Phase 3 Study to Evaluate the Efficacy and Safety of Pergoveris® in Assisted Reproductive Technology (ESPART): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Phase 3 Study to Evaluate the Efficacy and Safety of Pergoveris® in Assisted Reproductive Technology (ESPART) (NCT02047227), a randomized phase 3 study. The primary outcome is Number of Oocytes Retrieved.

2. Clinical Question

The trial therefore addresses a specific treatment-cycle question in the registry-defined population, rather than a general comparison for all people receiving assisted reproductive technology. Its primary endpoint concerns the number of oocytes obtained, while ongoing pregnancy, live birth, clinical pregnancy, biochemical pregnancy and embryo implantation are listed as secondary outcomes. Results are posted for some, but not all, of those secondary measures.

3. Trial Design

The registry identifies this as a Phase 3, randomized, controlled, single-blind, multicentre, parallel-arm study. It compared Pergoveris with GONAL-f for multifollicular development in poor ovarian responders receiving an assisted reproductive technology treatment cycle. The record lists overall enrollment as 939 and marks the study as completed. The study start is listed as 2014-01-31 and primary completion as 2015-08-31.

The posted primary-outcome groups include 462 participants assigned to Pergoveris and 477 assigned to GONAL-f. The primary analysis population is described as the intent-to-treat set, including all randomized subjects who received at least 1 dose of either study treatment. That definition is narrower than simply “all randomized”: subjects who did not receive a dose are not included under the registry’s stated definition. The record does not explain how many, if any, randomized subjects were excluded under that condition.

Random allocation is a design feature intended to make the treatment groups comparable on average before treatment, including with respect to characteristics that may not have been measured. It does not guarantee identical groups in a particular trial. The registry does not report the sequence-generation procedure, allocation concealment, stratification factors, or details of the single-blind arrangement. The term multicentre is reported, but the number and locations of participating centres are not reported in the registry record.

The study’s sponsor is Merck KGaA, Darmstadt, Germany. The registry gives the primary endpoint and its analysis population, but the full protocol and the trial’s original sample-size rationale are not reported in the registry record. The known planning assumptions accompanying the posted sample-size illustrations are distinct from those original design assumptions.

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

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Intent-to-treat (ITT) analysis set included all subjects randomized who received at least 1 dose of GONAL-f or Pergoveris.

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: All Randomized Subjects Set included all subjects who were randomized.

6. Primary Endpoint

The registry lists the primary outcome as Number of Oocytes Retrieved, measured over At approximately 34 to 38 hours after r-hCG administration (Day 113).

The primary endpoint is the number of oocytes retrieved, summarized as a mean and standard deviation for each treatment group. The registry defines the measurement as the mean number retrieved on the day of ovum pick up. It gives the timeframe as approximately 34 to 38 hours after r-hCG administration (Day 113). Oocyte retrieval is described as a technique used in in-vitro fertilization to remove oocytes from the ovary so that fertilization can occur outside the body.

As a general statistical explanation, a count of retrieved oocytes is a direct measure of ovarian response during the treatment cycle. Comparing group means gives a compact summary of the average count, while the standard deviation describes how widely individual counts vary around each group mean. Neither summary shows the full distribution of counts or how many participants had a particular value.

This endpoint is used to assess multifollicular development, the treatment objective named in the trial description. It is an intermediate treatment-cycle outcome, not a direct measure of whether a pregnancy continued or resulted in a live birth. Those later outcomes are listed separately as secondary endpoints, and the registry reports some of them independently. A difference in the mean number of retrieved oocytes should therefore be interpreted as a difference in the measured retrieval outcome, not as proof of a difference in pregnancy or live-birth outcomes.

7. Results Posted in the Registry

• GONAL-f: mean 3.60 versus 3.30 on control; difference 0.30 (95% CI -0.05 to 0.65), standardized effect d = 0.11; z-test p = 0.096.

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.

Ongoing Pregnancy Rate

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • GONAL-f: 12.4% (59/477) versus 11.0% (51/462) on control; risk difference 1.3 percentage points (95% CI -2.8 to 5.4), risk ratio 1.12 (0.79–1.59); unadjusted two-proportion z-test p = 0.526.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between GONAL-f and the control. The interval gives the range of effects that remain compatible with the data.

Live Birth Rate

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • GONAL-f: 11.7% (56/477) versus 10.6% (49/462) on control; risk difference 1.1 percentage points (95% CI -2.9 to 5.2), risk ratio 1.11 (0.77–1.59); unadjusted two-proportion z-test p = 0.581.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between GONAL-f and the control. The interval gives the range of effects that remain compatible with the data.

Clinical Pregnancy Rate

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • GONAL-f: 16.8% (80/477) versus 14.1% (65/462) on control; risk difference 2.7 percentage points (95% CI -1.9 to 7.3), risk ratio 1.19 (0.88–1.61); unadjusted two-proportion z-test p = 0.252.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between GONAL-f and the control. The interval gives the range of effects that remain compatible with the data.

The posted secondary results show a broadly consistent descriptive pattern: the recorded rates are numerically higher in the GONAL-f group for the pregnancy outcomes with results posted. That direction is also consistent with the direction of the observed primary group means, but the outcomes measure different stages of the treatment pathway. The primary endpoint is oocyte retrieval, whereas the secondary outcomes concern pregnancy or birth. A pattern across different endpoints does not by itself show that a difference in retrieval translates into a difference in later outcomes.

The recalculated unadjusted confidence intervals for ongoing pregnancy, live birth and clinical pregnancy each include zero for the risk difference. The corresponding unadjusted p-values do not establish a clear difference between groups. These analyses therefore describe uncertain comparisons, not proof that the groups have the same outcomes. The registry lists embryo implantation and biochemical pregnancy among the secondary outcomes, but results for those measures are not reported in the registry record.

Unadjusted secondary results generally carry less weight than a prespecified primary analysis when many outcomes are examined. As a general statistical explanation, testing multiple outcomes increases the chance that at least one result appears notable by chance if no adjustment or hierarchy is used. The registry record does not report a multiplicity adjustment, a testing hierarchy, or a plan specifying which secondary outcomes were confirmatory. The reported secondary p-values should therefore be read as unadjusted comparisons, not as independent confirmation of the primary result.

The primary and secondary measures also differ in timing and clinical meaning. The registry specifies distinct follow-up timeframes for the pregnancy outcomes, and outcomes later in the treatment pathway are not interchangeable with the retrieval count. Their joint pattern can help describe the trial results, but it does not remove uncertainty in any individual comparison or substitute for a documented analysis plan.

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

• Ongoing Pregnancy Rate
• Live Birth Rate
• Embryo Implantation Rate
• Clinical Pregnancy Rate
• Biochemical Pregnancy Rate

9. Statistical Interpretation

10. Statistical Methods Explained

The posted evidence includes group means and standard deviations, an unadjusted recalculation of the mean difference, and a sponsor-reported regression analysis. These are related summaries, but they do not answer precisely the same statistical question. The unadjusted result is calculated from the posted group summaries; the regression result is the analysis reported by the registry. The registry does not provide the full statistical analysis plan or enough model details to reproduce that regression.

What does a mean difference describe?

As a general statistical explanation, a group mean is the arithmetic average of participants’ outcome values. The mean difference compares those averages in a stated direction. In the recalculation, the stated contrast is GONAL-f versus Pergoveris, and the estimate is 0.3 oocytes. A positive value in that contrast means the observed mean is higher in the GONAL-f group. This is a summary of observed group averages; it is not an estimate adjusted for participant characteristics or centre unless such adjustment is explicitly part of the method.

The standard deviations, 2.71 for Pergoveris and 2.82 for GONAL-f, describe the spread of participant outcomes within their respective groups. Standard deviation is not the same as the uncertainty in a group mean. The uncertainty of a mean difference also depends on the amount of information in each group and on the variability of the outcome. The recalculation reports standard errors of 0.126 and 0.129 for the respective group means, and a pooled standard deviation of 2.766 for the comparison.

How should the confidence interval be read?

As a general explanation, a confidence interval expresses the sampling uncertainty around an estimate under the method used to construct it. The recalculated 95% interval for the mean difference extends from -0.054 to 0.654. Because the interval includes zero, the results do not rule out a zero mean difference under that calculation. The interval also includes nonzero differences in either direction. It is not a probability statement that the true effect has a 95% chance of lying within these particular endpoints; rather, the confidence level describes the long-run behavior of the interval procedure under its assumptions.

The interval is unadjusted and based on posted summary statistics. It does not incorporate an adjustment for multiple outcomes or covariates. The registry’s own interval is based on its Poisson regression model and therefore reflects that model’s analysis rather than the simple summary-statistic calculation. The registry does not report the regression specification, covariates, model diagnostics, or the precise contrast coding.

Why use a Poisson regression for this outcome?

As a general statistical explanation, Poisson regression is commonly used for count outcomes and relates the expected count to predictors through a model. Depending on the specification, such a model can estimate a treatment contrast while accounting for covariates. The registry reports a Poisson Regression Model and labels its parameter as a net mean difference. It does not report which covariates were included or whether the model included other adjustments, so those features should not be assumed.

As a general point, count data may show greater variability than a basic Poisson model allows, a condition often called over-dispersion. If the model’s variance assumptions do not match the data, standard errors and confidence intervals can be affected. The registry record does not report a diagnostic assessment of over-dispersion, an alternative variance method, or the distributional checks performed. This is a question about model assessment, not evidence that over-dispersion did or did not occur in this trial.

What does the p-value contribute?

As a general explanation, a p-value measures how unusual data at least as inconsistent with a specified null model would be if that null model and the analysis assumptions were correct. It is not the probability that the null hypothesis is true, nor does it measure the size or clinical importance of a difference. The unadjusted recalculation reports p = 0.096; the regression analysis reports p = 0.054. Their difference is expected to be possible when methods and estimands differ. The registry does not describe whether the reported model was prespecified in the analysis plan.

The primary outcome is accompanied by a single reported analysis in the registry record. A count of analyses in the registry does not establish how many analyses were planned in the protocol or whether any were conducted but not reported. That information is not reported in the registry record.

Why do analysis populations matter?

The registry defines the primary intent-to-treat analysis set as all randomized subjects who received at least 1 dose of GONAL-f or Pergoveris. In general, an intent-to-treat approach aims to preserve the comparison created by random allocation, while the exact definition used in practice matters for deciding who is included. Because the posted definition includes a treatment-receipt condition, it should be read as written rather than assumed to include every randomized participant. The handling of missing primary endpoint values and any additional analysis populations are not reported in the registry record.

11. Confidence Intervals and Precision

The primary recalculation estimates a mean difference of 0.3 oocytes, with a 95% confidence interval from -0.054 to 0.654. The interval is informative because it gives a range of values compatible with the data under the calculation’s assumptions, rather than presenting the point estimate alone. It includes zero and extends to values on both sides of zero. Thus, the observed data do not precisely establish the direction of a mean difference in the population.

The group standard deviations, 2.71 for Pergoveris and 2.82 for GONAL-f, are large relative to the observed separation between the group means. As a general statistical explanation, substantial within-group variability makes group averages less distinct and can widen uncertainty around comparisons. The posted standard errors for the group means are 0.126 and 0.129. They describe uncertainty in the estimated means, not the spread of individual outcomes.

The model-reported estimate is -0.235, with a 95% confidence interval from -0.474 to 0.003. Its interval also reaches across zero, although it is based on a different method and parameter label from the unadjusted recalculation. The registry does not explain the model contrast coding, so the two estimates should not be treated as directly interchangeable or their signs compared without qualification.

Precision is distinct from clinical importance. A confidence interval can describe statistical uncertainty, but the registry does not report a threshold for a clinically important difference in the number of oocytes. It also does not report the original sample-size justification. Therefore, the interval should be used to understand the range of estimates compatible with the analysis, not to declare equivalence or a clinically meaningful benefit.

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 illustrations are planning calculations based on the observed effect, not the original design rationale for this trial. The stated assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. Each planning row considers detecting the stated fraction of the observed effect. The trial’s own original sample-size assumptions are not reported in the registry record.

For the primary continuous outcome, the listed figures are participants per arm. The 80% power column describes the number of participants per arm that the planning calculation estimates would be needed to have an 80% chance of detecting the specified effect under the stated assumptions, if that effect and those assumptions were correct. The 90% power column describes the corresponding planning target for a 90% chance. They are planning targets, not descriptions of the trial’s observed results.

The rows consider 100% of the observed effect, 75% of the observed effect, and 50% of the observed effect. Smaller effects are harder to distinguish from random variation, so the planning calculations require more participants to detect them at the same stated power. The values are illustrations rather than a retrospective claim about what the trial could detect in practice. This is not observed power.

The assumptions used for the illustration do not establish that the original investigators chose the same significance level, allocation, model, or target effect. The protocol’s sample-size calculation and any allowances for missing outcomes are not reported in the registry record. For these primary and posted secondary outcome calculations, the figures are participants per arm, not total event counts.

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 = 0.300; pooled standard deviation σ = 2.766.
2. n = 2 × (1.96 + 0.842)² × 2.766² / 0.300² = 1,334.9, rounded up to 1,335 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 <- 0.3                      # observed difference in means
sigma <- 2.76642711756487                     # 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(2047227)
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 = 1,335 (80% power), 1,787 (90% power)
 75% of observed effect: n per arm = 2,374 (80% power), 3,177 (90% power)
 50% of observed effect: n per arm = 5,340 (80% power), 7,148 (90% power)
power.t.test: 1335.8 
Simulated power at n = 1,335 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 reports a Poisson regression analysis for the primary outcome: a net mean difference of -0.235, a 95% confidence interval from -0.474 to 0.003, and p = 0.054. The separate recalculation from the posted group summaries is unadjusted: it estimates a mean difference of 0.3 for GONAL-f versus Pergoveris, with a 95% confidence interval from -0.054 to 0.654 and p = 0.096.

These results should be kept distinct. The recalculation compares the posted group summaries without the reported regression model; the registry analysis uses Poisson regression and labels its parameter a net mean difference. The registry record does not specify model covariates or contrast coding, so it does not establish exactly why the estimates differ in sign or magnitude. They should not be combined into a single estimate or treated as duplicate calculations of an identical estimand.

The registry record does not flag a discrepancy between the posted result and the recalculation. The sponsor’s reported analysis remains the trial’s own model-based result; the recalculation is a simple unadjusted check based on summary data, not a replacement for the trial analysis. Neither result alone demonstrates equivalence, and both confidence intervals include zero.

15. Multiplicity, Missing Data and Analysis Populations

The registry record identifies a primary endpoint and lists secondary outcomes, but it does not report a multiplicity strategy. In particular, no adjustment method, testing hierarchy, or distinction between confirmatory and exploratory secondary analyses is reported in the registry record. The posted secondary comparisons are unadjusted. When several outcomes are tested, the chance of at least one apparently unusual result can increase; that is a general statistical concern, not a claim that any particular result in this trial is spurious.

The primary analysis population is defined as the intent-to-treat set of randomized subjects who received at least 1 dose of GONAL-f or Pergoveris. The registry does not report the number of randomized subjects excluded from that definition, whether all included subjects had primary outcome data, or how missing outcome values were handled. Imputation, sensitivity analyses for missing data, and additional analysis populations are not reported in the registry record.

The registry describes the trial as randomized, single-blind and multicentre, but it does not give the randomization sequence method, allocation concealment procedures, or details about who was blinded. It also does not report whether outcome assessors were blinded or whether the blind was maintained throughout follow-up. These omissions limit assessment of potential bias beyond what can be inferred from the design labels.

The sponsor-reported Poisson regression result is presented without a full model specification. Covariates, model diagnostics, handling of over-dispersion, and the exact contrast coding are not reported in the registry record. The registry also does not report interim analyses, stopping rules, protocol deviations, or the statistical analysis plan. Their absence from the registry description should not be interpreted as evidence that they were or were not part of the trial; they are simply 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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: abortion missed (0.6%), abortion (0.4%), abortion spontaneous (0.2%), premature labour (0.2%), imminent abortion (0.2%).

The registry reports serious adverse events for 8 of 462 participants in the Pergoveris group and 17 of 477 in the GONAL-f group. It also reports other adverse events for 111 of 462 and 151 of 477, respectively. These are counts in the posted safety record; they should be read as descriptive event summaries rather than as proof that treatment caused or prevented an event. The registry does not provide an adjusted comparative safety analysis in the record.

The listed serious-event terms and percentages are Abortion missed (0.6%), Abortion (0.4%), Abortion spontaneous (0.2%), Premature labour (0.2%), Imminent abortion (0.2%) and Premature baby (0.2%). The record does not give enough detail in these listings to determine the group-specific distribution of each term, the number of recurrent events, or the relationship of each event to treatment. The terms should not be interpreted as causal findings.

Deaths are not reported in the registry record. The adverse-event figures do not by themselves provide information about severity beyond the serious-event classification, duration, resolution, or exposure-adjusted rates. Those details are not reported in the registry record.

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

The results describe a randomized comparison of Pergoveris and GONAL-f in poor ovarian responders receiving assisted reproductive technology treatment. For the primary outcome, the observed mean number of oocytes retrieved was numerically higher in the GONAL-f group. The unadjusted recalculation has a confidence interval that includes zero, and the registry’s Poisson regression result also has a confidence interval reaching across zero. These findings leave uncertainty about the direction and size of a population mean difference.

This does not establish that the treatments are equivalent, nor does it establish that one treatment is better. A failure to rule out zero is not proof of no difference. Equivalence would require a prespecified margin and an appropriate analysis; such a margin is not reported in the registry record. The observed means and model estimate also do not, on their own, determine whether a difference would matter clinically, because a threshold for clinical importance is not reported.

The primary endpoint measures oocyte retrieval, not pregnancy or live birth. The registry posts several later pregnancy outcomes, but their comparisons also remain uncertain and they are secondary outcomes. A result on oocyte count should not be treated as a substitute for a result on pregnancy or birth, and a descriptive pattern across endpoints does not establish a causal pathway.

The safety counts provide a summary of reported events in each group, not a complete account of risk or evidence that a treatment caused a particular event. The registry record does not report deaths, detailed event-level follow-up, or a comparative safety model. Interpretation is further limited by missing information on missing-data handling, multiplicity, regression specification and the trial’s original sample-size rationale.

The appropriate conclusion is bounded: the registry reports the observed group summaries and its own model-based analysis, while the unadjusted recalculation offers a transparent comparison of those summaries. Together they show uncertain estimates rather than a definitive separation between treatments on the primary endpoint.

18. Why This Trial Matters Statistically

This trial illustrates why a reported point estimate should not be read in isolation. The raw group means and the simple unadjusted comparison offer one view of the primary outcome, while the registry’s Poisson regression offers a model-based view. Because the registry does not give enough detail to align the model’s contrast coding and specification with the recalculation, the methods should be described separately rather than blended into a single result.

It also demonstrates the distinction between statistical evidence and clinical importance. Confidence intervals describe uncertainty, and p-values summarize compatibility with a null model under stated assumptions; neither determines whether a difference matters to patients or clinicians. The registry does not report a clinically important threshold for oocyte counts, so interpretation should remain tied to the measured outcome and the uncertainty around it.

The trial’s outcome sequence is statistically important as well. Oocyte retrieval is an early treatment-cycle measure, while pregnancy and birth are later outcomes. Their numerical patterns may be considered together, but they address different questions and should not be substituted for one another. The secondary results are unadjusted, and the registry does not document how multiple comparisons were handled.

Finally, registry transparency affects how confidently an analysis can be evaluated. The record identifies the primary population and reports an analysis method, but it does not give the original sample-size rationale, detailed model specification, missing-data methods, or multiplicity plan. Recognizing these limits is part of interpreting the reported estimates; it is not a reason to disregard the data that are posted.

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.

Several limitations specific to this registry record constrain interpretation. The participant group is described as poor ovarian responders aligned with ESHRE criteria, but detailed eligibility rules and the number of participating centres are not reported in the registry record. Randomization, single blinding and multicentre conduct are stated, while the sequence-generation method, concealment procedures and details of who was blinded are not reported.

The registry defines the primary analysis set as randomized subjects who received at least 1 dose of either treatment. It does not report whether any randomized subjects were outside this set, the extent of missing primary endpoint data, or how missing values were handled. Other analysis populations and sensitivity analyses are not reported in the registry record.

The sponsor-reported model is identified as Poisson regression, but its covariates, diagnostics, variance specification and contrast coding are not described. This limits direct comparison with the unadjusted mean-difference recalculation. The record does not report the trial’s original sample-size assumptions, a clinically important effect threshold, or the basis for the chosen enrollment target.

Several secondary outcomes are listed, but results are posted for only some of them. The registry does not describe a multiplicity adjustment or testing hierarchy. Safety summaries give event counts and selected serious-event terms, but not event-level details, exposure-adjusted rates, or deaths; deaths are not reported in the registry record. These limitations concern the available registry information and do not establish that unreported procedures or analyses were absent from the full trial documentation.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Phase 3 Study to Evaluate the Efficacy and Safety of Pergoveris® in Assisted Reproductive Technology (ESPART) — NCT02047227.
• Registry record last updated 2017-08-24; status last verified 2017-07.
• Results first posted 2017-01-24.
• 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

The registry record identifies NCT02047227, a completed Phase 3 randomized, controlled, single-blind, multicentre, parallel-arm study sponsored by Merck KGaA, Darmstadt, Germany. It compared Pergoveris with GONAL-f for multifollicular development in poor ovarian responders undergoing assisted reproductive technology treatment. Overall enrollment is listed as 939.

The primary endpoint was the mean number of oocytes retrieved at approximately 34 to 38 hours after r-hCG administration (Day 113), in an intent-to-treat analysis set defined as randomized subjects receiving at least 1 dose. The record reports group means and standard deviations, a Poisson regression analysis, secondary pregnancy outcomes and adverse-event summaries.

The unadjusted recalculation and the registry-reported model result both have confidence intervals that include zero. The record does not report the original sample-size rationale, detailed regression specification, missing-data methods or multiplicity plan. The posted estimates therefore describe the trial results but leave important aspects of analysis and precision uncertain.