A Study of UCB0942 in Adult Patients With Highly Drug-resistant Focal Epilepsy: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of UCB0942 in Adult Patients With Highly Drug-resistant Focal Epilepsy (NCT02495844), a randomized phase 2 study. The primary outcome is 75 % Responder Rate During the 2-week Inpatient Period.

2. Clinical Question

The clinical question was whether the percentage of participants reaching a substantial reduction in focal seizure frequency differed between the treatment groups during the 2-week inpatient period. The registry defines a responder as a participant whose focal seizure frequency fell by 75 % or greater compared with the Baseline Period. During that inpatient period, one group received placebo and the other received UCB0942; after the inpatient period, participants in the placebo group were to receive UCB0942.

The comparison therefore focuses on the initial inpatient period, when the groups received different assigned treatments. Later outcomes need a different interpretation because participants in the original placebo group were to switch to UCB0942. The registry record does not state a more detailed primary hypothesis or a clinically meaningful threshold for the difference between groups.

3. Trial Design

This was a phase 2, randomized, parallel-group, placebo-controlled study, with quadruple masking recorded in the registry. The official title describes a double-blind study of efficacy, safety, tolerability, and pharmacokinetic profile. The brief summary says the study took place across multiple centers in Europe. The registry lists 55 participants as enrolled and marks the study completed.

The treatment sequences were UCB0942/UCB0942 and Placebo/UCB0942. The placebo group was scheduled to receive UCB0942 after the 2-week inpatient period. That sequence matters when interpreting results after the inpatient period: those results do not represent a continued comparison of UCB0942 against placebo in the same way as the primary endpoint.

The primary endpoint was assessed during the 2-week inpatient period. The registry lists 27 participants for the Placebo/UCB0942 FAS group and 26 for the UCB0942/UCB0942 FAS group for that result. These are the reported endpoint analysis counts, not a complete account of how participants were randomized, treated, or included in every analysis. The exact sequence of randomization, allocation concealment procedures, dose schedules, and detailed eligibility criteria is not reported in the registry record.

Enrollment is listed as 55. The registry also reports different analysis counts for some outcomes and for safety, so those counts should not be treated as interchangeable. The reason for each difference in analysis population is not reported in the registry record. The record gives the study’s status and dates, but does not describe the protocol’s sample-size justification, planned interim analyses, or full statistical analysis plan.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Before assignment: Participant Flow refers to the Randomized Set.

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.

Population: Baseline Characteristics refer to the Safety Set consisting of all subjects in the Randomized Set who received at least 1 dose of Investigational Medicinal Product (IMP).

6. Primary Endpoint

The registry lists the primary outcome as 75 % Responder Rate During the 2-week Inpatient Period, measured over During the 2-week Inpatient Period.

The primary endpoint is the 75 % responder rate during the 2-week inpatient period. It is a binary outcome expressed as the percentage of participants who met the responder definition. A participant qualified as a responder if focal seizure frequency during that period was reduced by 75 % or greater compared with the Baseline Period.

This endpoint turns each participant’s change in seizure frequency into a yes-or-no classification, then summarizes the proportion meeting the threshold in each group. That makes the result interpretable as a responder rate, but it does not show how large the reduction was among responders or how seizure frequency changed for participants who did not meet the threshold. Those details would require other outcome summaries.

General statistical context: a threshold-based endpoint can make a group contrast easier to communicate, but it also depends on the chosen threshold and observation window. Participants whose changes fall on opposite sides of the cutoff can be classified differently even when their underlying changes are similar. The registry lists other seizure-frequency and responder outcomes as secondary endpoints, but the primary endpoint is specifically tied to the inpatient period.

The endpoint addresses the proportion meeting the specified response definition, not every aspect of efficacy, safety, tolerability, or pharmacokinetics. The registry record does not report a rationale for selecting the 75 % cutoff or explain how the endpoint was chosen for the primary analysis.

7. Results Posted in the Registry

• UCB0942/UCB0942 (FAS): 30.8% (8/26) versus 11.1% (3/27) on control; risk difference 19.7 percentage points (95% CI -1.7 to 41.0), risk ratio 2.77 (0.82–9.31); unadjusted two-proportion z-test p = 0.078.

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.

75 % Responder Rate During the Last 4 Weeks of the Outpatient Maintenance Period

• 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:
  • UCB0942/UCB0942 (FAS): 29.2% (7/24) versus 33.3% (9/27) on control; risk difference -4.2 percentage points (95% CI -29.6 to 21.3), risk ratio 0.88 (0.39–1.99); unadjusted two-proportion z-test p = 0.749.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between UCB0942/UCB0942 (FAS) and the control. The interval gives the range of effects that remain compatible with the data.

75 % Responder Rate During the On-UCB0942 Overall Period

• 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:
  • UCB0942/UCB0942 (FAS): 25.9% (7/27) versus 25.9% (7/27) on control; risk difference 0.0 percentage points (95% CI -23.4 to 23.4), risk ratio 1.00 (0.41–2.46); unadjusted two-proportion z-test p = 1.000.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between UCB0942/UCB0942 (FAS) and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes extend the study’s description beyond the inpatient primary endpoint. They include responder-rate outcomes at later periods, seizure-frequency summaries, and seizure-free rates. The record lists 12 posted secondary outcomes, but only some of the secondary responder-rate comparisons have numerical results and recalculations detailed in the available data. Conclusions about the full set of secondary endpoints cannot be drawn from a subset of reported comparisons.

Across the posted secondary responder results, the pattern does not show a consistent separation between the original treatment-sequence groups. During the last 4 weeks of the outpatient maintenance period, the reported responder percentages were 33.3% for Placebo/UCB0942 and 29.2% for UCB0942/UCB0942. During the On-UCB0942 Overall Period, both groups had a reported responder rate of 25.9%. These later outcomes should not be treated as straightforward replications of the primary comparison, because the placebo group was scheduled to receive UCB0942 after the inpatient period.

The primary endpoint concerns the inpatient period, when the groups differed in assigned treatment. The later secondary outcomes address different time windows and treatment histories. They therefore add context but do not replace the primary result or resolve its uncertainty. The registry’s 75% responder outcomes and other listed secondary measures also assess different aspects of seizure response, and their results should not be collapsed into one overall effect.

General statistical context: when many outcomes are examined, some estimates can look unusual by chance even if there is no underlying effect for those outcomes. If results are treated as exploratory without adjustment for multiple comparisons, individual p-values carry less confirmatory weight. The registry record does not report a multiplicity adjustment or a hierarchy for interpreting secondary endpoints. Accordingly, unadjusted secondary results are best treated as descriptive and hypothesis-generating rather than as independent confirmation of the primary finding.

The change from placebo to UCB0942 in the original placebo group also affects what later comparisons mean. Those comparisons are between treatment sequences or exposure histories, not a continuing contrast of UCB0942 against placebo. This distinction is essential when relating the secondary outcome pattern to the primary inpatient result.

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

• Median Percent Change in Weekly Focal Seizure Frequency During the 2-week Inpatient Period
• Median Percent Change in Weekly Focal Seizure Frequency During the Outpatient Maintenance Period
• Median Percent Change in Weekly Focal Seizure Frequency During the On-UCB0942 Overall Period
• Seizure-free Rate (All Seizures) During the 2-week Inpatient Period
• Seizure-free Rate (All Seizures) During the Last 4 Weeks of the Outpatient Maintenance Period
• Seizure-free Rate (All Seizures) During the On-UCB0942 Overall Period
• 75 % Responder Rate During the Last 4 Weeks of the Outpatient Maintenance Period
• 75 % Responder Rate During the On-UCB0942 Overall Period

9. Statistical Interpretation

10. Statistical Methods Explained

The primary endpoint is binary at the participant level: each participant either met the 75 % responder definition during the inpatient period or did not. The analysis then compares the groups’ proportions of responders. The registry reports the percentages and group counts, and the sponsor’s analysis is described as logistic regression with an odds ratio as its effect measure.

What does an odds ratio describe?

General statistical context: odds are the probability of an event divided by the probability of its non-occurrence. An odds ratio compares those odds between groups. An odds ratio of 1 corresponds to equal odds under the comparison; values above or below that reference indicate different estimated odds. An odds ratio is not the same as a difference in percentages or a ratio of probabilities. In particular, readers should not translate the reported odds ratio directly into a percentage-point change.

The registry reports an odds ratio of 4.14 for the logistic regression analysis, with a 95% confidence interval from 0.90 to 19.06 and p = 0.068. The record does not identify the model’s covariates or provide further details of its specification. Consequently, the reported result can be described as a logistic-regression odds ratio, but the record does not support a more specific account of adjustment or model choices.

How does the unadjusted recalculation differ?

The unadjusted recalculation uses the posted counts and compares the observed proportions directly. Its effect measure is the risk difference, reported as 0.197, with a confidence interval from -0.017 to 0.410 and p = 0.078. This is a simple comparison of the group proportions, not the sponsor’s logistic-regression analysis. The recalculation does not add covariates or otherwise reproduce an analysis whose model details are not reported. The registry does not identify a discrepancy between its reported analysis and the recalculation.

General statistical context: an unadjusted comparison can be transparent and easy to relate to observed proportions, while a regression model can estimate a contrast using its specified model structure. The estimates may differ because the effect measures and methods differ. Their uncertainty intervals and p-values should be interpreted in the context of the method that produced them. A result from one method should not be presented as though it were the result of the other.

What does a confidence interval contribute?

General statistical context: a confidence interval is an uncertainty summary for an estimate under the chosen method and its assumptions. It is not a range that guarantees the true value lies within it, and it is not a measure of the probability that the treatment effect has a particular value. A wide interval indicates that the data are compatible with a broad range of effect sizes. When an interval includes the no-difference value for its measure, the data do not sharply distinguish that value from other values in the interval.

For the risk difference, the no-difference value is zero; for the odds ratio, it is one. The reported intervals span those reference values. This means the data do not rule out no difference under either stated interval, while also allowing estimates in the direction of a higher response in the UCB0942/UCB0942 group. The interval widths indicate that the amount of information is limited for establishing the size of the contrast precisely.

What do the p-values say?

General statistical context: a p-value is calculated relative to a null hypothesis and a statistical model. It summarizes how compatible the observed data are with that reference, subject to the method’s assumptions. It does not measure the probability that a treatment is effective, nor the size or clinical importance of an observed difference. The sponsor analysis reports p = 0.068, and the unadjusted recalculation reports p = 0.078. The registry does not state the decision threshold for the primary analysis, so these values should be reported as estimates of statistical evidence rather than used to infer an unreported decision rule.

For a binary endpoint, the observed percentages are also shaped by the number of participants. A small number of outcomes can lead to meaningful movement in the estimated proportion when one participant’s classification changes. General statistical context: this is one reason that confidence intervals and counts are useful alongside percentages and p-values. In this trial, the endpoint analysis counts were 27 and 26, and the interval estimates are broad.

The registry labels the endpoint analysis groups as FAS. The exact definition of FAS, rules for inclusion, and handling of missing endpoint data are not reported in the registry record. Therefore, interpretation should not assume an unstated intention-to-treat definition or imputation procedure. The reported values are the results for the populations and methods identified in the record.

11. Confidence Intervals and Precision

The primary endpoint estimate is uncertain. The sponsor-reported odds ratio is 4.14, with a 95% confidence interval from 0.90 to 19.06. This interval spans the no-difference reference for an odds ratio and includes a wide range of possible contrasts. The point estimate alone therefore gives an incomplete picture: the data do not estimate the size of the odds contrast precisely.

The unadjusted risk difference is 0.197, with a confidence interval from -0.017 to 0.410. For a risk difference, zero represents no difference between the proportions. Because the interval includes zero and also extends into positive values, it is compatible with no difference as well as with a range of positive differences under the unadjusted method. It is not appropriate to treat the observed estimate as though it were known without uncertainty.

General statistical context: a 95% confidence interval is constructed using a method that, under its assumptions, would capture the target parameter in 95% of repeated samples. It does not mean there is a 95% probability that the parameter lies within this particular interval. Its practical role is to convey the precision of the estimate and the values not ruled out by the observed data and method.

The analysis groups were small, with 27 and 26 participants in the primary result. The broad intervals are consistent with limited precision. The trial’s total enrollment was 55, but not every outcome uses identical analysis counts. The record does not explain the reason for each change in denominator. Interpretations of precision should therefore remain tied to the specific endpoint and analysis population rather than treating enrollment as the denominator for every estimate.

The sponsor and unadjusted intervals arise from different effect measures and methods. Their limits should not be lined up as if they estimate the same quantity. Together, they indicate that the posted primary comparison is uncertain about the magnitude of any between-group difference.

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 planning illustrations are based on the observed effect and are expressed as participants per arm for 80% power and participants per arm for 90% power. They are not total events: the stated convention for this binary endpoint is participants per arm. The listed assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. Each planning row uses a stated fraction of the observed effect as the effect to detect.

The 80% power column describes the planned number of participants per arm under those assumptions for a design with an 80% chance of detecting the specified effect if that effect is the true one and the assumptions hold. The 90% power column describes the corresponding planning target for a 90% chance. These are design probabilities, not guarantees about the result of a future study. For a given assumed effect, the 90% target is larger because greater power requires more information.

The planning rows consider 100% of the observed effect, 75% of the observed effect, and 50% of the observed effect. The figures rise as the assumed effect becomes smaller: detecting a smaller contrast requires more participants per arm under the stated calculation method. The particularly large planning estimates for the secondary endpoint reflect the small observed contrast used in those calculations. They should not be read as estimates of the original study’s planned enrollment.

These are planning illustrations, not observed power. The trial’s own original sample-size assumptions and justification are not reported in the registry record. The illustrations also depend on the observed effect, which is uncertain in a small trial; they should not be treated as reliable forecasts of a future study’s effect or as evidence that the completed study was adequately or inadequately powered.

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 = 0.111; treatment proportion p₁ = 0.308; average p̄ = 0.209.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.5754 = 1.1278.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.5584 = 0.4699.
4. n = (1.1278 + 0.4699)² / (0.1966)² = 66.1, rounded up to 67 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 <- 0.111111111111111                          # control proportion
p1 <- 0.307692307692308                          # 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(2495844)
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 = 67 (80% power), 88 (90% power)
 75% of observed effect: n per arm = 108 (80% power), 144 (90% power)
 50% of observed effect: n per arm = 218 (80% power), 291 (90% power)
power.prop.test: 66.1 
Simulated power at n = 67 per arm: 0.816 (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. Analyses Reported by the Sponsor

The registry’s reported primary analysis is a logistic regression with an odds ratio of 4.14, a 95% confidence interval from 0.90 to 19.06, and p = 0.068. The separate recalculation is an unadjusted risk difference of 0.197, with a confidence interval from -0.017 to 0.410 and p = 0.078. The sponsor’s analysis and the recalculation describe different effect measures and use different methods.

The sponsor’s odds ratio compares odds, whereas the recalculated risk difference compares proportions directly. Their point estimates are therefore not on a common scale, and differences between them do not by themselves indicate a disagreement in the underlying data. The registry does not report the logistic model’s covariates or full specification, so no further explanation of the difference can be established from the record.

The registry does not flag a discrepancy between the sponsor analysis and the unadjusted recalculation. The sponsor’s result is the trial’s reported analysis; the recalculation is a transparent, simple comparison of posted summary counts. Both intervals include their respective no-difference reference values. They should be presented as complementary descriptions of the same primary endpoint, not as interchangeable tests or as competing estimates from identical analyses.

15. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when a study evaluates several endpoints, time windows, or treatment contrasts. General statistical context: as more hypotheses are examined, the chance of at least one apparently notable result from random variation can increase. A prespecified strategy, such as a hierarchy or an adjustment procedure, can help control how confirmatory claims are made across multiple analyses. The registry lists 13 outcome measures and 12 posted secondary outcomes, but a multiplicity strategy is not reported in the registry record.

The registry reports one primary endpoint analysis and a number of secondary outcomes, including seizure-frequency, seizure-free, and responder measures across different periods. It does not state whether secondary endpoints were formally ordered, whether tests were adjusted, or how the overall family of comparisons was defined. The record also does not report interim analyses. These omissions do not establish that such procedures were absent from the full protocol; they mean that their use cannot be confirmed from the registry record.

Missing-data handling is also not reported. The primary outcome counts are 27 and 26, while other endpoints have different analysis counts. The safety groups have their own at-risk counts. The reason for differing denominators, the amount and pattern of missing data, and any imputation or sensitivity analyses are not reported in the registry record. As a result, it is not possible to determine from these results alone whether missing observations could affect the estimated response proportions.

The endpoint groups are labeled FAS, but the registry record does not give a definition of that analysis set or state the exact inclusion rules. It should not be assumed that FAS has a particular operational definition beyond the label shown. Nor is a per-protocol analysis or a detailed intention-to-treat analysis described in the posted material.

General statistical context: prespecification matters because choices made after examining outcomes can affect false-positive risk and interpretation. Without a reported statistical analysis plan, readers cannot assess whether the endpoint definitions, analysis populations, modeling choices, and multiplicity procedures followed a prespecified strategy. The primary and secondary estimates remain interpretable as posted results, but the registry record does not support stronger claims about the analysis plan or confirmatory status of the multiple secondary comparisons.

16. 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: judgement impaired (3.6%), status epilepticus (3.6%), delirium (3.6%), dysphoria (3.6%).

The safety results use groups labeled Placebo/UCB0942 and UCB0942/UCB0942, with 27 and 28 participants at risk for serious adverse events, respectively. The registry reports 0 participants with serious adverse events in the Placebo/UCB0942 group and 2 in the UCB0942/UCB0942 group. It reports 0 deaths in each group. The other adverse-event counts are 26 of 27 and 27 of 28, respectively.

The registry lists Judgement impaired, Status epilepticus, Delirium, and Dysphoria among the serious adverse-event terms, each with a reported frequency of 3.6%. Those terms should be read as reported categories; the record does not establish that they represent distinct participants or explain their relationship to the group-level count of participants with serious adverse events.

General statistical context: safety counts from a small study can describe what was observed, but they are not precise estimates of uncommon risks. These results do not establish whether events were caused by a treatment. The registry record does not provide further detail about event timing, severity beyond the serious classification, relatedness assessments, or the specific safety-analysis methods. The at-risk counts differ from the primary endpoint analysis counts and should not be substituted for them.

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

The primary endpoint shows a higher observed 75 % responder rate during the inpatient period in the UCB0942/UCB0942 group than in the Placebo/UCB0942 group: 30.8% compared with 11.1%. The sponsor-reported logistic-regression analysis estimates an odds ratio of 4.14, while the unadjusted comparison estimates a risk difference of 0.197. These are different effect measures, and both confidence intervals include their respective no-difference values.

The findings do not establish the precise size of a treatment contrast. The intervals are broad, and the primary analysis groups include 27 and 26 participants. A point estimate that favors one group does not remove this uncertainty. Conversely, an interval that includes no difference does not prove that the groups are identical. The posted data are compatible with a range of effects under the stated analyses.

The endpoint has a specific meaning: reaching a 75 % or greater reduction in focal seizure frequency compared with the Baseline Period during the 2-week inpatient period. It does not summarize the magnitude of change for every participant, and it does not by itself describe outcomes over the full course of treatment. The group assigned to placebo during the inpatient period was scheduled to receive UCB0942 afterward, so later results do not maintain the same placebo comparison.

The registry also reports safety events, but the small groups and limited event counts do not support precise estimates of uncommon harms. The record does not report enough detail to establish causation for specific events or to reconstruct all analysis decisions. The results apply to the studied population, treatment sequences, endpoint, and observation periods; they should not be extended beyond those conditions without additional evidence.

These statistical results are not medical advice and do not determine an individual treatment decision. They describe the pattern and uncertainty in the trial’s posted data. Any broader clinical interpretation would need to consider the full evidence base, details not reported in the registry record, and clinical judgment.

18. Why This Trial Matters Statistically

This trial illustrates why clinical-trial interpretation should distinguish an observed difference from the precision of that estimate. The inpatient responder percentages differ between groups, but the sponsor-reported odds-ratio interval and the unadjusted risk-difference interval both include the reference value for no difference. Reading only the group percentages or the point estimate would leave out important uncertainty.

It also demonstrates why effect measures need to be named. An odds ratio from logistic regression is not interchangeable with an unadjusted risk difference from the observed proportions. The two summaries answer related but distinct questions and use different scales. Clear reporting of the measure and method helps prevent direct numerical comparisons that are not meaningful.

The treatment sequence is another important statistical feature. The primary endpoint compares groups during the inpatient period, while the placebo group was scheduled to change to UCB0942 afterward. Later outcomes consequently describe a different comparison than the primary endpoint. A reader needs the design sequence to understand what a group label means at each time point.

The registry also illustrates limits on interpretation when key statistical-plan details are unavailable. The record does not state the original sample-size justification, missing-data methods, multiplicity strategy, or full regression specification. The posted outcomes can still be described, but the absence of these details constrains assessment of prespecification and confirmatory strength.

Finally, small analysis groups make confidence intervals particularly important. Planning illustrations based on the observed effect show how required participant counts change with the effect assumed, but they are not a substitute for the trial’s original design rationale. Together, these points show how endpoint definitions, analysis methods, treatment sequence, and precision shape the meaning of a trial result.

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.
• Participant counts were reconstructed from posted percentages and arm sizes, so small rounding differences are possible.

Several limitations are specific to the information available for this study. The primary endpoint is based on a threshold response during a 2-week inpatient period. It does not, on its own, describe the full distribution of changes in seizure frequency or establish whether a result persisted over later periods. The registry lists other outcomes intended to address different time windows, but not all their numerical results are detailed in the available record.

The treatment sequence complicates interpretation of later outcomes. Participants assigned to placebo during the inpatient period were to receive UCB0942 afterward. Thus, outpatient and on-UCB0942 outcomes cannot be read as a continuing placebo-controlled comparison. The record does not provide enough detail to determine how treatment exposure, adherence, or other post-inpatient factors were handled in those analyses.

The primary endpoint groups are small, and their confidence intervals are broad. The study enrolled 55 participants, but the analysis counts vary by endpoint and by safety analysis. Reasons for those denominator differences, missing-data amounts, and missing-data handling are not reported in the registry record. The FAS label is not accompanied by a definition.

Important design and analysis details are also not reported, including the original sample-size justification, the complete statistical analysis plan, model covariates, multiplicity procedures, and whether interim analyses were conducted. The record does not provide a detailed rationale for the endpoint threshold or the clinical importance of a particular between-group contrast. These gaps limit conclusions about precision, prespecification, and the broader relevance of the estimated effects.

Safety summaries are descriptive and based on small at-risk groups. The registry does not provide additional details needed to assess event timing, relatedness, or how multiple serious-event terms map to participants. These limitations do not negate the posted results, but they set boundaries on what can be inferred from them.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Study of UCB0942 in Adult Patients With Highly Drug-resistant Focal Epilepsy — NCT02495844.
• Registry record last updated 2019-02-05; status last verified 2019-02.
• Results first posted 2019-02-05.
• 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 describes a completed phase 2, randomized, parallel, placebo-controlled study of UCB0942 in adults with highly drug-resistant focal epilepsy. The study used quadruple masking and enrolled 55 participants. The primary endpoint was the 75 % responder rate during a 2-week inpatient period, defined as a 75 % or greater reduction in focal seizure frequency compared with the Baseline Period.

The primary analysis groups reported responder rates of 11.1% for Placebo/UCB0942 and 30.8% for UCB0942/UCB0942. The sponsor-reported logistic regression gave an odds ratio of 4.14, a 95% confidence interval from 0.90 to 19.06, and p = 0.068. An unadjusted recalculation gave a risk difference of 0.197, with a confidence interval from -0.017 to 0.410 and p = 0.078. Both intervals include their no-difference reference values.

The placebo group was scheduled to receive UCB0942 after the inpatient period, so later outcomes reflect different treatment histories. The registry does not report several details needed to assess the original design assumptions, missing-data handling, multiplicity procedures, or full analysis specifications.