Study to Evaluate the Safety and Efficacy of USL261 in Patients With Increased Bouts of Seizure Activity in the EMU: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Study to Evaluate the Safety and Efficacy of USL261 in Patients With Increased Bouts of Seizure Activity in the EMU (NCT01999777), a randomized phase 3 study. The primary outcome is Number of Participants That Were Seizure-free.

2. Clinical Question

3. Trial Design

This was a Phase 3, randomized, triple-masked, parallel-group trial comparing USL261 with intranasal placebo. The registry gives an enrollment of 62. The posted primary outcome counts are 17 of 31 participants for USL261 and 12 of 31 for placebo. The record marks the study as completed, with a start date of 2013-11 and primary completion in 2015-08. UCB Biopharma S.P.R.L. is listed as the sponsor.

The study’s stated purpose was to evaluate the efficacy, safety, and tolerability of USL261 compared with intranasal placebo for intermittent bouts of increased seizure activity. Its official title specifies treatment in the epilepsy monitoring unit. The intervention descriptions identify USL261 as 5 mg intranasal midazolam and the control as intranasal placebo. The design is parallel, meaning participants were assigned to a study arm rather than being described as switching between treatments. The allocation is recorded as randomized.

Triple masking indicates that the registry classifies the study as masked at three levels, but it does not say which people or roles those levels represent. The details of sequence generation, concealment, and how masking was maintained are not reported in the registry record. The protocol’s original sample-size assumptions, eligibility rules beyond the stated condition and setting, and any planned interim analyses are also not reported in the registry record.

The primary endpoint was assessed during a six-hour Treatment Phase and is labeled as a count of participants. The registry also lists time to first seizure following treatment as a secondary outcome. Although the record indicates that outcome measures and analyses were posted, numerical results for that secondary outcome are not included in the registry record. Therefore, this account interprets the primary binary outcome and the available safety counts without extending conclusions to unreported secondary results.

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: 188 eligible subjects entered Pretreatment Observation during which they were monitored in EMU for seizure events. Only subjects who met entry criteria and presented with seizure events meeting the treatment decision criteria were eligible to enter the Treatment Phase. Participant flow represents subjects who took at least 1 dose of study drug.

Overall Study

Analysis population for the primary outcome, as defined in the registry: Intent 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.

Population: The Baseline Characteristics refers to the Safety Analysis Set which included all subjects who took at least 1 dose of study drug.

6. Primary Endpoint

The registry lists the primary outcome as Number of Participants That Were Seizure-free, measured over 6 hours.

The primary endpoint counts participants who were seizure-free during the six-hour Treatment Phase. It is a binary outcome at the participant level: each person either met the complete seizure-free definition or was classified as having a seizure outcome for this analysis. The registry identifies the analysis population as intent to treat and reports counts for both arms.

The definition is broader than simply having no seizure recorded. A participant had to complete the Treatment Phase without recorded seizures, premature study-drug discontinuation, rescue intervention for acute central respiratory depression adverse event, or alteration to background anti-epileptic drug therapy. If any of those conditions occurred, the participant was included in the seizure-free analysis with the outcome of seizure. This operational rule makes the endpoint reflect both seizure observations and specified events that prevented a participant from satisfying the full treatment-phase definition.

A binary endpoint gives a direct comparison of how many participants in each arm met the stated criterion. It can be summarized as counts, proportions, and a between-group contrast such as a risk difference. Such summaries are useful for reading the observed group results, but the counts alone do not establish how precisely the underlying treatment-group difference is estimated. The registry does not provide additional detail about the rationale for selecting this endpoint or the original endpoint-specific analysis plan.

7. Results Posted in the Registry

• USL261: 54.8% (17/31) versus 38.7% (12/31) on control; risk difference 16.1 percentage points (95% CI -8.4 to 40.6), risk ratio 1.42 (0.82–2.45); unadjusted two-proportion z-test p = 0.203.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary endpoint is binary because each participant is classified as meeting or not meeting the seizure-free definition during the six-hour Treatment Phase. The record reports the number meeting the definition in each randomized group and the group sizes. A participant who did not satisfy every element of the definition was counted as having a seizure outcome, including when the reason was premature discontinuation, rescue intervention for acute central respiratory depression adverse event, or a change to background anti-epileptic drug therapy. This classification should be kept in mind when interpreting the endpoint as a count of seizure-free participants.

What does a risk difference describe?

A risk difference compares the proportion with the endpoint in one group with the proportion in the other group by subtraction. It is expressed on the same proportion scale as the endpoint. A positive estimate in this comparison corresponds to a greater observed seizure-free proportion in the USL261 group. The unadjusted recalculation reports a risk difference of 0.161. This is a calculated contrast from the posted group counts, not an adjusted estimate from the sponsor’s analysis. The registry’s own analysis should be distinguished from this simple recalculation because different procedures can produce somewhat different estimates or p-values.

General statistical explanation: a risk difference is often easier to interpret in absolute terms than a relative measure, because it describes the gap between group proportions. It does not, on its own, show whether the estimate is precise or whether the observed contrast is likely to persist beyond the trial participants. The confidence interval and the trial design provide context for that uncertainty.

What does the confidence interval add?

The recalculated interval runs from -0.084 to 0.406. It includes zero, which represents no risk difference, and spans both negative and positive values. Thus, the observed data do not tightly distinguish a difference in one direction from no difference or a difference in the other direction. The interval is broad relative to the point estimate, so considerable uncertainty remains about the magnitude and direction of the underlying contrast.

General statistical explanation: a confidence interval is a range produced by a method that reflects sampling uncertainty under its assumptions. It is not a guarantee that the true effect lies in the particular calculated interval, and it does not account automatically for every source of bias, measurement error, or departures from the analysis plan. The interval level and full calculation details are not reported in the registry record, so no nominal coverage level should be inferred.

How should the p-values be read?

The unadjusted recalculation reports p = 0.203. The registry reports a Wald asymptotic superiority analysis, with its p-value recorded to four decimal places. The two values are not identical. A p-value summarizes the compatibility of the observed data with a null hypothesis under the specified test; it is not an effect size and is not the chance that the null is true. A larger p-value does not establish equivalence, and a smaller p-value would not by itself establish clinical importance.

General statistical explanation: a Wald test uses an estimate and its estimated uncertainty to assess a null value, often by comparing their ratio on a standardized scale. The asymptotic label means the reference distribution is justified through an approximation that is expected to work under suitable conditions. With relatively small group counts, approximations can be sensitive to the method and assumptions. The record identifies the sponsor’s approach as Wald asymptotic but does not provide enough detail to reconstruct every implementation choice.

Why are the group counts important?

The outcome counts are 17 of 31 for USL261 and 12 of 31 for placebo. These are the observed results in the posted groups. They support a direct description of the trial data, but they do not eliminate random variation. In a binary outcome, the uncertainty in estimated proportions and their contrast depends in part on the numbers of participants and events. The modest counts make it especially important to interpret the interval rather than relying on the point estimate alone.

The trial is recorded as randomized, which supports a comparison based on assigned groups, but the registry does not report the detailed randomization procedure or allocation concealment process. The primary population is labeled intent to treat. The record does not provide a complete participant flow or enough detail to verify how all post-randomization events were handled beyond the endpoint rule. No adjustment variables or covariate-adjusted primary estimate are stated in the registry record. Accordingly, the recalculation is unadjusted and should not be presented as the sponsor’s adjusted or definitive estimate.

Multiplicity is relevant when several outcomes or analyses are examined, because repeated testing can increase the chance of at least one apparently notable result. A secondary outcome, time to first seizure following treatment, is listed, but its numerical result and any multiplicity adjustment are not reported in the registry record. No claim is made about whether the sponsor planned a hierarchy or adjustment. The interpretation of the primary result is therefore limited to the posted primary endpoint and reported analysis.

10. Confidence Intervals and Precision

The recalculated risk difference is 0.161, with a confidence interval from -0.084 to 0.406. The interval crosses zero and stretches across values in both directions. That means the observed data leave open the possibility of no underlying difference, as well as a difference favoring either group. The count in the USL261 group is higher in the observed trial results, but the interval shows why that count contrast should not be treated as a precise estimate of the effect.

Precision concerns the range of values that remain compatible with the data under the calculation method. A narrower interval would indicate less sampling uncertainty for the contrast; a wider interval indicates more. Here, the interval is wide enough that the direction and magnitude are not tightly determined. The registry record does not state the confidence level for this recalculated interval, so a nominal level is not assigned in this explanation.

General statistical explanation: confidence intervals describe uncertainty in an estimate under a statistical model and its assumptions. They do not capture every concern about trial conduct, endpoint definition, missing observations, or selection of analyses. Nor do they state the probability that a particular value is true. The interval should be considered with the design, the group counts, and the registry’s own analysis rather than as a stand-alone verdict.

The interval does not establish that the treatments are equivalent: equivalence requires a predefined margin and a design and analysis suited to that question, and such a margin is not reported in the registry record. It also does not establish a clinically important difference, because the registry does not specify a threshold for clinical importance. The safest statistical reading is that the observed groups differ in their counts, while the precision of the estimated contrast remains limited.

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 planning entries are participants per arm, not total events. They are illustrations based on the observed effect, not the trial’s original design calculation. For 100% of the observed effect, the entries are 150 participants per arm for 80% power and 200 participants per arm for 90% power. For 75% of the observed effect, they are 265 participants per arm for 80% power and 353 participants per arm for 90% power. For 50% of the observed effect, they are 590 participants per arm for 80% power and 789 participants per arm for 90% power.

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 the stated fraction of the observed effect. The trial’s own original sample-size assumptions are not reported in the registry record. These illustrations therefore should not be read as the study’s enrollment target rationale or as a reconstruction of the protocol plan.

In plain terms, the 80% power column gives the estimated number of participants per arm under those assumptions for a design with 80% power to detect the specified effect. The 90% power column gives the corresponding planning number for 90% power. Higher planned power requires more participants under the same assumptions. The entries also show that smaller effects need more participants: planning for 75% or 50% of the observed effect requires larger groups than planning for the full observed effect.

These are planning illustrations, not observed power. They do not indicate the probability that the completed trial’s result is correct or that a treatment works. Their values depend on using the observed effect as the starting point; that estimate itself is uncertain, as reflected in the broad confidence interval. A planning figure derived from an uncertain observed estimate should be treated as a scenario, not a promise about what a future study would show.

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 = 0.387; treatment proportion p₁ = 0.548; average p̄ = 0.468.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.7056 = 1.3830.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6964 = 0.5861.
4. n = (1.3830 + 0.5861)² / (0.1613)² = 149.0, rounded up to 150 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 <- 0.387096774193548                          # control proportion
p1 <- 0.548387096774194                          # 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(1999777)
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 = 150 (80% power), 200 (90% power)
 75% of observed effect: n per arm = 265 (80% power), 353 (90% power)
 50% of observed effect: n per arm = 590 (80% power), 789 (90% power)
power.prop.test: 149 
Simulated power at n = 150 per arm: 0.821 (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. Analyses Reported by the Sponsor

The registry’s own reported result is a Wald asymptotic superiority analysis comparing USL261 with placebo. Its p-value is recorded to four decimal places, whereas the simple unadjusted recalculation gives p = 0.203. They are not identical, so they should not be blended into a single result or described as if they came from the same calculation. The registry does not flag a discrepancy.

The recalculation is a straightforward comparison from the posted participant counts and is unadjusted. The sponsor’s stated method is a Wald asymptotic test, but the details necessary to reproduce every part of that analysis are not reported in the registry record. Differences in statistical procedure or implementation can lead to different p-values. The available record does not establish which particular calculation detail accounts for the difference.

Both results should be interpreted in context: the observed seizure-free count is higher in the USL261 group, while the recalculated interval includes zero and is broad. The registry’s own analysis remains the study’s reported analysis; the recalculation is supplementary and does not replace it. Neither p-value establishes clinical importance, equivalence, or a treatment effect with certainty.

14. Multiplicity, Missing Data and Analysis Populations

The primary analysis population is identified as intent to treat. The endpoint definition assigns a seizure outcome when a participant has premature study-drug discontinuation, rescue intervention for acute central respiratory depression adverse event, or alteration to background anti-epileptic drug therapy, as well as when seizures are recorded. That rule gives some information about how these circumstances affect endpoint classification. It does not explain the handling of every possible missing observation or every reason a participant might lack an evaluable outcome.

Further details about missing-data procedures, imputation, exclusions, and sensitivity analyses are not reported in the registry record. The posted group counts are available, but a full account of participant flow and the analysis dataset cannot be inferred from those counts alone. The record also does not provide enough information to assess whether all randomized participants contributed to every analysis in the same way.

The registry lists time to first seizure following treatment as a secondary outcome. Numerical results for that outcome, the full set of planned secondary analyses, and any hierarchy among endpoints are not reported in the registry record. A multiplicity adjustment or other strategy for handling repeated testing is also not reported in the registry record. It would therefore be inappropriate to assume that secondary outcomes were adjusted, that they were not adjusted, or that a particular testing sequence was prespecified.

General statistical explanation: when researchers test multiple outcomes or contrasts, the chance of observing an apparently unusual result somewhere among the tests can increase. Prespecified outcome hierarchies or multiplicity adjustments can address that issue, but the registry entries do not identify such a plan. The interpretation consequently centers on the posted primary endpoint and does not use unreported secondary findings to strengthen or weaken its statistical evidence.

The registry identifies an intent-to-treat population, but does not state the detailed estimand or handling rules for all intercurrent events beyond the endpoint’s explicit classification. The distinction matters: the endpoint counts a participant as not seizure-free after certain events even if no seizure is recorded. Readers should understand the result as performance on this composite operational definition, not as a count based solely on seizure occurrence.

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 non-serious adverse event terms

The serious adverse event counts are zero in the USL261 group and zero in the placebo group. Each group lists 31 participants at risk for serious adverse events. The registry also records zero deaths in each group. For other adverse events, the posted counts are 16 for USL261 and 17 for placebo, with 31 at risk in each group.

These are descriptive counts in the registry record. They do not show the severity, timing, duration, or clinical context of each other adverse event, and the listed details do not provide a breakdown by event term. No serious adverse event terms are listed. The counts do not establish that serious events cannot occur, nor do they establish that the treatments have equivalent safety profiles. A zero count means none were recorded in these trial groups under the registry’s reporting, not that risk is absent.

The record does not provide the detailed adverse-event collection rules, attribution assessments, exposure time, or a fuller account of all safety analyses. These figures should therefore be read as a limited summary of the reported serious-event and other-event counts, rather than as a comprehensive comparison of safety.

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

The primary result is a comparison of the number of participants who met a specific seizure-free definition during the six-hour Treatment Phase. The posted counts are 17 of 31 for USL261 and 12 of 31 for placebo. In the observed groups, more participants assigned to USL261 met that endpoint. The recalculated risk difference is 0.161, but its interval extends from -0.084 to 0.406, crossing zero and indicating substantial uncertainty about the underlying contrast.

The result does not establish that USL261 is effective or ineffective, and it does not establish that the treatments are equivalent. The study’s own reported Wald asymptotic analysis and the unadjusted recalculation are not identical, and the registry record does not give all details needed to reproduce the sponsor’s calculation. The distinction between a test result and the size or importance of an effect is essential: a p-value alone does not quantify benefit, harm, or clinical relevance.

The endpoint is not simply a record of seizure occurrence. Participants who discontinued study drug prematurely, received rescue intervention for acute central respiratory depression adverse event, or had background anti-epileptic drug therapy altered were classified as having a seizure outcome for the seizure-free analysis. Thus, the reported counts answer the question defined by this composite rule. They should not be recast as counts of participants with no seizure alone.

The record lists time to first seizure following treatment as a secondary outcome, but its numerical result is not reported in the registry record. Safety counts show no reported serious adverse events or deaths in either group, and other adverse events were recorded in both. Those descriptive counts do not prove the absence of risk or provide a complete safety comparison. The registry also does not report a threshold for clinically meaningful difference or a detailed account of missing-data methods and multiplicity procedures.

The appropriate interpretation is therefore narrow: the trial records a higher observed count meeting the seizure-free definition in the USL261 group, but the available estimate is imprecise and compatible with a range of underlying effects. These data do not, by themselves, settle the clinical importance, durability, or generalizability of the observed contrast.

17. Why This Trial Matters Statistically

This trial is statistically instructive because it places a straightforward binary outcome beside a small set of observed group counts and an uncertain between-group estimate. The difference in the observed counts can be stated plainly, but the confidence interval demonstrates why the point estimate alone is not enough. Statistical interpretation requires attention to both the direction of the observed result and the range of effects still compatible with the data.

The endpoint also illustrates how an outcome definition shapes the question being answered. Here, seizure-free status required more than the absence of a recorded seizure: premature discontinuation, specified rescue intervention, and alteration of background anti-epileptic drug therapy also prevented a participant from meeting the definition. That operational choice can be important to how readers understand and compare the resulting counts.

The registry supplies both a sponsor-reported Wald asymptotic test and enough group counts for an unadjusted recalculation. Their p-values are not identical, which highlights that a summary statistic depends on the analysis method and that a reanalysis should be labeled rather than substituted for the reported result. The interval crossing zero and the limited precision reinforce the value of reporting estimates with uncertainty rather than presenting a test result alone.

The planning illustrations make a further statistical point: sample-size calculations depend on the effect targeted. Smaller target effects require larger planned groups under the stated assumptions. Because these calculations use the observed effect, they are scenarios rather than evidence about the completed trial’s power or the original design rationale.

Together, the endpoint definition, uncertainty interval, analysis distinction, and planning assumptions show why a clinical trial result should be read as a structured estimate with limitations, not as a binary verdict.

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 posted primary outcome counts are modest, and the recalculated confidence interval is broad. The observed difference therefore has limited precision, with the interval spanning values on both sides of no difference. The registry entries do not provide a threshold for clinically meaningful change, so statistical summaries cannot determine whether any part of that range would matter clinically.

The endpoint’s definition combines seizure recording with treatment-phase completion and specified events. Premature study-drug discontinuation, rescue intervention for acute central respiratory depression adverse event, or alteration of background anti-epileptic drug therapy is counted as a seizure outcome. This may be appropriate to the endpoint as defined, but it means the reported measure should not be interpreted as seizure occurrence alone.

Although the study is recorded as randomized and triple-masked, the registry record does not describe the randomization procedure, allocation concealment, or which roles were masked. It does not provide detailed eligibility criteria or a full participant flow. The primary population is labeled intent to treat, but handling of missing data, sensitivity analyses, and all post-randomization events is not reported in the registry record.

The sponsor’s reported test is identified as Wald asymptotic, while the simple unadjusted recalculation gives a different p-value. The full analysis specification needed to reproduce the sponsor’s result is not included in the registry record. The secondary outcome is listed without numerical results, and multiplicity procedures are not reported in the registry record.

The sample-size illustrations use the observed effect and stated planning assumptions; they are not the original trial design calculation. Original sample-size assumptions and any interim-analysis plan are not reported in the registry record. Safety information is also limited to summary counts and does not include event-level detail, exposure duration, or attribution. These limitations constrain interpretation of efficacy, safety, precision, and generalizability.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Study to Evaluate the Safety and Efficacy of USL261 in Patients With Increased Bouts of Seizure Activity in the EMU — NCT01999777.
• Registry record last updated 2019-10-10; status last verified 2019-10.
• Results first posted 2018-11-27.
• 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

The registry record describes a completed Phase 3 randomized, triple-masked, parallel-group study sponsored by UCB Biopharma S.P.R.L. It compares 5 mg intranasal midazolam, identified as USL261, with intranasal placebo for intermittent bouts of increased seizure activity in the epilepsy monitoring unit. Enrollment is recorded as 62.

The primary outcome counts participants meeting a seizure-free definition during the six-hour Treatment Phase. The posted counts are 17 of 31 for USL261 and 12 of 31 for placebo. The endpoint treats specified discontinuation, rescue intervention, and background therapy changes as failures to meet seizure-free status. The recalculated risk-difference interval crosses zero, and the sponsor’s reported Wald asymptotic p-value is not identical to the unadjusted recalculation.

Time to first seizure following treatment is listed as a secondary outcome, but its numerical results are not reported in the registry record. Serious adverse events and deaths are recorded as zero in both groups; other adverse-event counts are 16 and 17. Details about missing-data methods, multiplicity, and the original sample-size rationale are not reported in the registry record.