Medical Therapy Versus Radiofrequency Endometrial Ablation in the Initial Treatment of Menorrhagia: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Medical Therapy Versus Radiofrequency Endometrial Ablation in the Initial Treatment of Menorrhagia (NCT01165307), a randomized phase 4 study. The primary outcome is Menstrual Blood Loss (MBL) as Measured by Pictorial Blood Loss Assessment Chart (PBLAC)..

2. Clinical Question

The clinical question concerns two different initial-treatment strategies for menorrhagia: a medical approach and a device-based procedure. The registry describes a randomized comparison, so assignment to the treatment groups was intended to be determined by random allocation rather than by participant or clinician choice. The study was open label, meaning masking was not used.

The primary outcome was a score based on menstrual products used and the amount of staining, saturation, or clotting recorded during a menstrual cycle. The reported group summaries are medians with interquartile ranges. At 12 months, the medical-therapy group had a median PBLAC score of 15, with an interquartile range from 0 to 131. The radiofrequency-ablation group had a median of 0, with an interquartile range from 0 to 4. The registry reports a Wilcoxon (Mann-Whitney) comparison with p = 0.003.

This comparison addresses the posted endpoint and follow-up time, not every possible consequence of either treatment. The study also listed quality-of-life, blood, and cost outcomes, but their presence in the registry does not establish that a particular result was favorable or clinically important. Those interpretations depend on their own posted results, which are not specified in the primary-outcome details.

3. Trial Design

The registry describes a prospective randomized trial with a parallel-group design and no masking. It lists the study as phase 4 and completed, with enrollment of 77. The recorded study start was 2009-08 and primary completion was 2015-09. These dates describe the registry record; they do not establish the timing of each participant’s treatment or follow-up.

Participants were assigned to medical therapy or radiofrequency endometrial ablation. The medical-therapy arm used monthly packets of estradiol 30mcg / levonorgestrel 150mcg monophasic oral contraceptive pills. Participants unable to tolerate or unwilling to take those pills were to receive naproxen sodium pills. This means the medical arm could include more than one medication regimen; the posted group result is not a separate comparison of those regimens.

The procedure arm used NovaSure radiofrequency endometrial ablation. The registry description says the procedure was to take place within 4 weeks of randomization and at any time during the menstrual cycle, without endometrial pre-treatment. It also mentions suction curettage for endometrial thinning in 50% of cases included in the description. Further procedural details are not reported in the registry record available for this summary.

The primary outcome was measured at 12 months following initial treatment. The analysis population was restricted to participants who completed the 12-month visit: 19 in the medical-therapy group and 31 in the ablation group. Those analysis counts are not the same thing as the enrollment figure, and the registry entries do not explain the reasons participants were not included in the primary analysis.

The registry lists secondary outcomes covering quality of life, hemoglobin, ferritin, and direct and indirect medical costs. It does not report a full statistical analysis plan, details of allocation concealment, or a sample-size justification. The planning assumptions described for the separate sample-size illustration are not evidence of the trial’s original design rationale.

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: 77 subjects were initially consented, but 10 subjects were not randomized because 4 declined participation after consent, 4 reported spontaneous improvement during initial evaluation, 1 had an abnormal Pap smear indicating treatment, and 1 had a polypoid mass on hysteroscopy.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Only subjects who completed the 12 month visit were included in the analysis.

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 Menstrual Blood Loss (MBL) as Measured by Pictorial Blood Loss Assessment Chart (PBLAC)., measured over Measured at 12 months following initial treatment.

The primary endpoint was menstrual blood loss measured at 12 months after initial treatment using the Pictorial Blood Loss Assessment Chart, or PBLAC. The chart assigns scores according to the staining or saturation of tampons and sanitary napkins and the size of clots. The total score is formed by adding the scores recorded for products used during a menstrual cycle.

The registry describes a PBLAC score of 100 or greater as abnormal bleeding, corresponding to menorrhagia defined as greater than 80 mL of menstrual blood loss. It describes a score of 75 or less as normal bleeding and a score of 0 as amenorrhea, or absence of menstruation. These thresholds give clinical context to the scale, but the primary group comparison is based on the reported score distributions rather than a separately reported count above a threshold.

A score is useful because it provides a structured way to record menstrual bleeding using a pictorial tool, rather than relying only on a general description such as light or heavy bleeding. It is a scored measure, not a direct laboratory measurement of blood volume. The endpoint is reported as a median and interquartile range for each group. Only participants completing the 12-month visit were included in the analysis, as stated in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary result combines descriptive summaries with a rank-based group comparison. Each method answers a different question: the medians and interquartile ranges describe the observed scores, while the Wilcoxon (Mann-Whitney) test evaluates evidence of a difference between the groups’ outcome distributions.

What does a median tell us?

For an ordered set of observations, the median is the middle value. It is less affected by very high or very low observations than an arithmetic mean can be, which can make it useful when scores are skewed or contain extreme values. The registry reports a median PBLAC score of 15 for medical therapy and 0 for radiofrequency ablation. These are summaries of the analyzed participants; they do not imply that each person’s score was close to the group median.

The interquartile range runs from the first quartile to the third quartile, covering the central half of the observed distribution. The reported ranges are 0 to 131 for medical therapy and 0 to 4 for radiofrequency ablation. An interquartile range is a measure of spread, not a confidence interval. It describes variation among observed scores and does not, by itself, quantify uncertainty in the estimated median or in a between-group effect.

Why use a Wilcoxon (Mann-Whitney) test?

The Wilcoxon (Mann-Whitney) test is a nonparametric method that compares the ranks of observations across two independent groups. Rather than requiring a normal distribution of the measured outcome, it assesses whether values in one group tend to rank higher or lower than values in the other. This can be appropriate for outcomes that are skewed or otherwise unsuitable for a simple normal-theory comparison.

A common shorthand is to describe this test as a comparison of medians, but that description needs care. The test is fundamentally about ranks or distributions. It can be interpreted as a location or median comparison under additional conditions, such as distributions with similar shapes. If group distributions differ in spread or shape, the test result is not simply a test of whether the two medians differ. The registry reports the test method and p-value, but does not provide further distributional diagnostics or a detailed model.

What does p = 0.003 mean?

A p-value measures how unusual data at least as incompatible with the null hypothesis as the observed data would be, assuming the null hypothesis and the test’s assumptions. The reported p = 0.003 is evidence against the null hypothesis of no rank-based group difference. It does not give the probability that the null hypothesis is true. It does not measure the magnitude of the difference, the probability that an individual will have a particular score, or the clinical value of either treatment.

The registry does not report a confidence interval for the primary between-group comparison. A confidence interval would express uncertainty around an estimated effect and show which effect sizes remain compatible with the data under the chosen method. Without such an interval or a reported effect estimate, the p-value and group summaries provide only a partial account of precision. The group medians and interquartile ranges describe the observed data, but they do not substitute for an interval around a comparative effect.

Why does over-dispersion matter in general?

In general, over-dispersion means that observed outcome variability is greater than a statistical model expects. If a model assumes less variability than is present, its standard errors may be too small and its confidence intervals too narrow, making evidence appear more precise than it is. Rank-based methods avoid some distributional assumptions used by particular mean-based models, but they do not make differences in spread irrelevant. Distribution shape and variability still matter when interpreting what a rank test means.

The posted primary analysis is a Wilcoxon (Mann-Whitney) test, not a regression model with a reported dispersion parameter. The registry does not report a model-based assessment of over-dispersion for this endpoint, so it would not be appropriate to claim that such a check was performed or that a correction was applied.

What do the analysis counts and planning assumptions mean?

The primary analysis included participants who completed the 12-month visit: 19 assigned to medical therapy and 31 assigned to ablation. A completer analysis describes those with the required follow-up, but can be vulnerable to bias if the likelihood of completing the visit is related to treatment or outcome. The registry record does not report a method for correcting for missing outcomes, or whether other analysis populations were used for the primary result.

The separate sample-size illustration uses planning assumptions of a two-sided significance level of 0.05, equal allocation, and a normal approximation, with effects defined as a stated fraction of the observed effect. Such calculations are planning illustrations, not observed power and not evidence of the original trial’s sample-size rationale. The registry does not report that rationale, and no sample-size illustration should be read as a result measured in trial participants.

10. Analyses Reported by the Sponsor

The registry reports the sponsor’s primary comparison as a Wilcoxon (Mann-Whitney) analysis, with p = 0.003. The posted group summaries are a median PBLAC score of 15 (interquartile range 0 to 131) for medical therapy and 0 (interquartile range 0 to 4) for radiofrequency ablation. Together, these describe the observed 12-month scores and the reported rank-based comparison.

No recalculated comparative effect estimate, confidence interval, or recalculated p-value is stated in the registry record available for this analysis. It is therefore not possible to describe a numerical agreement between a recalculation and the sponsor’s test from these registry entries. The registry does not flag a discrepancy. The sponsor’s reported analysis remains the reported inferential result; it should not be replaced with an unreported calculation.

Even where a simple recalculation is available from summary statistics, it is generally an unadjusted analysis and may not reproduce the original method or account for participant-level data, ties, or other analytic choices. General caution of this kind does not establish that a recalculation was performed for this comparison. The registry record does not report additional details that would allow the sponsor’s test to be independently reconstructed.

11. Multiplicity, Missing Data and Analysis Populations

The registry lists secondary outcomes covering the Short Form-12 health survey, the Menorrhagia Multi-Attribute Scale, hemoglobin, ferritin, and direct and indirect medical costs. It records 11 secondary outcomes and 12 outcome measures overall. When a study examines multiple outcomes or multiple comparisons, the chance of at least one apparently unusual result can increase. Statistical approaches can account for multiplicity, but the registry record does not report a multiplicity adjustment, a prespecified hierarchy of outcomes, or a complete plan for selecting which results to emphasize.

The primary endpoint is identified as menstrual blood loss measured by PBLAC at 12 months. The record reports one primary comparison using the Wilcoxon (Mann-Whitney) method. It does not report interim analyses or stopping rules, and those details should not be inferred from the completed status or the posted p-value. The statistical analysis plan is not reported in the registry record.

The primary analysis population consisted only of participants who completed the 12-month visit. The registry does not report why some enrolled participants were not included in that outcome analysis, whether missing outcomes differed between treatment groups, or whether outcome values were imputed. It also does not report sensitivity analyses assessing the effect of missing data. These omissions matter because a completer-only result may not represent the experience of all randomized participants if completion is associated with treatment or outcome.

The registry entries do not specify whether the primary test followed an intention-to-treat principle, whether participants were analyzed according to assigned group, or whether any per-protocol analysis was also conducted. Random allocation is a feature of the design; it does not by itself answer how all post-randomization exclusions or missing observations were handled. Without those details, interpretation should stay anchored to the stated analysis population and the posted result.

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 0% reporting threshold.

Most frequent non-serious adverse event terms

The registry reports no serious adverse events affecting participants in either group: 0 among 33 at risk in medical therapy and 0 among 34 at risk in radiofrequency ablation. It reports other adverse events affecting 12 among 33 at risk in medical therapy and 1 among 34 at risk in radiofrequency ablation. These are reported event counts, not a comprehensive description of event types or severity.

Deaths are not reported in the registry record. The list of top serious adverse event terms is empty, but that should not be interpreted as a detailed account of all possible events. The serious-event counts and the “other” event counts are separate categories; the latter should not be relabeled as serious events.

Safety counts need context. The groups may differ in follow-up exposure or event ascertainment, and the record does not provide details on event definitions, timing, relatedness, or missing safety information. The trial is listed as unmasked, which is relevant context for how outcomes may have been observed or reported, but does not establish that reporting was biased. These figures describe what the registry reports and are not, by themselves, a complete comparative safety analysis.

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

The result shows that, among participants completing the 12-month visit, the observed PBLAC score distributions differed between the medical-therapy and radiofrequency-ablation groups. The ablation group had a reported median score of 0 and an interquartile range of 0 to 4; the medical-therapy group had a median of 15 and an interquartile range of 0 to 131. The registry’s Wilcoxon (Mann-Whitney) comparison returned p = 0.003.

This supports a statistical distinction between the observed group distributions under the test’s assumptions. It does not establish that every participant in one group had a lower score than every participant in the other. The medians and interquartile ranges show that individual outcomes varied, particularly in the medical-therapy group. The result also does not tell how a particular participant’s score would change under a different treatment assignment.

The p-value is not a measure of clinical importance. The registry does not report a confidence interval or a comparative effect estimate for this primary test, so the precision of the between-group difference cannot be fully assessed from the p-value alone. The score thresholds described in the record provide context for interpreting PBLAC values, but the primary comparison is not reported as a comparison of the proportion of participants exceeding a threshold.

The analysis was limited to 12-month visit completers. The record does not explain the missing outcomes or show whether the result changes under other assumptions about them. This makes the analysis population an important boundary on the conclusion. It also means that the result should not be presented as a complete account of outcomes for all randomized participants.

Finally, this primary result does not answer every question raised by the study. It does not by itself establish comparative effects on quality of life, hemoglobin, ferritin, direct or indirect costs, or adverse events. Nor does it establish a broader treatment recommendation. Those questions require their own results and careful consideration of the study population, follow-up, uncertainty, and analysis methods.

14. Why This Trial Matters Statistically

This trial is statistically informative because it compares two initial-treatment strategies using random allocation and a patient-relevant measure of menstrual blood loss. The endpoint is a scored measure with a bounded lower value and a potentially wide spread, reflected in the difference between the reported interquartile ranges. Describing the data with medians and interquartile ranges makes the distributional summaries more visible than a mean alone would.

The trial also illustrates why a small p-value and an effect estimate are not interchangeable. The Wilcoxon (Mann-Whitney) result is reported as p = 0.003, while the record does not give a confidence interval for a comparative effect. The p-value provides evidence about compatibility with a null hypothesis; it does not quantify how precisely the treatment contrast is estimated or how meaningful it is to participants.

The analysis population is another central statistical issue. Restricting the primary comparison to participants who completed the 12-month visit leaves open questions about post-randomization missingness and how it may affect interpretation. Randomization helps make treatment groups comparable at assignment, but missing outcome data can compromise that advantage if completion relates to outcome or treatment.

Finally, the registry lists outcomes beyond bleeding, including quality of life, blood measures, and costs. This range emphasizes the distinction between a primary statistical result and a full evaluation of a treatment strategy. Multiple outcomes raise questions about prespecification and multiplicity, while the absent analysis-plan details limit the ability to assess those choices. The registry result is most useful when its endpoint, population, uncertainty, and reporting limits are kept together.

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.

The most direct limitation in interpreting the primary result is that it includes only participants who completed the 12-month visit. The registry does not report why the remaining enrolled participants were not part of the primary analysis, whether noncompletion differed by group, or whether missing outcomes were handled with imputation or sensitivity analyses. The observed comparison therefore describes completers rather than necessarily all randomized participants.

The reported primary summaries do not include a confidence interval or comparative effect estimate. The p-value gives evidence from a rank-based test but does not show the precision of a treatment contrast. The wide upper end of the medical-therapy interquartile range also indicates substantial spread in that group’s observed scores, so a single median cannot capture all participant experiences.

The record identifies a randomized, parallel, unmasked design, but does not report further details about allocation concealment, the full statistical analysis plan, or interim analyses. The medical-therapy strategy also allowed naproxen sodium for participants unable to tolerate or unwilling to take oral contraceptive pills. Consequently, the group comparison is between assigned treatment strategies as described, not a clean comparison of one uniform medication against the procedure.

The registry lists secondary outcomes but does not report a multiplicity plan in the record. It also does not report the reasons for the observed analysis counts or a full accounting of outcome availability. These gaps limit assessment of selective reporting and the robustness of results across analysis populations.

Safety reporting is similarly limited: counts are available for serious and other adverse events, but event definitions, timing, relatedness, and deaths are not fully reported. The record’s primary result should therefore be interpreted as an endpoint-specific comparison, not as a complete assessment of clinical outcomes, costs, or safety.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Medical Therapy Versus Radiofrequency Endometrial Ablation in the Initial Treatment of Menorrhagia — NCT01165307.
• Registry record last updated 2017-01-20; status last verified 2017-01.
• Results first posted 2016-11-15.
• 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

NCT01165307 is a completed phase 4 randomized, parallel-group trial comparing medical therapy with NovaSure radiofrequency endometrial ablation for initial treatment of menorrhagia. The trial was unmasked and recorded enrollment of 77.

The primary endpoint was PBLAC menstrual blood loss at 12 months. The analysis included participants who completed that visit: 19 in the medical-therapy group and 31 in the ablation group. Their reported medians were 15 and 0, with interquartile ranges of 0 to 131 and 0 to 4, respectively. The registry reports a Wilcoxon (Mann-Whitney) comparison with p = 0.003.

The registry also lists quality-of-life, hemoglobin, ferritin, and cost outcomes. Further details on missing outcomes, multiplicity, and the statistical analysis plan are not reported in the registry record.