 2010 required women to have a serum haemoglobin of more than 8 g/dL; Kiseli 2016 used a cut‐off of more than 10 g/dL; whilst Goshtasebi 2013 and Goshtasebi 2015 required women to have a serum haemoglobin of more than 10.5 g/dL

Three studies excluded women with a self‐reported history of irregular menstrual bleeding (Bonnar 1996; Freeman 2011; Lukes 2010); whilst Jaisamrarn 2006; and Preston 1995 tested mid‐luteal progesterone to confirm that HMB was ovulatory.

All studies except Zhang 2008 mentioned excluding women with an underlying pelvic aetiology of their HMB, although several did not elucidate how they excluded pelvic pathology (i.e. by history, examination, ultrasound, hysteroscopy or endometrial biopsy, or a combination of these).

Fathima 2012 included women with leiomyomata. Goshtasebi 2013 and Kriplani 2006 excluded women found to have uterine leiomyomata, whilst Freeman 2011 and Lukes 2010 only excluded women with fibroids thought to warrant surgical management. Goshtasebi 2015 excluded women with fibroids greater than 3 cm in diameter, and Kiseli 2016 excluded women with fibroids that were greater than 2 cm or indented the uterine cavity on ultrasound.

Several studies mentioned adenomyosis: Fathima 2012 included women with adenomyosis, whilst Kriplani 2006 specifically excluded women thought to have adenomyosis.

Bonnar 1996, Edlund 1995, Freeman 2011, Lukes 2010 and Preston 1995 required a negative Pap smear within 0 to 12 months of trial entry. Kiseli 2016 excluded women with malignant cervicovaginal pathology.

The following studies excluded women who were taking medications that might affect their menstrual pattern (such as anticoagulants, aspirin, or NSAIDs/COX‐2 inhibitors during the menstrual phase of their cycle): Edlund 1995; Freeman 2011; Goshtasebi 2013; Goshtasebi 2015; Jaisamrarn 2006; Lukes 2010; Preston 1995; Zhang 2008. Several studies listed hormonal contraception as an exclusion criterion (Edlund 1995; Fathima 2012; Freeman 2011; Goshtasebi 2013; Goshtasebi 2015; Jaisamrarn 2006; Kriplani 2006; Lukes 2010; Preston 1995; Zhang 2008). Kiseli 2016 excluded women with a history of having taken medications for menorrhagia previously.

With regard to non‐gynaecological and non‐haematological co‐morbidities: Bonnar 1996, Edlund 1995, Jaisamrarn 2006 and Zhang 2008 excluded women with renal/hepatic dysfunction, and Preston 1995 those with renal dysfunction. Bonnar 1996, Edlund 1995, Fathima 2012, Freeman 2011, Jaisamrarn 2006, Lukes 2010 and Zhang 2008 mentioned excluding women with either a history of, or definitive proof of, a coagulopathy or fibrinolytic disorder. Kiseli 2016 measured participants' coagulation profile.

The following studies excluded women who reported a history of venous thromboembolism: Bonnar 1996; Edlund 1995; Freeman 2011; Goshtasebi 2013; Goshtasebi 2015; Jaisamrarn 2006; Lukes 2010 and Zhang 2008. Freeman 2011 also excluded women with a history of arterial thrombosis (i.e. ischaemic heart disease, acute myocardial infarction, stroke/cerebrovascular accident, transient ischaemic attack); Kiseli 2016 excluded women with coronary artery disease. The following studies excluded women with a history of coagulopathy/fibrinolytic disorder: Bonnar 1996; Edlund 1995; Fathima 2012; Freeman 2011; Goshtasebi 2015; and Lukes 2010.

Fathima 2012, Freeman 2011, Kiseli 2016, and Kriplani 2006 excluded women with thyroid disease, whilst Freeman 2011 also excluded women with hyperprolactinaemia.

There were several co‐morbidities that were used as an exclusion criterion by only one study: Bonnar 1996 excluded women with inflammatory bowel disease, peptic/intestinal ulceration; Fathima 2012 excluded women with a history of diabetes, hypertension, tuberculosis, "malignancies" or hypothalamic pituitary dysfunction; Kiseli 2016 excluded women with hypertension or diabetes; Lukes 2010 excluded women with a history of sub‐arachnoid haemorrhage, "endocrinopathy" or ocular disease.

Goshtasebi 2013 simply reports excluding women with any "history of chronic diseases".

Interventions

The thirteen studies (twelve parallel group, and one cross‐over trial) used various antifibrinolytic formulations and dosage regimens, as detailed below.

Tranexamic acid, used by 12 of the 13 studies, is a synthetic analogue of the amino acid lysine. Kabi (used in Edlund 1995) is a pro‐drug of TXA.

Dosage

The majority of studies used regular dose TXA (ranging from 3 g/day to 4 g/day) (Bonnar 1996; Callender 1970; Fathima 2012; Jaisamrarn 2006; Kiseli 2016; Lukes 2010; Preston 1995; Zhang 2008).

Four other studies used low‐dose TXA (ranging from 2 g/day to 2.4 g/day) (Edlund 1995; Goshtasebi 2013; Goshtasebi 2015; Kriplani 2006).

Freeman 2011 compared low‐dose (1.95 g/day) to regular‐dose TXA (3.9 g/day) to placebo.

Treatments

Four studies compared antifibrinolytic treatment to placebo: Callender 1970 (a cross‐over trial); Edlund 1995; Freeman 2011; and Lukes 2010 (all parallel group studies). No studies compared antifibrinolytic treatment to no treatment.

Nine studies compared antifibrinolytic treatment to other medical therapies: Bonnar 1996; Fathima 2012; Goshtasebi 2013; Goshtasebi 2015; Jaisamrarn 2006; Kiseli 2016; Kriplani 2006; Preston 1995; and Zhang 2008. Of note: some studies compared antifibrinolytic treatment to two different alternative medical options.

6 assessed progestogens (short or long course) (Goshtasebi 2013; Jaisamrarn 2006; Kiseli 2016; Kriplani 2006; Preston 1995; Zhang 2008)
2 assessed NSAIDs (Bonnar 1996; Jaisamrarn 2006)
1 assessed ethamsylate (Bonnar 1996)
2 assessed herbal remedies (one assessed Safoof Habis, and another pomegranate flower) (Fathima 2012; Goshtasebi 2015)
1 assessed levonorgestrel intrauterine system (LIUS)

Of the six studies that compared TXA to progestogens, four — Jaisamrarn 2006, Kiseli 2016, Preston 1995 and Zhang 2008 — compared TXA to short‐course progestogen (e.g. days 19 to 26 of the menstrual cycle only), whilst two — Goshtasebi 2013 and Kriplani 2006 — compared TXA to long‐course progestogens (e.g. from days 5 to 26).

One study compared antifibrinolytic treatment to a Unani formulation called Safoof Habis (Fathima 2012). Unani is a type of traditional medicine widely practised in South‐East Asia. Safoof Habis is made up of: Teen Ahmer (silicate of alumina and iron oxide); Sange Jarahat (hydrated magnesium silicate); and Raal Sufaid (Vateria indica Linn, which is a species of plant in the Disterocarpaceae family, endemic to India). In Fathima 2012 the treatment arm (Safoof Habis) was made up of equal parts of all three components, and given in 5 g doses twice per day, from days 1 to 5.

One study compared TXA (500 mg four times per day on days 1 to 5) to Punica granatum Linn (pomegranate flower) (500 mg four times per day on days 1 to 5) (Goshtasebi 2015).

Outcomes

Objective assessment of women's menstrual blood loss

Five parallel group studies confirmed the participants' HMB with objective testing such as alkaline haematin testing (Bonnar 1996; Edlund 1995; Freeman 2011; Lukes 2010; Preston 1995). The alkaline haematin method of quantifying MBL was developed by Hallberg and Nilsson in the 1960s, and involves women collecting their menstrual pads, then sending them to the lab for analysis (for extraction and measurement of the amount of blood) (Hallberg 1964).

The one cross‐over study used the Oxford total body counter as a way of quantifying participants' MBL (Callender 1970).

Subjective assessment of women's menstrual blood loss

Seven studies — Fathima 2012, Goshtasebi 2013, Goshtasebi 2015, Jaisamrarn 2006, Kiseli 2016, Kriplani 2006 and Zhang 2008 — assessed women’s menstrual bleeding by Pictorial Blood Assessment Chart (PBAC) (Higham 1990).

The PBAC involves women recording the number of pads/tampons used, and documenting the degree of soiling. The PBAC has its limitations: for instance it is binary in nature and there is no international consensus on the cut‐off level for the definition of HMB. Some authors consider 100 points in the PBAC to be equivalent to 80 mL of MBL (i.e. HMB); some 150 points. Also, the PBAC does not allow for a volumetric correlation between the patient's PBAC score, and the volume of MBL.

Self‐reported improvement in HMB

The gold standard for diagnosing HMB is patient's self‐reported assessments of their own menstrual loss. Six studies asked women to self‐report heavy menstrual bleeding, using a questionnaire (Bonnar 1996; Callender 1970; Edlund 1995; Freeman 2011; Goshtasebi 2013; Goshtasebi 2015).

Quality of life measures

Eight studies reported quality of life measures. Freeman 2011, Goshtasebi 2013, Goshtasebi 2015 and Lukes 2010 used the Menorrhagia Impact Questionnaire (with lower scores representing better quality of life). In addition, Goshtasebi 2013 and Goshtasebi 2015 also used the SF‐36 quality of life questionnaire (with higher scores representing better quality of life). Jaisamrarn 2006 used a questionnaire which included six questions relating to the impact of HMB on impairment of social life, work performance, tiredness, productivity, appetite and depression; (these data could not be used because measures of variation were not reported). Kiseli 2016 used the World Health Organization's Quality of Life‐Short Form, Turkish version (WHOQOL‐BREF TR), in which patients report limitations in physical health, psychological status, social support, and limitations relating to their environment. Preston 1995 used a 5‐point scale for quality‐of‐life assessments, which evaluated general health, flooding and leakage, abdominal pain, limitation on social activities, and effect on sex life. Zhang 2008 used a 6‐item questionnaire, but it was not clear whether it was validated, or what specific items were assessed.

Adverse events

Eleven studies reported adverse events (other than thromboembolic events): Bonnar 1996; Callender 1970; Edlund 1995; Freeman 2011; Goshtasebi 2013; Goshtasebi 2015; Kiseli 2016; Kriplani 2006; Lukes 2010; Preston 1995; and Zhang 2008. The adverse events reported included: abdominal pain, allergic reaction, anxiety, back pain, bloating, breast tenderness, depression, chest pain, diarrhoea, dizziness, dysmenorrhoea, dyspepsia, excess hair growth, headache, intermenstrual bleeding, menstrual cramps, mood changes, myalgia, nausea, ocular events (lenticular opacities, blurred vision), rash, vaginal dryness, vertigo, vomiting, weight gain.

Excluded studies

Nine studies were excluded from this version of the review, for the following reasons.

1/9 was not a RCT (Muse 2010)
1/9 only included women who had a proven coagulopathy (Kouides 2009)
1/9 compared women taking tranexamic acid to women taking a combination of tranexamic and mefenamic acids (Najam 2010)
4/9 were cross‐over trials that did not provide data at the end of the first phase of the study, before participants were crossed over (Andersch 1988; Nilsson 1967; NCT01428713; Vermylen 1968)
2/9 were excluded when no response was received from the authors after several attempts to contact them (Moghtadaei 2012; Tabatabaei 2013)

Risk of bias in included studies

Please see Figure 2 and Figure 3 for summaries of the risk of bias for the included studies.

Methodological quality graph: review authors' judgements about each methodological quality item presented as percentages across all included studies.

Methodological quality summary: review authors' judgements about each methodological quality item for each included study.

Allocation

Nine studies were rated as being at low risk of selection bias related to sequence generation, as they used computer randomization or a random numbers table. The other four studies did not describe the method used and were rated at unclear risk of this bias.

Six studies were at low risk of selection bias related to allocation concealment. Seven studies did not describe the method used, and were at unclear risk of allocation bias.

Blinding

We did not consider that blinding was likely to influence findings for the primary review outcome (MBL), where this was measured by the objective alkaline haematin method; however, where women used the PBAC, knowledge of their treatment may have influenced their assessment of blood loss. In addition, for adverse effects and subjective secondary outcomes (such as quality of life), blinding status could also potentially affect findings.

Six studies were deemed to be at low risk of this bias. Two studies did not describe the method used, so were deemed to be at unclear risk of detection bias. Five studies were deemed to be at high risk of this bias.

Incomplete outcome data

Seven studies analyzed all or most (> 95%) of the women randomized and we judged them to be at low risk of attrition bias. Three studies were at unclear risk of attrition bias. Three studies were at high risk of attrition bias.

Selective reporting

We rated most studies as at low risk of selective reporting bias. However, the abstract of Zhang 2008 only reported results where significant differences were seen between experimental and control groups, potentially leading to overly‐optimistic conclusions; this trial was reported as having a high risk of reporting bias.

Other potential sources of bias

We found no potential sources of within‐study bias in 11 studies. Two studies were thought to be at unclear risk of within‐study bias: Goshtasebi 2015, as baseline factors were only reported for women who completed the study (therefore, baseline comparability is unknown); Lukes 2010, due to imbalances between the two groups at baseline.

Effects of interventions

See: Table 1; Table 2; Table 3; Table 4; Table 5; Table 6

Summary of findings for the main comparison. Antifibrinolytics compared to no treatment or placebo for heavy menstrual bleeding.

Antifibrinolytics compared to no treatment or placebo for heavy menstrual bleeding
Population: women with heavy menstrual bleeding  Setting: gynaecology outpatient departments; one study simply said ''clinical sites''  Intervention: antifibrinolytics  Comparison: no treatment or placebo
Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect  (95% CI)	№ of participants  (studies)	Quality of the evidence  (GRADE)	Comments
	Risk with no treatment or placebo	Risk with antifibrinolytics
Menstrual blood loss: mean loss  Assessed with: alkaline haematin method  Follow‐up: range 3 months to 6 months	The mean menstrual blood loss: mean loss ranged from 206 mL to 252 mL	MD 53.2 mL lower  (62.7 lower to 43.7 lower)	‐	565  (4 RCTs)	⊕⊕⊕⊝  MODERATE ¹
Menstrual blood loss: improvement rates  Assessed with a variety of methods  Follow‐up: range 3 months to 6 months	109 per 1000	363 per 1000  (200 to 662)	RR 3.34  (1.84 to 6.09)	271  (3 RCTs)	⊕⊕⊕⊝  MODERATE ²
Adverse events (any)	836 per 1000	990 per 1000	RR 1.05 (0.93 to 1.18)	297 (1 RCT)	⊕⊕⊝⊝ LOW³	Most of these adverse events were mild to moderate in severity.
Thromboembolic events	Only one thromboembolic event occurred in the two studies that reported this outcome.
*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; RR: Risk ratio
GRADE Working Group grades of evidence  High quality: we are very confident that the true effect lies close to that of the estimate of the effect  Moderate quality: we are moderately confident in the effect estimate — the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different  Low quality: our confidence in the effect estimate is limited — the true effect may be substantially different from the estimate of the effect  Very low quality: we have very little confidence in the effect estimate — the true effect is likely to be substantially different from the estimate of effect

Antifibrinolytics compared to progestogens for heavy menstrual bleeding
Population: women with heavy menstrual bleeding  Setting: any  Intervention: antifibrinolytics  Comparison: progestogens
Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect  (95% CI)	№ of participants  (studies)	Quality of the evidence  (GRADE)	Comments
	Risk with progestogens	Risk with antifibrinolytics
Menstrual blood loss: mean loss (overall)  Assessed with: PBAC¹¹  Follow‐up: range 2 months to 3 months	The mean PBAC score ranged from 114 to 209 pts	MD 12.22 pts lower  (30.8 lower to 6.36 higher)	‐	312  (3 RCTs)	⊕⊝⊝⊝  VERY LOW ^{1 2}	Two additional trials (at low risk of bias) had skewed data and are displayed in Additional Table 1. They both found a significant benefit for TXA.
Menstrual blood loss: improvement rates (overall)  Assessed with: patient assessment  Follow‐up: range 2 months to 3 months	463 per 1000	701 per 1000 (607 to 833)	RR 1.54  (1.31 to 1.80)	422  (5 RCTs)	⊕⊕⊝⊝  LOW ⁶	Overall effect combining short‐ and long‐course progestogens.
Adverse events (any)	319 per 1000	210 per 1000 Need to add CI here	RR 0.66 (0.46 to 0.94)	349 (4 RCTs)	⊕⊕⊝⊝ LOW¹⁰
Thromboembolic events	No thromboembolic events were diagnosed in either group
*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; RR: Risk ratio
GRADE Working Group grades of evidence  High quality: we are very confident that the true effect lies close to that of the estimate of the effect  Moderate quality: we are moderately confident in the effect estimate — the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different  Low quality: our confidence in the effect estimate is limited — the true effect may be substantially different from the estimate of the effect  Very low quality: we have very little confidence in the effect estimate — the true effect is likely to be substantially different from the estimate of effect

Antifibrinolytics compared to NSAIDs for heavy menstrual bleeding
Population: women with heavy menstrual bleeding  Setting: all  Intervention: antifibrinolytics  Comparison: NSAIDs
Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect  (95% CI)	№ of participants  (studies)	Quality of the evidence  (GRADE)	Comments
	Risk with NSAIDs	Risk with antifibrinolytics
Menstrual blood loss: mean loss  Assessed with: alkaline haematin method  Follow‐up: mean 3 months	The mean menstrual blood loss: mean loss was 148 mL	MD 73 mL lower  (123.35 lower to 22.65 lower)	‐	49  (1 RCT)	⊕⊕⊝⊝  LOW ^{1 2}
Menstrual blood loss: improvement rates  Assessed with: patient questionnaire  Follow‐up: range 2 months to 3 months	608 per 1000	869 per 1000  (717 to 1000)	RR 1.43  (1.18 to 1.74)	161  (2 RCTs)	⊕⊕⊝⊝  LOW ³
Adverse events (any)	The total number of adverse events per group were not measured. Individual adverse events were uncommon.
Thromboembolic events	These were not measured in the study.
*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; RR: Risk ratio
GRADE Working Group grades of evidence  High quality: we are very confident that the true effect lies close to that of the estimate of the effect  Moderate quality: we are moderately confident in the effect estimate — the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different  Low quality: our confidence in the effect estimate is limited — the true effect may be substantially different from the estimate of the effect  Very low quality: we have very little confidence in the effect estimate — the true effect is likely to be substantially different from the estimate of effect

Antifibrinolytics compared to ethamsylate for heavy menstrual bleeding
Population: women with heavy menstrual bleeding  Setting: all  Intervention: antifibrinolytics  Comparison: ethamsylate
Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect  (95% CI)	№ of participants  (studies)	Quality of the evidence  (GRADE)	Comments
	Risk with ethamsylate	Risk with antifibrinolytics
Menstrual blood loss: mean loss  Assessed with: alkaline haematin method  Follow‐up: mean 3 months	The mean menstrual blood loss: mean loss was 175 mL	MD 100 mL lower  (141.82 lower to 58.18 lower)	‐	53  (1 RCT)	⊕⊕⊝⊝  LOW ^{1 2}
Menstrual blood loss: improvement rates  Assessed with: patient assessment  Follow‐up: mean 3 months	444 per 1000	693 per 1000  (422 to 1000)	RR 1.56  (0.95 to 2.55)	53  (1 RCT)	⊕⊝⊝⊝  VERY LOW ^{2 3}
Withdrawal from treatment because of adverse events	148 per 1000	115 per 1000	RR 0.78 (0.19 to 3.15)	53 (1 RCT)	⊕⊝⊝⊝  VERY LOW ^{2 3}
Thromboembolic events	Thromboembolic events were not measured in the study.
*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; RR: Risk ratio
GRADE Working Group grades of evidence  High quality: we are very confident that the true effect lies close to that of the estimate of the effect  Moderate quality: we are moderately confident in the effect estimate — the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different  Low quality: our confidence in the effect estimate is limited — the true effect may be substantially different from the estimate of the effect  Very low quality: we have very little confidence in the effect estimate — the true effect is likely to be substantially different from the estimate of effect

Antifibrinolytics compared to herbal medicines for heavy menstrual bleeding
Patient or population: heavy menstrual bleeding  Setting: all  Intervention: antifibrinolytics  Comparison: herbal medicines
Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect  (95% CI)	№ of participants  (studies)	Quality of the evidence  (GRADE)	Comments
	Risk with herbal medicines	Risk with antifibrinolytics
Menstrual blood loss: mean loss ‐ after 3 months Rx  Assessed with: PBAC⁴	The mean PBAC score ranged from 51 to 143 pts	MD 23.90 pts lower (31.2 lower to 15.88 lower)		121  (2 RCTs)	⊕⊕⊝⊝  LOW ¹	One trial assessed a 'unani' formulation vs TXA and the other a pomegranate extract vs TXA
Menstrual blood loss: mean loss ‐ after 3 months' follow‐up from end of Rx  Assessed with: PBAC⁴  Follow‐up: mean 3 months	The mean PBAC score was 71.3 pts	MD 10.4 pts lower  (19.2 lower to 1.6 lower)	‐	45  (1 RCT)	⊕⊝⊝⊝  VERY LOW ^{2 3}	PBAC scores evaluated 3 months after Rx was completed in both groups.
Rates of improvement	This outcome was not reported.
Adverse events (any)	85 per 1000	191 per 1000	RR 2.25 (0.74 to 6.80)	94 (1 RCT)	⊕⊝⊝⊝  VERY LOW ^{2 3}
Thromboembolic events	The study did not measure any thromboembolic events.
*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; RR: Risk ratio; OR: Odds ratio
GRADE Working Group grades of evidence  High quality: we are very confident that the true effect lies close to that of the estimate of the effect  Moderate quality: we are moderately confident in the effect estimate — the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different  Low quality: our confidence in the effect estimate is limited — the true effect may be substantially different from the estimate of the effect  Very low quality: we have very little confidence in the effect estimate — the true effect is likely to be substantially different from the estimate of effect

Antifibrinolytics compared with levonorgestrel for heavy menstrual bleeding
Patient or population: heavy menstrual bleeding Settings: all Intervention: antifibrinolytics Comparison: levonorgestrel intrauterine system (IUS)
Outcomes	*Illustrative comparative risks (95% CI)**		Relative effect  (95% CI)	No of Participants  (studies)	Quality of the evidence  (GRADE)	Comments
	Assumed risk	Corresponding risk
	Risk with levonorgestrel IUS	Risk with TXA
Menstrual blood loss: median difference in PBAC score ‐ after 6 months Rx Assessed with: PBAC⁴ Follow‐up: outcomes measured at end of treatment	−252.0 (IQR 124.5) ¹	−126.5 (IQR 104.5)		42 (1 RCT)	⊕⊝⊝⊝  VERY LOW^{2 3}
Menstrual blood loss: improvement in mean blood loss (PBAC score < 100) Assessed with: PBAC⁴ Follow‐up: outcomes measured at end of treatment	850 per 1000	364 per 1000 (204 to 655)	RR 0.43 (0.24 to 0.77)	42 (1 RCT)	⊕⊝⊝⊝  VERY LOW^{2 3}
Adverse events (any)	500 per 1000	455 per 1000	RR 0.83 (0.25 to 2.80)	42 (1 RCT)	⊕⊝⊝⊝  VERY LOW^{2 3}
Thromboembolic events	The study did not measure any thromboembolic events.
The basis for the assumed risk* (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (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; IQR: Interquartile Range; OR: Odds Ratio; RR: Risk Ratio
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.

Study	Comparison	Outcome	Intervention group			Control group			Finding (e.g. P value for difference between groups, as reported in primary study)
Study	Comparison	Outcome	Intervention	n	Result (mean (SD))	Intervention	n	Result (mean (SD))
Jaisamrarn 2006	TXA vs progestogens or NSAIDs	PBAC score	TXA 3 g daily	56	204.4 (SD 255.7)	Norethisterone 10 mg daily (luteal phase)	56	298.7 (SD 141.3)	P < 0.0001
Jaisamrarn 2006	TXA vs progestogens or NSAIDs	PBAC score	TXA 3 g daily	56	204.4 (SD 255.7)	Mefenamic acid 1.5 mg daily	56	278.3 (SD 164.2)	P < 0.001
Preston 1995	TXA vs luteal phase MPA	Mean loss by alkaline haematin method (end score)	TXA 4 g daily	25	97 (SD 89)	Norethisterone (21 participants) 5 mg taken twice daily on days 19 to 26 of cycle.	21	208 (SD 135)	P = 0.001

Median difference in PBAC score
Study	Comparison	N (TXA)	Median (TXA)	IQR (TXA)	N (LIUS)	Median (LIUS)	IQR (LIUS)	P value
After 6 months Rx
Kiseli 2016	TXA vs. LIUS	22	‐126.5	104.5	20	‐252	124.5	0.002

Date	Event	Description
6 June 2018	Amended	Correction of detail in Abstract and PLS
19 April 2018	Amended	Correction to title of plain language summary
19 April 2018	Amended	Correction of author address
7 November 2017	New search has been performed	Methods section updated to current Cochrane standards.
7 November 2017	New citation required but conclusions have not changed	Six new studies were added (Fathima 2012, Freeman 2011, Goshtasebi 2013, Goshtasebi 2015, Kiseli 2016, and Zhang 2008); and 3 previously excluded studies were included (Jaisamrarn 2006,Kriplani 2006, and Lukes 2010). Three previously included studies were found not to meet inclusion criteria and have been excluded (Andersch 1988; Nilsson 1967; Vermylen 1968). With the addition of new studies and exclusion of others, the conclusions of this review have not changed.
6 November 2008	Amended	Converted to new review format.
30 August 2000	New citation required and conclusions have changed	Substantive amendment

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Menstrual blood loss: mean loss	4	565	Mean Difference (IV, Fixed, 95% CI)	‐53.20 [‐62.70, ‐43.70]
1.1 Objective assessment of mean loss	4	565	Mean Difference (IV, Fixed, 95% CI)	‐53.20 [‐62.70, ‐43.70]
2 Menstrual blood loss: improvement rates	3	271	Risk Ratio (M‐H, Fixed, 95% CI)	3.34 [1.84, 6.09]
3 Quality of life scores (change from baseline)	2		Mean Difference (IV, Random, 95% CI)	Subtotals only
3.1 Limitations in social/leisure activities	2	365	Mean Difference (IV, Random, 95% CI)	0.52 [0.31, 0.74]
3.2 Limitations in physical activities	2	365	Mean Difference (IV, Random, 95% CI)	0.55 [0.34, 0.77]
3.3 Limitations in work in or outside the home	1	187	Mean Difference (IV, Random, 95% CI)	0.55 [0.30, 0.80]
4 Adverse events	3		Risk Ratio (M‐H, Fixed, 95% CI)	Subtotals only
4.1 Any adverse event	1	297	Risk Ratio (M‐H, Fixed, 95% CI)	1.05 [0.93, 1.18]
4.2 Gastrointestinal side effects	2	244	Risk Ratio (M‐H, Fixed, 95% CI)	1.07 [0.55, 2.07]
4.3 Headache	2	244	Risk Ratio (M‐H, Fixed, 95% CI)	1.09 [0.82, 1.45]
4.4 Uterine cancer	1	55	Risk Ratio (M‐H, Fixed, 95% CI)	1.38 [0.06, 32.36]
4.5 Vaginal dryness	1	55	Risk Ratio (M‐H, Fixed, 95% CI)	0.15 [0.01, 3.60]
4.6 Dysmenorrhoea	2	244	Risk Ratio (M‐H, Fixed, 95% CI)	1.18 [0.90, 1.55]
4.7 Viral URTI	2	486	Risk Ratio (M‐H, Fixed, 95% CI)	1.19 [0.58, 2.47]
4.8 Fatigue	2	486	Risk Ratio (M‐H, Fixed, 95% CI)	1.65 [0.68, 3.97]
4.9 Musculoskeletal pain	2	486	Risk Ratio (M‐H, Fixed, 95% CI)	1.40 [0.82, 2.39]
4.10 Arthralgia	2	486	Risk Ratio (M‐H, Fixed, 95% CI)	1.78 [0.72, 4.41]
4.11 Myalgia	2	486	Risk Ratio (M‐H, Fixed, 95% CI)	2.52 [0.78, 8.16]
4.12 Nasal congestion	1	297	Risk Ratio (M‐H, Fixed, 95% CI)	6.77 [0.40, 113.42]
4.13 Sinusitis	2	486	Risk Ratio (M‐H, Fixed, 95% CI)	2.82 [0.84, 9.51]
4.14 Multiple allergies	2	486	Risk Ratio (M‐H, Fixed, 95% CI)	1.78 [0.68, 4.65]
4.15 Throat irritation	1	297	Risk Ratio (M‐H, Fixed, 95% CI)	1.02 [0.22, 4.79]
4.16 Anaemia	2	486	Risk Ratio (M‐H, Fixed, 95% CI)	1.89 [0.72, 4.99]
4.17 Abdominal discomfort	1	189	Risk Ratio (M‐H, Fixed, 95% CI)	0.82 [0.30, 2.27]
4.18 Cough	1	189	Risk Ratio (M‐H, Fixed, 95% CI)	0.86 [0.28, 2.61]
4.19 Insomnia	1	189	Risk Ratio (M‐H, Fixed, 95% CI)	0.62 [0.21, 1.84]
4.20 Dyspepsia	1	189	Risk Ratio (M‐H, Fixed, 95% CI)	0.23 [0.06, 0.84]
4.21 Migraine	1	189	Risk Ratio (M‐H, Fixed, 95% CI)	2.15 [0.46, 10.09]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Menstrual blood loss: mean PBAC score	3	312	Mean Difference (IV, Fixed, 95% CI)	‐12.22 [‐30.80, 6.36]
1.1 Luteal phase NET	1	128	Mean Difference (IV, Fixed, 95% CI)	‐25.0 [‐65.54, 15.54]
1.2 Long course (20‐25 days) MPA	2	184	Mean Difference (IV, Fixed, 95% CI)	‐8.82 [‐29.73, 12.09]
2 Median difference in PBAC score			Other data	No numeric data
2.1 After 6 months Rx			Other data	No numeric data
3 Percentage difference in PBAC scores			Other data	No numeric data
3.1 After 6 months Rx			Other data	No numeric data
4 Menstrual blood loss: improvement rates	5	422	Risk Ratio (M‐H, Fixed, 95% CI)	1.54 [1.31, 1.80]
4.1 Luteal phase	4	328	Risk Ratio (M‐H, Fixed, 95% CI)	1.66 [1.34, 2.05]
4.2 Long course (20 to 25 days)	1	94	Risk Ratio (M‐H, Fixed, 95% CI)	1.32 [1.08, 1.61]
5 Quality of life ‐ improvement (luteal phase MPA)	1		Risk Ratio (M‐H, Fixed, 95% CI)	Subtotals only
5.1 General health	1	44	Risk Ratio (M‐H, Fixed, 95% CI)	1.67 [0.76, 3.64]
5.2 Abdominal pain	1	44	Risk Ratio (M‐H, Fixed, 95% CI)	1.88 [0.68, 5.19]
5.3 Limitation of social activities	1	44	Risk Ratio (M‐H, Fixed, 95% CI)	1.48 [0.85, 2.60]
5.4 Sex life	1	44	Risk Ratio (M‐H, Fixed, 95% CI)	3.06 [0.99, 9.46]
6 Quality of life ‐ SF36 (long course MPA)	2		Mean Difference (IV, Fixed, 95% CI)	Subtotals only
6.1 Physical functioning	2	132	Mean Difference (IV, Fixed, 95% CI)	0.79 [‐0.67, 2.25]
6.2 Role physical	1	90	Mean Difference (IV, Fixed, 95% CI)	4.0 [‐5.63, 13.63]
6.3 Bodily pain	1	90	Mean Difference (IV, Fixed, 95% CI)	‐2.80 [‐11.61, 6.01]
6.4 General health	1	90	Mean Difference (IV, Fixed, 95% CI)	5.0 [‐2.49, 12.49]
6.5 Vitality	1	90	Mean Difference (IV, Fixed, 95% CI)	1.40 [‐6.62, 9.42]
6.6 Social functioning	2	132	Mean Difference (IV, Fixed, 95% CI)	0.23 [‐1.76, 2.22]
6.7 Role emotional	1	90	Mean Difference (IV, Fixed, 95% CI)	‐0.5 [‐12.47, 11.47]
6.8 Mental health	2	132	Mean Difference (IV, Fixed, 95% CI)	0.36 [‐1.07, 1.79]
6.9 Environmental domain	1	42	Mean Difference (IV, Fixed, 95% CI)	1.20 [‐0.17, 2.57]
6.10 Environmental domain ‐ TR	1	42	Mean Difference (IV, Fixed, 95% CI)	0.90 [‐0.39, 2.19]
7 Quality of life ‐ HMB score	2	218	Std. Mean Difference (IV, Fixed, 95% CI)	‐0.06 [‐0.32, 0.21]
7.1 Luteal phase progestogen	1	128	Std. Mean Difference (IV, Fixed, 95% CI)	0.0 [‐0.35, 0.35]
7.2 Long course progestogen	1	90	Std. Mean Difference (IV, Fixed, 95% CI)	‐0.14 [‐0.56, 0.27]
8 Adverse events (short and long course progestogens)	6	1905	Risk Ratio (M‐H, Fixed, 95% CI)	0.67 [0.52, 0.86]
8.1 Any adverse event	4	349	Risk Ratio (M‐H, Fixed, 95% CI)	0.66 [0.46, 0.94]
8.2 Gastrointestinal events	2	140	Risk Ratio (M‐H, Fixed, 95% CI)	0.58 [0.22, 1.50]
8.3 Headache	4	294	Risk Ratio (M‐H, Fixed, 95% CI)	0.79 [0.44, 1.40]
8.4 Dysmenorrhoea	2	158	Risk Ratio (M‐H, Fixed, 95% CI)	0.58 [0.29, 1.19]
8.5 Weight gain	2	88	Risk Ratio (M‐H, Fixed, 95% CI)	2.04 [0.31, 13.46]
8.6 Allergic reaction	1	94	Risk Ratio (M‐H, Fixed, 95% CI)	2.76 [0.12, 66.07]
8.7 Intermenstrual bleeding	1	94	Risk Ratio (M‐H, Fixed, 95% CI)	0.08 [0.00, 1.47]
8.8 Giddiness	1	94	Risk Ratio (M‐H, Fixed, 95% CI)	0.31 [0.03, 2.84]
8.9 Breast tenderness	2	136	Risk Ratio (M‐H, Fixed, 95% CI)	0.18 [0.01, 3.73]
8.10 Mood changes	1	94	Risk Ratio (M‐H, Fixed, 95% CI)	0.31 [0.01, 7.34]
8.11 Muscle pain	1	112	Risk Ratio (M‐H, Fixed, 95% CI)	3.0 [0.12, 72.10]
8.12 Bloating	1	42	Risk Ratio (M‐H, Fixed, 95% CI)	0.18 [0.01, 3.59]
8.13 Nausea	1	42	Risk Ratio (M‐H, Fixed, 95% CI)	2.73 [0.31, 24.14]
8.14 Rash	1	42	Risk Ratio (M‐H, Fixed, 95% CI)	2.74 [0.12, 63.63]
8.15 Spotting	1	42	Risk Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
8.16 Excess hair growth	1	42	Risk Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
8.17 Depression	1	42	Risk Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]

Methods	Multi‐centre, randomized, double‐blind, placebo‐controlled, parallel‐group study Randomization: method of randomization not stated Blinding: participants, investigators and assessors all blinded Duration of trial: 2 control cycles pre‐treatment, followed by 3 cycles of treatment
Participants	Country: USA Number of participants: 304 randomized, 272 analyzed (106 in low‐dose TXA arm; 103 in high‐dose in control arm; 63 in placebo arm) Age: 18 to 49 years old Inclusion criteria: 'history of cyclic HMB', confirmed by alkaline haematin method during the 2 pre‐treatment cycles; normal pelvic examination and Pap smear; normal transvaginal ultrasound Exclusion criteria: 'clinically significant disease', anovulatory dysfunctional uterine bleeding, metrorrhagia, menometrorrhagia, polymenorrhoea, endometrial polyps, endometrial hyperplasia, endometrial carcinoma, cervical carcinoma, myocardial infarction, ischaemic disease, stroke, transient ischaemic attack, VTE, or coagulopathy; abnormality on electrocardiography; serum prolactin > 30 μg/L; uncontrolled hypothyroidism; severe anaemia (Hb < 8 g/dL); history of bilateral oophorectomy or hysterectomy; women who were pregnant, breast‐feeding, planning to become pregnant during the study, or became pregnant during the study; women currently taking anticoagulants, aspirin, dong quai, aminocaproic acid, hydroxychloroquine, or hormonal contraceptives Note: women with fibroids were only excluded if they were thought to require surgical management
Interventions	First intervention arm: oral TXA 0.65 g 3 times daily for up to 5 consecutive days during menstruation versus Second intervention arm: oral TXA 1.3 g 3 times daily for up to 5 consecutive days during menstruation versus Control arm: placebo tablets, 3 times daily, for up to 5 consecutive days during menstruation
Outcomes	MBL: alkaline haematin method (change scores) Subjective improvements in MBL: MIQ (i.e. QoL) Side effects/adverse drug effects: "adverse events monitoring" (conducted at each study visit); physical examination; electrocardiograph; vital signs; and laboratory evaluation
Notes	Funded by Ferring Pharmaceuticals (a drug company)
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Details of method of randomization not provided
Allocation concealment (selection bias)	Unclear risk	Not stated
Blinding (performance bias and detection bias)   All outcomes	Low risk	Participants, investigators and assessors blinded
Incomplete outcome data (attrition bias)   All outcomes	Low risk	Minimal drop‐outs. 294/304 randomized women included in safety analysis, 272/304 (90%) in efficacy analysis
Selective reporting (reporting bias)	Low risk	All pre‐specified outcomes reported
Other bias	Low risk	Groups appeared comparable at baseline

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Menstrual blood loss: mean loss	1	53	Mean Difference (IV, Fixed, 95% CI)	‐100.0 [‐141.82, ‐58.18]
2 Menstrual blood loss: improvement rates	1	53	Risk Ratio (M‐H, Fixed, 95% CI)	1.56 [0.95, 2.55]
3 Withdrawal from treatment because of adverse events	1	53	Risk Ratio (M‐H, Fixed, 95% CI)	0.78 [0.19, 3.15]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Median difference in PBAC score			Other data	No numeric data
1.1 After 6 months Rx			Other data	No numeric data
2 Percentage difference in PBAC scores			Other data	No numeric data
2.1 After 6 months Rx			Other data	No numeric data
3 Improvement in mean blood loss (PBAC score < 100)	1	42	Risk Ratio (M‐H, Fixed, 95% CI)	0.43 [0.24, 0.77]
4 Quality of life scores	1		Mean Difference (IV, Fixed, 95% CI)	Subtotals only
4.1 Physical domain	1	42	Mean Difference (IV, Fixed, 95% CI)	0.80 [‐0.78, 2.38]
4.2 Psychological domain	1	42	Mean Difference (IV, Fixed, 95% CI)	‐0.40 [‐1.85, 1.05]
4.3 Social domain	1	42	Mean Difference (IV, Fixed, 95% CI)	0.20 [‐1.55, 1.95]
4.4 Environmental domain	1	42	Mean Difference (IV, Fixed, 95% CI)	1.0 [‐0.06, 2.06]
4.5 Environmental domain TR	1	42	Mean Difference (IV, Fixed, 95% CI)	0.80 [‐0.29, 1.89]
5 Adverse events	1		Odds Ratio (M‐H, Fixed, 95% CI)	Subtotals only
5.1 Any adverse event	1	42	Odds Ratio (M‐H, Fixed, 95% CI)	0.83 [0.25, 2.80]
5.2 Headache	1	42	Odds Ratio (M‐H, Fixed, 95% CI)	1.67 [0.34, 8.10]
5.3 Bloating	1	42	Odds Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
5.4 Nausea	1	42	Odds Ratio (M‐H, Fixed, 95% CI)	7.36 [0.36, 151.91]
5.5 Weight gain	1	42	Odds Ratio (M‐H, Fixed, 95% CI)	2.86 [0.11, 74.31]
5.6 Rash	1	42	Odds Ratio (M‐H, Fixed, 95% CI)	2.86 [0.11, 74.31]
5.7 Spotting	1	42	Odds Ratio (M‐H, Fixed, 95% CI)	0.11 [0.01, 2.30]
5.8 Breast tenderness	1	42	Odds Ratio (M‐H, Fixed, 95% CI)	0.29 [0.01, 7.51]
5.9 Excess hair growth	1	42	Odds Ratio (M‐H, Fixed, 95% CI)	0.29 [0.01, 7.51]
5.10 Depression	1	42	Odds Ratio (M‐H, Fixed, 95% CI)	0.29 [0.01, 7.51]

Methods	RCT: computer‐generated randomization list to 1 of 3 groups  No blinding mentioned
Participants	Country: UK No: 76 Age: 35 to 46 years Inclusion criteria: women reporting HMB confirmed to have > 80 mL per cycle loss, normal cervical smear 3 to 12 months before commencing the study Exclusion criteria: organic causes of menorrhagia found at hysteroscopy or endometrial biopsy, previous renal or hepatic impairment, VTE, inflammatory bowel disease, peptic or intestinal ulceration, coagulation or fibrinolytic disorders
Interventions	Ethamsylate 500 mg 4 times daily for days 1 to 5 of cycle (27 participants) TXA 1 g 4 times daily for days 1 to 5 of cycle (26 participants) MFA 500 mg 3 times daily for days 1 to 5 of cycle (23 participants) Duration: 3 placebo cycles and 3 treatment cycles
Outcomes	MBL: objective measurement (alkaline haematin method), duration of blood loss (days), participant's estimate of blood loss, number of sanitary towels used (end scores and change scores) Dysmenorrhoea Side effects
Notes	Clarification of data sought from authors and reply received Funded by the Health Research Board of Ireland and Pharmacia (a drug company)
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomized "by a computer generated randomization list"
Allocation concealment (selection bias)	Unclear risk	Unclear
Blinding (performance bias and detection bias)   All outcomes	Unclear risk	No details; blinding not mentioned
Incomplete outcome data (attrition bias)   All outcomes	Low risk	5 withdrew after randomization (2 in MFA group, 2 in ethamsylate group, 1 in TXA group) and lost to follow‐up
Selective reporting (reporting bias)	Low risk	All pre‐determined outcomes were reported and were relevant
Other bias	Low risk	Groups appeared comparable at baseline

Methods	Cross‐over trial, with 3 control cycles, followed by 3 cycles of either intervention or placebo Thereafter, women were 'crossed over' to the other arm, for a further 3 cycles Both participants and carers were blind to treatment allocation
Participants	Country: UK No: 20 (16 were analyzed) Age: 33 to 48 years Inclusion criteria: HMB either as described by the participant or participants presenting with iron deficiency anaemia presumed to be due to HMB Exclusion criteria: significant clinical abnormality (from gynaecological examination) or significant histological abnormality (from dilatation and curettage)
Interventions	TXA 1 g 4 times daily on days 1 to 4 of cycle versus placebo 3 cycles of either TXA or placebo, followed by 3 cycles of the other
Outcomes	MBL by total body counter: 2 μg to 4 μg of Cu ⁵⁹Fe given intravenously and total body count measured at 2‐weekly intervals throughout the study Blood loss estimated from loss of radioactivity multiplied by the total blood volume (end scores and change scores) Duration of bleeding in days and number of pads used Side effects reported
Notes	Several women were able to detect the active treatment Side effects reported, but not systematically Source of funding was not stated
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of sequence generation not described
Allocation concealment (selection bias)	Low risk	Remote allocation: the tablets were either TXA 0.5 g per tablet or a placebo of similar size and appearance The order of treatments was randomized, and neither the participants nor those conducting the study knew the identity of tablets A and B
Blinding (performance bias and detection bias)   All outcomes	Low risk	Both participants and carers blind to treatment allocation
Incomplete outcome data (attrition bias)   All outcomes	High risk	4 withdrawals after allocation (all apparently in the placebo group (40%)): these were not included in the final analysis
Selective reporting (reporting bias)	Low risk	A priori outcomes reported and were relevant
Other bias	Low risk	Groups appeared comparable at baseline

Methods	RCT; randomization performed using computer‐generated list and sealed envelopes 2:2:1 randomization with greater numbers in the active groups Double dummy technique of administration of tables
Participants	Country: Sweden Age: > 18 years Number: 91 randomized, but results complete for only 68 women Inclusion criteria: > 80 mL per cycle blood loss, regular cycles, normal‐sized uterus on clinical examination Exclusion criteria: renal or hepatic impairment, clinical pelvic pathology or cervical intra‐epithelial neoplasia, concomitant disease or medication affecting menstruation, VTE, haematological or coagulation disorders, dilatation and curette within the previous 2 months, inability to comply with the protocol
Interventions	Kabi 2161 1200 mg twice daily for days 1 to 5 of cycle (26 participants) Kabi 2161 600 mg 4 times daily for days 1 to 5 of cycle (28 participants) Placebo 2 tablets 4 times daily for days 1 to 5 of cycle (14 participants) Duration: 3 cycles
Outcomes	MBL: objective (alkaline haematin method) as absolute measurement and relative change from baseline, duration of loss (days), number of sanitary towels used; and participant's subjective assessment (end scores and change scores) Side effects reported
Notes	Authors contacted for clarification of data and additional data received Source of funding not stated
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer generated list 2:2:1
Allocation concealment (selection bias)	Low risk	Sealed envelopes
Blinding (performance bias and detection bias)   All outcomes	Low risk	Double blind, with double‐dummy technique
Incomplete outcome data (attrition bias)   All outcomes	Low risk	4 exclusions post randomization; 19 lost to follow‐up
Selective reporting (reporting bias)	Low risk	All a priori outcomes were reported and were relevant
Other bias	Low risk	Groups appeared comparable at baseline

Methods	RCT: randomization with lottery method Only participants were blinded Duration of trial: 1 control cycle; 3 consecutive treatment cycles; followed up for 3 cycles post‐treatment
Participants	Country: India Number of participants: 52 (35 in intervention arm; 17 in control arm) Women aged 18 to 45 years old Inclusion criteria: 'history of menorrhagia' (no details given); Hb > 8 g%; presence of pelvic pathology (e.g. endometriosis, pelvic inflammatory disease, uterine fibroids, polyps/adenomyosis) acceptable Exclusion criteria: use of COCP, IUS, diabetes, hypertension, tuberculosis, severe anaemia (not defined), ischaemic heart disease, thyroid dysfunction, blood dyscrasias, and malignancy
Interventions	Intervention arm: oral Safoof Habis, a Unani formulation (a type of traditional medicine widely practised in South‐East Asia), which was made up of Teen Ahmer (silicate of alumina and iron oxide), Sange Jarahat (hydrated magnesium silicate) and Raal Sufaid (Vateria indica Linn, which is a species of plant in the Disterocarpaceae family, endemic to India) Safoof Habis was made up of equal parts of all 3 components, and given in 5 g doses BD, from days 1 to 5 versus Control arm: oral TXA 500 mg thrice daily, from days 1 to 5
Outcomes	MBL: menstrual pictogram (modified PBAC technique); duration of menstruation; subjective assessment of menstrual flow (end scores and change scores) Dysmenorrhoea: visual analogue scale Hb concentration Side effects/adverse drug effects: recorded with participant's diary; 'evaluated with biochemical tests'
Notes	Power calculations not performed The authors deny receiving any funding
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	"Lottery method"
Allocation concealment (selection bias)	Unclear risk	Unclear whether allocation was concealed; participants only were blinded
Blinding (performance bias and detection bias)   All outcomes	Unclear risk	Unclear; participants only blinded
Incomplete outcome data (attrition bias)   All outcomes	Unclear risk	13% drop‐out rate; reasons not provided
Selective reporting (reporting bias)	Low risk	All pre‐specified outcomes reported
Other bias	Low risk	Groups appeared comparable at baseline

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 to 45 years old Inclusion criteria: reported regular HMB; BMI 19 to 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 to 5 of menses MPA 5 mg twice daily, for days 5 to 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 (performance bias and detection bias)   All outcomes	High risk	Unable to blind participants, due to different regimens It is unclear whether or not personnel were blinded Potential knowledge of treatment may have influenced the primary outcome of MBL which was measured by PBAC
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 pre‐specified outcomes were reported
Other bias	Low risk	Groups appeared comparable at baseline

Methods	RCT with parallel group design Method of randomization: equal allocation ratio in blocks of 6 Double blinding with participants and nurses both blinded
Participants	Country: Iran Age: 20 to 49 years Number of participants: 94 (76 completed the trial) Inclusion criteria: self‐reported HMB, Hb ≥ 10.5 g/dL, BMI of 10 kg/m² to 29 kg/m² Exclusion criteria: history of VTE, chronic illnesses, other diseases known to interfere with menstrual bleeding (such as fibroids > 3 cm), history of iron supplementation, anticoagulant agents, COCP or other hormonal drug use, IUS in situ
Interventions	(1) TXA 500 mg 4 times daily from days 1 to 5 of cycle (2) Extract of Punica granatum Linn (from pomegranate) 4 times daily from days 1 to 5 of cycle Duration of trial: 3 consecutive menstrual cycles
Outcomes	PBAC (end scores) QoL scores (MIQ, SF36)
Notes	Funded by Tarbiat Modares University research office
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Equal allocation ratio, blocked randomization
Allocation concealment (selection bias)	Low risk	"Sequentially numbered and sealed opaque envelopes"
Blinding (performance bias and detection bias)   All outcomes	Low risk	Double blinding (participants and nurses blinded)
Incomplete outcome data (attrition bias)   All outcomes	Unclear risk	Significant drop‐out 76/94 (85%) randomized women included in analysis, similar drop‐out rates in each group
Selective reporting (reporting bias)	Low risk	All pre‐specified outcomes were reported
Other bias	Unclear risk	Baseline characteristics only reported for women who completed the study

Methods	Multi‐centre RCT Central computer‐generated randomization schedule used for allocation Described as open label trial
Participants	Country: Thailand (4 sites) Mean age: 34.6 years Number of participants: 169 women were randomized and 167 completed the trial Inclusion criteria: aged between 18 and 45 years, regular menstrual cycle (21 to 35 days), serum progesterone during 5 to 9 days before menstruation of ≥ 5.0 ng/mL, PBAC score > 130 during run‐in phase, no contraindication to treatment drugs, normal renal and liver function, normal pelvic examination Exclusion criteria: concomitant diseases, organic disease, VTE, haemorrhagic or fibrinolytic disorder, hormone therapy during last 3 months, taking any medication that might affect MBL, need or desire for contraception, need for iron supplementation, inability to comply and no consent
Interventions	(1) TXA 3 g daily on days 1 to 5 of cycle (2) MFA 1.5 mg daily on days 1 to 5 of cycle (3) NET 10 mg daily on days 19 to 26 of cycle, for 2 consecutive cycles
Outcomes	MBL using PBAC (end scores) Cure rate (success rate) (defined as PBAC ≤ 130) Adverse events QoL using a 'standardized questionnaire' Acceptability of treatment Hb Duration of menstruation
Notes	Unpublished copy of trial sighted, also conference abstract Source of funding not stated
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer‐generated randomization scheme
Allocation concealment (selection bias)	Low risk	Centralised randomization scheme separate from study investigators
Blinding (performance bias and detection bias)   All outcomes	High risk	Described as "open label" Potential knowledge of treatment may have influenced the primary outcome of MBL as this was measured by PBAC
Incomplete outcome data (attrition bias)   All outcomes	Low risk	Minimal drop‐outs
Selective reporting (reporting bias)	Low risk	All prespecified outcomes reported
Other bias	Low risk	Groups appeared comparable at baseline

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 pre‐treatment; 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 ultrasound 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 daily from day 1 for 4 days versus NET 5 mg thrice daily from day 14 to 23 of the cycle versus levonorgestrel 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) World Health Organization QoL‐Short Form (Turkish version), in which women report limitations in physical health, psychological status, social support, and "relating to their environment"
Notes	The authors deny receiving any funding
*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 (performance bias and detection bias)   All outcomes	High risk	''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
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)	Low risk	All prespecified outcomes reported
Other bias	Low risk	Groups appeared comparable at baseline and no other potential bias was identified

PERMALINK

Antifibrinolytics for heavy menstrual bleeding

Alison C Bryant‐Smith

Anne Lethaby

Cindy Farquhar

Martha Hickey

Abstract

Background

Objectives

Search methods

Selection criteria

Data collection and analysis

Main results

Authors' conclusions

Plain language summary

Summary of findings

Background

Description of the condition

Description of the intervention

How the intervention might work

Why it is important to do this review

Objectives

Methods

Criteria for considering studies for this review

Types of studies

Types of participants

Types of interventions

Types of outcome measures

Primary outcomes

Secondary outcomes

Search methods for identification of studies

Electronic searches

Searching other resources

Data collection and analysis

Selection of studies

1.

Data extraction and management

Assessment of risk of bias in included studies

Measures of treatment effect

Unit of analysis issues

Dealing with missing data

Assessment of heterogeneity

Assessment of reporting biases

Data synthesis

Subgroup analysis and investigation of heterogeneity

Sensitivity analysis

Overall quality of the body of evidence: 'Summary of findings' table

Results

Description of studies

Results of the search

Included studies

Study design and setting

Participants

Interventions

Dosage

Treatments

Outcomes

Objective assessment of women's menstrual blood loss

Subjective assessment of women's menstrual blood loss

Self‐reported improvement in HMB

Quality of life measures

Adverse events

Excluded studies

Risk of bias in included studies

2.

3.

Allocation

Blinding

Incomplete outcome data

Selective reporting

Other potential sources of bias

Effects of interventions

Summary of findings for the main comparison. Antifibrinolytics compared to no treatment or placebo for heavy menstrual bleeding.

Summary of findings 2. Antifibrinolytics compared to progestogens for heavy menstrual bleeding.

Summary of findings 3. Antifibrinolytics compared to NSAIDs for heavy menstrual bleeding.

Summary of findings 4. Antifibrinolytics compared to ethamsylate for heavy menstrual bleeding.

Summary of findings 5. Antifibrinolytics compared to herbal medicines for heavy menstrual bleeding.

Summary of findings 6. Antifibrinolytics compared to levonorgestrel intrauterine system.

1. Antifibrinolytics versus placebo or no treatment

Primary outcomes

Methods	Single‐centre, prospective RCT Randomization by "computerised randomization table" Blinding not possible, given the different dosage schedules of the 2 arms of the study
Participants	Country: India Age: 15 to 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, and unwilling to trial medical management
Interventions	TXA 500 mg 4 times daily from day 1 for 5 days versus MPA 10 mg twice daily from day 5 to 25 of the cycle Treatment for 3 cycles, then subjects 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
*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 (performance bias and detection bias)   All outcomes	High risk	No evidence of blinding; different dosage schedules make blinding impossible Potential knowledge of treatment may have influenced the primary outcome of MBL which was measured by PBAC
Incomplete outcome data (attrition bias)   All outcomes	High risk	Authors stated that there were minimal drop‐outs 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 pre‐specified outcomes were reported
Other bias	Low risk	Groups appeared comparable at baseline

Methods	Multi‐centre (40 clinical sites), double‐blind, parallel‐group study Randomization via permuted block randomization scheme, with a block size of 8 (5 allocated to intervention arm, and 3 to placebo arm) Blinding: participants, investigators, sponsor, statisticians, clinical data management staff, and clinical monitors were all blinded Duration of trial: 2 control cycles pre‐treatment; 6 cycles of treatment
Participants	Country: USA Age: 18 to 49 years old Number of participants: 196 women randomized Inclusion criteria: history of at least 3 days of HMB over at least 4 of their last 6 cycles; confirmed during 2 cycles before treatment phase commenced; normal pelvic examination and Pap smear; no "clinically important" findings on transvaginal ultrasound; willingness to use non‐hormonal contraception during the trial Exclusion criteria: history of significant medical problem; severe anaemia (Hb < 8 g/dL); pregnant/lactating; endometrial abnormalities; cervical carcinoma; anovulatory dysfunctional uterine bleeding; glaucoma; ocular hypertension; use of anticoagulants; aspirin; dong quai; aminocaproic acid; hydroxychloroquine Note: uterine fibroids were only an exclusion criteria if thought to require surgical management
Interventions	TXA 1.3 g 3 times daily (123 participants), commenced with the onset of HMB, for up to 5 days versus placebo (73 participants) 3 times daily, commenced with the onset of HMB, for up to 5 days
Outcomes	Objective measurement of MBL: alkaline haematin method (change scores) Subjective improvements in MBL: MIQ; occurrence of large blood stains Hb and ferritin concentrations Side effects
Notes	11 of 12 authors report receiving funding from 1 (or more) drug company/companies
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Permuted block randomization schedule
Allocation concealment (selection bias)	Low risk	Allocation managed by an independent group ("Fisher Clinical Services")
Blinding (performance bias and detection bias)   All outcomes	Low risk	Participants, investigators and assessors were all blinded
Incomplete outcome data (attrition bias)   All outcomes	Low risk	28% participants (55/196) withdrew, but proportions were similar in each study arm, and 95% were included in a modified intention‐to‐treat analysis
Selective reporting (reporting bias)	Low risk	All pre‐specified outcomes reported
Other bias	Unclear risk	Imbalances between groups in baseline MBL

Methods	Double‐blind, placebo‐controlled RCT Randomization by computer‐generated numbers, which were placed in sealed envelopes 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 to 4 and days 19 to 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 menstrual loss over 2 cycles > 80 mL per cycle; no hormone therapy within 3 months; no medication which may affect MBL; confirmed to be ovulating; and 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 cycles
Interventions	2 months of placebo (to assess eligibility) followed by 2 months of either TXA (25 participants) 1 g taken 4 times daily on days 1 to 4 or NET (21 participants) 5 mg taken twice daily on days 19 to 26 of cycle
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	Originally 103 women were recruited but 57 were excluded during the placebo cycles because of lack of objective menorrhagia (> 80 mL per cycle), anovulation or lack of compliance with protocol Additional data were provided by the author Funded by Pharmacia (a drug company)
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	"Randomization was by computer generated numbers"
Allocation concealment (selection bias)	Low risk	Adequate: "opaque sealed, consecutively numbered envelopes"
Blinding (performance bias and detection bias)   All outcomes	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
Incomplete outcome data (attrition bias)   All outcomes	Low risk	4 women did not complete the trial, but all were included in the results
Selective reporting (reporting bias)	Low risk	All pre‐specified outcomes were reported
Other bias	Low risk	Groups appeared comparable at baseline

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 thrombo‐embolus
Interventions	TXA 1 g 3 times daily during days 1 to 5 (69 participants) versus NET 5 mg twice daily on days 19 to 26 (59 participants) Administered for 2 consecutive cycles
Outcomes	MBL (PBAC) Length of menstrual period 6‐item QoL questionnaire collected in the second week before, during and after each treatment cycle and a third (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 (performance bias and detection bias)   All outcomes	High risk	Treatment and control groups were given their pills on different days of the menstrual cycle
Incomplete outcome data (attrition bias)   All outcomes	Low risk	It appears that there is minimal drop‐out
Selective reporting (reporting bias)	High risk	The 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	Groups of participants appeared similar at baseline

Study	Reason for exclusion
Andersch 1988	Cross‐over trial which did not report data from the first phase of the study.
Chamberlain 1991	Treatment group received ethamsylate.
Drosdal 1993	Participants included women with IUD in situ.
Harrison 1976	Treatment group received ethamsylate.
Kasonde 1975	Participants had intrauterine coil devices in situ.
Kouides 2009	Participants had a proven coagulopathy.
Moghtadaei 2012	Author was contacted for further information about their study, to judge whether it should be included/excluded. No reply was received.
Muse 2010	There was no control group.
Najam 2010	Compared women taking tranexamic acid to those taking a combination of tranexamic acid and mefenamic acid.
NCT01190150	Dose‐finding RCT of tranexamic acid in girls aged 12 to 16 years.
NCT01428713	Data not provided for the end of the first phase of the cross‐over study.
NiIsson 1965	Treatment group received ethamsylate.
Nilsson 1967	Cross‐over trial which did not report data from the first phase of the study.
Petersen 1983	Treatment was given for only 1 month. Criterion specified in our protocol was that we would include only studies using at least 2 months of treatment.
Tabatabaei 2013	Author was contacted for further information about their study, to judge whether it should be included/excluded. No reply was received.
Vermylen 1968	Cross‐over trial which did not report data from the first phase of the study.
Westrom 1970	Participants had intrauterine coil devices in situ.
Ylikorkala 1983	Participants had intrauterine coil devices in situ.