Treatment of vaginal bleeding irregularities induced by progestin-only contraceptives

Combined estrogen–progestin contraceptive pills can be used to treat bleeding irregularities in users of progestin-only contraceptive implants. The effectiveness of combined pills for treatment of bleeding irregularities in DMPA users has not been evaluated. Ibuprofen or mefenamic acid may also be used as short-term treatments, but use of mifepristone is not recommended.

RHL Commentary by Brache V


This 2007 Cochrane review was conducted with the objective of evaluating preventive and therapeutic interventions for treating bleeding irregularities associated with progestin-only contraceptive methods. Nineteen randomized controlled trials were included with a total of 2290 participants: 10 trials were conducted among Norplant users, six among depot-medroxyprogesterone acetate (DMPA) users, and three among progestin-only pill users. Most of the studies were designed to assess short-term therapeutic interventions for women experiencing bleeding irregularities, while a few were designed to test prophylactic interventions. Although the primary outcome studied in almost all trials was the effect of treatment on bleeding episodes, this outcome was reported in many different ways: cessation of bleeding, number of bleeding days during treatment, number of bleeding days after treatment, number of days of treatment required to stop bleeding, etc. The diversity in outcome measures and statistical reporting resulted in the fact that very few studies could be combined for this analysis.

Estrogen (used in various formulations) was the most frequently evaluated treatment (six trials), followed by non-steroidal anti-inflammatory drugs (NSAIDS) (four trials) and antiprogestins (four trials). Estrogen–progestin combinations, progestin alone, venotonic drugs, vitamin E, vitamin E combined with aspirin, and a selective estrogen receptor modulator were evaluated in one or two trials.

Although the results were inconsistent, estrogen seemed to have a beneficial effect in stopping a current bleeding episode in Norplant and DMPA users. Ethinyl-estradiol (50 µg/day) seemed to be more effective than the less potent 100 µg/day transdermal patch and estrone sulfate. However, ethinyl-estradiol was associated with gastric intolerance. Use of transdermal patch by DMPA users did not lead to an improvement in DMPA continuation rates. There were insufficient data to evaluate the effect of estrogen treatment in minipill users.

The combined contraceptive pill (250 µg levonorgestrel and 50 µg ethinyl-estradiol) was also effective in stopping a bleeding episode. Use of a progestin-only pill (levonorgestrel 30 µg twice a day) produced moderate improvements in long-term (one-year) bleeding patterns, but the study had potential sources of bias, and the way in which it was presented made it impossible to compare its data with data on single episodes of treatment.

With regard to NSAIDS, ibuprofen and mefenamic acid improved bleeding patterns in small clinical trials among Norplant users, but among DMPA users improvements did not reach statistical significance. Aspirin failed to have a beneficial effect in a large trial among Norplant users.

Antiprogestins, mainly mifepristone, have been used with some success in reducing bleeding; however, there is concern that by antagonizing the effect of the progestin in the contraceptive, the efficacy of the method may be potentially compromised.

A single trial that evaluated a selective estrogen receptor modulator, tamoxifen, reported improvement in irregular bleeding during treatment, unacceptable bleeding after treatment, and discontinuation rates of Norplant use during the study.

Vitamin E showed no beneficial effect on bleeding patterns in a larger trial among Norplant users, although a previous small pilot study, had reported reduction in number of bleeding days during treatment.

The only study that evaluated the effectiveness of a capillary-protecting venotonic drug (registered in France, containing extract of Ruscus aculeatus with ascorbic acid and hesperidin methyl chalcone) during noresthisterone acetate minipill use, showed a small reduction in the number of bleeding days during treatment.

An additional study (1) on the effect of mifepristone alone, mifepristone with ethinyl-estradiol, doxycycline and a placebo on irregular bleeding patterns during Implanon use was published after the publication of this Cochrane review. Mifepristone plus ethinyl-estradiol and doxycycline were more effective than placebo in terminating an episode of bleeding in women.

The results of this review are somewhat limited due to the lack of standardization in research methodology and diversity of treatment outcomes. However, it seems clear that some of the treatments reported were successful, mainly in stopping an ongoing bleeding episode, yet no medium- or long-term beneficial effects were found. Furthermore, many of the studies were conducted in a small number of women: those findings need to be reproduced in larger trials. In the few studies with long-term follow-up, improvement in continuation rates of the contraceptive were not seen.


2.1. Magnitude of the problem

Progestin-only long-acting methods are very popular in the developing world owing to their unique attributes: high effectiveness and safety; ease of use with discretion, long-acting action, and no need for daily action by the user. Over 30% of contraceptive users in South-East Asia and Southern Africa use injectables or implants; the corresponding figure in Latin America is about 6%. In some countries of Latin America, such as El Salvador, Nicaragua and Peru, over 20% of users use injectables or implants (1). The prevalence of injectables is far higher than that of implants.

Two factors that impede the achievement of even higher prevalence rates of progestin-only implants are the up front higher cost of implants and the need for training in their insertion and removal. But without doubt, bleeding irregularities associated with progestin-only long-acting methods (implants as well as injectables) are the foremost reason for dissatisfaction with these methods, as shown by the fact that bleeding disturbance is the main reported reason for discontinuation of the methods. Although bleeding irregularities with these methods are not associated with adverse health effects (generally there is no reduction in haemoglobin levels, and some studies even show improvement), they impact the quality of life of women adversely, interfering with their sexual lives and, in some countries, with religious activities, causing discomfort and increased expenses related continuous sanitary protection.

Nonetheless, the continuation rates for these methods have been consistently higher in almost all clinical trials than those observed with other contraceptive methods. In most of these settings, counselling prior to initiation of use has been the key factor for higher degrees of satisfaction with the contraceptive methods. Women who are informed about these irregularities prior to use, tolerate them better, and women who know beforehand that they will not be able to deal with bleeding irregularities, choose some other form of contraception. For many women, the benefits of these methods outweigh by far the associated disadvantages. Yet, there is no question, that reduction in bleeding irregularities would have a dramatic impact on the acceptability of these methods.

Unfortunately, much of the data presented in this review gives little hope that any of the tested interventions will bring significant improvements in disturbed bleeding patterns in users of progestin-only methods. The problem is that probably the most valued attribute of these long-acting methods is that they do not require a daily action by the user; many women are not willing to have to take additional medication (on a daily basis) to reduce the bleeding. The experience at our centre has been that we have had difficulty in enrolling implant users with bleeding irregularities in our studies of therapeutic interventions that require daily intake of another pill. Women prefer to discontinue use of the method rather than take additional medication over a prolonged period of time.

2.2. Applicability of the results

The studies included in this review were conducted in a diversity of regions of the world, and the results were quite similar throughout the regions. However, what may vary from region to region is the tolerability of bleeding disturbances, and hence the perceived need for treatment. Furthermore, it has been well established, that bleeding patterns also vary from women to women and with duration of use of the method. Ethnic differences in the metabolism of steroids have been reported. In Norplant users, for instance, bleeding improves with time as levonorgestrel levels slowly decrease and cyclic ovarian function is re-established. On the other hand, in the case of DMPA, irregular bleeding decreases over time but amenorrhea becomes more frequent. Weight has also been correlated with bleeding patterns: amenorrhea is more frequent in implant users with lower body mass index due to a greater exposure to the progestin and hence greater endometrial suppression.

2.3. Implementation of the intervention

The review found that the evaluated treatments yielded limited benefits, and some treatments even caused frequent, albeit mild, side-effects. For implant users, it may be possible to provide combined estrogen–progestin contraceptive pills without making any organizational changes in the family planning clinic as such pills are widely available. However, combined pills have not been evaluated for bleeding disturbances in DMPA users.

Similarly short-term use of ibuprofen or mefenamic acid may be easy to implement as both these drugs are available for sale over the counter.

I would not recommend the use of mifepristone until the issue of whether it compromises contraceptive effectiveness is resolved. Furthermore, mifepristone is not readily available in most settings

As bleeding is the main complaint of users of long-acting progestin-only contraceptives, it is likely that any treatment that offers relief by stopping or reducing the duration of bleeding will be welcome. However, stronger evidence of their effectiveness is needed.


A large multicenter randomized controlled trial comparing the most promising treatments is needed. Combined estrogen–progestin treatment (with a low dose of ethinyl-estradiol), anti-inflammatory drugs, tamoxifen, doxycycline (although not included in the Cochrane review) and placebo should be compared for their effectiveness in stopping prolonged bleeding, which is the most disturbing problem. The outcomes of greatest significance to the users will be: number of days between initiation of treatment and cessation of bleeding and the non-bleeding interval after starting treatment. The effects of treatment over continuation of use, side-effects and subject satisfaction with the treatment are other important outcomes to be measured.

I would not propose to include 50 µg ethinyl-estradiol orally in the comparative trial, in spite of this drug being the subject of the largest number of publications according to the Cochrane review, because it is an excessively large dose, which is not as readily available as other alternatives. The estrogen–progestin combinations evaluated in the Cochrane review also used high-dose combined oral contraceptives (50 µg ethinyl-estradiol). Likewise, these pills are no longer available, therefore, low-dose ethinyl-estradiol combined pills should be used.

I believe that only short-term treatments to control episodes of prolonged bleeding should be the priority, because women who use long-acting methods choose them because of they dislike taking pills on a daily basis. They may be willing to accept short periods of medication for their irregular bleeding, but regimens that require long-term use of pills will very likely not be accepted by these women.

Treatment of prolonged bleeding is more relevant for implant users than injectable users because early discontinuation of use of an implant means the loss of a what is expected to be a long lasting investment for the woman and for the family planning programme. In addition, it is known that bleeding disturbances decline over time. This means that there is a good chance that there will not be need for repeated courses of treatment. Thus, any large randomized comparative study should be done among implant users.


  • Weisberg E, Hickey M, Pamer D, O’Connor V, Salamonsen LA, Findlay JK et al. A pilot study to assess the effect of three short-term treatments on frequent and/or prolonged bleeding compared to placebo in women using Implanon. Hum Reprod 2006; Jan;21(1):295-302; DOI:10.1093/humrep/dei273.

This document should be cited as: Brache V. Treatment of vaginal bleeding irregularities induced by progestin-only contraceptives: RHL commentary (last revised: 11 September 2007). The WHO Reproductive Health Library; Geneva: World Health Organization.


About the author