Treatment of Irregular Uterine Bleeding Caused by Progestin-Only Contraceptives: A Systematic Review

Abstract

Progestins have shown to be highly effective, adequately safe contraceptives and a real alternative in the cases with contraindications to estrogen use. This review assesses the effectiveness of treatments for managing uterine bleeding due to the use of progestin-only contraceptives. A systematic review of databases such as Embase, LILACS, CINAHL, and PubMed, with a result of 701 studies published in English between 2005 and 2022, was carried out. 21 randomized clinical trials were analyzed. There were effective non-hormonal alternatives for managing progestin-related bleeding. While several treatments show promise, results must be replicated by studies with longer follow-up periods. Differences came from the five new studies and their good methodological quality—specifically, blinding and assessment of population losses. Benefits were the stop of uterine bleeding and a shortened duration of these episodes, with studies generally reporting satisfaction among their patients. There were few adverse effects seen, with the most common being headaches and nausea, which should be addressed by future studies. These results are expected to be of use for health professionals and midwives working on contraceptive management.

1. Introduction

The use of contraceptives brought great changes to women’s health, with a myriad of benefits beyond birth control [1,2]; however, some hormonal components can produce unintended effects, which could compromise the use and continuity of the treatment and leave many women unsatisfied or even without protection [3,4].

Progestins have shown to be highly effective, adequately safe contraceptives and a real alternative in the cases with contraindications to estrogen use. Moreover, progestins can be administered in several ways. Progestin-only contraceptives, such as progestin-only pills (POP), medroxyprogesterone acetate (DMPA) in any dosage or route of administration, levonorgestrel (LNG) and etonogestrel (ENG) subdermal implants, and levonorgestrel intrauterine devices (LNG-IUD) [5], are valued for generating less adverse effects than other contraceptives. However, unintended effects have been reported, such as heavier uterine bleeding, which seems to be significant enough for many women to abandon its use. Among patients who discontinued their use of long-acting reversible contraception methods (LARC), the most common reason was cramps for intrauterine devices (IUD), and irregular or frequent bleeding for implants [6]. Other adverse effects include acne, hirsutism, headaches, weight gain, and spotting or light bleeding—the latter being very prevalent [7,8,9]. Most oral contraceptives ameliorate acne and hirsutism; however, using progestin-only contraceptives –especially those of first or second generation– can trigger or worsen acne, due to their androgenic effect [10]; thus, the interest on addressing bleeding patterns and taking them out of the side effects.

This review aims to evaluate the effectiveness of treatments for uterine bleeding caused by progestin-only contraceptives.

2. Materials and Methods

2.1. Design

We conducted a systematic review based on the recommendations written in The Cochrane Handbook for Systematic Reviews of Interventions [11] and PRISMA [12,13]. Protocol was archived in PROSPERO (CDR 42021235946).

2.2. Searching Process

A professional librarian helped searching in four electronic databases (PubMed, CINAHL, LILACS, Embase), covering the period between 2005 and 2022. Studies prior to 2005 were not included due to evidence of follow-up losses in most of them. Filters were complete text, randomized clinical trials, controlled studies, and crossover. The Boolean operators were “OR” and “AND”, and the language used in the search equation was English. Terms were Contraceptive agents, Female, Medroxyprogesterone Acetate, Levonorgestrel, Contraception, Long-acting reversible contraception, Bleeding, Anti-Inflammatory Agents, Non-Steroidal, Adverse effects, Progestin, and Progestin-only contraceptives. Search terms were adjusted for each database.

Figure 1 shows the searching process in further detail.

2.3. Study Eligibility

Articles had to present data on treatments for bleeding due to the use of progestin as contraceptives in women: their effectiveness, adverse effects, and satisfaction.

Inclusion criteria were: women of fertile age, progestin-only contraceptive users, published articles or available summary, randomized clinical trials, and cross-over studies. Exclusion criteria were: studies focused on women with dysfunctional uterine bleeding -according to the International Federation of Gynecology and Obstetrics’ PALM-COEIN Classification of Abnormal Uterine Bleeding-, and/or with cardiovascular pathologies [14].

Study selection and data extraction were done independently by two researchers. Whenever disagreements arose, the whole reviewing team discussed them, and a third researcher intervened when needed. COVIDENCE^® was used for study selection, and a questionnaire already tested on two studies was used for data extraction.

First, we selected articles based on their titles/abstracts and then did a second selection in duplicate. Discrepancies were resolved through discussion and consensus. Works to be included in the global review were determined after reviewing their full text in duplicate.

2.4. Data Extraction

We summarized the data with standardized questionnaires that were specifically created for this project, collecting information on quote, place, target population, study design, sample size, key quantitative/qualitative results, and rigor of the study (using the risk of bias tool from the evidence project). Results were presented as a narrative in order to highlight the main findings on the effectiveness and adverse effects of the treatments and on patients’ satisfaction.

2.5. Methodological Quality Assessment

All RCTs were at low risk of bias of random sequence generation (100%, 21/21). Besides, most of the studies were at low risk of bias of selective outcome reporting (95%, 20/21), allocation concealment (90%, 19/21), blinding of participants and personnel (76%, 16/21), and other bias (81%, 17/21). On the other hand, 48% (10/21) were at high risk of incomplete outcome data, while the bias of blinding of outcome assessment had a considerable, unclear risk (33%, 7/21). Generally, most studies had an acceptable validity (low risk of bias).

Figure 2 and Figure 3 show details of risk of bias of published studies.

3. Results

701 studies were identified; after the selection process was done, 21 randomized clinical trials were included, done in 9 different countries (Thailand, Australia, Egypt, United States of America, Brazil, Malaysia, Denmark, Ireland, and Norway)—10 of them were done in USA, 2 in Australia, 4 in Thailand, 2 in Egypt, 1 in Brazil, 1 in Malaysia, and one was a multicenter study covering Norway, Denmark, and Ireland. Studies were done in hospitals, clinics, and university health centers, between 2005 [15] and 2022 [16,17,18]. 5 studies included users of LNG-IUD as their method of birth control [17,19,20,21,22], 5 studies had users of injectable DMPA [23,24,25,26,27], and 11 trials included users of implants [5,15,16,18,28,29,30,31,32,33,34].

There were several treatments featured in the different trials, but most of them were orally administered. Some treatments used were tamoxifen (oral route) [15,16,22,28], doxycycline (oral) [23,29,33], estrogen (oral) [34], estradiol acetate (vaginal) [26], estradiol (patch) [20], ibuprofen (oral) [34], celecoxib (oral) [24], ulipristal acetate (oral) [5,17], levonorgestrel + ethinylestradiol (oral) [30,31], naproxen (oral) [20], valdecoxib (oral) [25], tranexamic acid (oral) [19,27,32], mefenamic acid (oral) [18], ulipristal acetato (CDB-2914) (oral) [21], and mifepristone (oral) [29,33]. Study duration varied considerably; some trials showed results after 10 days [15], while others did so after 180 days. Most studies seemed to cover just one treatment; however, some of them, like the ones by Weisberg et al. [29,33] included more than one. Also, most used placebos for the comparison—except for Upawi et al. [18], and Warner et al. [21].

Table 1. Results of the included studies listed in alphabetical order.

Article	Treatment Description/Plan	Primary Outcome(s)	Secondary Outcome(s) Satisfaction and Adverse Effects
Abdel Aleem, 2005	Two daily oral doses of 10 mg tamoxifen during 10 days -- Two daily oral doses of placebo during 10 days	At the end of the treatment, a high percentage of tamoxifen users reported a stop in the bleeding as compared to the control group (88% and 68%, respectively; p = 0.016). First month: 85.7% vs. 75.5%; second month: 34.7% vs. 23.9%. Difference is statistically significant (p < 0.05). There were no differences between tamoxifen and placebo users in the third month. 30 days after starting the therapy, days with bleeding were 5.0, 4.3, and 3.6 for patients receiving ethinylestradiol, ibuprofen, and placebo, respectively. These differences were not statistically significant. Mean number of days with no bleeding during this period was 1.1, 1.8, and 2.8 for ethinylestradiol, ibuprofen, and placebo, respectively. There was a statistically significant difference between ethinylestradiol and placebo in spotting reduction (p = 0.04). Mean number of days with no bleeding in the first 30 days after ending the treatment was 6.1, 6.2, and 6.4 for ethinylestradiol, ibuprofen, and placebo, respectively, and these results presented no statistically significant differences.	Satisfaction: The beneficial effect of tamoxifen was seen in women’s improved lifestyles and their general satisfaction with the treatment. One and two months after the treatment ended, a significantly higher percentage of tamoxifen users were satisfied with the treatment than in the control group (85.7 and 75.5% versus 34.7% and 23.9%, respectively; p < 0.005). The first group also stated a high performance in their household chores, religious duties, and sex life as compared to the latter group. There were no differences in the third month. Adverse effects: Headache, gastrointestinal disorders, dizziness, easy fatigue, hot flashes.
Abdel Aleem, 2012	An oral dose of 100 mg doxycycline every twelve hours, during 5 days -- Placebo every 12 h, during 5 days	The number of bleeding and spotting days was 14.5 ± 5.1 in the doxycycline group and 13.6 ± 4.6 in the placebo group. This demonstrates that there was no significant difference between the two study groups regarding the primary outcome (women who stopped bleeding within 10 days of starting treatment; 22/34 vs. 25/34 0.88 (0.64, 1.21)) Doxycycline treatment caused no statistically significant differences in the number of bleeding or spotting days in the 3 months following treatment as compared to the placebo group. In the third month following treatment, the mean number of bleeding days was 7.2 ± 4.4 in the doxycycline group, and 7.3 ± 4.6 in the placebo group. The mean number of spotting days was 3.7 ± 2.8 and 3.6 ± 2.7 in the two groups, respectively. The relative risks [RR] of having a bleeding-free interval equal to or less than 15 days in the first 3 months following treatment are summarized. There was no statistically significant difference between the treatment and control groups. In the third month following treatment, RR was 0.90 and 95% CI was 0.64–1.28.	Satisfaction: Not reported. Adverse effects: Two women in the doxycycline group complained about slight nausea during treatment, while only one in the control group did so. Another patient in the doxycycline group complained about diarrhoea.
Archer, 2008	Daily 20 mcg EE (oral tablet) during 10 days -- Two daily doses of placebo during 5 days — Two daily oral doses of 400 mg ibuprofen during 5 days -- A daily dose of placebo during 10 days	The primary outcome measure was the number of days of vaginal (endometrial) spotting/bleeding (S/B). The mean number of S/B days during Days 1–5 of treatment was 1.1, 1.8 and 1.2 for ethinylestradiol, ibuprofen, and placebo, respectively. These results were not statistically different among groups. From day 6 to day 10, there were no differences found in mean S/B days between treatments: 1.4, 0.9 and 1.0 for ethinylestradiol, ibuprofen, and placebo, respectively (see Table 2). There were no differences in mean S/B days within each treatment (ethinylestradiol, ibuprofen, and placebo) when days 1–5 were compared to days 6–10. The mean number of S/B days was not different among treatments during days 1 to 15 after initiation of therapy. The mean number of S/B days was not different among treatments during days 1 to 15 after initiation of therapy.	Satisfaction: Not reported. Adverse effects: No effect on cervical mucus or sperm was found between treatments.
Buasang, 2009	Daily oral dose of 200 mg celecoxib during 5 days -- Daily oral dose of placebo during 5 days	The percentage of subjects whom bleeding was stopped within 7 days after initial treatment was significantly higher in the celecoxib group than in the placebo group (70% vs. 0%; p < 0.001). The mean duration of bleeding-free interval was significantly longer in celecoxib than placebo group (24.0 + 1.65 days vs. 10.0 + 6.50 days; p < 0.001). The mean duration of bleeding days was significantly shorter in celecoxib than placebo group (5.0 + 1.65 vs. 19.0 + 6.50 days; p < 0.001). The mean duration of bleeding-free interval was significantly longer in celecoxib than placebo group (24.0 + 1.65 days vs. 10.0 + 6.50 days; p < 0.001).	Satisfaction: Patients’ satisfaction in the celecoxib group was significantly higher than in the placebo group (80% vs. 30%; p < 0.001). Adverse effects: There was no detectable adverse effect in any group.
Cohen, 2019	Two daily doses of 10 mg tamoxifen during 7 days -- Two daily doses of placebo during 7 days	40% reduction in B/S days over 30 days Compared to placebo users, participants who received tamoxifen had slightly decreased mean B/S days after initiating study drug (16.8 ± 9.0 days vs. 12 ± 5.8 days, p = 0.08). This equated to mean 4.8 fewer B/S days (95% confidence interval—10.1 to 0.5 days, p = 0.08).	Satisfaction: The study groups reported similar satisfaction with their bleeding patterns. Adverse effects: There were no severe events reported.
Dempsey, 2010	24.8 mg estradiol acetate (vaginal ring) -- No placebo	For each additional day of bleeding and/or spotting reported, women were 3% less likely to receive a second injection (OR 0.97, 95% CI 0.94–0.99). Those who received a second injection of DMPA had, on average, 15 fewer bleeding days during the first follow-up period than those who discontinued (p = 0.02).	Satisfaction: The majority reported being somewhat or very satisfied with the ring. Adverse effects: Reported side effects, including breast tenderness, nausea, mood swings, headaches, vaginal discharge, weight change, and acne, did not differ between randomization groups. The frequency of these reported side effects also did not predict continuation.
Fava, 2020 (rev)	5 mg ulipristal acetate (pills) during 5 days -- 5 mg placebo during 5 days	The number of days before bleeding stopped and the days free of bleeding were statistically similar between the two groups (p > 0.05), although there was a trend toward fewer days before bleeding stopped (3.3 ± 4.5 days vs. 6.6 ± 8.5 days; p = 0.399) and more days free of bleeding (19.6 ± 14.0 days vs. 11.7 ± 9.5 days; p = 0.177) among users of ulipristal acetate than in users of placebo. Furthermore, 30, 60, and 90 days after initiation of treatment, the mean number of bleeding days was not statistically different between groups (p > 0.05), although there was a trend toward fewer bleeding days among the ulipristal acetate group compared with the placebo group. Endometrial thickness at baseline compared with at 90 days was similar among both treatment groups (4.7 ± 1.2 days vs. 6.2 ± 4.5 days; p = 0.618). One IUS was removed from a participant in the ulipristal acetate group.	Satisfaction: Not reported. Adverse effects: In general, the main adverse effect reported during the treatment period was headaches. No other adverse events were reported.
Guiahi, 2015	150 mcg levonorgestrel -- 30 mcg ethinylestradiol -- Two daily doses of placebo during 14 days	Participants randomized to receive OCPs were 11.7 times more likely (95% CI 1.9–70.2) to have a temporary interruption of bleeding during the study drug period: 14 (87.5% ± 16.2%) in the OCP group compared with 6 (37.5% ± 23.7%) in the placebo group (p <0.01). The median number of days to the start of the temporarily interrupted episode (if one occurred) was 5.0 for the intervention group compared with 9.0 for the placebo group (p = 0.05). Regardless of whether a temporary interruption of bleeding was achieved, the median number of days without bleeding during the 14-day therapy period was higher in the treatment group than in the placebo group (9.0 vs. 3.0, p = 0.03).	Satisfaction: Not reported. Adverse effects: Not reported.
Madden, 2012	Naproxeno 500 mg twice daily x 5 days x 4 week Two daily doses of 500 mg naproxen for the first 5 days of each 4 week period -- Post-insertion patch of 0.1 mg estradiol every week -- Oral placebo	The naproxen group was more likely to be in the lowest quartile of bleeding and spotting days compared with placebo (naproxen: 42.9%, placebo: 16.3%; p = 0.03). The median number of bleeding and spotting days was 27.5 (range: 5–83 days) in the naproxen group, 44 (range: 2–82 days) in the estradiol group, and 32 (range: 9–84 days) in the placebo group. With the use of nonparametric testing, the distribution of bleeding and spotting days was not significantly different in the naproxen group compared with the placebo group (p = 0.15) or in the estradiol group compared with the placebo group (p = 0.10). In the multivariable analysis, the naproxen group had a 10% reduction in bleeding and spotting days (adjusted relative risk: 0.90; 95% CI: 0.84–0.97) compared with placebo. More frequent bleeding and spotting was observed in the estradiol group (adjusted relative risk: 1.25; 95% CI: 1.17–1.34). The administration of naproxen resulted in a reduction in bleeding and spotting days compared with placebo.	Satisfaction: Women in the estradiol group were more likely to be dissatisfied with their bleeding pattern at 4 weeks; 39.5% of the women in the estradiol group were dissatisfied with their bleeding, compared with 9.5% and 11.6% of women in the naproxen and placebo groups, respectively (p = 0.01). Satisfaction with bleeding patterns improved over time in all groups. At 12 weeks, 85% of women reported being “somewhat” or “very” satisfied with their bleeding pattern, and satisfaction did not differ by treatment arm (p = 0.12). Satisfaction with the LNG-IUS was 94% at 12 weeks in all 3 study groups. There were few serious adverse events reported during the treatment period. One participant in the naproxen arm had her LNG-IUS removed because of complaints of chest pain. Adverse effects: Not reported.
Madden, 2012	3 daily valdecoxib tablets (20 mg) during 5 days -- 2 daily placebo tablets during 5 days	The mean duration of bleeding-free days in a period of 28 days was 17.8 ± 8.2 in the valdecoxib group and 11.5 ± 10.4 days in the placebo group. There were significant differences between both groups (p < 0.05) 77.3% treated with valdecoxib stopped bleeding, whereas 33.3% in the placebo group stopped bleeding within 7 days after the initial treatment. The mean value of the number of treatment days required for stopping the bleeding episode and the mean value for the bleeding-free interval (within 28 days of the follow-up period) in the valdecoxib-treated group were 1.7 ± 0.9 and 18.6 ± 6.0 days, respectively (p > 0.05).	Satisfaction: Not reported. Adverse effects: Not reported.
Phupong, 2006	4 daily doses of 2 250 mg tranexamic acid capsules during 5 days -- Same doses of placebo	During the follow-up period (4 weeks after the initiation of the treatment), 58.8% in the tranexamic acid group and 76.5% in the placebo group stopped bleeding (p = 0.12). The percentage of women whose bleeding stopped within 7 days after the initiation of the treatment was significantly higher after tranexamic acid treatment than placebo (64.7% vs. 35.3%, p = 0.015). The mean duration of bleeding and spotting days was not significantly different between the two groups during the 28 days of the follow-up period (mean: 15.4 and 12.7 days, respectively; p = 0.182).	Satisfaction: Not reported Adverse effects: There were no detectable adverse effects in any group.
Sordal, 2013	Three daily oral doses of 500 mg tranexamic acid while bleeding or spotting -- Three daily oral doses of 500 mg mefenamic acid while bleeding or spotting -- Placebo (lactose and magnesium stearate)	The mean value of bleeding or spotting days was lower in the tranexamic acid group than in the mefenamic acid and placebo groups. The mean reduction in bleeding and/or spotting days was 6 days (95% CI: −14.0–1.0; p = 0.049) for the tranexamic acid treatment, and 3 days (95% CI: −11.0–5.0; p = 0.229) for the mefenamic acid treatment.	Satisfaction: There were no significant differences in satisfaction with the use of IUD. In all groups, more than 85% of the women declared to be satisfied. This shows that satisfaction with the studied drug does not influence general satisfaction with the levonorgestrel intrauterine system. Continuation rates after 16 weeks were 96.7%, 96.6%, and 98.3% for the tranexamic acid, mefenamic acid, and placebo groups, respectively. Adverse effects: Gastrointestinal adverse effects were more common in the tranexamic acid group, while headaches were less frequent in the mefenamic acid gorup. Two women from the tranexamic acid group suffered a severe adverse event. One of them had an asymptomatic case of intra-abdominal migration of a copper IUD, which was apparently spontaneously expelled, while the other suffered from endometritis and had to be hospitalized. The endometritis was completely cured with antibiotics and removal of the levonorgestrel intrauterine system.
Sethong, 2009	4 daily doses of 250 mg tranexamic acid during 5 days -- 4 daily doses of 250 mg placebo during 5 days	Of the subjects, 88% treated with tranexamic acid stopped bleeding, whereas 8.2% in the placebo group stopped bleeding within seven days after initiation of the treatment, thus a significant difference (p < 0.05). The mean bleeding-free interval during 28 days for the tranexamic acid and placebo were 20.6 days and 7.5 days, respectively. This was also a significant difference between the two groups (p < 0.05).	Satisfaction: Not reported. Adverse effects: Not reported.
Warner, 2010	Daily oral dose of 50 mg progesterone receptor modulator CDB-2914 for three consecutive days, in separate treatments starting 21, 49, and 77 days after LNG-IUS insertion. -- No placebo	After the first treatment, there was a statistically significant difference in bleeding/spotting that favoured CDB-2914, for an amount of 3 days (difference: 210.6%, p = 0.011), but in the 64 days following the third treatment, women from the CDB-2914 group reported more days with bleeding/spotting, at 6 days (±9.5%, p = 0.022). In this last period, the difference in the highest amount of bleeding/spotting-free days was not statistically significant (24 days for CDB-2914 versus 26 days for placebo). During the treatment and follow-up stages (from the 21st to the 166th day), both groups showed a reduction in bleeding/spotting, with a highly significant statistical difference (p = 0.0001).	Satisfaction: Patients’ dissatisfaction with treatment. Vaginal bleeding acceptability was favorable and similar between the two groups, with 92% of each group reporting less bleeding than before insertion (57/60 in the CDB-2914 group, and 48/52 in the placebo group). Some women found that bleeding patterns were “more inconvenient” than before, with a higher percentage of them being in the CDB-2914 group [39% (22/56) versus 19% (9/48); p = 0.036]. Both groups also had favorable and similar responses to general health and heaviness of the latest menstruation as compared to before LNG-IUS. Both groups reported very similar results in the last 4 weeks, experiencing 13 out of the 14 symptoms appearing in the ending questionnaire; however, women from the CDB-2914 group were more likely to declare that “they were liable to gain weight” [23% (14/61) vs. 4% (2/55); 95% CI: 7–32%, p = 0.003]. Adverse effects: Not reported.
Weisberg, 2006	Two doses of 25 mg mifepristone for 1 day + two daily doses of placebo for 4 days -- Two daily doses of 100 mg doxycycline during 5 days -- Two doses of 25 mg mifepristone for 1 day + 2 daily doses of 10 μg ethinylestradiol for 4 days -- Two daily doses of placebo during 5 days	Two active treatment regimens were significantly more effective than placebo in terminating a bleeding episode. Both mifepristone in combination with ethinylestradiol and doxycycline alone were significantly more effective than mifepristone alone or placebo in stopping a bleeding episode. Mifepristone with ethinylestradiol and doxycycline alone were equally effective in stopping bleeding. Mifepristone with ethinylestradiol terminated a bleeding episode after a mean duration of 4.2 days (CI 3.5–5.2), compared to 4.8 days (CI 3.9–5.8) for doxycycline 5.9 days (CI 4.8–7.2) for mifepristone alone, and 7.5 days (CI 6.1–9–1) for placebo.	Satisfaction: Not reported. Adverse effects: Women in each treatment group reported a wide range of side effects and this did not differ according to the treatment used or between those taking active treatment and those taking placebo. Side effects occurred in 23 women on placebo, 28 women using mifepristone and ethinylestradiol, 19 women using doxycycline, and 21 women using mifepristone alone. These side effects were usually minor and did not result in women stopping treatment. The most common side effect was nausea, occasionally accompanied by vomiting which occurred in 8 women on placebo, 4 using mifepristone plus ethinylestradiol, 2 using doxycycline, and 5 using mifepristone. There were no serious adverse events during the study.
Weisberg, 2009	Two daily doses of 100 mg doxycycline during 5 days -- Two daily doses of 100 mg doxycycline + one daily dose of 20 ug ethinylestradiol during 5 days -- Two doses of 25 mg mifepristone for 1 day, then a daily oral dose of 20 ug ethinylestradiol for 4 days -- Two doses of 25 mg mifepristone for 1 day + two daily doses of 100 mg doxycycline for 5 days -- Two daily doses of placebo during 5 days	Two active treatment regimens were significantly more effective than the placebo in terminating a bleeding episode. Mifepristone plus ethinylestradiol and mifepristone plus doxycycline terminated a bleeding episode significantly more effectively within a mean of 4.0 (CI 3.5–4.6) and 4.4 (CI 3.8–5.2) days, respectively, compared with 6.4 (CI 4.8–8.6) for doxycycline plus ethinylestradiol, 6.4 (CI 4.4–9.2) for doxycycline alone, and 6.4 (CI 5.1–8.0) days for placebo (p < 0.0008 and 0.0108, respectively). Mifepristone combined with either ethinylestradiol or doxycycline was equally effective in stopping a bleeding episode. By 8 days after starting the first treatment, those who had stopped bleeding were: all women in the mifepristone and ethinylestradiol group; 97% in the mifepristone plus doxycycline group; 65.6% in the doxycycline group; 60% in the doxycycline plus ethinylestradiol group, and 70.3% in the placebo group. The percentage reductions in the mean number of bleeding and spotting days in the 90-day treatment phase when compared with pre-treatment were 17.7% for placebo, 22% for mifepristone plus ethinylestradiol, 25.7% for mifepristone plus doxycycline, and 5.5% for doxycycline alone, while there was an increase of 4.5% for doxycycline plus ethinylestradiol.	Satisfaction: Around 75% of women changed their perceptions on their bleeding patterns after the treatment, with no significant differences among groups (62% in the mifepristone plus ethinylestradiol group, 75% in the mifepristone plus doxycycline group, 70% in the doxycycline group, 82% in the doxycycline plus ethinylestradiol group, and 71% in the placebo group; p = 0.729). Before treatment, 85.3% of Implanon users felt their bleeding was “too long”; after treatment, 49.5% felt that the interval between bleeding episodes was “too short”. Adverse effects: There were no severe adverse events caused by the study drug. Women from all groups reported a wide range of minor secondary effects, with the most common being nausea and headaches. Nausea was present in 8 women with doxycycline, in 3 with doxycycline plus ethinylestradiol, in 6 with placebo, and in 5 with doxycycline plus mifepristone. Only one of them (in the placebo group) suspended the treatment due to nausea. Headaches were more common in women using doxycycline and ethinylestradiol; 9 women from this group reported them, in contrast to 2 in the placebo group, 4 in the doxycycline plus mifepristone group, and 3 in the doxycycline alone group. Less common secondary effects were vomit, diarrhoea, heartburn, stomach cramps or abdominal distension, vaginal discharge or itching, acne, extreme fatigue, and hot flashes.
Weisberg, 2009	Group A received COCP (20 mcg ethinylestradiol/150 mcg desogestrel) for 42 consecutive days Group B received three daily doses of 500 mg mefenamic acid for two courses, 5 days each and 21 days apart -- No placebo	Comparing the efficacy between COCP and NSAID in treating the bleeding irregularities in etonogestrel implant users, the mean duration of bleeding days was 15.63 ± 2.55 for COCP, and 14.75 ± 2.33 for NSAID (p = 0.103). More women in the COCP group stopped bleeding in the 7 days after starting treatment than in the NSAID group (p < 0.05). After 42 days of treatment, a small number of women in both groups presented a recursive, unacceptable bleeding pattern after suspending treatment for 7 days or more. This was more frequent in the COCP group (p < 0.05). The mean duration of bleeding within the 90 days of treatment was 7.29 ± 3.16 for COCP, and 10.5 ± 4.14 for NSAID (p < 0.05).	Satisfaction: Not reported. Adverse effects: A wide range of minor side effects, the most common of which were nausea and headaches.
Edelman, 2020	Two daily doses of 10 mg tamoxifen during 7 days -- Same for placebo	After the first treatment, women with tamoxifen experienced 9.8 (95% CI: 4.6–15.0) more consecutive days of amenorrhea than the placebo group. Those in the tamoxifen group also experienced more total days of no bleeding (amenorrhea or spotting) in the first 90 days [median: 73.5 (range 24–89) vs. 68 (range 11–81); p = 0.001]. They also had a longer time to restart bleeding or spotting after their first treatment as compared to the placebo group [Kaplan-Meier median: 12 days (range 1–56) vs. 6 days (range 1–38); p < 0.001]. There was a 1-day difference between groups in how fast bleeding stopped after the first treatment [median: 5 (range 1–21) vs. 6 (range 1–26); p = 0.029].	Satisfaction: Women were satisfied with their implant during the whole study. Bleeding pattern satisfaction and acceptability was low and similar for both groups at the beginning of the study. Satisfaction was significantly higher in the tamoxifen group than in the placebo group after the initial treatment [median: 71 (range 8.5–100) vs. 31 (range 0–100); p < 0.001] and at the end of the first 90-day reference interval [median: 62 (range 16–100) vs. 46 (range 0–100); p = 0.023]. After completing the open-label phase, satisfaction with bleeding was not significantly different between the treatment groups [median: 67.5 (range 0–100) vs. 54.25 (range 0.5–9.8); p = 0.129]. Acceptability showed similar trends, with no significant differences between groups. Adverse effects: Few participants reported treatment-related adverse events, while no serious adverse events occurred during the study. More women taking tamoxifen reported fluid retention (12 vs. 1), headache (19 vs. 1), and mood changes (13 vs. 2).
Hou, 2016	Monophasic oral contraceptive (30 mcg ethinylestradiol + 150 mg levonorgestrel) -- Placebo (28 capsules)	All 12 COC users and 75% of placebo users noted improvement in bleeding (p = 0.09). More women noted “significant” with COC use (n = 11 [92%]) than placebo (n = 5 [42%]) (p = 0.03). No woman in the COC group complained of worse bleeding, compared to two (17%) placebo users (p = 0.48). All 12 COC users and 75% of placebo users stopped bleeding during the first four weeks of treatment. COC users and placebo users who stopped bleeding did so after a median of 1 day (range 1–9) and 4.5 days (range 1–28), respectively (p = 0.63). All COC users and 9 placebo users returned for their 3-month follow-up visit. 12 out of 16 women who had used COCP for two months or more during the study saw better bleeding patterns, while 5 placebo users who kept their usealso saw improvements (p = 0.54). Surprisingly, 3 out of 15 women using COCP during the study did not report any improvement in their bleeding at the three-month point. 6 COCP users (38%) and 3 women under no treatment (60%) reported to still suffer from unscheduled bleeding, which is almost half of the patients (9 out of 21; p = 0.61).	Satisfaction: Women who expressed during registration their wish to have their implant removed were more likely to ask for removal in the following phases of the study (3/5 [60%] versus 1/17 (6%); p = 0.03). Adverse effects: Adverse effects seen in the first four weeks affecting more than one participant were headaches (n = 10), nausea or vomit (n = 3), cramps (n = 3), and breast tenderness (n = 4). There were no differences between groups for these events or for the average weight changes during the first four weeks.
Simmons, 2017	Two daily doses of 10 mg tamoxifen during 7 days -- Same for placebo	Women in the tamoxifen group reported more days of baseline bleeding than the placebo group (p < 0.05, two-sample T test) and a slightly shorter duration of implant use (p > 0.05). Women using tamoxifen (10.5 ± 9.0) had less days with bleeding/spotting than placebo users (15.5 ± 8.5) in the 30 days after drug activation (with a median difference of 5 days, p = 0.05), with most of this difference being due to less days with bleeding (−9.9, −0.05). Days of bleeding: 5.0 ± 6.2 (tamoxifen users) vs. 8.6 ± 7.0 (placebo users (p = 0.006). Days of bleeding/spotting during the study (out of 180): 65.6 (tamoxifen users) vs. ± 37.5 (placebo users) 46.9 ± 28.1 Mean difference: 18.7 (−4.7, 42.1; p = 0.011).	Satisfaction: Satisfaction with the bleeding pattern was equivalent and low for both groups at baseline. Satisfaction data were available for 24 subjects in each group after the first course of the study drug, and for 22 tamoxifen and 21 placebo users at their final study visit. Although satisfaction increased in both groups after the first course of the study drug, it was significantly higher for tamoxifen users (mean VAS 70.3 mm for tamoxifen vs. 49.3 mm for placebo; p = 0.02). By 180 days, bleeding satisfaction was higher than baseline but equivalent for both groups (mean 61 ± 24.7 mm for tamoxifen vs. 53.6 ± 33.3 mm for placebo; p = 0.39). Adverse effects: Tamoxifen was well tolerated, with no serious adverse effects and side effects equivalent to placebo.
Zigler, 2018	15 mg ulipristal acetate during 7 days -- Same for placebo	At the end of the 30-days follow-up, the mean of days with bleeding was 12.0 (IQR 6–21) for the placebo group and 7.0 (IQR 4.5–11) for the ulipristal acetate group (p = 0.002). 3 women using placebo (9.7%) and 11 using ulipristal acetate (34.4%) stopped bleeding at day 10 (p = 0.03).	Satisfaction: Satisfaction with bleeding patterns was recorded at the start and at the end of the study. Both groups presented similar satisfaction levels at the beginning, while at the end, the ulipristal acetate users were more likely to be satisfied with their bleeding patterns than the placebo users (71.9% vs. 26.7%, respectively; p < 0.001). 27 women from the ulipristal acetate group declared to be “not happy at all” or “somewhat unhappy” with their bleeding. No patient declared their dissatisfaction levels. Most women from both groups stated they wanted their bleeding to be eliminated (63.3% from the ulipristal acetate group and 67.7% from the placebo group; p = 0.72). At the end of the study, more ulipristal acetate users wanted to keep their implant in comparison to placebo users (89.7% vs. 63.3%; p = 0.03). Adverse effects: There were few side effects reported during the treatment period, including headache (19.4% in placebo group, 9.4% in ulipristal acetate group), and nausea and vomiting (9.7% in placebo group, 3.1% in ulipristal acetate group). Groups were similar in side effects, with the majority experiencing no side effects (placebo group: 74.2%, ulipristal acetate group: 81.3%; p = 0.5). We did not encounter any serious adverse effects. All participants from both groups felt their medication was easy to use. All ulipristal acetate users stated they would use the medication again for bleeding, which was significantly different from placebo users (66.7%; p = 0.001).

includes the main characteristics of the studies included in this review.

Regarding the treatments for bleeding that showed a statistically significant improvement:

3.1. Non-Steroidal Anti-Inflammatory Drugs (NSAIDs)

The cyclooxygenase-2 inhibitors show significant progress for managing bleeding [24,25]. Patients ingesting valdecoxib [25] in 20 mg tablets during 5 days had a higher average amount of days without bleeding in a period of 28 days than the placebo group, which was a significant result. In the case of celecoxib [24], the percentage of women whose bleeding stopped within 7 days was significantly higher in the group who received a daily oral dose of 200 mg during 5 days (p = 0.001). Also, this group reported a significantly higher satisfaction than the placebo group, with 80% versus 30% (p = 0.001).

3.2. Selective Estrogen Receptor Modulator (Tamoxifen)

Four studies covered the use of tamoxifen [15,16,22,28], with three of them showing significant results [15,16,28]. One study used two daily doses of 10 mg during 10 days [15], while the other two used the same two daily doses of 10 mg but for 7 days [16,28].

Women who used tamoxifen in the study by Abdel Aleem et al. [15] spoke of less bleeding during and after the treatment, and even of discontinuing the use of implants (RR 0.41 (95% CI: 0.21–0.80); RR 0.38 (95% CI: 0.18–0.88); RR 0.20 (95% CI: 0.05–0.87), respectively). Follow-ups in the first and second months showed significant differences between the tamoxifen and the placebo groups; however, there were no significant differences between both groups after three months.

Other observations include an improved lifestyle and general satisfaction with the treatment, as well as diverse side effects. Said satisfaction led patients to keep using this contraceptive method, as mentioned by Abdel Aleem et al. in 2012 [23].

Edelman et al. [16] observed that, after the first treatment with Implanon^®, the group who received it averaged 9.8 more consecutive days of amenorrhea and more days with no bleeding (amenorrhea or spotting) in the first 90 days [median: 73.5 (range: 24–89) versus 68 (range: 11–81); p = 0.001]. The placebo group showed a similar effect after the first active use of tamoxifen in the open phase of the study. After the first 90 days, the women who received tamoxifen reported a higher satisfaction [median: 62 (range: 16–100)] than those treated with placebo [median: 46 (range: 0–100); p = 0.023).

In Simmons et al. [28], the women who received the drug -users of etonogestrel implants- had less days with bleeding or spotting than those receiving a placebo in the 30 days after activation of the study drug (mean difference: 5 days; p = 0.05), with most of the difference being attributable to less days of bleeding. Both groups had their first day with no bleeding in similar moments after starting the study (median of 5 days for tamoxifen, and of 6 days for placebo), but women who received tamoxifen reported more consecutive days without bleeding (mean difference: 15.2 days; p = 0.02).

3.3. Antifibrinolytics

Tranexamic acid was used in three studies [19,27,32]. Sørdal et al. [19] gave three daily doses of 500 mg during the bleeding or spotting episode; Senthong [27] used four daily doses of 250 mg during 5 days, and Phupong et al. [32] administered a dose of two 250 mg capsules four times a day for five days.

During the follow-up period (4 weeks after the start of the treatment), bleeding stopped for 58.8% of women in the tranexamic acid group and 76.5% of women in the placebo group (p = 0.12). The percentage of women whose bleeding stopped within 7 days after starting the treatment was significantly higher in the tranexamic acid group than in the placebo group (64.7% versus 35.3%; p = 0.015). The average duration of days with bleeding and spotting was not significantly different between the two groups in the 28 days of the follow-up period (medians of 15.4 and 12.7 days, respectively; p = 0.182).

3.4. Antibiotics, Matrix Metalloproteinase Inhibitors

Three studies used doxycycline [23,29,33]. Abdel Aleem et al. [23] did not observe significant changes in bleeding. On the other hand, Weisberg et al. [33] reported that doxycycline was more effective than the placebo in stopping a hemorrhage episode at 90 days, as well as a decrease of 10.4% in average bleeding, less spotting, and fewer days with bleeding. Doxycycline managed to stop bleeding in 4.8 days on average (95% CI:3.9–5.8). However, this result could not be replicated in the broader study in other 90 days.

Weisberg et al. [29] used five different interventions in their study, but only three of them compared doxycycline with a placebo. One of these only consisted on doxycycline for five days; other used doxycicline with mifepristone, and the last one used doxycycline with ethinylestradiol. All of them were administered to users of Implanon^®.

Doxycycline took 6.4 days (95% CI: 4.4–9.2) on average to stop bleeding; on the other hand, combinations of doxycycline with ethinylestradiol and with mifepristone were tested as well, with an average result of 6.4 days (95% CI: 4.8–8.6) and 4.4 days (95% CI: 4.38–5.2), respectively. The only combination that presented statistically significant results was that of doxycycline with mifepristone (p = 0.008).

3.5. Selective Progesterone Receptor Modulator (SPRM)—UPA Versus Placebo

One of the most important findings of this review was the use of ulipristal acetate (UPA) by two trials—Fava et al. [17] and Zigler & McNicholas [5], who used daily doses of 5 mg during five days and 15 mg during seven days, respectively.

In Fava et al. [17], UPA users reported it took less time to stop their bleeding, and their period with no bleeding was longer than what placebo users informed; however, the differences were not significant. Zigler & McNicholas [5] reported less days of bleeding, which was statistically significant—as Fava et al. did not report something similar [17]. In the latter, the number of days before the bleeding stopped and the number of days with no bleeding were statistically similar for both groups (p < 0.05); although the users of LNG-IUD who received a daily dose of 5 mg UPA during 5 days reported a bigger decrease in abnormal bleeding than those receiving a placebo, the results were not statistically significant.

On the other hand, Warner et al. [21] observed that the prophylactic use of CDB-2914 (antiprogestin) led to a statistically significantly lower percentage of days with bleeding/spotting; however, this effect disappeared in the third cycle, and the percentage of days with bleeding was then lower in the placebo group.

3.6. Combined Contraceptives (COCP)

Three studies used COCPs to treat genital bleeding [18,30,31].

Upawi et al. [18] found that women with etonogestrel implants who received COCP had a significantly lower median number of days with bleeding and spotting than women who received NSAIDs (7.29–3.16 vs. 10.57–4.14 days; p < 0.05). In addition, more women in the COCP group saw their bleeding stop in the week after starting treatment than those in the NSAID group (p < 0.05).

Guiahi et al. [30] reported an 11.7-fold higher probability (95% CI: 1.9–70.2) of experiencing a temporary interruption in bleeding during the study drug administration period: 14 (87.5% ± 16.2%) in the COCP group compared with 6 (37.5% ± 23.7%) in the placebo group (p < 0.01).

Finally, Hou et al. [31] reported that all COCP users and 75% of placebo users stopped bleeding during the first four weeks of treatment. COCP users and placebo users who stopped bleeding did so after a median of 1 day (range 1–9) and 4.5 days (range 1–28), respectively (p = 0.63).

3.7. Progesterone Receptor Antagonists

Weisberg et al. [33] tested three alternative treatments and placebo in women with implants: two doses of 25 mg mifepristone on the first day, followed by two daily doses of placebo during 4 days; two doses of 25 mg mifepristone on the first day, followed by a dose of 20 mcg ethinylestradiol (EE) in the morning and a dose of placebo in the night during 4 days; two daily doses of 100 mg doxycycline during 5 days, and two daily doses of placebo during 5 days.

The treatments with mifepristone plus ethinylestradiol and with doxycycline only were significantly more effective than the treatment with mifepristone only and the placebo in stopping bleeding. Mifepristone with ethinylestradiol stopped bleeding after a median of 4.2 days (95% CI: 3.5–5.2); doxycycline, in 4.8 days (95% CI: 3.9–5.8); mifepristone, in 5.9 days (95% CI: 4.8–7.2), and the placebo, in 7.5 days (95% CI: 6.1–9.1).

In Weisberg et al. [29], the patients were users of Implanon^® and all four treatments included placebo: two doses of 25 mg mifepristone on day 1, followed by a daily dose of 20 mcg ethinylestradiol during 4 days; two daily doses of 100 mg doxycycline during 5 days; two doses of 25 mg mifepristone on day 1, followed by two daily doses of 100 mg doxycycline during 5 days; two daily doses of 100 mg doxycycline and one daily dose of 20 mcg ethinylestradiol during 5 days, and two daily doses of placebo during 5 days.

The study showed that the treatments with mifepristone plus ethinylestradiol and with mifepristone plus doxycycline were significantly more efficient in stopping hemorrhagic episodes, with medians of 4.0 (CI 3.5–4.6) and 4.4 (CI 3.8–5.2) days, respectively, than doxycycline plus ethinylestradiol (median: 6.4 days; CI 4.8–0.86), doxycycline only (median: 6.4 days; CI 4.4–9.2), and the placebo (median: 6.4 days; CI 5.1–8.0) (p = 0.0008 and 0.0108, respectively).

As for other results, it is important to highlight:

3.8. Participant Satisfaction

Only 13 studies reported the participants’ satisfaction [5,15,16,19,20,21,22,23,24,26,28,29]. These reports are varied and based on estimated scales or on a dichotomous scale -percentage of satisfied versus unsatisfied women-, as seen in the study by Abdel Aleem et al. [15], who obtained satisfactory results up to the second month and reported insignificant differences by the third month. The participants’ decision to keep using the same contraceptive method was also considered as satisfaction, as mentioned by Abdel Aleem et al. [23]. Madden et al. [20] carried out a more precise assessment of satisfaction, based on weekly progression, with the results at four weeks showing that 39.5% of the estradiol group was not satisfied with their bleeding, in contrast to 9.5% of the naproxen users and 11.6% of the placebo group, respectively (p = 0.01). Satisfaction with the bleeding patterns improved in all groups after time.

At 12 weeks, 85% of women said to be “somewhat” or “very” satisfied with their bleeding patterns, and this satisfaction did not depend on the type of treatment (p = 0.12). Satisfaction with LNG-IUD was 94% at 12 weeks in all three study groups. Sørdal et al. [19] reported that satisfaction with the method was unrelated to satisfaction with the treatment, the latter being 85%. Weisberg et al. [29] observed that around 75% percent of women changed their perception of their bleeding patterns after the treatment, but this did not significantly vary among groups. In the study by Hou et al. [31], women who stated their wish to get their implants removed when registering were more likely to ask for removal later during the study (3 out of 5 [60%] versus 1 out of 17 [6%]; p = 0.03).

3.9. Adverse Effects

Data on adverse effects was not as detailed as on satisfaction. Four trials did not report them [24,25,30,32], while the other 17 studies just mentioned them, with the most common effects being headache and nausea. Warner et al. [21] observed weight gain in the control group, with a significant difference.

Table 2. Other background information on the included studies, sorted by year.

Author/Year/ Country	Objective	Study Design/Duration/ Follow-Up	Initial Population/Dropout	General Characteristics of the Population (Age, Parity, BMI)	Progestin/Baseline Characteristics
Abdel Aleem, 2005 Egypt	To evaluate the possible role of tamoxifen (selective estrogen receptor modulators, SERM) in treating bleeding irregularities associated with Norplant contraceptive use.	Double-blinded randomized controlled trial Three months At 10 days, at the end of the first month, at the end of the second month, and at the end of the third month	100 women Tamoxifen: 50 Placebo: 50 Dropout: 7%	Age: Tamoxifen: 32.48 ± 5.6/Placebo: 32.28 ± 6.1 Parity: Tamoxifen: 4.34 ± 1.7/Placebo: 4.70 ± 1.8 BMI: Tamoxifen: 24.79 ± 4.0/Placebo: 24.53 ± 3.3	Norplant® Bleeding was defined as current episode of bleeding/spotting 8 days, or bleeding-free interval 15 days. Users were on their first year of Norplant®.
Nathirojanakun, 2006 Thailand	To evaluate the efficacy of valdecoxib and placebo for controlling irregular uterine bleeding in depot-medroxyprogesterone acetate (DMPA) users.	Double-blind, placebo-controlled study 28 days Not reported	51 women Valdecoxib: 25 Placebo: 26 Dropout: 9.8%	Age: Valdecoxib: 24.1 ± 6.59/Placebo: 25.1 ± 7.35 Parity: Valdecoxib: 1.2 ± 0.62/Placebo: 1.3 ± 0.54 BMI: Valdecoxib: 21.22 ± 3.13/Placebo: 21.06 ± 3.06	DMPA Having 8 days or more of a bleeding or spotting prior to obvious bleeding. DMPA: For 12 meses
Phupong, 2006 Thailand	To evaluate the effects of tranexamic acid and placebo on controlling irregular uterine bleeding secondary to Norplant use.	Prospective randomized, double-blind, placebo-controlled trial 4 weeks 1 week post-treatment, and then at 4 weeks.	68 women Tranexamic acid: 34 Placebo: 34 Dropout: 0%	Age: Tranexamic acid: 27.0 ± 5.7/Placebo: 29.1 ± 6.1 Parity: Tranexamic acid: 12 (35.3)/Placebo: 10 (29.4) Tranexamic acid: 18 (52.9)/Placebo: 16 (47.1) Tranexamic acid: 4 (11.8)/Placebo: 8 (23.5) BMI: Tranexamic acid: 22.7 ± 3.9/Placebo: 21.4 ± 3.0	Norplant® Bleeding disturbances were defined as bleeding or spotting for eight or more continuous days or a current bleeding episode initiated after a bleeding-free interval of 14 days or less. Norplant inserted 3–36 months before enrollment.
Weisberg, 2006 Australia	To compare three treatments with placebo on the duration and recurrence of frequent and/or prolonged bleeding in Implanon users.	Randomized, double-blind, placebo-controlled trial 180 days At 90 days	179 women Mifepristone: 44 Doxycycline: 45 Mifepristone + ethinylestradiol: 45 Placebo: 45 Dropout: 6.1%	Age: Mifepristone: 29.9 ± 1.007/Placebo: 28.9 ± 0.995/Mifepristone + ethinylestradiol: 27.6 ± 0.995/Doxycycline: 28.8 ± 0.995 Parity: Only pregnancies are reported. BMI: Only weight (kilograms) is reported.	Implanon® Women who had used Implanon for contraception for ≥3 months and were experiencing prolonged or frequent bleeding patterns were recruited at four Australian sites either through clinics or by advertisement. They were enrolled into the treatment phase, provided they had met one of the WHO criteria for prolonged or frequent bleeding.
Archer, 2008 USA	To evaluate ethinylestradiol or ibuprofen compared to placebo on spotting and bleeding and a postcoital test in women using the levonorgestrel subcutaneous implant.	Multicenter prospective randomized study Once a month during one year or until the spotting and bleeding had stopped for 2 months.	106 women Estrogen: 20 Ibuprofen: 42 Placebo: 44 Dropout: 0%	Age: not reported Parity: not reported BMI: Only weight	LNG subcutaneous implant Implant has more than 1 month, with spotting or bleeding lasting 8 or more consecutive days, or more than 10 out of 14 days in the previous 3 weeks, plus no treatment against spotting or bleeding in the last month.
Buasang, 2009 Thailand	To evaluate the efficacy of valdecoxib and placebo for controlling irregular uterine bleeding in depot-medroxyprogesterone acetate (DMPA) users.	Randomized, double-blind, placebo-controlled trial. 28 days At 4 weeks	40 women Celecoxib: 20 Placebo: 20 Dropout: 0%	Age: Celecoxib: 34.10 ± 7.40/Placebo: 29.50 ± 9.84 Parity: Celecoxib: 2.20 ± 1.50/Placebo: 1.50 ± 0.51 BMI: Celecoxib: 21.90 ± 3.82/Placebo: 23.30 ± 2.45	Jadelle® Current bleeding on the day of participation Users of Jadelle® for 9–13 months.
Weisberg, 2009 Australia	To compared four treatments against a placebo in Implanon users and tested whether repeated treatment improved subsequent bleeding patterns.	Randomized, double- blind, placebo-controlled trial. 90 days 90 days	204 women Doxycycline Doxycycline + ethinylestradiol Mifepristone + ethinylestradiol Mifepristone + doxycycline Placebo: 40 Dropout: 13.9%	Age: Doxycycline: 28.6 ± 6.8/ Doxycycline + ethinylestradiol: 28.8 ± 6.0/ Placebo: 28.8 ± 5.8/ Mifepristone + ethinylestradiol: 29.1 ± 7.2/ Mifepristone + doxycycline: 29 ± 6.0/ Non-randomized: 28 ± 6.3 Weight: Doxycycline: 69.5 ± 14.3/ Doxycycline + ethinylestradiol: 69.3 ± 12.4/ Placebo: 71.5 ± 21.9/ Mifepristone + ethinylestradiol: 72.8 ± 26.2/ Mifepristone + doxycyline: 72.5 ± 17.5/ Non-randomized: 70.6 ± 19.0	Implanon® Daily menstrual record for at least 90 days, then they could enroll in the treatment phase, as long as they met one of the WHO criteria for frequent bleeding/spotting (Belsey & Pinol, 1997), which is defined as a bleeding/spotting episode that lasts more than 10 days, or more than 4 bleeding/spotting episodes within 90 days.
Sethong, 2009 Thailand	To evaluate the efficacy of tranexamic acid and placebo for controlling irregular uterine bleeding in depot-medroxyprogesterone acetate (DMPA) users.	Prospective randomized trial double-blind, placebo-controlled study 28 days Not reported	99 women Tranexamic acid: 50 Placebo: 49 Dropout: 1%	Age: Tranexamic acid: 29.6± 7.09/Placebo: 28.69 ± 7.671 Parity: Tranexamic acid: 1.32 ± 0.74/Placebo: 1.12 ± 0.60 BMI: Tranexamic acid: 23.11 ± 2.76/Placebo: 22.5 ± 2.07	DMPA Being DMPA users for a period of 1–18 months; having 8 days or more with bleeding or spotting.
Dempsey, 2010 USA	To hypothesize that the estradiol vaginal ring will be acceptable to participants and will decrease bleeding during the first 3 months of DMPA use and increase continuation at 3 months.	Prospective, randomized, controlled, non-blinded trial 6 months 6 months	71 women Vaginal ring: 35 Not intervened: 36 Placebo: no Dropout: At 3 months: DMPA: 36.1%, Vaginal ring: 25.7% At 6 months: DMPA: 56%, Vaginal ring: 51%	Age: 18–20 Total: 28 (39)          DMPA: 14 (39), Ring: 14 (40) 21 or older Total: 43 (61) DMPA: 61 (22), Ring: 21 (60) Parity: None Total: 27 (38)            DMPA: 11 (31), Ring: 16 (46) One Total: 29 (41)             DMPA: 16 (44), Ring: 13 (47) Two or more Total: 15 (21)   DMPA: 9 (25), Ring: 6 (17) BMI: ≤ 25 Total: 32 (58)            DMPA: 20 (65), Ring: 12 (50) 25–29.9 Total: 16 (29)    DMPA: 9 (29), Ring: 7 (29) ≥30 Total: 11 (13)            DMPA: 2 (6), Ring: 5 (21)	DMPA Hadn’t used DMPA before, suffering from oligomenorrhea.
Warner, 2010 USA	The aim of the clinical trial reported here was to determine whether intermittent administration of SPRM CDB-2914	Double-blind, randomized, controlled trial and placebo-controlled 6 months (all phases included) At 1, 3, and 6 months	136 women CDB-2914: 69 Placebo: 67 Dropout: 14.7%	Age: CBD-2914: 36.9 (6.5)/Placebo: 35.8 (7.0) Parity: 0 CBD-2914: 14 (20)/Placebo: 20 (30) 1 CBD-2914: 18 (26)/Placebo: 13 (19) 2 CBD-2914: 30 (43)/Placebo: 27 (40) 3 CBD-2914: 6 (9)/Placebo: 7 (10) 4 CBD-2914: 1 (1)/Placebo: 0 (0) BMI: Not reported	LNG-IUS initiating use of an LNG-IUS for contraception, with regular menstrual cycles lasting 17–42 days, menstrual periods lasting less than 11 days.
Abdel Aleem, 2012 Egypt	To examine the effect of DOX compared with placebo in stopping a current bleeding episode during DMPA use as welL as the effect of drug treatment on the bleeding pattern in the 3 months following the treatment.	Double-blind, randomized, placebo-controlled trial 3 months (including follow-up) At the 1st, 2nd, and 3rd months.	68 women Doxycycline: 34 Placebo:34 Dropout: 14.7%	Age: Doxycycline: 29.6 ± 7.3 Placebo: 30.0 ± 5.7 Parity: The study only asks about living children. BMI: Not reported.	DMPA Users of DMPA as contraception for at least 1 month, complaining of a bleeding episode (defined as ≥8 bleeding or spotting days or bleeding-free interval of 15 days or less). The treatment package was started if the bleeding or spotting episode had lasted for the last 8 or more days in one continuum or when having bleeding or spotting occur within 15 days from the last bleeding episode.
Madden, 2012 USA	To evaluate whether oral naproxen or transdermal estradiol decreases bleeding and spotting in women initiating the levonorgestrel-releasing intrauterine system (LNG-IUS) Mirena®	Randomized controlled trial 16 weeks Calls at 4, 8, and 16 weeks + 1 visit at 12 weeks	129 women Naproxen: 44 Transdermal estradiol: 42 Placebo: 43 Dropout: 17.8%	Age: <21: Estradiol: 5 (11.4%)/Naproxen: 5 (11.9%)/Placebo: 4 (9.3%) 21–29: Estradiol: 30 (68.2%)/Naproxen: 28 (66.7%)/Placebo: 30 (69.8%) ≥30: Estradiol: 9 (20. 5%)/Naproxen: 9 (21.4%)/Placebo: 9 (20.9%) Parity: 0: Estradiol: 20 (45.4%)/Naproxen: 22 (53.4%)/Placebo: 22 (51.1%) 1–2: Estradiol: 23 (52.3%)/Naproxen: 14 (33.3%)/Placebo: 18 (41.9%) ≥3: Estradiol: 1 (2.3%)/Naproxen: 6 (14.3%)/Placebo: 3 (7.0%) BMI: <25: Estradiol: 12 (27.3%)/Naproxen: 15 (35.7%)/Placebo: 22 (51.2%) 25–29.9: Estradiol: 10 (22.7%)/Naproxen: 15 (35.7%)/Placebo: 9 (20.9%) ≥30: Estradiol: 22(50.0%)/Naproxen: 12 (28.6%)/Placebo: 12 (27.9%)	Mirena® Not reported.
Sordal, 2013 Denmark, Ireland, Norway	To investigate whether counseling plus tranexamic acid or the NSAID, mefenamic acid, was superior to counseling plus placebo in the management of initial bleeding or spotting after levonorgestrel-releasing intrauterine system placement	Prospective, double-blind, three-arm, randomized placebo-controlled, phase IV 9 months (30 days for assessment) At 30 days	187 women Tranexamic acid: 63 Mefenamic acid: 63 Placebo: 61 Dropout: 10.69%	Age: Tranexamic acid: 31.56 ± 7.2/Mefenamic acid: 32.96 ± 6.7/Placebo:33.16 ± 7.0 Parity: Tranexamic acid: 1.46 ± 1.1/Mefenamic acid: 1.76 ± 1.0 BMI: Tranexamic acid: 25.056 ± 4.54/Mefenamic acid: 24.766 ± 3.74/Placebo: 25.076 ± 4.63	LNG-IUS Regular menstrual cycles (21–35 days).
Guiahi, 2015 USA	To understand whether using oral contraceptive pills (OCPs) results in temporary interruption of bleeding for etonogestrel contraceptive implant users during a 14-day course.	Double-blind, randomized controlled trial 14 days Weekly for a month, or until the first bleeding	32 women OCPs: 16 Placebo: 16 Dropout: 1%	Age: OCPs: 22.46 ± 4.8/Placebo: 21.46± 3.0 Parity: Only pregnancies are reported. BMI: OCPs: 23.26 ± 2.9/Placebo: 23.36± 3.5	Etonogestrel implant users who reported bothersome bleeding with a current bleeding–spotting episode of at least 7 consecutive days.
Hou, 2016 USA	Estimate symptom improvement rate of women with bleeding complaints using the etonogestrel contraceptive implant when started on continuous combined oral contraceptives.	Double-blind randomized, multicenter placebo-controlled trial 28 days Up to 3 months	26 women COCs: 13 Placebo: 13 Dropout: 30.7%	Age: COCs: 25.4 ± 3.8/Placebo: 25.8 ± 5.0 Parity: 0: COCs: 8 (62%)/Placebo: 9 (69%) 1: COCs: 1 (8%)/Placebo: 3 (23%) 2: COCs: 2 (15%)/Placebo: 1 (8%) BMI: COCs: mean: 26.5, range: 18.9–38.9. Placebo: mean 23.1, range: 21.0–43.9.	Etonogestrel implants Significantly abnormal bleeding or heavy flow.
Simmons, 2017 USA	To evaluate whether a short course of tamoxifen reduces bleeding/spotting days compared to placebo in ENG implant users.	Randomized, double-blind, placebo-controlled trial 17 months 14 days after ending the treatment, then at 180 days or a last visit.	56 women Tamoxifen: 28 Placebo: 28 Dropout: 39.2%	Age: Tamoxifen: 23.9 ± 5.7/Placebo: 25.4 ± 4.5 Parity (Nulliparous): Tamoxifen: 18 (64)/Placebo: 19 (68) BMI: Tamoxifen: 27.2 ± 7.3/Placebo: 28.4 ± 11.3	Etonogestrel implant 15–45-years-old women using the ENG 68-mg subdermal implant) for at least 30 days. Frequent or prolonged bleeding or spotting during the past month. Frequent was defined as bleeding/spotting (B/S) more than every 24 days, and prolonged was defined as B/S every day in a row for 14 days or longer, consistent with prior study populations.
Cohen, 2019 USA	To determine if a course of oral tamoxifen following the placement of a levonorgestrel 52 mg intrauterine system (IUS) reduces bleeding/spotting days over 30 days	Randomized, double-blind, placebo-controlled clinical trial 16 months At 30 days	37 women Tamoxifen: 18 Placebo: 19 Dropout: 8.1% for outcome 1 and 18.9% for outcome 2	Age: Tamoxifen: 30.6 ± 7.6/Placebo: 28.8 ± 6.1 Parity (Nulliparous): Tamoxifen: 13 (76.5)/Placebo: 15 (88.2) BMI: Tamoxifen: 28.9 ± 8.0/Placebo: 23.7 ± 3.3	LNG-IUS 15–45 years old women initiating levonorgestrel 52 mg IUS for contraception.
Zigler, 2019 USA	To evaluate if ulipristal acetate reduces the number of bleeding days in etonogestrel implant users in a 30-day period as compared to placebo.	Randomized, double-blind, placebo-controlled trial. 8 months Participants received a weekly call, to fill out a follow-up survey, during 4 weeks.	65 women Ulipristal acetate: 32 Placebo:33 Dropout: <10%	Age: Ulipristal acetate: 26.4 ± 6.2/Placebo: 25.5 ± 6.3 Parity: 0: Ulipristal acetate: 14 (43.7)/Placebo: 17 (54.8) 1: Ulipristal acetate: 5 (15.6)/Placebo: 7 (22.6) 2: Ulipristal acetate: 7 (21.9)/Placebo: 3 (9.7) +3: Ulipristal acetate: 6 (18.8)/Placebo: 4 (12.9) BMI: Ulipristal acetate: 29.6 ± 8.2/Placebo: 27.8 ± 7.0	Etonogestrel subdermal implant (68 mg) 18–45 years old English-speaking women, with an implant for more than 90 days and less than 3 years before sign-up, as well as being unsatisfied with their bleeding pattern (more than 1 bleeding episode in the last 24 days).
Upawi, 2020 Malaysia	To evaluate whether unacceptable bleeding among the etonogestrel implant users could be better alleviated using combined oral contraceptive pills (COCP) or nonsteroidal anti-inflammation drug (NSAID).	Randomized controlled trial 18 months At 90 days	86 women COCP: 43 Mefenamic acid: 43 Placebo: No Dropout: 2.3%	Age: COCP: 32 (29–35)/NSAID: 33 (30–35) Parity: 0: COCP: 0 (0)/NSAID: 1 (2.4) 1: COCP: 8 (19.05)/NSAID: 4 (9.5) 2: COCP: 16 (38.1)/NSAID: 16 (38.1) 3: COCP: 10 (23.8)/NSAID: 11 (26.2) 4: COCP: 5 (11.9)/NSAID: 8 (19.0) 5: COCP: 2 (4.8)/NSAID: 2 (4.8) 6: COCP: 1 (2.4)/NSAID: 0 (0) BMI: COCP: 26.55 ± 2.66 NSAID: 25.81 ± 3.31	Implanon® Women with etonogestrel implants and abnormal menstrual cycles—defined as frequent or prolonged bleeding. Frequent bleeding was experiencing more than 4 bleeding/spotting episodes in a period of 90 days. Prolonged bleeding was having one or more bleeding/spotting episodes of more than 10 days in the same period.
Fava, 2020 Brazil	To assess the efficacy of ulipristal acetate (UPA) for reducing abnormal bleeding among women using the 52-mg levonorgestrel intrauterine system (LNG-IUS).	Double-blind, randomized, placebo-controlled pilot study 30 days At 90 days	30 women Ulipristal acetate: 15 Placebo: 15 Dropout: 16.6%	Age: 20–29: Ulipristal acetate: 4 (26.7%)/Placebo: 3 (30%) 30–39: Ulipristal acetate: 8 (53.3%)/Placebo: 4 (40.0%) ≥40: Ulipristal acetate: 3 (20.0%)/Placebo: 3 (30.0%) Parity: Nulliparous: Ulipristal acetate: 3 (20.0%)/Placebo: 3 (30%) ≥1: Ulipristal acetate: 12 (80.0%)/Placebo: 7 (70.0%) BMI: 20–24.9: Ulipristal acetate: 3 (21.4%)/Placebo: 4 (40.0%) 25–29.9: Ulipristal acetate: 7 (50.0%)/Placebo: 4 (40.0%) ≥30: Ulipristal acetate: 4 (28.6%)/Placebo: 2 (20.0%)	LNG-IUS Women aged 18–45 years who had been using the 52-mg LNG-IUS for at least 1 year but less than 4 years, reporting complaints of abnormal uterine bleeding that was prolonged (more than 14 days of bleeding) or frequent (less than 24 days between menstrual periods) in the previous month.
Edelman, 2020 USA	To evaluate whether a short course of tamoxifen decreases bothersome bleeding in etonogestrel contraceptive implant users.	Double-blind, randomized control trial 2 years and 3 months At 90 days	112 women Tamoxifen: 57 Placebo: 55 Dropout after the first treatment: 7.1% Dropout at the sixth treatment: 73%	Age: Tamoxifen: 23.4 ± 4.5/Placebo: 24.5 ± 5.1 Parity: Only nulliparity is reported. BMI Tamoxifen: 27.4 ± 7.4/Placebo: 27.2 ± 7.5	Etonogestrel implant Women aged 15–45 years using the etonogestrel 68-mg subdermal contraceptive implant for at least 30 days, who met criteria for frequent or prolonged bleeding or spotting during the previous month. Defined ‘frequent’ as two or more independent bleeding or spotting episodes, and ‘prolonged’, as seven or more consecutive days of bleeding or spotting in a 30-day interval.

shows more details.

4. Discussion

This review aimed to update the management of bleeding associated to the use of progestins as contraceptives, with the 2013 systematic review by Abdel Aleem [35] as its reference and adds to recent literature [36]. That review covered 33 randomized clinical trials (RCT) from 1970 to 2012, and its objective was to evaluate prophylaxis and treatment of abnormal bleeding associated to the use of progestin-only contraceptives. Said review concluded that, while the bleeding stopped in some of the patients using progestin-only contraceptives, the interventions lacked medium- and long-term effectiveness.

In the current review, one of the main differences seen in the diagnosis is the new classification introduced by the Federation of Gynecology and Obstetrics (FIGO) in 2018 [37], which differentiates the types of bleeding by their origin. For this reason, this review did not include studies whose causes could be found in the PALM-COEIN classification (polyp; adenomyosis; leiomyoma; malignancy and hyperplasia; coagulopathy; ovulatory dysfunction; endometrial; iatrogenic, and not yet classified). Current scientific evidence still sees abnormal bleeding caused by progestin use as a reproductive health challenge, and keeps searching for solutions or better management.

The process of collecting and analyzing the articles selected for this review showed a wide variety of data: country of origin, methodologies, treatment plans, follow-ups, etc. The results, whose main objective was to find out whether bleeding stopped or not, were unevenly reported, which complicates comparative analysis.

Patients’ ages were one of the similarities found among the studies. 82% of the trials presented this data in age ranges, with the minimum age being 17.5 years in Nathirojanakun et al. [25], and the maximum, 30.5 years, described by Warner et al. [21].

The most used progestin contraceptive was the implant, seen in 12 trials [5,15,16,18,24,28,29,30,31,32,33,34], followed by DMPA, with 4 studies [23,25,26,27], and IUD, with 5 [17,19,20,22,27].

Follow-up length varied among studies, lasting between 10 days [15,23] and 180 days [16,28]. Long-term effectiveness was seemingly better in longer follow-ups; however, not all studies with longer follow-ups could demonstrate that treatments can keep an extended, sustained effectiveness. On the other hand, loss to follow-up ranged between 0% and 36%, with Dempsey et al. [26] having the biggest loss.

Regarding the use of placebos, 18 out of 21 trials used them to assess treatment effectiveness. Madden et al. [20] did not use placebo, as one of their interventions included the estradiol patch, which does not have an equivalent, and the same happened in Dempsey et al. [26], whose treatment was based on the vaginal ring, and in Upawi et al. [18], who used a comparator with an active principle.

Regarding the most common treatments seen in the reviewed trials, NSAIDs appeared in 5 studies [19,20,24,25,34]. Reviewed NSAIDs (valdecoxib, celecoxib, ibuprofen, naproxen, and mefenamic acid) were administered to women who were already using other types of contraceptives, such as DMPA, IUD, and implants. These studies demonstrated a general decrease in bleeding and, in users of DMPA who received valdecoxib or celecoxib, an increase in days without bleeding.

COCPs were evaluated in three studies with users of etonogestrel implants, with one of them [30,31] using 150 mg levonorgestrel and 30 mcg ethinylestradiol, the second using 150 mg levonorgestrel and 20 mcg ethinylestradiol, and the third using 150 mcg desogestrel [18]. Two studies [18,30] showed favorable data for the effect of COCPs; in the third study, the group who used contraceptives reported a significant improvement [31].

This review covered three trials where estrogen was used, with two of them comparing it with NSAIDs and placebos [20,34] and the other, with only placebo [26]. Dempsey et al. [26] and Madden et al. [20] reported the benefits of its use against bleeding, while Archer et al. [34] did not find significant differences in the use of estrogen patches by IUD users when compared with placebo.

Tamoxifen, a selective estrogen receptor modulator, was tested in four trials [15,16,22,28]. All of them administered two daily doses of 10 mg, for 10 days in the study by Abdel Aleem et al. [15] and for 7 days in the other three studies [16,22,28].

Abdel Aleem et al. [15], Edelman et al. [16], and Simmons et al. [28] found significant benefits to the use of tamoxifen versus placebo. Abdel Aleem et al. [15] could demonstrate that the bleeding stopped within 7 days of treatment, lasting for two months; however, the effect did not last to the third month of follow-up. Both Edelman et al. [16] and Simmons et al. [28] proved a reduction in days with bleeding when compared to placebo, with a statistically significant difference. In contrast, Cohen et al. [22] showed that tamoxifen users bleed slightly less days than placebo users; nonetheless, these differences were not statistically significant.

Three studies used tranexamic acid. Phupong et al. [32] administered two daily doses of 500 mg during 5 days, while Sørdal et al. [19] gave the same dosis three times a day. The third study, by Senthong [27], had a treatment plan of four daily doses of 250 mg for 5 days. In the studies whose objective was to stop the bleeding -Phupong et al. [32] and Senthong [27]-, it did so 7 days after starting the treatment, with a statistically significant result. In the case of Sørdal et al. [19], the mean of days with bleeding was 6, which is lower than for placebo and the comparator (mefenamic acid); however, this difference was not statistically significant.

Doxycycline appears in three trials [23,29,33], in two daily doses of 100 mg during 5 days. Weisberg et al. [29] also added two other plans: one with doxycycline and a daily dose of 20 mcg ethinylestradiol, and one with doxycycline and 25 mg mifepristone. The use of doxycycline by Abdel Aleem et al. [23] did not have a significant impact in stopping bleeding in the first 10 days; the relative risk to stop a bleeding episode in those 10 days was 0.88 (CI 0.64–1.21) for the group receiving the treatment, and there were no significant differences between the treatment group and the placebo group in the amount of days with bleeding and/or spotting in the three months after the treatment. Weisberg et al. [33], who compared the treatment at 90 days with the pre-treatment, did not observe differences in the days with bleeding or in the bleeding episodes as well. In the study by Weisberg et al. [29], doxycycline alone and combined with ethinylestradiol did not have a statistically significant effect on stopping bleeding; however, doxycycline with mifepristone was significantly more effective.

One of the most interesting findings of this review was the use of ulipristal acetate (UPA) by two trials [5,17]. Zigler and McNicholas [5] reported a statistically significant decrease in the days with bleeding after a daily dose of 15 mg during 7 days; however, Fava et al. [17] did not report something similar—there was a decrease in abnormal bleeding among LNG-IUD users who received a daily dose of 5 mg during 5 days in comparison with placebo users, but it was not statistically significant.

On the other hand, Warner et al. [21] observed a statistically significant decrease in the number of days with bleeding/spotting after administering a prophylactic plan of CDB-2914 (antiprogestin); however, the effect is lost in the third cycle, where the number of days with bleeding is lower in the placebo group.

With regard to the certainty of the evidence, it can be noted that, in most of the included studies, a favorable trend toward a reduction in bleeding days was observed in the intervention group. However, the magnitude and duration of the effect varied considerably between studies, as did the definitions of the primary outcome. The overall certainty of the evidence for this outcome can be considered moderate, due to limitations in blinding, small sample sizes, and risk of detection bias in some studies. However, the consistency in the direction of the effect favors active intervention over placebo in several comparisons.

Regarding treatment satisfaction, this was assessed heterogeneously, both in terms of time and method (visual analog scales, structured interviews, or dichotomous scales). In some cases, significantly higher levels of satisfaction were reported in the intervention groups, associating them with improved quality of life, performance in daily activities, and perceived control over bleeding. Therefore, the certainty of the evidence for this outcome is low, mainly due to imprecision and heterogeneity in the measurement of self-reported outcomes.

Most studies reported few serious adverse effects, the most common being nausea, headaches, fluid retention, and mood changes. These reactions were generally mild and did not lead to treatment discontinuation. Based on the consistency of these findings and the absence of serious effects, the certainty of the evidence on safety can be considered high to moderate.

This review adds new studies and findings to the contributions made by Abdel Aleem in 2013 [35], updating the scarce state-of-the-art knowledge on the management of bleeding caused by progestin-only contraceptives.

Lastly, while this review offers updated insights into managing bleeding associated with progestin-only contraceptives, certain limitations remain. Variations in study design, diagnostic criteria, outcome definitions, and follow-up periods make it difficult to directly compare results and limit the strength of the conclusions. Excluding studies where bleeding causes were classified under the PALM-COEIN system also narrows the scope of the findings, and many of the trials reviewed involved small samples, short follow-ups, or notable loss to follow-up. These factors lower the certainty of the evidence and leave unanswered questions about long-term effectiveness and safety. Moving forward, progress in this field will require larger clinical trials, standardized outcome measures, and extended follow-up to better guide practice and address the real needs of patients.

5. Conclusions

Treatments for bleeding are diverse—there are not only hormone, antagonist, or partial agonist therapies, but also others that decrease or avoid progestin-related bleeding. This review shows other non-hormonal alternatives for the management of bleeding caused by progestin, which supplement existing alternatives.

Combined oral contraceptives retain their higher effectiveness on stopping the bleeding over NSAIDs, while doxycycline is a good non-hormonal alternative. Another option, born from this review, is the use of selective progesterone receptor modulator and antifibrinolytics, which decreased the amount of days with bleeding. Besides, selective progesterone receptor modulators and NSAIDs showed a higher amount of days without bleeding in the cycle. Another combination treatment is mifepristone with ethinylestradiol, which stopped hemorrhages faster than doxycycline or placebo.

While several treatments are promising, their results must be first replicated by studies with better follow-ups—as most studies in this review had short-term interventions, longer treatments could not be standardized, leaving an opening for further research.

Regarding the quality of the studies, this review showed similar results to Abdel Aleem’s 2013 systematic review [35]. Differences are most likely due to the new five studies added and their better methodological quality; specifically, blinding and assessment of population losses.

On the topic of benefits for women -stopping uterine bleeding and its duration-, studies that reported patient satisfaction generally reported that women were satisfied with the management of the bleeding episodes. There were few reported adverse effects, with the most common being headaches and nausea.

Hopefully, these results will be useful for health professionals and midwives’ work on contraceptive management.