Safety and Efficacy of CDB-2914 in Comparison to Levonorgestrel for Emergency Contraception: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Safety and Efficacy of CDB-2914 in Comparison to Levonorgestrel for Emergency Contraception (NCT00551616), a randomized phase 3 study. The primary outcome is Number of Participants With Pregnancy.

2. Clinical Question

The broader study purpose refers to pregnancy prevention up to 5 days after unprotected intercourse, while the posted primary outcome specifies a timeframe of within 72 hours. These descriptions should be kept distinct: the primary comparison reported in the registry concerns the stated 72-hour timeframe. The record also names actual pregnancy rate compared with expected pregnancy rate and menstrual bleeding patterns as secondary outcomes. Detailed results for the actual-versus-expected pregnancy outcome are not reported in the registry record.

3. Trial Design

The registry describes a completed phase 3 study sponsored by HRA Pharma. It began in 2007-04, had a primary completion date of 2009-04, and recorded enrollment of 2221. The design is randomized, single masked, and parallel. These labels identify broad design features; the registry record does not specify who was masked or provide further details about the randomization procedure.

The two intervention groups were CDB-2914 and levonorgestrel. In the primary outcome analysis, the CDB-2914 group included 1104 participants and the levonorgestrel group included 1117. The endpoint is identified as the number of participants with pregnancy, with an intention-to-treat population and a timeframe of within 72 hours of unprotected intercourse. The record does not report the original sample-size justification or detailed intervention regimens.

The primary comparison is between participant counts in the two randomized groups. The posted risk difference is oriented as levonorgestrel relative to CDB-2914: the listed group rates are 0.027 and 0.018, respectively, and the estimated difference is 0.009. The estimate and its interval describe the observed comparison, but the registry does not report the protocol’s planned analysis details or decision criteria.

Menstrual bleeding patterns are also listed as a secondary outcome, measured in days within the menstrual cycle of the unprotected intercourse. The registry lists actual pregnancy rate compared with expected pregnancy rate as another secondary outcome, but its detailed results are not reported in the registry record. The masking roles, allocation-concealment procedures, study sites, and further design details 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

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Intention to treat

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 With Pregnancy, measured over Within 72 hours of unprotected Intercourse.

The primary endpoint is the number of participants with pregnancy within 72 hours of unprotected intercourse, among women described as receiving emergency contraception in that timeframe. It is recorded as a binary outcome: each participant either had a pregnancy event recorded or did not. The posted arm results therefore give event counts alongside the numbers of participants in each group.

A binary endpoint is useful for a direct clinical comparison because it expresses how many participants experienced the event of interest in each randomized group. The risk difference summarizes the absolute contrast between the group rates, with the comparison oriented as levonorgestrel relative to CDB-2914. Its sign indicates which group has the higher posted rate; the interval conveys how uncertain that estimated contrast is.

The endpoint is specifically tied to the registry’s stated 72-hour timeframe. Although the study purpose also mentions pregnancy prevention up to 5 days after unprotected intercourse, the posted primary outcome description does not use that broader timeframe. The registry identifies the primary analysis population as intention to treat, but details about handling participants with missing pregnancy outcomes are not reported in the registry record.

This endpoint measures pregnancy events, not every aspect of emergency contraception or menstrual experience. Menstrual bleeding patterns are listed separately as a secondary outcome. The primary endpoint alone cannot describe the full range of outcomes recorded in the study.

7. Results Posted in the Registry

• Levonorgestrel: 2.7% (30/1117) versus 1.8% (20/1104) on control; risk difference 0.9 percentage points (95% CI -0.4 to 2.1), risk ratio 1.48 (0.85–2.59); unadjusted two-proportion z-test p = 0.165.

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.

Menstrual Bleeding Patterns

• 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 deviations.
• Result:
  • Levonorgestrel: mean -1.20 versus 2.10 on control; difference -3.30 (95% CI -3.97 to -2.63), standardized effect d = -0.41; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Levonorgestrel than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The registry names actual pregnancy rate compared with expected pregnancy rate and menstrual bleeding patterns as secondary outcomes. Detailed results for the actual-versus-expected pregnancy comparison are not reported in the registry record. The posted menstrual bleeding analysis concerns a different outcome from the primary pregnancy endpoint and reports a between-group difference in the timing measure.

These secondary results add context about outcomes beyond the primary pregnancy comparison, but they do not answer the primary question in place of the pregnancy endpoint. A difference in menstrual timing is not itself evidence about the between-group pregnancy contrast. Similarly, the named actual-versus-expected outcome cannot be interpreted without its detailed results and analysis information.

Secondary findings generally carry less confirmatory weight when the registry does not report a prespecified testing hierarchy or multiplicity adjustment. Assessing multiple outcomes can increase the chance that at least one unadjusted result appears notable by chance. The registry’s posted unadjusted secondary result should therefore be treated as a distinct finding whose role in the study’s overall evidence is uncertain, rather than as a replacement for the primary result.

The number of outcomes and analyses recorded does not, by itself, show how the investigators planned to prioritize them. Details of the protocol’s hierarchy, whether outcomes were prespecified in the form reported, and any correction for multiple comparisons are not reported in the registry record. The secondary results may be informative, but their interpretation should remain separate from the primary estimate and account for that missing context.

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

• Actual Pregnancy Rate Compared to Expected Pregnancy Rate
• Menstrual Bleeding Patterns

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is binary: it counts participants with a pregnancy event. For a binary outcome, a group’s event rate is the number with the event relative to the number included in that group’s analysis. The registry posts the event counts, group denominators, and rates for CDB-2914 and levonorgestrel. These summaries describe the observed participants; they do not by themselves explain how missing outcome information was handled.

What does the risk difference express?

A risk difference subtracts one group’s event rate from the other group’s event rate. In the posted comparison, the reference group is CDB-2914 and the comparison is levonorgestrel, so a positive estimate corresponds to a higher observed rate in the levonorgestrel group. The estimated risk difference is 0.009. The measure is an absolute contrast, rather than a relative ratio: it describes the separation between the two rates on the rate scale.

Direction depends on the order of subtraction. A risk difference with the opposite group order would have the opposite sign, even though it describes the same underlying contrast. It is therefore important to read the group order together with the estimate. The posted interval also uses that same orientation.

How should the confidence interval be read?

A confidence interval is a range of values compatible with the observed data under the method used to calculate it. It communicates uncertainty around an estimate; it does not identify a range containing a fixed proportion of individual participant outcomes. The primary risk-difference interval runs from -0.004 to 0.021 and includes zero. Zero represents no difference in the two group rates under this measure. The interval therefore does not isolate a single direction of difference.

The unadjusted p-value for the primary comparison is 0.165. 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 does it measure the size or clinical importance of a difference. The registry calls this result unadjusted; a multiplicity adjustment is not reported in the registry record.

What does the secondary mean difference represent?

Menstrual bleeding patterns are reported as a continuous outcome in days, with group means and standard deviations. The mean difference compares the average values, ordered as levonorgestrel relative to CDB-2914. The posted estimate is -3.3 days, with a 95% confidence interval from -3.97 to -2.63. Its unadjusted p-value is <0.001. A mean difference describes a shift in group averages; it does not say that every participant experienced that shift. Standard deviations describe spread among participant values, whereas the confidence interval describes uncertainty in the estimated group contrast.

The record reports the standard deviations for the two arms and a pooled standard deviation used in the comparison. In general, pooling combines within-group variability for a comparison under assumptions about the groups and analysis. The registry does not report a detailed statistical analysis plan explaining the choice of method or any diagnostics for its assumptions. The posted mean comparison should therefore be understood through its reported estimate and interval rather than as evidence about every aspect of the distribution.

Why do assumptions matter?

Statistical methods rely on assumptions about how observations are generated and how the outcome behaves. For a binary endpoint, the analysis treats participants’ event status as the unit of outcome. If observations are dependent, or if the analyzed participants differ from the randomized population because of missing outcomes, uncertainty may not be represented as intended. The registry identifies the primary population as intention to treat, but further missing-data procedures and analysis details are not reported in the registry record.

For continuous outcomes, distributional shape and variability can affect how well a method’s uncertainty estimates behave. General statistical practice examines these features, but diagnostic results are not reported in the registry record. No unsupported claim about particular diagnostics, adjustments, interim analyses, or protocol methods can be made from the posted information.

There are several outcomes and comparisons in the registry record. When many hypotheses are assessed, the chance of at least one apparently unusual result can increase if each result is considered separately. A multiplicity strategy may control or otherwise address that issue, but such a strategy is not reported in the registry record. The unadjusted secondary result should be interpreted with this limitation, not as an automatically confirmatory finding.

11. Confidence Intervals and Precision

The primary risk-difference estimate is 0.009, and its 95% confidence interval extends from -0.004 to 0.021. The interval includes zero, so the posted data do not determine a single direction for the underlying difference. Its lower end is negative and its upper end is positive. The observed point estimate is positive, but it should not be interpreted without the uncertainty around it.

Confidence intervals communicate precision as well as direction. A narrower interval would represent a more tightly estimated contrast under the analysis method; a wider interval leaves a broader range of values compatible with the data. The primary interval spans values on both sides of no difference, so it allows for more than one interpretation of the underlying risk difference. It does not establish that the groups are equivalent, and it does not establish that they differ.

The registry also reports an unadjusted p-value of 0.165 for the primary comparison. This is consistent with the uncertainty reflected by an interval that includes zero, but the two summaries serve different explanatory roles: the interval shows a range of estimated contrasts, while the p-value summarizes evidence relative to a null model. Neither quantifies the clinical importance of a possible difference.

For menstrual bleeding patterns, the reported mean-difference interval is -3.97 to -2.63 days. That interval is distinct from the primary pregnancy interval and concerns a different outcome and measurement scale. The statistical precision of one endpoint does not transfer to another. More generally, confidence intervals depend on the analysis assumptions, and the registry does not provide enough detail to assess every aspect of those assumptions or the handling of missing 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 planning figures are illustrations based on the observed effects, not a report of the trial’s original design calculation. For the binary pregnancy comparison, the entries are participants per arm. The menstrual bleeding comparison is also expressed as participants per arm; these are not total event counts. No time-to-event planning figures are reported.

The stated 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 a stated fraction of the observed effect. The entries give the planned number of participants per arm for 80% power and for 90% power under those assumptions. In plain terms, power is the chance that a specified analysis would detect an effect of the size being planned for, if that effect were the true effect and the assumptions held. The 90% column uses a higher target probability of detecting that effect than the 80% column.

The planning illustrations consider 100%, 75%, and 50% of the observed effect. Smaller effects are harder to distinguish from random variation, so they require more participants under the same assumptions. A curve expressing the same planning relationship should be read as showing how required sample size changes as the effect to detect changes, not as a prediction of the results that a future study would observe.

These figures are not observed power. They are forward-looking planning illustrations tied to the observed effect and stated assumptions. The original trial’s sample-size justification and design assumptions are not reported in the registry record, so the illustrations should not be used to infer what the trial originally planned.

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

Worked calculation for this trial

1. Control proportion p₀ = 0.018; treatment proportion p₁ = 0.027; average p̄ = 0.022.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.2097 = 0.4109.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.2096 = 0.1764.
4. n = (0.4109 + 0.1764)² / (0.0087)² = 4,514.2, rounded up to 4,515 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 <- 0.0181159420289855                          # control proportion
p1 <- 0.026857654431513                          # 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(551616)
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 = 4,515 (80% power), 6,043 (90% power)
 75% of observed effect: n per arm = 7,645 (80% power), 10,234 (90% power)
 50% of observed effect: n per arm = 16,342 (80% power), 21,877 (90% power)
power.prop.test: 4514.2 
Simulated power at n = 4,515 per arm: 0.802 (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.

14. Multiplicity, Missing Data and Analysis Populations

The registry lists more than one outcome, including the primary pregnancy endpoint and secondary outcomes concerning actual versus expected pregnancy rate and menstrual bleeding patterns. It also records two analyses. When several outcomes or comparisons are assessed, interpreting each unadjusted p-value in isolation can overstate the strength of the overall evidence. A planned hierarchy or multiplicity procedure can clarify which comparisons are confirmatory and how uncertainty from multiple testing is handled.

A multiplicity strategy is not reported in the registry record. The posted primary comparison has an unadjusted p-value, and the menstrual bleeding comparison also has an unadjusted p-value. Without a reported testing plan, the role of these results in a family of planned hypotheses cannot be established. The actual-versus-expected pregnancy outcome is named, but detailed results are not reported in the registry record.

The primary analysis population is identified as intention to treat. This label indicates the stated population for the primary outcome, but the record does not describe how participants with missing pregnancy status were handled, whether any participants were excluded from the analysis, or what sensitivity analyses were performed. The posted group denominators do not supply those details by themselves. Missing-data methods and the amount of missing outcome information are not reported in the registry record.

Other details that could shape interpretation—including the prespecified statistical model, any covariate adjustment, interim analyses, and the trial’s original sample-size assumptions—are not reported in the registry record. Their absence should not be filled by guessing. The posted estimates and uncertainty measures remain interpretable as registry results, but the missing analysis context limits how fully they can be matched to the trial’s original protocol.

15. 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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: urinary tract infection (0.1%), right contact lense related corneal ulcer (0.1%), dizziness (0.1%), vomiting blood stained fluid (0.1%), molar pregnancy (0.1%).

The registry reports 3 participants with serious adverse events in the CDB-2914 group, among 1104 at risk, and 4 in the levonorgestrel group, among 1117 at risk. It reports 0 deaths in each group. These are event-affected participant counts as recorded in the registry; they do not, by themselves, describe the nature, timing, duration, or relatedness of each event.

The record separately reports other adverse events affecting 596 participants in the CDB-2914 group and 572 in the levonorgestrel group. These other-event figures should not be combined with the serious-event figures, because the registry presents them as separate categories. The serious-event terms listed include urinary tract infection, right contact lense related corneal ulcer, dizziness, vomiting blood stained fluid, molar pregnancy, and ruptured ovarian cyst; each is listed as 0.1%. The record does not specify the arm attribution for those terms in the information posted.

Counts are descriptive and do not establish whether an event was caused by an intervention. Further definitions, event narratives, exposure-adjusted analyses, and details of seriousness assessment are not reported in the registry record. The figures should be read as the recorded safety summaries, not as a complete account of comparative safety.

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

The primary registry result is a comparison of pregnancy counts in the two randomized groups within the stated 72-hour timeframe. The observed risk-difference estimate favors a higher recorded rate in the levonorgestrel group in this particular comparison, but the confidence interval includes zero and extends across values in both directions. The unadjusted p-value is 0.165. Together, these statistics describe an uncertain estimate rather than a definitive conclusion about the direction or size of the underlying difference.

The result does not establish that the groups have identical pregnancy rates. An interval that includes zero is not proof of equivalence; equivalence requires a prespecified margin and an analysis designed to assess it, and such a margin is not reported in the registry record. Nor does the observed point estimate alone establish a reliable difference. Its uncertainty must be considered alongside the point estimate.

The secondary menstrual bleeding finding addresses a different outcome, and its result cannot be substituted for the primary pregnancy comparison. The registry also names an actual-versus-expected pregnancy outcome without detailed results. The safety summaries are descriptive counts; they do not establish causation or provide a complete account of event severity and exposure.

The registry’s broad study purpose mentions prevention of pregnancy up to 5 days after unprotected intercourse, while the primary endpoint is specified within 72 hours. The primary result should be interpreted according to its recorded timeframe rather than extended beyond it. Further participant characteristics, missing-outcome methods, and the full statistical plan are not reported in the registry record. These limits matter when assessing how broadly the posted comparisons apply.

17. Why This Trial Matters Statistically

This trial illustrates why a point estimate should be read together with its confidence interval. The primary risk-difference estimate has a positive direction, but its interval includes zero and extends to values in the opposite direction. Reading only the point estimate would conceal this uncertainty; reading only the p-value would not show the range of contrasts compatible with the estimate.

It also illustrates the distinction between a primary outcome and secondary outcomes. Pregnancy is the stated primary endpoint, while menstrual bleeding patterns and actual pregnancy rate compared with expected pregnancy rate are listed as secondary outcomes. A secondary result can add context, but it does not resolve uncertainty in a different primary outcome. The absence of a reported multiplicity strategy further limits the confirmatory weight that can be assigned to unadjusted secondary findings.

The record also makes clear that study design labels and posted summaries do not supply every detail needed for a full statistical appraisal. The trial is recorded as randomized, single masked, and parallel, and the primary population is identified as intention to treat. But the masking roles, missing-data procedures, original sample-size justification, and detailed analysis plan are not reported in the registry record. Statistical interpretation must distinguish what is recorded from what remains unknown.

Finally, planning illustrations based on observed effects are not the same as the original design calculation or a measure of the study’s observed power. They show how sample-size requirements change under stated assumptions and effect sizes. Keeping those concepts separate helps prevent retrospective planning figures from being mistaken for evidence about the strength of the trial’s results.

18. 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 timeframe is within 72 hours of unprotected intercourse, whereas the study purpose also refers to prevention up to 5 days. The posted primary result should therefore not be assumed to cover the broader timeframe. The registry does not provide enough detail to determine how outcomes outside the stated primary timeframe relate to the reported comparison.

Participant characteristics beyond the described population are not reported in the registry record. The record also does not detail study locations, dosing regimens, masking roles, or the allocation procedure. Those omissions limit assessment of how the recorded study population and conduct relate to other settings or populations.

Methods for missing outcomes, protocol deviations, and sensitivity analyses are not reported in the registry record. Although the primary population is identified as intention to treat, the label alone does not explain how every participant’s outcome contributed to the analysis. A detailed multiplicity plan and the trial’s original sample-size justification are also not reported. These gaps limit a full appraisal of the prespecified inferential framework.

The actual-versus-expected pregnancy outcome is listed among the secondary outcomes, but its detailed results are not reported in the registry record. The adverse-event summaries provide counts, but not detailed narratives, timing, or attribution. Finally, the unadjusted secondary result should be interpreted cautiously because the record does not explain whether a hierarchy or correction for multiple outcomes was planned.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Safety and Efficacy of CDB-2914 in Comparison to Levonorgestrel for Emergency Contraception — NCT00551616.
• Registry record last updated 2022-05-03; status last verified 2022-04.
• Results first posted 2022-03-16.
• 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.

22. Record Summary

NCT00551616 is a completed phase 3 randomized, single-masked, parallel study sponsored by HRA Pharma. The record lists enrollment of 2221 and compares CDB-2914 with levonorgestrel for emergency contraception. The primary outcome is the number of participants with pregnancy within 72 hours of unprotected intercourse, analyzed in an intention-to-treat population.

The posted primary comparison includes 20 pregnancy events among 1104 participants in the CDB-2914 group and 30 among 1117 in the levonorgestrel group. The risk difference is 0.009, with a 95% confidence interval from -0.004 to 0.021 and an unadjusted p-value of 0.165. Menstrual bleeding patterns are also reported as a secondary outcome; detailed results for actual pregnancy rate compared with expected pregnancy rate are not reported in the registry record.

The registry records serious adverse event counts of 3 and 4, with 0 deaths in each group. Further details about missing data, multiplicity, masking roles, and the original sample-size justification are not reported in the registry record.