Study in Women With Idiopathic Menorrhagia to Determine the Reduction in Menstrual Blood Loss (MBL) After Treatment With the Levonorgestrel-releasing Intrauterine System (IUS): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Study in Women With Idiopathic Menorrhagia to Determine the Reduction in Menstrual Blood Loss (MBL) After Treatment With the Levonorgestrel-releasing Intrauterine System (IUS) (NCT00360490), a randomized phase 3 study. The primary outcome is The Change in Absolute Value From Baseline Menstrual Blood Loss (MBL) to the End-of-study MBL (Cycle 6).

2. Clinical Question

The study was designed to compare two active treatments over 6 cycles, rather than compare treatment with no intervention. Its primary outcome concerned menstrual blood loss, including intermenstrual bleeding as well as withdrawal bleeding. Baseline blood loss was a composite measured during screening cycles, while the end-of-study measure was taken during Cycle 6 of treatment.

The clinical question is therefore specific to change in measured blood loss over the treatment period in this study population. It is not a question about every possible effect of either treatment, nor does the primary endpoint alone establish how participants experienced treatment or how results apply to people outside the enrolled population. The registry record does not report additional population details beyond the condition and analysis descriptions.

3. Trial Design

This was a Phase 3, randomized, open-label, parallel-group, active-control study sponsored by Bayer. The official title describes a multicenter study. The registry lists no masking. The study status is completed; its recorded start was 2006-07 and its primary completion was 2008-06.

The planned comparison assigned participants to either the levonorgestrel intrauterine system or oral medroxyprogesterone acetate. The intrauterine system had an initial release rate of 20µg per 24 hours and was used for 6 cycles. MPA was taken orally at 10mg per tablet on 10 consecutive days of each cycle, also for 6 cycles. Both groups therefore had an active treatment, and the registry describes a parallel-group rather than crossover design.

Enrollment was 165. The primary analysis description identifies 160 available participants: 79 in the levonorgestrel group and 81 in the MPA group. The Full Analysis Set is described as containing 165 subjects. The registry’s primary-analysis population description also notes missing diary data for one participant in the levonorgestrel group and two in the MPA group, and states that two additional levonorgestrel participants were unavailable for analysis. The reason given for those two participants is incomplete in the registry record.

Randomization can help balance measured and unmeasured participant characteristics across groups, but it does not ensure that the groups are identical. Open-label treatment means participants and study personnel were not masked to assignment. The registry record does not report the randomization procedure, allocation concealment details, site count, or any additional procedures intended to reduce bias. These details should not be inferred from the study title or design labels.

The registry record does not report the sample-size justification or enough information to assess how the planned enrollment was selected. It also does not report an interim-analysis plan. Those omissions limit what can be concluded about planning and monitoring, but they do not change the definition of the posted endpoint or the reported comparison.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Before assignment: 807 subjects screened; 165 subjects randomized to either Levonorgestrel Intrauterine System (LNG IUS) or Medroxyprogesterone acetate (MPA) for treatment of idiopathic menorrhagia; 145 subjects completed study.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Among of the 165 Full Analysis Set (FAS) subjects, a total of 160 subjects (79 for LNG IUS and 81 for MPA) were available for the analysis. One subject in LNG IUS, and two subjects in MPA group were not available due to the missing of diary data, and 2 subjects in LNG IUS group were not available due to the invalid baseline data.

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

6. Primary Endpoint

The registry lists the primary outcome as The Change in Absolute Value From Baseline Menstrual Blood Loss (MBL) to the End-of-study MBL (Cycle 6), measured over Baseline and up to 6 months.

The primary endpoint was the change in absolute menstrual blood loss from baseline to end of study, measured in milliliters. The end-of-study measure was taken during Cycle 6. The registry describes baseline blood loss as the composite menstrual blood loss during each screening-phase cycle and says that each cycle’s measure included intermenstrual bleeding in addition to withdrawal bleeding.

A change from baseline provides a way to describe how the measured outcome at follow-up differs from the participant’s starting measurement. Negative values in the posted results indicate a decrease from baseline; more negative group medians describe larger decreases in the middle of the observed group distributions. The primary endpoint is reported as a median with a full range, rather than as a mean with a standard deviation. That choice describes the central observation using the middle value and also conveys the observed extremes through the range.

Menstrual blood loss was the direct outcome named in the study’s purpose, so it addresses the study’s stated efficacy question. It is nonetheless a measurement over specified cycles, not a complete account of all experiences, benefits, or harms. The registry record does not report the measurement instrument or additional validation information in the primary-outcome description.

The primary analysis included 160 participants, with 79 in the levonorgestrel group and 81 in the MPA group. The reported median change was -128.8 mL in the levonorgestrel group, with a full range from -393.6 to 1242.0, and -17.7 mL in the MPA group, with a full range from -271.5 to 78.6. These ranges indicate substantial variation in observed changes within each group. A group median alone does not show how every participant changed.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Statistical methods determine which feature of the data is compared and how uncertainty is assessed. The registry’s primary endpoint is a change in menstrual blood loss, summarized within each treatment group by its median and full range. The comparison between groups uses a Wilcoxon (Mann-Whitney) test. The posted secondary percentage-change outcomes use group means and standard deviations, with reported two-sided t-tests. These methods describe different summaries and should not be treated as interchangeable.

What does the median describe?

In general, the median is the middle observation after values are ordered. It is less influenced by extreme observations than a mean, which can make it useful when data are skewed or contain unusually large values. The registry reports the primary outcome as a median, but does not report a measure of spread around that median beyond the full range. A full range gives the lowest and highest observed values; it does not show how observations are distributed between those endpoints. It is also sensitive to extreme observations and is not an estimate of uncertainty in the median.

For the primary endpoint, the medians are -128.8 mL and -17.7 mL. The negative signs indicate decreases from baseline under the stated change measure. These summaries locate the centers of the observed distributions, but do not mean that all participants had changes close to the group median. The ranges, from -393.6 to 1242.0 in the levonorgestrel group and from -271.5 to 78.6 in the MPA group, show that individual changes extended well beyond the center of each group.

What does the Wilcoxon comparison test?

In general, the Wilcoxon (Mann-Whitney) test compares the relative ordering of observations between two independent groups. It is often described as a test of distributions or ranks. It does not automatically test a difference in medians in every setting: interpreting it specifically as a median comparison requires assumptions about the shapes of the group distributions. The registry gives the method and p-value but does not report additional details about assumptions, diagnostics, or the exact estimand used for the primary test.

The reported p-value is p < 0.001. In general, a p-value measures how unusual the observed data, or results more extreme, would be under a specified null model and test procedure. It is not the probability that the null hypothesis is true, nor does it measure the probability that the result arose by chance in an everyday sense. It also does not state whether an observed difference is important to participants. The registry record does not report a confidence interval for the primary group comparison, so the precision of that contrast is not directly expressed through an interval.

How should the secondary t-tests be read?

The secondary percentage-change results are summarized by means and standard deviations. In general, a mean is the arithmetic average and can be affected by unusually large or small observations. A standard deviation describes spread among individual observations around the mean; it is not a confidence interval for the mean. For the Cycle 6 percentage-change outcome, the posted group means are -70.75 and -21.54, with standard deviations 88.31 and 35.79. For Cycle 3, the means are -61.71 and -11.11, with standard deviations 41.76 and 42.46.

The registry reports two-sided t-tests for these outcomes, with p < 0.001 for each. A two-sided test allows evidence in either direction to count against the null hypothesis. The reported Cycle 6 estimate is -49.2 with a 95% confidence interval from -70.2 to -28.3; the Cycle 3 estimate is -50.6 with a 95% confidence interval from -63.8 to -37.4. The comparison estimates are reported using the levonorgestrel group relative to MPA. In general, a confidence interval describes a range of effect estimates compatible with the data and the method’s assumptions; it expresses precision, not the probability that the true effect lies within the particular interval.

Additional calculations for these secondary summaries are labeled as recalculated comparisons and are simple, unadjusted calculations from the posted summary statistics. Their contrast direction is MPA relative to the levonorgestrel group, so their signs differ from the sponsor-reported contrast direction. The recalculated estimates and intervals are close to the sponsor’s reported results after rounding, and the registry record does not flag a discrepancy. Such recalculations do not replace the sponsor’s analysis or account for unreported adjustments, model choices, or individual-level data.

Why can over-dispersion matter?

In general, over-dispersion means that the observed variability is greater than a model assumes. When a method assumes less variability than the data contain, uncertainty can be understated and statistical evidence can appear stronger than it should. The primary outcome has wide observed ranges, but a range alone cannot establish whether a particular model is over-dispersed. The primary test is rank-based, and the registry does not report a model-based variance assumption for it. For the secondary t-tests, means and standard deviations are posted, but the registry does not report checks of distributional assumptions or alternative analyses. These limits make it inappropriate to claim that over-dispersion was present or absent.

What do the planning illustrations mean?

General sample-size calculations estimate how many participants per arm would be needed under specified assumptions to detect a chosen effect with a chosen power. The posted planning illustrations use the observed secondary effects and assume a two-sided significance level of 0.05, equal allocation, and a normal approximation. Their rows consider planning for 100%, 75%, and 50% of the observed effect, with separate targets for 80% and 90% power.

These are planning illustrations, not the trial’s original design calculation and not observed power. Observed power is not a substitute for interpreting the effect estimate and its uncertainty after results are known. A calculation based on an observed effect may be particularly optimistic if that effect is imprecisely estimated. The registry record does not report the original sample-size justification or its assumptions. The illustrations apply to the stated secondary comparisons and should not be read as evidence that the trial was planned or powered in that way.

10. Analyses Reported by the Sponsor

The sponsor-reported primary analysis used a Wilcoxon (Mann-Whitney) method and reported p < 0.001. The registry does not give a sponsor-reported effect estimate or confidence interval for that primary comparison, so the posted group medians and ranges should be read as descriptive summaries rather than as a sponsor-reported interval estimate of the treatment contrast.

For the two posted secondary percentage-change outcomes, the sponsor reports two-sided t-tests. The Cycle 6 reported mean difference is -49.2, with a 95% confidence interval from -70.2 to -28.3 and p < 0.001. The Cycle 3 reported mean difference is -50.6, with a 95% confidence interval from -63.8 to -37.4 and p < 0.001. These reported estimates use the levonorgestrel group relative to MPA.

The unadjusted recalculations use the reverse contrast, MPA relative to the levonorgestrel group, which reverses the signs. For Cycle 6, the recalculated estimate is 49.21 with a 95% interval from 28.23 to 70.19; for Cycle 3, it is 50.6 with a 95% interval from 37.55 to 63.65. The recalculated p-values are each < 0.001. This sign change reflects the contrast direction, not a different substantive pattern. The recalculated estimates and intervals are close to the sponsor’s results after rounding, and the registry record does not flag a discrepancy. Where a discrepancy is flagged, the sponsor’s own analysis takes precedence; no such discrepancy is flagged for these comparisons.

These recalculations use posted summary statistics and are not a substitute for analysis of participant-level data. The registry record does not report whether the sponsor’s secondary analyses incorporated adjustments beyond the methods stated in the results.

11. Multiplicity, Missing Data and Analysis Populations

The registry record identifies a primary endpoint and lists multiple secondary outcomes, but does not report a statistical plan describing how these outcomes were ordered or how multiplicity was handled. When several hypotheses are tested, the chance of at least one apparently strong result can increase. In general, adjustment procedures can control that risk, while a prespecified outcome hierarchy can clarify which results are primary and which are supportive. Without those details, secondary p-values should not be treated as though they necessarily preserve a single overall error rate.

The primary endpoint analysis was available for 160 participants from the Full Analysis Set of 165: 79 in the levonorgestrel group and 81 in the MPA group. The registry description notes that one levonorgestrel participant and two MPA participants were unavailable because of missing diary data. It also notes that two additional levonorgestrel participants were unavailable, but the recorded explanation is incomplete. The registry record does not report a full account of missingness for each outcome or a missing-data sensitivity analysis.

Missing outcome data can affect interpretation if the availability of measurements is related to the outcome or differs systematically across groups. In general, an analysis restricted to participants with available measurements can be biased when the missingness process is informative. That possibility cannot be assessed from the registry summaries alone. The record does not report whether values were imputed, whether an analysis included all randomized participants, or whether alternative assumptions about missing values were examined.

The 160 participants available for the primary analysis are not the same number as the Full Analysis Set. That distinction matters when interpreting the result: the reported primary summaries describe those with available endpoint data, while the registry describes a larger analysis set. The record does not provide enough detail to determine how all participants were handled across the complete set of outcomes.

Other elements of the statistical plan are also not reported in the registry record, including the original sample-size justification, any interim analyses, and the intended multiplicity procedure. The planning illustrations based on observed secondary effects do not fill those gaps; they are not the trial’s design justification. These missing details limit assessment of the completeness and prespecification of the analyses, without changing the values reported for the primary endpoint.

12. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 2% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: endometriosis (1.2%).

The registry reports serious adverse events for 82 participants at risk in the levonorgestrel group and 83 at risk in the MPA group. One participant in the levonorgestrel group and zero participants in the MPA group were reported as affected by a serious adverse event. Endometriosis is listed as the leading serious-event term, at 1.2%. These figures describe events recorded during the study; they do not by themselves establish that treatment caused an event or that a between-group difference is meaningful.

The registry also reports other adverse events affecting 67 of 82 participants in the levonorgestrel group and 52 of 83 in the MPA group. These counts are separate from the serious-event counts and should not be conflated with them. The registry record does not provide enough detail in these summaries to assess event severity beyond the serious/non-serious grouping, timing, duration, or causal attribution for each event.

Deaths are not reported in the registry record. The listed serious-event counts are small, so they provide limited information about uncommon harms. In general, absence of a reported event in a study does not establish that an event cannot occur, and adverse-event summaries should be interpreted alongside completeness of follow-up and event definitions. The registry record does not report additional details that would support broader safety conclusions.

13. What the Result Does, and Does Not, Mean

The posted data show a larger decrease in median menstrual blood loss from baseline to Cycle 6 in the levonorgestrel group than in the MPA group. The primary comparison is reported as p < 0.001 using a Wilcoxon (Mann-Whitney) test. This supports evidence of a difference between the observed group distributions under that method. The group ranges are wide, however, and individual participant changes varied substantially.

The result does not mean that every participant in the levonorgestrel group had a decrease, or that every participant in the MPA group had a smaller decrease. Group medians summarize the center of the observed distributions and cannot represent every individual outcome. The primary result also does not quantify the precision of the between-group effect with a confidence interval; such an interval is not reported for the primary comparison.

The comparison is between two active treatments, over 6 cycles, among women with idiopathic menorrhagia who entered this study. It does not answer how either treatment compares with no treatment, and it should not be generalized automatically to other populations, treatment durations, or outcomes. The study was open-label, and the registry record does not report masking or additional procedures that would address the potential influence of knowing the assigned treatment.

The results concern measured menstrual blood loss. They do not alone describe all effects that matter to participants, nor do they establish the cause of any individual outcome. Secondary blood-loss results have a broadly similar direction, but the record does not report a multiplicity plan for interpreting the full set of outcomes. Unadjusted secondary results should therefore be treated as supportive and exploratory context rather than as multiple independent confirmations.

The safety summaries report one participant with a serious adverse event in the levonorgestrel group and none in the MPA group, with other adverse events also reported. These counts do not establish causation and offer limited information about uncommon events. The posted efficacy and safety data should be read as evidence from this particular trial and its recorded analyses, not as a complete account of benefits and harms.

14. Why This Trial Matters Statistically

This trial illustrates why statistical interpretation depends on the outcome summary as well as the test result. The primary endpoint is reported with medians and full ranges, and the comparison uses a rank-based method. That combination keeps attention on the distribution of observed changes rather than assuming that a mean difference alone captures the data. At the same time, wide ranges show why a central summary should not be mistaken for a uniform response across participants.

The secondary results provide a useful contrast in methods: percentage changes are summarized by means and standard deviations and analyzed with two-sided t-tests. The reported confidence intervals communicate precision for those secondary mean differences, while the primary comparison has no reported confidence interval. Readers can therefore see that statistical significance and precision are distinct considerations, and that the registry gives different levels of detail for different outcomes.

The record also demonstrates how the direction of a contrast affects its sign. The sponsor’s secondary estimates compare levonorgestrel with MPA, while the recalculations use the reverse order. A positive estimate in one presentation and a negative estimate in the other can describe the same between-group pattern when the contrast is reversed. Clear naming of the comparison direction is essential to avoid mistaking notation for disagreement.

Finally, the trial highlights limits in registry reporting. The primary-analysis population is smaller than the Full Analysis Set, the reason for some unavailable primary data is incomplete, and the registry record does not report a multiplicity plan, interim-analysis plan, or original sample-size justification. These details matter for assessing analysis completeness and planning, but they should not be guessed. Distinguishing reported results from missing design information is an important part of responsible interpretation.

15. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

Several limitations specific to this trial affect how broadly the posted findings can be interpreted. The primary endpoint is available for 160 participants, whereas the Full Analysis Set is described as 165. The registry notes missing diary data for participants in both groups and two additional levonorgestrel participants unavailable for the primary analysis, but the explanation for those two is incomplete. The record does not report imputation methods or sensitivity analyses addressing missing outcomes.

The primary outcome is summarized by medians and full ranges, but no confidence interval for the primary comparison is reported. A range communicates the observed extremes, not the precision of a group contrast. The wide ranges also show considerable variation in observed changes. The registry summaries do not provide participant-level distributions or additional analyses that would clarify how the extreme values affected interpretation.

The study was open-label, and the registry lists no masking. This design feature is especially relevant when outcomes involve participant-recorded diaries, although the record does not quantify any resulting measurement effect. The primary outcome description says that diary data contributed to availability, but the measurement instrument and procedures are not detailed in the registry record.

The registry record does not report the original sample-size justification, interim analyses, or a multiplicity procedure for the many outcomes. It also does not provide enough information to assess the detailed statistical plan, allocation procedures, or analysis assumptions. The recalculated planning illustrations are based on observed secondary effects and cannot establish the original design rationale or observed power.

Finally, the registry provides limited detail for interpreting safety events, including their timing and causal attribution. The primary findings address a defined population, endpoint, and 6-cycle treatment period; wider conclusions require evidence beyond these registry summaries.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Study in Women With Idiopathic Menorrhagia to Determine the Reduction in Menstrual Blood Loss (MBL) After Treatment With the Levonorgestrel-releasing Intrauterine System (IUS) — NCT00360490.
• Registry record last updated 2013-12-09; status last verified 2013-11.
• Results first posted 2009-11-19.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

19. Record Summary

NCT00360490 was a completed Phase 3 randomized, open-label, parallel-group active-control study sponsored by Bayer. It compared a levonorgestrel-releasing intrauterine system with oral medroxyprogesterone acetate in women with idiopathic menorrhagia. Both treatments were given for 6 cycles, with the system’s initial release rate at 20µg per 24 hours and MPA taken at 10mg per tablet on 10 consecutive days of each cycle.

The primary endpoint was absolute change in menstrual blood loss from baseline to Cycle 6. The medians were -128.8 mL in the levonorgestrel group and -17.7 mL in the MPA group, with a reported Wilcoxon (Mann-Whitney) result of p < 0.001. The primary analysis included 79 participants in the levonorgestrel group and 81 in the MPA group. Secondary percentage-change analyses also reported p < 0.001, with sponsor estimates and confidence intervals.

The registry record does not report a primary confidence interval, the original sample-size justification, an interim-analysis plan, or a multiplicity procedure. It also provides incomplete detail about some unavailable primary endpoint data. These omissions constrain assessment of precision and analysis planning; the posted results remain specific to the recorded population, outcomes, and treatment period.