Flibanserin in Men Receiving Androgen Suppression for Prostate Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Flibanserin in Men Receiving Androgen Suppression for Prostate Cancer (NCT04743934), a randomized phase 2 study. The primary outcome is Number of Participants Who Reported Attempting Sexual Intercourse at Least 3 Times in the Prior Month.

2. Clinical Question

The trial focuses on a specific reported behavior during androgen suppression, rather than on sexual interest as an abstract or continuous measure. Its primary outcome divides reported attempts into 0-2 or 3+ in the prior month. The registry’s brief summary describes the broader purpose as investigating whether flibanserin promotes sexual interest in men with prostate cancer receiving androgen suppression. The outcome posted for comparison is the count of participants meeting the 3+ threshold.

Two secondary outcomes are listed: Sexual Quality of Life (QoL) and frequency of patient-reported Grade 3+ adverse events. Results for those outcomes are not reported in the registry record. The available primary outcome therefore addresses the threshold-based intercourse-attempt question; it does not by itself describe every aspect of sexual interest or quality of life.

3. Trial Design

This is a randomized, parallel-group phase II trial comparing flibanserin plus ADT with placebo plus ADT. The registry lists an enrollment of 44 and a completed status. The recorded start date is 2021-07-02, and the primary completion date is 2025-03-01. These record dates describe the trial timeline; they do not establish when every participant entered or completed follow-up.

The arm descriptions specify flibanserin 100mg by mouth once daily at bedtime alongside ADT, or placebo at 100mg by mouth once daily at bedtime alongside ADT. ADT is part of both arms, so the posted comparison concerns adding flibanserin rather than comparing ADT with no ADT. The registry lists placebo as the comparator and identifies the allocation as randomized.

The registry’s masking field says quadruple masking, while its brief summary calls the study double-blinded. The record does not identify which participant, clinician, investigator, or assessor roles were masked, or explain the difference between those descriptions. No further details about randomization procedures, allocation concealment, treatment duration beyond the outcome timeframe, adherence, or deviations from assigned treatment are reported in the registry record.

The primary outcome timeframe is baseline up to 12 weeks. The record lists 44 enrolled participants, while the primary outcome arm entries list 22 participants for flibanserin plus ADT and 19 for placebo plus ADT. The registry does not explain the relationship between those arm counts and enrollment or give a participant-flow account. That limits interpretation of how many enrolled participants contributed to the posted primary outcome.

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:

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 Number of Participants Who Reported Attempting Sexual Intercourse at Least 3 Times in the Prior Month, measured over Baseline up to 12 weeks.

The primary endpoint counts participants who reported attempting sexual intercourse at least 3 times in the prior month. Reports are grouped as 0-2 attempts or 3+ attempts, and the outcome is assessed from baseline up to 12 weeks. The posted arm counts are 1 participant in the flibanserin-plus-ADT arm and 3 participants in the placebo-plus-ADT arm; the corresponding arm entries list 22 and 19 participants, respectively.

A threshold-based count makes the endpoint a binary outcome for each participant: meeting the 3+ category or not meeting it. Comparing the number in that category can be useful when the clinical question is framed around reaching a defined level of reported activity. It also yields an outcome that can be compared between randomized groups using methods for categorical data.

This measure captures the number crossing the chosen threshold, not the full distribution of attempts, the reasons for attempts or no attempts, or sexual interest in all its dimensions. The registry’s description also does not state how reports were collected, how missing reports were handled, or whether the threshold was assessed consistently at every point in the timeframe. Those measurement and analysis details are not reported in the registry record.

7. Results Posted in the Registry

• Flibanserin + ADT: 4.5% (1/22) versus 15.8% (3/19) on control; risk difference -11.2 percentage points (95% CI -29.8 to 7.3), risk ratio 0.29 (0.03–2.54); unadjusted two-proportion z-test p = 0.226.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is categorical because each participant is classified according to whether reported attempts reached the 3+ threshold in the prior month. The registry describes comparing the number of participants in the two groups who fall into that category. It names a two-group chi-square test as the planned comparison, with Fisher’s exact test as an alternative if the assumptions for the chi-square test are not tenable. The registry does not say which of these tests produced the posted comparison.

Why use a chi-square or Fisher’s exact test?

In general, a chi-square test evaluates whether the distribution across categorical outcome groups differs between treatment arms. Its usual approximation works best when the expected counts in the categories are adequate. When counts are sparse, Fisher’s exact test can be used to calculate a comparison without relying on the same large-sample approximation. The choice between the tests depends on the data and the assumptions; the registry does not report the specific test used for this result or the expected cell counts considered.

The observed outcome counts are sparse: 1 participant met the threshold in the flibanserin-plus-ADT arm and 3 did so in the placebo-plus-ADT arm. Sparse counts are one reason to pay close attention to the interval estimates, not only the p-value. A small change in the event count can materially change an estimate when there are few events. That general property helps explain why the reported intervals are broad; it does not establish that a particular test was used.

What does the risk difference express?

The risk difference subtracts the proportion meeting the outcome threshold in one arm from the proportion in the other, according to the comparison’s stated direction. The reported estimate is -0.112. In this comparison, the negative sign indicates a lower estimated proportion in the flibanserin arm than in the placebo arm. The confidence interval from -0.298 to 0.073 describes uncertainty around that estimate. Because the interval includes zero, the data are compatible with no difference as well as with differences in either direction.

A risk difference is an absolute measure, expressed on the probability scale. It is often directly interpretable as the estimated difference in the share of participants meeting the endpoint. The posted arm proportions are 0.045 and 0.158. The risk difference and those proportions describe the same outcome comparison from different angles; they should not be treated as separate replications.

How do relative measures differ?

The risk ratio compares the outcome proportions as a ratio, while the odds ratio compares the outcome odds. Their reported point estimates are 0.288 and 0.254, respectively. The risk-ratio interval is 0.033 to 2.542, and the odds-ratio interval is 0.024 to 2.676. For ratios, the no-association value is 1; both intervals include that value. Both relative intervals are wide, reinforcing that the estimate is uncertain. Relative measures can appear numerically far from 1 while still being imprecise when outcome counts are limited.

The comparison also gives an unadjusted p-value of 0.226. In general, a p-value describes how unusual data at least as discrepant as those observed would be under a specified null model and its assumptions. It is not the probability that the null hypothesis is true, nor a measure of the size or clinical importance of an effect. A p-value alone does not convey the range of effects compatible with the data, which is why the estimate and confidence interval matter as well.

What can be concluded from the analysis record?

The registry lists zero outcome analyses. It does post arm counts and comparison estimates, but it does not provide a full analysis report. The statistical description names the categorical tests, yet the record does not identify the test actually applied, give a prespecified analysis plan, or describe any adjusted model. It is therefore not possible to determine from the registry whether the reported comparison follows a particular model beyond its displayed unadjusted form.

General statistical interpretation also depends on how participants were assigned, measured, and included in the analysis. Randomization is listed, but details of sequence generation and concealment are not reported in the registry record. The analysis population, handling of missing outcome reports, and treatment of participants who stopped or did not adhere are also not reported in the registry record. These omissions matter because the estimated comparison can depend on which participants contribute data and how unobserved outcomes are handled. No assumptions about those procedures should be inferred from the displayed counts.

10. Confidence Intervals and Precision

The risk-difference estimate is -0.112, and its 95% confidence interval ranges from -0.298 to 0.073. A confidence interval is a range of values compatible with the data and the statistical method, under its assumptions. It communicates uncertainty around the estimate; it is not a guarantee that the true value lies in the interval, nor does it assign a probability to a particular treatment effect after the data are observed.

Here the interval spans zero, so it includes no difference between arms. It also includes values on both sides of zero, leaving uncertainty about the direction of the underlying difference. The breadth of the interval indicates limited precision. The relative-effect intervals are also broad: 0.033 to 2.542 for the risk ratio and 0.024 to 2.676 for the odds ratio. Each includes the no-association value for its measure.

The outcome counts are 1 and 3, so relatively few participants met the threshold. In general, estimates based on sparse outcomes tend to be less stable, and their intervals can be wide. The posted data support a range of possible effects rather than a narrowly determined effect size. The p-value of 0.226 is consistent with that uncertainty, but it cannot replace the interval or describe precision by itself.

The registry does not report confidence intervals for the arm-specific proportions, a prespecified precision target, or a formal assessment of whether the study was large enough for a desired precision. The interval should therefore be read as uncertainty around the posted comparison, not as evidence that the study achieved a particular precision goal.

11. 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 figures are planning illustrations based on fractions of the observed effect, not the trial’s original design calculation. For this binary outcome, the figures are participants per arm, rather than a total number of events. The planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is set to the stated fraction of the observed effect.

The rows correspond to planning for 100%, 75%, and 50% of the observed effect. The 80% power column gives the estimated number of participants per arm for a hypothetical study to have an 80% chance of detecting the specified effect under those assumptions, if that effect were the true one. The 90% power column gives the corresponding per-arm planning figure for a 90% chance. These are design calculations, not estimates of the probability that the observed result is true.

Smaller effects are harder to distinguish from random variation, so planning to detect a smaller fraction of the observed effect requires more participants per arm. The listed figures increase as the effect used for planning becomes smaller, and the 90% power target calls for more participants per arm than the 80% target. These values are not observed power: they do not measure the probability that this completed trial detected an effect, and they do not change the uncertainty of the results already posted.

The registry does not report the trial’s own original sample-size assumptions or justification. The planning values should therefore be used only to understand the consequences of the stated hypothetical assumptions, not to infer the rationale for the enrollment target.

12. 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, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.158; treatment proportion p₁ = 0.045; average p̄ = 0.102.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.4274 = 0.8377.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.4199 = 0.3534.
4. n = (0.8377 + 0.3534)² / (-0.1124)² = 112.2, rounded up to 113 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
p0 <- 0.157894736842105                          # control proportion
p1 <- 0.0454545454545455                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^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 (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(4743934)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 113 (80% power), 150 (90% power)
 75% of observed effect: n per arm = 225 (80% power), 301 (90% power)
 50% of observed effect: n per arm = 560 (80% power), 749 (90% power)
power.prop.test: 112.2 
Simulated power at n = 113 per arm: 0.833 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-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.

13. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary outcome and two secondary outcomes: Sexual Quality of Life (QoL) and frequency of patient-reported Grade 3+ adverse events. Results for the secondary outcomes are not reported in the registry record. The record also lists zero outcome analyses, so it does not provide a complete set of inferential results across the listed outcomes.

When a study tests several outcomes or performs multiple comparisons, the chance of at least one apparently unusual result can increase. General approaches to multiplicity include prespecifying a hierarchy of outcomes or adjusting the interpretation of tests across a family of comparisons. The registry does not report whether any such strategy was planned or used, whether there were additional analyses, or how secondary outcomes were to be interpreted alongside the primary outcome. No multiplicity adjustment should be assumed.

The handling of missing outcome data is not reported in the registry record. The record does not state whether participants with missing reports were excluded, imputed, or treated in another way. It also does not provide the number of missing reports by arm, reasons for missingness, or sensitivity analyses. The listed arm counts cannot by themselves establish how missing data were handled.

The analysis population is not reported in the registry record. It is not stated whether the comparison included all randomized participants, only participants with available outcome data, or another defined group. The record also does not describe adherence-based exclusions, treatment discontinuation, crossover, or protocol deviations. These details are relevant to interpreting a randomized comparison, but their absence does not establish that any particular issue occurred.

14. Safety Results

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

Deaths

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 serious adverse event entries show 0 affected participants among 22 at risk in the flibanserin-plus-ADT arm and 0 among 19 at risk in the placebo-plus-ADT arm. The death count is 0 in each arm. These are the serious-event and death counts posted in the registry; they do not demonstrate that the risk of a serious event is absent.

For the other adverse-event category, the record lists 10 affected participants among 22 at risk in the flibanserin-plus-ADT arm and 6 among 19 at risk in the placebo-plus-ADT arm. The registry does not provide event terms for those entries in the posted summary, so the counts should not be read as describing a particular symptom, severity distribution, or causal relationship. The secondary outcome specifically naming patient-reported Grade 3+ adverse events does not have results reported in the registry record.

These counts are descriptive. No safety comparison, confidence interval, exposure-adjusted rate, or attribution analysis is reported in the registry record. With no serious events recorded in either arm, the data do not estimate how the arms might differ in the frequency of uncommon serious events. The arm denominators and follow-up details also limit interpretation of the counts.

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

The primary outcome comparison records fewer participants meeting the 3+ attempts threshold in the flibanserin-plus-ADT arm than in the placebo-plus-ADT arm: 1 versus 3. The estimated risk difference is -0.112. The estimate points toward a lower proportion meeting the threshold in the flibanserin arm, but its 95% confidence interval, -0.298 to 0.073, includes zero and spans values in both directions. The unadjusted p-value is 0.226. Taken together, the posted statistics leave substantial uncertainty about the size and direction of the difference.

This result is not evidence that the arms are identical. A result with an interval including zero can arise when the data are compatible with a range of effects, including no difference. Conversely, the directional point estimate alone does not establish that flibanserin caused a lower threshold rate. The registry record does not report enough analysis detail to support stronger causal or clinical claims than the randomized comparison and its uncertainty allow.

The endpoint represents a thresholded report of intercourse attempts, not every component of sexual interest, sexual function, or quality of life. The registry’s broader study description concerns sexual interest, but the primary outcome does not directly measure that entire construct. Secondary outcomes on Sexual Quality of Life (QoL) and Grade 3+ adverse events are listed without posted results, so they cannot be used to broaden the interpretation.

The observed counts are based on the arm entries listed in the primary outcome record. The record does not explain the relationship between those entries and the enrollment figure, and it does not report missing-data methods or the analysis population. Those limitations, together with the small number of threshold events, make the result best understood as an imprecise comparison rather than a definitive estimate.

16. Why This Trial Matters Statistically

This trial illustrates how a randomized design can yield a comparison that remains statistically uncertain when the outcome is uncommon in the observed groups. Randomization supports a between-arm comparison in principle, but it does not guarantee that a modest trial will estimate an effect precisely. Here, the reported threshold counts are 1 and 3, and the risk-difference interval includes both zero and values on either side.

The example also shows why effect estimates and confidence intervals deserve attention alongside a p-value. The unadjusted p-value is 0.226, but that single statistic does not say how large the difference might be or how uncertain its direction is. The risk difference and relative measures each provide a different scale for describing the same comparison, while their broad intervals make the limited precision visible.

There is a practical distinction between an outcome that is easy to classify and one that fully captures the research question. The 3+ threshold creates a clear binary endpoint for a categorical test, but it does not preserve the detailed frequency of attempts or measure all aspects of sexual interest. That distinction is important when interpreting what the statistical result represents.

The record also underscores the value of complete analysis reporting. The planned test options are described, but the test actually used, missing-data procedures, analysis population, and multiplicity strategy are not reported in the registry record. Without those details and the secondary outcome results, readers have a narrower basis for evaluating the full analysis than the study’s broad aims might suggest.

17. 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 primary outcome has only 1 participant meeting the threshold in the flibanserin-plus-ADT arm and 3 in the placebo-plus-ADT arm. Such sparse counts make estimates sensitive to individual observations and contribute to broad confidence intervals. The risk-difference interval includes zero and values in both directions, so the direction and size of the underlying difference remain uncertain.

The endpoint reduces reported attempts to two categories, 0-2 and 3+. This simplifies comparison but does not describe the full range of reported activity or every dimension of sexual interest. The registry does not report how the measure was collected, how often participants were assessed within the stated timeframe, or how incomplete reports were treated.

The registry lists an enrollment of 44, while the primary outcome entries list 22 participants in one arm and 19 in the other. It does not explain the relationship between enrollment and these outcome counts or provide a participant-flow account. This observation alone does not establish why the counts differ, but it limits the ability to trace who contributed to the outcome analysis.

Other details that would aid interpretation are not reported in the registry record: the analysis population, missing-data handling, the test actually used, any adjusted analysis, multiplicity procedures, and results for the listed secondary outcomes. The masking descriptions also differ: the masking field says quadruple, while the brief summary says double-blinded, without clarifying the roles masked. These gaps constrain assessment of the analysis and the breadth of conclusions that can be drawn.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Flibanserin in Men Receiving Androgen Suppression for Prostate Cancer — NCT04743934.
• Registry record last updated 2026-07-20; status last verified 2026-06.
• Results first posted 2026-07-20.
• 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.

21. Record Summary

The registry record describes a completed randomized, parallel-group phase II trial of flibanserin plus ADT versus placebo plus ADT in men with prostate cancer. Enrollment is listed as 44. The primary outcome counts participants reporting attempts at sexual intercourse at least 3 times in the prior month, from baseline up to 12 weeks.

The posted counts are 1 of 22 participants in the flibanserin arm and 3 of 19 in the placebo arm. The risk difference is -0.112, with a 95% confidence interval from -0.298 to 0.073 and an unadjusted p-value of 0.226. The interval includes zero and reflects substantial uncertainty.

Two secondary outcomes are listed, but their results are not reported in the registry record. The record does not report the analysis population, missing-data procedures, or which of the described categorical tests was used. Serious adverse events and deaths are recorded as 0 in each arm; other adverse events are listed for 10 and 6 participants, respectively.