Estradiol Vaginal Softgel Capsules in Treating Symptoms of Vulvar and Vaginal Atrophy in Postmenopausal Women: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Estradiol Vaginal Softgel Capsules in Treating Symptoms of Vulvar and Vaginal Atrophy in Postmenopausal Women (NCT02253173), a randomized phase 3 study. The primary outcome is Co-Primary Efficacy Endpoint - Vaginal Superficial Cells.

2. Clinical Question

The primary endpoint was a laboratory measure from vaginal cytologic smears, rather than a direct measure of symptoms. The registry describes the assessment as change from baseline to Week 12 and reports least-squares means for the arms. It identifies a mixed-model repeated-measures analysis for the endpoint and notes that the number analyzed depended on having both baseline and Week 12 data.

The record reports three separate comparisons with placebo, one for each estradiol dose. The posted arm summaries show least-squares mean changes of 17.5 for 4mcg, 16.72 for 10mcg, and 23.2 for 25mcg, compared with 5.63 for placebo. The recalculated unadjusted mean differences and confidence intervals quantify the contrasts in those summaries; the registry also reports its own mixed-model p-values.

The clinical question is limited to this endpoint, its specified time frame, and these randomized groups. The registry record does not report a primary symptom endpoint in the registry record used for this explanation, nor does this cell measure alone establish the size or importance of any change in symptoms. The results should therefore be read as evidence about the recorded cytologic outcome, not as a complete account of all outcomes relevant to participants.

3. Trial Design

The registry describes a Phase 3, randomized, triple-masked, parallel-group, placebo-controlled, multicenter trial. The official title identifies TX-004HR and says the study evaluated safety and efficacy in postmenopausal women with moderate to severe symptoms of vulvar and vaginal atrophy. The sponsor is TherapeuticsMD. Enrollment is listed as 764, and the record status is completed.

Four arms are listed: estradiol vaginal softgel capsules at 4mcg, 10mcg, and 25mcg, and a placebo vaginal softgel capsule. Random allocation and a parallel model mean the record describes participants being assigned to a group and followed in that group, rather than all participants receiving each intervention in sequence. Triple masking is the registry's masking classification; the record does not specify which roles were masked.

The listed study start is 2014-09 and the primary completion is 2015-10. The primary efficacy measure is the change from baseline to 12 Weeks in the percentage of vaginal superficial cells. The registry specifies a mixed-model repeated-measures approach and cautions that the number analyzed varied according to availability of both baseline and Week 12 data. Arm counts attached to efficacy summaries are not necessarily the same as the counts for safety summaries.

There are 48 outcome measures and 141 analyses in the record, with 43 secondary outcomes posted. These counts describe the scope of recorded outcomes and analyses; they do not disclose how the analysis hierarchy or multiplicity strategy was determined. The original sample-size justification, allocation ratio, interim-analysis plan, and details of the masking procedures are not reported in the registry record.

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.

Overall Study

Analysis population for the primary outcome, as defined in the registry: For the statistical test utilized (MMRM), the number of subjects analyzed varied based on if they had both Baseline and Week 12 data so it would not necessarily match the overall number of subjects in the MITT Population at Baseline.

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.

Population: The modified intent to treat (MITT) population was the primary efficacy population, defined as all ITT subjects who received the treatment to which they were randomized, had baseline values for all co-primary variables, and had at least one post-baseline value for any of the four co-primary variables at any visit.

6. Primary Endpoint

The registry lists the primary outcome as Co-Primary Efficacy Endpoint - Vaginal Superficial Cells, measured over Baseline and 12 Weeks.

The primary endpoint is the change from baseline to Week 12 in the percentage of vaginal superficial cells, measured by vaginal cytologic smear. Superficial cells are one component of the vaginal cell composition. The endpoint is expressed as a percentage, and the registry reports least-squares mean changes with standard errors for each arm.

In general, a cellular measure can serve as a biological or tissue-level outcome that is evaluated consistently across study groups. The registry record does not explain the rationale for selecting this endpoint or establish how a particular change corresponds to a change in symptoms. It should therefore be interpreted as a cytologic measure, not as a substitute for a patient-reported outcome.

The reported timeframe is baseline and 12 Weeks. The endpoint description specifies comparison with placebo. For the primary comparison, the recalculated unadjusted mean differences are 11.87 for 4mcg, 11.09 for 10mcg, and 17.57 for 25mcg. Their confidence intervals are 7.603 to 16.14, 6.826 to 15.35, and 13.32 to 21.82, respectively. These are simple recalculations from posted summary statistics, not the sponsor's model-based effect estimates.

The registry says that the mixed-model analysis included participants depending on whether both baseline and Week 12 data were available. Thus the analysis count can differ from the overall count in the MITT population at baseline. The specific handling of missing values and the precise analysis population used for each reported test are not reported in the registry record.

7. Results Posted in the Registry

• Estradiol 4mcg Vaginal Softgel Capsule: mean 17.50 versus 5.63 on control; difference 11.87 (95% CI 7.60 to 16.14), standardized effect d = 0.59; z-test p = <0.001.
• Estradiol 10mcg Vaginal Softgel Capsule: mean 16.72 versus 5.63 on control; difference 11.09 (95% CI 6.83 to 15.35), standardized effect d = 0.55; z-test p = <0.001.
• Estradiol 25mcg Vaginal Softgel Capsule: mean 23.20 versus 5.63 on control; difference 17.57 (95% CI 13.32 to 21.82), standardized effect d = 0.87; z-test p = <0.001.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Secondary Efficacy Endpoints - Vaginal Superficial Cells

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard errors.
• Result:
  • Estradiol 4mcg Vaginal Softgel Capsule: mean 31.35 versus 6.05 on control; difference 25.30 (95% CI 21.15 to 29.45), standardized effect d = 1.24; z-test p = <0.001.
  • Estradiol 10mcg Vaginal Softgel Capsule: mean 31.93 versus 6.05 on control; difference 25.88 (95% CI 21.74 to 30.02), standardized effect d = 1.27; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Estradiol 4mcg Vaginal Softgel Capsule than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Estradiol 10mcg Vaginal Softgel Capsule than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Secondary Efficacy Endpoints- Vaginal Superficial Cells

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard errors.
• Result:
  • Estradiol 4mcg Vaginal Softgel Capsule: mean 18.41 versus 5.43 on control; difference 12.98 (95% CI 8.74 to 17.22), standardized effect d = 0.64; z-test p = <0.001.
  • Estradiol 10mcg Vaginal Softgel Capsule: mean 16.88 versus 5.43 on control; difference 11.45 (95% CI 7.20 to 15.70), standardized effect d = 0.57; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Estradiol 4mcg Vaginal Softgel Capsule than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Estradiol 10mcg Vaginal Softgel Capsule than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Secondary Efficacy Endpoints - Vaginal Parabasal Cells

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard errors.
• Result:
  • Estradiol 4mcg Vaginal Softgel Capsule: mean -40.23 versus -7.00 on control; difference -33.23 (95% CI -38.00 to -28.46), standardized effect d = -1.42; z-test p = <0.001.
  • Estradiol 10mcg Vaginal Softgel Capsule: mean -44.42 versus -7.00 on control; difference -37.42 (95% CI -42.17 to -32.67), standardized effect d = -1.60; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Estradiol 4mcg Vaginal Softgel Capsule than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Estradiol 10mcg Vaginal Softgel Capsule than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The posted secondary outcomes include vaginal superficial cells at Week 2 and Week 6, vaginal parabasal cells at Week 2, and vaginal pH at listed time points. Across the posted cell outcomes, the arm summaries and recorded comparisons show differences from placebo for the estradiol groups represented in those analyses. The broad pattern is directionally consistent with the primary superficial-cell endpoint: the changes in cell composition are not limited to the Week 12 primary assessment.

These secondary outcomes provide context for the primary cytologic result, but they do not independently establish the meaning of the primary measure for symptoms. The registry includes a substantial number of outcome measures and analyses, and only a subset of the secondary outcomes is detailed in the registry record described for this statistical account. The complete prespecified hierarchy and analysis plan are not reported in the registry record.

Several of the posted secondary comparisons have reported mixed-model p-values, while the recalculated contrasts are unadjusted. In general, when many outcomes or comparisons are tested, the chance of obtaining at least one small p-value by chance can increase unless the analysis accounts for multiplicity. The record does not report whether the secondary analyses were adjusted for multiple testing or designated as confirmatory versus exploratory.

For that reason, secondary results carry less inferential weight than a clearly prespecified primary analysis with a documented multiplicity strategy. They can help describe whether related outcome measures show a broadly coherent pattern, but they should not be treated as a set of independent confirmations. The absence of a reported testing hierarchy and adjustment plan limits how strongly any individual secondary p-value can be interpreted.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Secondary Efficacy Endpoints - Vaginal Superficial Cells
• Secondary Efficacy Endpoints- Vaginal Superficial Cells
• Secondary Efficacy Endpoints - Vaginal Superficial Cells
• Secondary Efficacy Endpoints - Vaginal Parabasal Cells
• Secondary Efficacy Endpoints - Vaginal Parabasal Cells
• Secondary Efficacy Endpoints - Vaginal Parabasal Cells
• Secondary Efficacy Endpoints - Vaginal pH
• Secondary Efficacy Endpoints - Vaginal pH

9. Statistical Interpretation

10. Statistical Methods Explained

The primary endpoint is continuous: it is a percentage-based measure summarized by least-squares means. The registry identifies a mixed-model repeated-measures analysis for the primary endpoint and calls the reported analysis method “Mixed Models Analysis.” In general, a repeated-measures model can use observations taken at more than one time while accounting for the fact that measurements from the same participant are related. The specific model formula, covariance structure, and adjustment variables are not reported in the registry record.

What is a least-squares mean?

In general, a least-squares mean is a model-based group mean. Depending on the model, it may account for specified factors or repeated measurements rather than being a simple arithmetic average of all recorded observations. The registry labels the arm values as least-squares means, but the precise model specification and covariates are not reported in the registry record. The posted standard errors describe uncertainty around those estimates under the analysis used.

What does a mean difference represent?

The recalculated comparisons subtract the posted placebo mean from the posted mean for an estradiol arm. A positive estimate means that the posted least-squares mean change was numerically higher in the estradiol arm than in the placebo arm for this endpoint. The recalculated differences are 11.87, 11.09, and 17.57 for the 4mcg, 10mcg, and 25mcg comparisons. They are unadjusted calculations from the posted summaries, not model coefficients taken from the sponsor's analysis.

General statistical interpretation: the standard error describes the estimated variability of an estimate across repeated samples under the model. It is not the standard deviation of individual participant measurements. The registry reports standard errors for the arm summaries. For the recalculations, approximate standard deviations are inferred from those standard errors and arm counts; that approximation is why these results should not be treated as a re-fit of the mixed model.

How should the p-values be read?

A p-value, in general, measures how incompatible the observed data are with a specified null hypothesis under the analysis assumptions. It is not the probability that the null hypothesis is true, nor a measure of clinical importance. The registry reports mixed-model p-values of < 0.001 for each primary comparison. The unadjusted recalculations report p < 0.001. These are distinct results from distinct procedures and should not be conflated.

“Unadjusted” means that the recalculated comparisons do not incorporate an adjustment for multiple comparisons. The registry lists three primary arm-versus-placebo tests and many additional analyses. Whether and how the sponsor adjusted the primary tests, or controlled the overall type I error across outcomes, is not reported in the registry record. Therefore, the individual p-values should be interpreted with that uncertainty in mind.

General statistical interpretation: over-dispersion is a concern when observed variability exceeds what a model assumes. For continuous outcomes, the choice of variance model and the handling of repeated observations can affect standard errors and inferential results. The posted endpoint is analyzed with a mixed-model approach, but the model details needed to assess the variance assumptions are not reported in the registry record. No separate over-dispersion assessment is reported.

The registry notes that the number analyzed varied according to availability of baseline and Week 12 data. In general, missing observations can affect estimates if the chance of missingness is related to outcomes or treatment group. The record does not describe the missing-data mechanism, imputation procedure, sensitivity analyses, or the number contributing to each model-based comparison. These omissions limit independent assessment of robustness.

Finally, sample-size illustrations based on observed effects are not part of the trial's original analysis. They are planning calculations under stated assumptions and should not be read as retrospective evidence that the trial had a particular probability of detecting an effect. The study's original sample-size justification and planned statistical analysis details beyond the listed mixed-model method are not reported in the registry record.

11. Confidence Intervals and Precision

The primary recalculated comparisons have confidence intervals of 7.603 to 16.14 for 4mcg versus placebo, 6.826 to 15.35 for 10mcg versus placebo, and 13.32 to 21.82 for 25mcg versus placebo. All three intervals exclude zero, and their lower and upper limits indicate a range of plausible positive differences under the unadjusted recalculation. The estimates are 11.87, 11.09, and 17.57, respectively.

In general, a confidence interval conveys both the estimated direction and the precision of an estimate. A narrower interval indicates less statistical uncertainty under the calculation's assumptions; a wider interval indicates more. The interval is not a guarantee that the true effect lies within its limits, and its interpretation depends on the model and method used. The registry record does not identify a confidence level for these recalculated intervals, so no level should be inferred from the posted information.

The arm-level standard errors are close in size: 1.542, 1.54, and 1.529 for the estradiol groups, and 1.537 for placebo. The recalculation derives approximate pooled standard deviations from the reported standard errors and arm counts. Because these are reconstructed quantities, the resulting intervals are approximate and unadjusted, rather than intervals from the sponsor's mixed model.

The sponsor reports p-values for the comparisons but does not post an effect estimate or confidence interval in the analysis entries represented in the record. Therefore, the recalculated intervals add an estimate-scale description, while the sponsor's tests remain the reported trial analysis. Precision around a cytologic difference does not by itself tell how important that difference is to symptoms or other outcomes.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The sample-size display gives planning illustrations based on the observed comparison effects. For these continuous outcomes, the figures are participants per arm, not total events. The stated assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. Each planning scenario targets a stated fraction of the observed effect: the full observed effect, 75% of it, or 50% of it.

The 80% power column gives the illustrative number of participants per arm for an 80% chance of detecting the assumed effect under those planning assumptions, if that effect and the assumptions were correct. The 90% power column gives the corresponding per-arm planning number for a 90% chance. Power is a design probability under a specified model and effect, not a measure of the probability that the result is true.

Smaller effects generally require more participants to detect, all else equal. Accordingly, the planning figures increase when the scenario assumes a smaller fraction of the observed effect. The display is useful for seeing how a proposed effect size and a desired power level influence a future study's planned size. It does not describe the trial's actual design rationale, because the original sample-size justification and original design assumptions are not reported in the registry record.

These calculations are planning illustrations, not observed power. They use the observed effect as an input and therefore should not be used as a retrospective judgment of whether the completed trial was adequately powered. The registry does not report the study's original target power or its prespecified minimum effect of interest.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = 11.870; pooled standard deviation σ = 20.132.
2. n = 2 × (1.96 + 0.842)² × 20.132² / 11.870² = 45.2, rounded up to 46 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
delta <- 11.87                      # observed difference in means
sigma <- 20.1315640919486                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(2253173)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 46 (80% power), 61 (90% power)
 75% of observed effect: n per arm = 81 (80% power), 108 (90% power)
 50% of observed effect: n per arm = 181 (80% power), 242 (90% power)
power.t.test: 46.1 
Simulated power at n = 46 per arm: 0.798 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Analyses Reported by the Sponsor

The registry's own primary analysis is described as Mixed Models Analysis, with p-values of < 0.001 for each of the three estradiol-dose comparisons with placebo. The sponsor analysis entries do not include a numerical effect estimate or confidence interval in the record represented.

The recalculated comparisons use the posted least-squares mean changes and standard errors to estimate unadjusted mean differences and confidence intervals. They report differences of 11.87 for 4mcg, 11.09 for 10mcg, and 17.57 for 25mcg, with unadjusted p-values of <0.001 for each. The recalculation is approximate and does not reproduce the mixed-model analysis or its potentially model-based handling of repeated measures and available data.

The record's discrepancy flag is false, so the posted statistical summaries do not indicate a discrepancy between the posted values and the recalculation that would require a special precedence warning. The sponsor's own analysis remains the primary trial analysis; the recalculated results are a supplementary, unadjusted way to express the posted arm summaries as contrasts. Their different p-value thresholds should not be treated as a conflict, because the calculations and reported analyses are not identical.

Details of the sponsor's model specification, covariate adjustment, and any multiplicity procedure are not reported in the registry record. The registry's p-values should therefore be cited as reported, while the recalculated estimates should be labeled as unadjusted approximations.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists one primary endpoint assessed across three estradiol-dose comparisons with placebo, along with many outcome measures and analyses. The record reports mixed-model p-values for the primary comparisons and several secondary comparisons. It does not report a family-wise error control procedure, a false-discovery procedure, a gatekeeping strategy, or another multiplicity plan. Whether a particular endpoint or contrast was prespecified as confirmatory is also not reported in the registry record.

In general, multiplicity matters because repeated testing creates more opportunities for a small p-value to occur even when there is no underlying effect. The effect of that issue depends on which hypotheses belong to the same testing family and on the adjustment method. Without a stated hierarchy or adjustment procedure, the individual reported p-values are best read as results for their respective comparisons, not as proof that the overall collection of tests met a particular error-rate standard.

The registry states that the primary mixed-model analysis included participants based on having both baseline and Week 12 data, so the number analyzed could vary and need not match the MITT population at baseline. The record does not report the exact analysis count for each model comparison in the registry record. It also does not report how missing values were handled, whether participants with incomplete follow-up were included through other model observations, or whether sensitivity analyses assessed missing-data assumptions.

These details matter because a model's results can depend on the observations included and on assumptions about missingness. In general, repeated-measures models can use available outcome observations under specified assumptions, but the registry's brief description is not enough to determine the assumptions or their adequacy in this trial. The analysis population definition, missing-data strategy, prespecified testing plan, and interim-analysis details beyond what is listed are not reported in the registry record.

16. 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 3% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: atrial fibrillation (0.5%), appendicitis (0.5%), endophthalmitis (0.5%), chronic obstructive pulmonary disease (0.5%), sinus node dysfunction (0.5%).

The registry's safety summaries list participants at risk and counts with serious adverse events and other adverse events separately for the four arms. Serious adverse events were recorded for 0 of 191 in the 4mcg group, 3 of 191 in the 10mcg group, 4 of 190 in the 25mcg group, and 1 of 192 in the placebo group. The registry lists other adverse events for 46 of 191, 41 of 191, 45 of 190, and 66 of 192, respectively.

The registry also lists serious-event terms, including atrial fibrillation, appendicitis, endophthalmitis, chronic obstructive pulmonary disease, sinus node dysfunction, and ankle fracture, each recorded as 0.5%. These listed terms should be read as the registry's reported event information; the record does not provide enough detail to establish timing, relatedness to study intervention, recurrence, or clinical course.

The safety denominators differ from some efficacy analysis counts, and those counts should not be combined. The record does not report deaths in the safety fields represented. These figures are descriptive and do not by themselves establish comparative safety, particularly without details of exposure, event adjudication, or statistical comparisons.

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

The registry results show that the posted Week 12 least-squares mean changes in vaginal superficial cells were higher in each estradiol arm than in the placebo arm. The recalculated unadjusted contrasts are positive, and their confidence intervals exclude zero. The sponsor's reported mixed-model tests also give p-values of < 0.001 for each comparison. These statements describe the recorded endpoint and the analyses reported for it.

They do not establish that the observed differences are clinically important. The registry record does not state how a particular change in superficial-cell percentage maps to symptom relief or to a threshold that participants would consider meaningful. Nor does a cytologic outcome by itself establish effects on every symptom or condition listed for the trial.

The result also does not remove uncertainty related to model specifications, missing observations, or the number of outcomes and comparisons. The sponsor's own analysis and the approximate unadjusted recalculations are not interchangeable. The recalculated effects and intervals are intended to make the contrast scale more interpretable, while the sponsor's mixed-model p-values are the reported trial tests.

Safety should be considered separately from the primary efficacy measure. The registry reports serious and other adverse event counts by group, but these descriptive counts alone do not establish a causal relationship or provide a complete assessment of comparative safety. The listed safety summaries should be interpreted alongside the limitations of the available record.

Overall, the data support a statistical description of differences in a specific cellular measure over the specified time frame. They do not, on their own, answer broader questions about symptom experience, long-term outcomes, or how these findings apply beyond the population and study conditions identified in the registry.

18. Why This Trial Matters Statistically

This trial illustrates why the numerical result and its statistical interpretation should be considered together. A primary endpoint can have an estimate, a confidence interval, and a p-value, yet each describes a different aspect of the evidence. The estimate describes the contrast on the endpoint's scale; the interval conveys uncertainty; and the p-value addresses compatibility with a null hypothesis under an analysis model.

The registry also distinguishes a model-based sponsor analysis from an unadjusted recalculation. That distinction matters because repeated-measures data and incomplete follow-up can affect how estimates and their uncertainty are calculated. A recalculation from arm summaries can be useful for understanding scale, but it does not reproduce the sponsor's model or the analysis population's full data structure.

The large set of recorded outcomes and analyses makes multiplicity relevant. Without a reported testing hierarchy or adjustment strategy, small p-values across multiple endpoints should not automatically be treated as independent confirmation. The trial therefore offers a useful example of why interpretation depends not only on the individual numerical result but also on the statistical plan and reporting detail.

Finally, the endpoint is a cellular measure. Statistical evidence for a difference in that measure is not equivalent to evidence about the importance of the change to participants. Separating outcome measurement, statistical precision, and clinical meaning is essential when reading registry results.

19. 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 registry record gives limited detail about the primary mixed model. Its full specification, including covariance structure, covariates, and assumptions, is not reported in the registry record. The sponsor's analysis entries give p-values but no effect estimate or confidence interval, so the recalculated contrast is useful context but cannot replace a complete model-based result.

The primary endpoint is a cytologic measure, and the record does not explain its relationship to symptom changes or specify a threshold for clinical importance. The available statistical account therefore cannot determine how the observed differences should be interpreted in terms of participant experience.

The registry notes that analysis counts depended on having both baseline and Week 12 data. Exact analysis counts for each primary comparison, the missing-data strategy, and sensitivity analyses are not reported in the registry record. This limits assessment of whether incomplete observations could have influenced the results.

The record contains numerous outcomes and analyses, but the prespecified hierarchy and multiplicity adjustment are not reported. The original sample-size justification and interim-analysis plan are not reported either. The planning illustrations based on observed effects cannot fill those gaps or establish the trial's original design assumptions.

Safety summaries are descriptive and do not provide, in the information represented here, details of event relatedness, exposure duration, or adjudication. The registry also does not report deaths in the safety fields represented. These limits call for a cautious reading of both efficacy and safety summaries.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Estradiol Vaginal Softgel Capsules in Treating Symptoms of Vulvar and Vaginal Atrophy in Postmenopausal Women — NCT02253173.
• Registry record last updated 2017-05-25; status last verified 2017-04.
• Results first posted 2017-05-25.
• 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.

23. Record Summary

NCT02253173 is a completed randomized, triple-masked, parallel-group, placebo-controlled, multicenter Phase 3 trial sponsored by TherapeuticsMD. It evaluated vaginal softgel capsules containing estradiol at 4mcg, 10mcg, and 25mcg against placebo in postmenopausal women with moderate to severe symptoms of vulvar and vaginal atrophy.

The primary endpoint was change from baseline to Week 12 in the percentage of vaginal superficial cells. The registry reports a mixed-model analysis and p-values of < 0.001 for each dose comparison with placebo. Unadjusted recalculations from posted summaries give positive mean differences, with confidence intervals excluding zero. The registry notes that analysis counts depended on availability of baseline and Week 12 data.

The record also includes secondary outcomes and descriptive adverse-event counts. Important details—including the original sample-size justification, full model specification, missing-data strategy, and multiplicity plan—are not reported in the registry record. The findings describe the recorded cellular endpoint and should not be taken alone as evidence of clinical importance or a complete account of safety.