Abstract

To address limitations that exist with existing definitions of menstrual bleeding changes that occur with contraceptive methods, we assembled a panel to develop new recommendations for standardization of bleeding data analyses associated with contraceptive use to better inform users, clinicians, investigators, pharmaceutical companies, and regulatory agencies. We propose three criteria for assessing bleeding outcomes: pattern, flow, and duration. The descriptors within each criterion depend on whether the contraceptive is designed to result in a predictable or unpredictable bleeding pattern. Predictable pattern outcomes quantify days of scheduled, unscheduled and no bleeding, while unpredictable pattern outcomes assess frequency. Flow is quantified based on patient comparisons to their typical flow when not using contraception, with spotting representing no menstrual products use. Duration of a prolonged bleeding and/or spotting episode is more than 7 days. Studies should assess bleeding characteristics for a minimum of 12 months for 21/7, 24/4, extended cycle or continuous regimens, two years for injectables, and the full duration of use for long-acting contraceptives. Describing pattern, flow and duration as independent categories allows a fuller understanding of the bleeding outcomes and better future assessments of acceptability and continuation. Standardization of outcomes permits better comparison between studies and data synthesis; standardization will also improve the ability of clinicians and patients to understand differences between products.

Keywords1. IntroductionAlthough contraceptives are approved by regulatory agencies based primarily on efficacy and safety, secondary outcomes, particularly changes in bleeding, influence contraceptive choice, user satisfaction, and continuation. Some societies, cultures, and individuals are less prone to accept menstrual changes; thus, bleeding complaints are one of the most common reasons for discontinuation [1Oral contraceptive discontinuation: a prospective evaluation of frequency and reasons., 2Long-acting injectable contraception with depot medroxyprogesterone acetate., 3Grunloh DS Casner T Secura GM Peipert JF Madden T. Characteristics associated with discontinuation of long-acting reversible contraception within the first 6 months of use., 4Polis CB Hussain R Berry A. There might be blood: a scoping review on women's responses to contraceptive-induced menstrual bleeding changes., 5Cohen R Sheeder J Teal SB. Predictors of discontinuation of long-acting reversible contraception before 30 months of use by adolescents and young women.]. Decrease in flow has led to wide acceptance of certain methods, such as the levonorgestrel 52 mg intrauterine device [[6]Kaunitz AM Bissonnette F Monteiro I Lukkari-Lax E Muysers C Jensen JT. Levonorgestrel-releasing intrauterine system or medroxyprogesterone for heavy menstrual bleeding: a randomized controlled trial.,[7]Mawet M Nollevaux F Nizet D Nizet D Wijzen F Gordenne V et al.Impact of a new levonorgestrel intrauterine system, Levosert(®), on heavy menstrual bleeding: results of a one-year randomised controlled trial.], yet less flow or amenorrhea may be poorly accepted, especially if unexpected [[8]Sivin I el Mahgoub S McCarthy T Mishell Jr, DR Shoupe D Alvarez F et al.Long-term contraception with the levonorgestrel 20 mcg/day (LNg 20) and the copper T 380Ag intrauterine devices: a five-year randomized study.,[9]Rowe P Farley T Peregoudov A Piaggio G Boccard S Landoulsi S et al.Safety and efficacy in parous women of a 52-mg levonorgestrel-medicated intrauterine device: a 7-year randomized comparative study with the TCu380A.].In 1986, the World Health Organization (WHO) proposed standardized terminology for and analysis of contraceptive-induced bleeding changes, known as the Belsey criteria [[10]Belsey EM Machin D d'Arcangues C. The analysis of vaginal bleeding patterns induced by fertility regulating methods. World Health Organization Special Programme of Research, Development and Research Training in Human Reproduction.]. These criteria used 90-day segments (called “reference periods”) as first suggested by Rodriguez et al. in 1976 [[11]Rodriguez G Faundes-Latham A Atkinson LE. An approach to the analysis of menstrual patterns in the critical evaluation of contraceptives.], with the goal that this time frame would be long enough to allow an overall description of the flow and frequency of bleeding episodes (Table 1). The definitions described flow as bleeding or spotting episodes that may or may not require menstrual products. Within each 90-day reference period, the authors recommended dividing the patterns of these episodes into amenorrhea or 1 of 4 specific categories defined as prolonged, frequent, infrequent, or irregular bleeding [[10]Belsey EM Machin D d'Arcangues C. The analysis of vaginal bleeding patterns induced by fertility regulating methods. World Health Organization Special Programme of Research, Development and Research Training in Human Reproduction.]. Belsey and colleagues [[12]The analysis of menstrual bleeding patterns: a review.] argued that 1 month or cycle is too short for “…assessments of events which occur at intervals often exceeding 30 days and does not allow reliable estimation of the average length and variability of events for the individual subjects.”Table 1World Health Organization Belsey definitions of bleeding patterns with contraceptive use [10]Belsey EM Machin D d'Arcangues C. The analysis of vaginal bleeding patterns induced by fertility regulating methods. World Health Organization Special Programme of Research, Development and Research Training in Human Reproduction.In 2005, a consensus panel convened to address apparent deficiencies in bleeding analyses in contraceptive studies [[13]Mishell Jr, DR Guillebaud J Westhoff C Nelson AL Kaunitz AM Trussell J et al.Combined hormonal contraceptive trials: variable data collection and bleeding assessment methodologies influence study outcomes and physician perception.]. The Belsey criteria were not being applied routinely in clinical trials. More importantly, because the criteria were not designed to describe planned cyclic bleeding, they did not apply well to combined hormonal contraceptives. Pharmaceutical companies and regulatory agencies generally did not follow these guidelines, and pharmaceutical companies often developed novel bleeding categories to maximize the positive aspects of their results. The 2005 panel provided recommendations based on best practice in trial design, data collection, and analysis focusing on combined hormonal methods, creating what is known today as the “Mishell” criteria [[14]Mishell Jr, DR Guillebaud J Westhoff C Nelson AL Kaunitz AM Trussell J et al.Recommendations for standardization of data collection and analysis of bleeding in combined hormone contraceptive trials.].More than thirty-five years since the Belsey criteria and 15 years since the Mishell criteria, we face the same difficulties. The Belsey criteria are commonly used for injectable and long-acting methods, and the Mishell criteria, when used, are applied to short-acting methods such as pills, rings, and patches [[10]Belsey EM Machin D d'Arcangues C. The analysis of vaginal bleeding patterns induced by fertility regulating methods. World Health Organization Special Programme of Research, Development and Research Training in Human Reproduction.−[14]Mishell Jr, DR Guillebaud J Westhoff C Nelson AL Kaunitz AM Trussell J et al.Recommendations for standardization of data collection and analysis of bleeding in combined hormone contraceptive trials.]. However, not all contemporary short-acting methods are designed to have predictable monthly bleeding. Whereas cyclic methods were most consistently used in a 21-day hormone/7-day hormone-free interval pattern decades ago, many formulations are now 24/4 or have extended use for 3 months or more. Several studies serve as examples of how contraceptive-induced menstrual bleeding changes are reported heterogeneously, even when evaluating the same product [15Baldaszti E Wimmer-Puchinger B Löschke K. Acceptability of the long-term contraceptive levonorgestrel-releasing intrauterine system (Mirena): a 3-year follow-up study., 16Bastianelli C Farris M Mirena Rapiti S. , 17Modesto W Bahamondes MV Bahamondes L. A randomized clinical trial of the effect of intensive versus non-intensive counselling on discontinuation rates due to bleeding disturbances of three long-acting reversible contraceptives.]. The lack of standardization among studies assessing these changes underscores that the current recommendations to describe the changes are inadequate [[4]Polis CB Hussain R Berry A. There might be blood: a scoping review on women's responses to contraceptive-induced menstrual bleeding changes.].

The authors have worked in this topic area over the last 2 or more decades. Three of the authors (MDC, CSV, DM) recognized many of the issues during a recent collaborative analysis and invited the fourth author to join the panel (CLW) based on expertise. The panel had initial conversations in May 2020 followed by a general outline of a proposal for funders to obtain support for time to develop the recommendations and publication-related costs. We obtained funding as unrestricted grants or gifts from multiple international pharmaceutical companies that manufacture contraceptives, expecting they would support use of the final output. We initiated meetings and writing of draft recommendations in May 2021. After follow-up meetings and correspondence to reach a consensus, we completed an initial set of recommendations in January 2022. After review by the Society of Family Planning Clinical Affairs Committee and peer review, this document represents our final recommendations.

Bleeding outcome assessments include a description of contraceptive-induced bleeding, acceptability, and therapeutic use in patients with heavy bleeding complaints. Our panel aimed to create contemporary recommendations specific to assessments of contraceptive-induced bleeding with the use of modern hormonal and nonhormonal contraceptives. We recognized that the Belsey and Mishell criteria were developed by and for researchers without consideration of how patients might view bleeding descriptors. While investigators have evaluated bleeding outcomes with contraceptive use and patient acceptance of these bleeding profiles, we found no research that focused on the descriptors. Both sets of previous criteria used bleeding definitions that failed to acknowledge a patient's subjective view; for example, one person may prefer to use a panty liner when spotting while someone else may not.

We developed these recommendations for contraceptive users, clinicians, investigators, pharmaceutical companies, and regulatory agencies. In addition, these recommendations are targeted to allow clinicians and patients to understand contraceptive-induced bleeding in a simpler way, to aid in counseling and comparison between products.

Our goal with these recommendations is to allow future studies to describe bleeding characteristics in a specific manner that facilitates clinicians’ and patients’ ability to make informed choices. Daily data collection is still needed for any analyses, so existing data sets from past studies will be amenable to most, but not all, of this proposed framework. Standardization of outcomes will allow better comparison between studies and data synthesis across studies in line with the Core Outcomes in Effectiveness Trials (COMET) initiative [[18]Williamson PR Altman DG Blazeby JM Clarke M Gargon E. The COMET (Core Outcome Measures in Effectiveness Trials) Initiative.,[19]

Williamson P, Clarke M. The COMET (Core Outcome Measures in Effectiveness Trials) Initiative: Its Role in Improving Cochrane Reviews. Cochrane Database Syst Rev. 2012;(5):ED000041. doi: 10.1002/14651858.ED000041.

]. Additionally, standardization will help inform communication among those working to manage bleeding complaints.2. Criteria considerationWe assume that contraceptive users do not want their bleeding outcomes to be worse than their norm. Some may be willing to accept a worse bleeding outcome as a trade-off for other positive features of a contraceptive, but we believe this is unlikely a preferred result. Typically, individuals want bleeding to be the same or better; for some, less or no bleeding may be considered an improvement. Importantly, everyone's baseline is different, and their individual assessment of baseline bleeding is not uniform. For example, amongst those with heavy menstrual bleeding (blood loss >80 mL based on alkaline hematin testing), 41% considered their flow moderate or scanty while 14% with blood loss [20]Hallberg L Högdahl AM Nilsson L Rybo G. Menstrual blood loss–a population study. Variation at different ages and attempts to define normality.].

Using consistent terminology across studies to describe bleeding outcomes is key. Confusion occurs when investigators or authors define terms differently. For decades, ‘bleeding patterns’ has commonly been used to describe all of the criteria related to contraceptive-induced bleeding. Here, we attempt to outline clearer terminology that differentiates 3 unique outcome criteria: pattern (referring specifically to when bleeding occurs or does not occur, i.e., frequency), flow, and duration.

The Belsey criteria include several terms used as pattern descriptors: amenorrhea, prolonged bleeding, frequent and infrequent bleeding, and irregular bleeding (Table 1). These pattern definitions are not mutually exclusive – e.g., a user could experience infrequent, prolonged episodes; yet, authors have erroneously analyzed bleeding outcomes as if the definitions are exclusive [[17]Modesto W Bahamondes MV Bahamondes L. A randomized clinical trial of the effect of intensive versus non-intensive counselling on discontinuation rates due to bleeding disturbances of three long-acting reversible contraceptives.]. Thus, the description needs to separate characteristics of the bleeding episode based on frequency (frequent, normal frequency, or infrequent) and duration (prolonged or not prolonged). The Belsey definition included irregular bleeding and did not define regular. However, because regularity (and by default, irregularity) of bleeding is a measure of frequency, the use of regular or irregular as a pattern is duplicative and is not included in these new recommendations.The Belsey definition of a prolonged bleeding episode (more than 14 days) seems excessively long, as users would not be likely to speak favorably about 14-day bleeding episodes, but negatively about 15-day episodes. A 2019 analysis of bleeding patterns with the etonogestrel implant found no difference in discontinuation rates with a definition of prolonged bleeding shortened to more than 7 days [[21]Mansour D Fraser IS Edelman A Vieira CS Kaunitz AM Korver T et al.Can initial vaginal bleeding patterns in etonogestrel implant users predict subsequent bleeding in the first 2 years of use?.]. In that study, investigators examined discontinuation rates among users with prolonged bleeding who had an otherwise favorable bleeding pattern (amenorrhea, infrequent bleeding, or normal frequency bleeding with no prolonged bleeding) in the first few months of use. Discontinuation rates did not differ with defining prolonged bleeding as more than 14 days (12/325 [3.7%]) or more than 7 days (5/211 [2.4%]), p = 0.46 (Fisher exact test). Of note, among implant users with an initial unfavorable pattern (prolonged or frequent bleeding), discontinuation rates with prolonged bleeding are defined as more than 14 days (27/212 [12.7%]) or more than 7 days (34/326 [10.4%]) were similar, p = 0.41. However, as the authors point out, the clinical trial data used in this assessment were more than 15 years old, and “more contemporary data may result in a different finding.” We agree that a shorter definition is more appropriate and aligned with real-life data.We evaluated other possible criteria published subsequent to the Belsey and Mishell guidelines that could be applied to new recommendations. The International Federation of Gynecology and Obstetrics (FIGO) created a classification system for causes of abnormal uterine bleeding in 2011 and updated the system in 2018 [[22]Munro MG Critchley HO Broder MS Fraser IS FIGO Working Group on Menstrual Disorders. FIGO classification system (PALM-COEIN) for causes of abnormal uterine bleeding in nongravid women of reproductive age.,[23]Munro MG Critchley HOD Fraser IS FIGO Menstrual Disorders Committee. The two FIGO systems for normal and abnormal uterine bleeding symptoms and classification of causes of abnormal uterine bleeding in the reproductive years: 2018 revisions.]. The authors defined abnormal uterine bleeding as “the symptom of disturbed menstrual bleeding” [[22]Munro MG Critchley HO Broder MS Fraser IS FIGO Working Group on Menstrual Disorders. FIGO classification system (PALM-COEIN) for causes of abnormal uterine bleeding in nongravid women of reproductive age.]. These criteria include frequency, regularity, duration, volume, and intermenstrual bleeding with the goal of defining normal and abnormal uterine bleeding in those not using contraception. The purpose of this FIGO classification is to consider when diagnostic testing or treatment may be indicated to identify pathologic conditions, which is different than our goal of defining bleeding outcomes for contraception users.Another example is an attempt to simplify the description of the bleeding patterns. In a reanalysis of data from the phase 3 studies of the etonogestrel implant, Mansour et al. [[20]Hallberg L Högdahl AM Nilsson L Rybo G. Menstrual blood loss–a population study. Variation at different ages and attempts to define normality.] classified the bleeding changes associated with this contraceptive into favorable (amenorrhea, infrequent bleeding, and normal frequency bleeding with no prolonged bleeding) or unfavorable (prolonged and/or frequent bleeding) based on the discontinuation rates associated with these patterns. Prolonged and/or frequent bleeding were associated with a higher etonogestrel implant discontinuation rate [[21]Mansour D Fraser IS Edelman A Vieira CS Kaunitz AM Korver T et al.Can initial vaginal bleeding patterns in etonogestrel implant users predict subsequent bleeding in the first 2 years of use?.−[24]Mansour D Korver T Marintcheva-Petrova M Fraser IS. The effects of Implanon on menstrual bleeding patterns.]. Which bleeding changes a contraceptive user will consider favorable/unfavorable or acceptable/not acceptable will depend on many factors such as age, culture, religion, life course, education, presence of symptoms associated with bleeding, and contraceptive options available [[4]Polis CB Hussain R Berry A. There might be blood: a scoping review on women's responses to contraceptive-induced menstrual bleeding changes.]. A scoping review on responses to contraception-induced bleeding changes concluded that substantial variability exists in how contraceptive users respond, including what they prefer and what they are willing to tolerate in exchange for the contraceptive benefits. Importantly, these responses are shaped by individual and social influences. For example, stated preferences for absent bleeding ranged from 0% to 65% across studies included in the review [[4]Polis CB Hussain R Berry A. There might be blood: a scoping review on women's responses to contraceptive-induced menstrual bleeding changes.]. Therefore, it is not possible to classify bleeding outcomes into positive/negative, acceptable/not acceptable, or favorable/unfavorable without knowing individual bleeding preferences during the contraceptive decision-making process. As such, we opted to not include this simplified system for categorizing outcomes.3. Bleeding criteria, descriptors, and definitionsThe recommended criteria include pattern, flow, and duration. The descriptors within each criterion depend on whether the contraceptive is designed to result in a predictable or unpredictable bleeding episodes. The expected pattern is a key to understanding bleeding outcomes and not whether the contraceptive is considered a short- or long-term method. The overall recommended descriptors for bleeding outcome criteria are reviewed for products designed to result in predictable episodes in Table 2 and unpredictable episodes in Table 3.

Table 2Recommended bleeding outcomes criteria and descriptors for contraceptive products with expected predictable bleeding episodes

Table 3Recommended bleeding outcomes criteria and descriptors for contraceptive products with expected unpredictable bleeding episodes

Products designed to create predictable bleeding, whether that bleeding is planned to occur monthly or less often, must be assessed in a similar manner to how a patient would evaluate the product. For example, a pill taken continuously for 11 to 12 weeks followed by a hormone-free week should have no bleeding or spotting during the 11 to 12 weeks of use and an expected bleed during the hormone-free week, because of the intention of the product design is to create predictable bleeding. However, a pill that is used daily without a hormone-free period is not expected to have predictable bleeding episodes and the evaluation methodology would be different. Product developers should be clear about the intended pattern to define bleeding outcomes appropriately and to ensure that clinicians and patients understand the intended pattern. Bleeding pattern evaluations of future contraceptive technologies should be assessed in this same manner, i.e., based on whether they are expected to have predictable or unpredictable bleeding episodes.

Flow or volume of bleeding is a separate criterion from pattern, as a person with any pattern can have no (absence of bleeding/spotting), light, moderate, or heavy flow. Pattern and flow have historically been described as separate entities for methods with expected predictable bleeding episodes.

We must also define a bleeding duration that would be considered prolonged, whether the contraceptive is intended to have predictable or unpredictable bleeding episodes. For products with a predictable pattern, this duration will also define if an expected withdrawal bleed is considered prolonged. The FIGO classification system for causes of abnormal uterine bleeding defined a normal duration of bleeding as 8 days or less based on historical studies that included hundreds of thousands of cycles throughout the entire reproductive lifespan to define normal menstrual length [[22]Munro MG Critchley HO Broder MS Fraser IS FIGO Working Group on Menstrual Disorders. FIGO classification system (PALM-COEIN) for causes of abnormal uterine bleeding in nongravid women of reproductive age.−[25]Fraser IS Critchley HO Munro MG Broder M. Can we achieve international agreement on terminologies and definitions used to describe abnormalities of menstrual bleeding?.]. The panel that created this system agreed that the duration could still be considered abnormal for a shorter duration if the individual considered the change unacceptable [[25]Fraser IS Critchley HO Munro MG Broder M. Can we achieve international agreement on terminologies and definitions used to describe abnormalities of menstrual bleeding?.]. We also considered a more recent study that evaluated 786 cycles in 130 participants 40 years and younger who reported regular menstrual cycles and were enrolling in barrier method contraceptive studies [[26]Creinin MD Keverline S Meyn LA. How regular is regular? An analysis of menstrual cycle regularity.]. The investigators reported an average length of menses of 5.2 ± 1.0 days; at 2 standard deviations, 95% of cycles will be 7 days duration or less. Accordingly, we considered both 7 and 8 days for our definition of prolonged bleeding and decided that 7 days would be more appropriate for defining contraceptive bleeding outcomes.All contraceptive products, whether they are designed to have a predictable or unpredictable bleeding pattern, will result in bleeding episodes, which are 1 or more consecutive days with bleeding or spotting. The Mishell criteria [[14]Mishell Jr, DR Guillebaud J Westhoff C Nelson AL Kaunitz AM Trussell J et al.Recommendations for standardization of data collection and analysis of bleeding in combined hormone contraceptive trials.] describe an episode as a bleeding or spotting event that is bordered by 2 full days without any bleeding or spotting, which we agree is a clear definition.3.1 Patterns

The bleeding pattern relates to when bleeding episodes occur. Currently available contraceptives expected to have predictable bleeding episodes include cyclic oral, transdermal, and transvaginal methods, combined injectable, and nonhormonal IUDs. Contraceptives that most commonly result in unpredictable bleeding episodes are progestin implants, hormonal IUDs, progestin-only injectable, and noncyclic progestin-only pills, patches, and rings.

3.1.1 Products with an expected predictable bleeding pattern

The reference period for evaluation of products with an expected predictable pattern is based on the product design. Currently available products are commonly designed to cause 28-day or 84-day patterns, although future products could be designed with other intended cycle durations.

Those using products with expected predictable patterns may experience scheduled and unscheduled bleeding episodes and, possibly, cycles without any bleeding or spotting. Expected bleeding should occur as determined by the method and cycle duration and described as detailed in Table 2.•Cyclic pills, patches, rings and combined injectable: Scheduled bleeding or spotting should start during the planned nonhormone use period and last no more than 7 days. Some studies have used tactics (online appendix Figure) to minimize how unscheduled bleeding is reported by including bleeding or spotting that started any time prior to the scheduled interval and continued into the interval as expected bleeding (“early withdrawal bleeding”), including some recent drug approval studies [[27]Masson S Franssen E Hilditch JR Powell MG. A clinical comparison of two triphasic oral contraceptives with levonorgestrel or norethindrone: a prospective, randomized, single-blind study.,[28]Creinin MD Westhoff CL Bouchard C Chen MJ Jensen JT Kaunitz AM et al.Estetrol-drospirenone combination oral contraceptive: North American phase 3 efficacy and safety results.]. Another tactic is defining bleeding or spotting that starts during the hormone-free interval but continues afterward for more than 7 days (“continued withdrawal bleeding”) as expected bleeding. Neither early nor continued withdrawal bleeding should be used in any bleeding descriptions or analysis plans as they do not meet expected bleeding for users, who would presume to bleed would start during the scheduled interval and not last more than a reasonable (7-day) duration. The proposed definition allows bleeding that starts during the expected interval and lasts no more than a total of 7 days to be considered as expected bleeding, thereby limiting the timing and duration of scheduled bleeding (Fig. 1).

Fig. 1Bleeding definitions for expected predictable bleeding pattern product using a 24-day hormone/4-d placebo pill example.* Scheduled bleeding can include any bleeding or spotting that starts during the scheduled bleeding interval and lasts no more than 7 days.

•

Nonhormonal IUDs: Scheduled bleeding is expected to occur monthly in those who experience monthly cycles and without a predictable pattern for those with previous irregular cycles. Most studies will only enroll participants with regular cycles so bleeding or spotting episodes should be reported as scheduled (monthly) and should not last for more than 7 days. All bleeding or spotting not starting during the expected interval would be reported as unscheduled bleeding. If a study includes participants with baseline unpredictable (irregular) cycles, then bleeding descriptions should be stratified to describe patterns experienced among those with baseline regular cycles and those without baseline regular cycles, with the latter being reported according to recommendations for products with an unpredictable pattern.

3.1.2 Products with an expected unpredictable bleeding patternProducts with an expected unpredictable bleeding pattern do not have scheduled bleeding episodes so the number of days of bleeding or spotting during a reference period should be used to standardize bleeding pattern description. We recognize that no ideal reference period duration exists; examples include the historical 90-day period defined by Belsey [[10]Belsey EM Machin D d'Arcangues C. The analysis of vaginal bleeding patterns induced by fertility regulating methods. World Health Organization Special Programme of Research, Development and Research Training in Human Reproduction.], 84 days (3, 28-day cycles), and 91 days (1 year [365 days] divided into approximately 4 equal time periods). Since these time frames are relatively similar, we see no benefit from changing from current recommendations to use a 90-day period, as maintaining the 90-day period will allow comparisons across studies. As such, we recommend that contraceptive bleeding descriptions with products that are expected to have an unpredictable bleeding pattern be