Types of studies

We only considered randomised controlled trials (RCTs) for inclusion. We excluded quasi‐randomised studies.

Types of participants

1. Premenopausal women with uterine fibroids, diagnosed by clinical manifestation and physical signs, and confirmed by ultrasound scanning, computed tomography (CT), or MRI, or a combination of more than one of the procedures.

2. Premenopausal women without any symptoms, but determined to have uterine fibroids during routine gynaecological examination and confirmed by imaging techniques.

3. We included premenopausal women with fibroid‐related symptoms and palpable uterine fibroids without confirmation by imaging technology, and compared them in sensitivity analyses.

We used the diagnostic criteria produced by the International Federation of Gynecology and Obstetrics (FIGO 2011).

This review did not include premenopausal women with uterine fibroids who planned to undergo surgery for uterine fibroids, either hysterectomy (abdominal, vaginal, or laparoscopic), myomectomy (laparotomy or laparoscopy), or resection for uterine fibroids.

Types of interventions

Experimental interventions included oral progestogens, depot medroxyprogesterone acetate (DMPA) intramuscular injections, or progestin‐releasing intrauterine devices (IUS). The control interventions included no treatment, placebo, medical therapy, or surgical procedures.

Types of outcome measures

Electronic searches

For the 2020 update of this review, we searched:

• The Cochrane Gynaecology and Fertility Group (CGF) Specialised Register of Controlled Trials; ProCite platform (searched 02 July 2020; Appendix 1);

• CENTRAL, via the Central Register of Studies Online; Web platform (searched 02 July 2020; Appendix 2). CENTRAL now contains records from CINAHL and the trial registries clinicaltrials.gov and the World Health Organisation International Trials Registry Platform search portal;

• MEDLINE Epub Ahead of Print, In‐Process & Other Non‐Indexed Citations; Ovid platform (searched from 1946 to 02 July 2020; Appendix 3);

• Embase; Ovid platform (searched from 1980 to 02 July 2020; Appendix 4);

• PsycINFO; Ovid platform (searched from 1806 to 02 July 2020; Appendix 5).

We combined the MEDLINE search with the Cochrane highly sensitive search strategy for identifying randomised trials, which appears in the Cochrane Handbook of Systematic Reviews of Interventions (Chapter 6, 6.4.11; Lefebvre 2011). We combined the Embase search with trial filters adapted from the Scottish Intercollegiate Guidelines Network (SIGN).

Searching other resources

We searched LILACS Web (Latin American and Caribbean Health Sciences Literature) and Google Scholar (searched 02 July 2020). We searched reference lists of included trials, and relevant review articles, and contacted experts in the field to obtain additional data. In liaison with the Information Specialist, we also handsearched relevant journals and conference abstracts that are not covered in the CGF register.

Selection of studies

For the 2020 update, Marian Showell (CGF Information Specialist) conducted the initial search. Two review authors (US, PL) independently conducted an initial screen of titles and abstracts using Covidence; we retrieved the full texts of all potentially eligible trials. Two review authors (US, PL) independently examined these full‐text articles for compliance with the inclusion criteria, and selected eligible trials. We resolved disagreements by discussion. If any reports required translation, we described the process used for data collection. We documented the selection process with a PRISMA flow chart (Figure 1).

Figure 1

---

---

Study flow diagram

Data extraction and management

For the 2020 update, two review authors (US, PL) independently extracted and assessed the quality of the data, using forms designed and pilot‐tested by the authors. We resolved discrepancies by discussion. We analysed included trials for criteria and methodological details as shown in Appendix 6. Where necessary, we sought additional information on trial methodology or actual trial data from the principal author of any trial that appeared to meet the eligibility criteria. In cases where trials presented results in graphs, and did not give any actual data, we extracted the data from the graphs. Where studies had multiple publications, we collated multiple reports under a single study ID with multiple references.

Assessment of risk of bias in included studies

Two review authors (US, PP) 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. We assigned judgements recommended in the Handbook (Section 8.5, Higgins 2011). We resolved any disagreement by discussion. We described all judgements fully, and presented the conclusions in the 'Risk of bias' table (Characteristics of included studies) and summary figures (Figure 2; Figure 3).

Figure 2

---

---

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

Figure 3

---

---

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

We used these definitions to assess risk of bias.

Measures of treatment effect

For continuous data (e.g. blood loss, uterine or fibroid volume), if all studies reported the same outcomes, we calculated the mean difference (MD) between treatment groups. If studies reported similar outcomes, on different scales, we calculated the standardised mean difference (SMD). We presented 95% confidence intervals (CI) for all outcomes.

For dichotomous data (e.g. uterine fibroid symptoms), we used the numbers of events in the control and intervention groups of each study to calculate risk ratios (RRs) with 95% CI.

Unit of analysis issues

Our review only included outcome data from RCTs that randomised individuals. The unit of analysis was women randomised, and the outcomes for each woman were collected and analysed. For a study with multiple intervention groups, we presented a separate comparison of each active intervention versus the one comparison group. The number of participants in the comparison was split between the intervention arms. Therefore, there was no issue about double counting the data from the comparison group in a single analysis.

Dealing with missing data

We contacted the study authors by email to request missing data. We analysed data on the basis of an intention‐to‐treat analysis. We imputed values for primary outcomes only. If these data were unobtainable, and imputation was not feasible, we analysed the available data.

Assessment of heterogeneity

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

Assessment of reporting biases

We ensured that our search was comprehensive, without any language restriction, and we were alert for duplication of data, to minimise the potential difficulty of detecting and correcting for publication bias and other reporting biases. In this review, we could not assess publication bias using the funnel plot because there were not enough studies.

In future updates, if the number of included trials is more than 10, we will assess potential publication bias using 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; (Sterne 2017)).

Data synthesis

We identified only one study that contributed data to each analysis, so we were unable to perform meta‐analysis. For future updates, if we include more studies, we will pool data in a meta‐analysis using a fixed‐effect model, provided data are homogeneous (I² < 50%, consistency of forest plot, and P value of Chi² > 0.10) for:

• dichotomous outcomes (e.g. uterine fibroids symptoms) – relative risks (RR) with 95% CIs;
• continuous outcomes (e.g. blood measures) – mean differences between the treatment arms at the end of follow‐up if all trials measured the outcome on the same scale; standardised mean differences if similar but different scales are used to measure the same outcome; both with 95% CI.

We will pool statistically heterogeneous results using a random‐effects model if trials are sufficiently similar. We will present pooled effect estimates with 95% CIs, and estimates of T² and I².

We used Review Manager 5 to conduct the statistical analyses and generate forest plots (Review Manager 2014).

Subgroup analysis and investigation of heterogeneity

We did not perform any subgroup analyses. If more data are available in future updates, we will perform the following subgroup analyses:

1. location of uterine fibroids (subserous, intramural, or submucous).

2. dose (low, medium, high, based on data).

Sensitivity analysis

We did not perform any sensitivity analyses. If more data are available in future updates, we repeat the analyses to explore the influence of the following factors on effect size:

1. study quality, such as allocation concealment, blinding, and numbers lost to follow‐up;

2. any very long duration or large studies, to establish how much they dominated the results;

3. use of different rating scales to assess symptom relief.

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

We generated a ’Summary of findings’ table using GRADEpro GDT software and Cochrane methods (GRADEpro GDT; Higgins 2011a). This table evaluates the overall quality of the body of evidence for the main review outcomes (blood loss, uterine fibroid volume, adverse events) for the main review comparison (progestogens or progestogen‐releasing intrauterine systems versus control interventions included no treatment, placebo, medical therapy, or surgical procedures). We prepared additional 'Summary of findings' tables for the main review outcomes for other important comparisons (levonorgestrel‐releasing intrauterine system (LNG‐IUS) versus low dose combined oral contraceptive, LNG‐IUS versus oral progestogen, and oral progestogen versus goserelin acetate). Two review authors (US, PP) independently judged the overall quality of the evidence for each main outcome, using GRADE criteria; risk of bias, consistency of effect, imprecision, indirectness, and publication bias. We resolved disagreements by discussion. We justified, documented, and incorporated our judgements about the quality of evidence (high, moderate, low, or very low) into the report of results for each outcome.