Inpatient, Dose-Ranging Study of Staccato Alprazolam in Epilepsy With Predictable Seizure Pattern: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Inpatient, Dose-Ranging Study of Staccato Alprazolam in Epilepsy With Predictable Seizure Pattern (NCT03478982), a randomized phase 2 study. The primary outcome is Percentage of Participants in Each Treatment Group Achieving Seizure Activity Cessation Within 2 Minutes and no Recurrent Seizure Within 2 Hours.

2. Clinical Question

3. Trial Design

This was a phase 2, randomized, parallel-group, quadruple-masked, placebo-controlled, dose-ranging inpatient study. The registry lists enrollment of 156 and identifies the study as completed. It describes adults aged 18 years and older with focal or generalized epilepsy and a documented history of predictable seizure episodes. The listed condition is epilepsy, and the sponsor is Engage Therapeutics, Inc.

Participants were to be admitted to a Clinical Research Unit or Epilepsy Monitoring Unit for a stay described as 2-8 days. The brief summary says one seizure event per participant would be treated with study medication. Staff caregivers assessed the seizure event through clinical observation, with video electroencephalogram confirmation. The primary endpoint concerns cessation within 2 minutes and no recurrence within 2 hours after dosing on the dosing day.

The double-blind groups were placebo, Staccato alprazolam 1.0 mg, and Staccato alprazolam 2.0 mg, each described as a single dose for inhalation. The record also lists an open-label Staccato alprazolam 1.0 mg group. Posted primary comparisons are between placebo and each double-blind dose group; the open-label group is not part of those reported comparisons.

The primary analysis population was modified intent-to-treat, defined in the record as participants who had a seizure event, received study drug, and had at least one evaluation after study drug administration. Enrollment and analysis-population counts should not be treated as interchangeable: the record lists enrollment of 156, while the primary results give group-specific denominators. The registry record does not report the original sample-size justification or the planned allocation ratio.

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 Enrolled Population.

Overall Study

Analysis population for the primary outcome, as defined in the registry: The Modified Intent-to-treat (mITT) population consisted of all participants who had a seizure event, received study drug, and had at least one evaluation after study drug administration.

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 refers to Intent-to-treat (ITT) population. ITT population included all participants who had a seizure event and received study drug during the Treatment Period.

6. Primary Endpoint

The registry lists the primary outcome as Percentage of Participants in Each Treatment Group Achieving Seizure Activity Cessation Within 2 Minutes and no Recurrent Seizure Within 2 Hours, measured over 2 hours post-dosing on dosing day.

The primary endpoint is the percentage of participants whose predictable seizure activity ceased within 2 minutes after study drug administration and did not recur within 2 hours. It is reported for the modified intent-to-treat population and uses clinical observation as the basis for the outcome. The timeframe is 2 hours post-dosing on the dosing day.

This is a binary participant-level outcome: each participant either meets the combined cessation-and-no-recurrence criterion or does not. Combining rapid cessation with absence of recurrence makes the endpoint more informative than cessation alone, because it captures whether seizure activity returns during the specified observation period. It does not, by itself, describe the exact time to cessation, the duration of any seizure activity, or outcomes beyond the stated observation window.

General statistical context: a percentage is readily interpretable as the share of participants meeting the endpoint, but it compresses individual variation into a single group summary. The registry also describes observation by staff caregivers and video electroencephalogram confirmation in the study summary. The primary endpoint description says results were based on clinical observation; the record does not provide additional detail on how observation and video electroencephalogram findings were reconciled for this endpoint.

7. Results Posted in the Registry

• Double-blind: Staccato Alprazolam 1.0 mg: 65.8% (25/38) versus 42.5% (17/40) on control; risk difference 23.3 percentage points (95% CI 1.8 to 44.8), risk ratio 1.55 (1.01–2.37); unadjusted two-proportion z-test p = 0.039.
• Double-blind: Staccato Alprazolam 2.0 mg: 65.8% (25/38) versus 42.5% (17/40) on control; risk difference 23.3 percentage points (95% CI 1.8 to 44.8), risk ratio 1.55 (1.01–2.37); unadjusted two-proportion z-test p = 0.039.

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.

Percentage of Participants With Use of Rescue Medication

• 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:
  • Double-blind: Staccato Alprazolam 1.0 mg: 15.8% (6/38) versus 7.5% (3/40) on control; risk difference 8.3 percentage points (95% CI -5.9 to 22.5), risk ratio 2.11 (0.57–7.82); unadjusted two-proportion z-test p = 0.252.
  • Double-blind: Staccato Alprazolam 2.0 mg: 7.9% (3/38) versus 7.5% (3/40) on control; risk difference 0.4 percentage points (95% CI -11.4 to 12.2), risk ratio 1.05 (0.23–4.90); unadjusted two-proportion z-test p = 0.948.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Double-blind: Staccato Alprazolam 1.0 mg and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Double-blind: Staccato Alprazolam 2.0 mg and the control. The interval gives the range of effects that remain compatible with the data.

The registry identifies secondary outcomes concerning seizure episode severity, use of rescue medication, and secondary generalization. The posted numerical secondary results include the rescue-medication outcome. For the other named secondary outcomes, results are not reported in the registry record. The available secondary information therefore does not support a broad assessment across all aspects of seizure severity, recurrence, or progression.

Across the rescue-medication comparisons, the estimates do not present a single clear pattern of a precise difference between dose groups and placebo: the reported contrast for the 1.0 mg dose has an interval that includes no difference, and the 2.0 mg contrast is very close to no difference with a wide interval. These results should be read as uncertain secondary findings, not as confirmation or contradiction of the primary endpoint. The primary endpoint measures seizure cessation and lack of recurrence during a defined observation window, whereas rescue medication is a separate outcome.

General statistical context: secondary endpoints can help describe other dimensions of outcome, but they commonly involve additional comparisons and may be less central to the trial's main question. Unadjusted secondary tests do not account for the increased chance of a small p-value when multiple outcomes or contrasts are considered. The registry record does not report a multiplicity adjustment for the secondary analyses, and no sponsor-reported secondary inferential analyses are listed. Consequently, secondary findings carry less weight than the designated primary analysis and should be regarded as exploratory in the absence of a documented testing plan.

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

• Percentage of Participants With Seizure Episode Severity Assessed by Seizure Episode Severity Scale
• Percentage of Participants With Use of Rescue Medication
• Percentage of Participants With Secondary Generalization (Evolution to a Complex Partial Seizure and/or a Generalized Tonic-Clonic Seizure)

9. Statistical Interpretation

10. Statistical Methods Explained

What does the percentage compare?

The endpoint is a binary outcome summarized as the percentage of participants meeting both conditions: seizure activity cessation within 2 minutes and no recurrence within 2 hours. Each group's percentage is based on its own analysis denominator. The primary double-blind comparison asks whether that percentage differs between a Staccato alprazolam dose group and placebo. The reported analyses use a chi-squared method and express the result as a difference in percentage.

General statistical context: a group percentage describes the observed frequency of an outcome in the analyzed group. A difference in percentages describes the contrast between two group frequencies, with the sign and size depending on which group is subtracted from which. In the sponsor's reported contrasts, the difference is expressed in favor of the active-dose group relative to placebo. A comparison is more informative than looking at either percentage alone because it directly describes the observed separation between groups.

What does the confidence interval add?

A confidence interval gives a range of values compatible with the observed data under the method used. It communicates uncertainty about the estimated contrast, rather than changing the point estimate itself. For each primary comparison, the sponsor reports a 95% confidence interval for the difference in percentage. Its lower and upper limits are 1.8 and 44.8. The interval indicates that the estimated contrast is not highly precise: values spanning a sizable range remain compatible with the data and analysis.

General statistical context: an interval that excludes no difference is often described as statistically significant at a corresponding two-sided threshold, but that does not convey the clinical importance of every value in the interval. Nor does it establish that the estimate will be reproduced in a new study. The interval is conditional on the outcome definition, included participants, data, and statistical method.

What do the p-values say?

The sponsor reports p = 0.039 for each primary comparison. The simple unadjusted recalculation gives p = 0.039. A p-value describes how unusual data at least as incompatible with a no-difference model would be, if that model and the test assumptions were correct. It is not the probability that the no-difference model is true, and it does not measure the magnitude or practical importance of the observed result.

The p-values are unadjusted in the recalculation. The registry record does not report whether or how the primary comparisons were adjusted for multiplicity. If multiple outcomes or contrasts are examined, the chance of at least one small p-value can be greater than for a single planned test. Without a reported plan, the primary p-values should not be interpreted as if the family of tests and its error control were fully documented.

Why use a chi-squared comparison?

The posted sponsor analyses name a chi-squared method for comparing group percentages. General statistical context: this method assesses whether observed counts across categorical outcome groups differ more than would be expected under a no-association model. Its usefulness depends in part on the pattern and size of the cell counts. The registry's reported method is the sponsor's analysis; the record does not provide enough detail to assess all assumptions or any additional model specifications.

The recalculated risk differences use the posted event counts and denominators to express the unadjusted absolute contrast. This is a transparent summary of the observed group difference, but it does not reproduce any adjustments or other features that may have been part of the sponsor's analysis. In this record, the recalculation agrees with the reported contrast and p-value to the precision stated.

What is the analysis population?

The primary endpoint population is modified intent-to-treat. The record defines it as participants who had a seizure event, received study drug, and had at least one evaluation after study drug administration. This population definition is important because the percentages describe those meeting the criteria, not necessarily every person enrolled. The registry record does not report how many otherwise eligible participants were excluded from this analysis or the reasons for any exclusions.

General statistical context: restricting an analysis to participants with post-treatment evaluation can affect interpretation if the chance of being included is related to treatment assignment or outcome. The posted definition alone does not show whether such selection occurred. Details on missing outcomes, imputation, sensitivity analyses, and adherence to randomized assignment are not reported in the registry record.

How should the estimates be used?

The sponsor's reported analysis is the primary inferential result recorded for the two double-blind dose comparisons. The unadjusted recalculation is useful as a consistency check, not as a separate confirmatory result. Neither a confidence interval nor a p-value removes uncertainty, and neither answers questions outside the endpoint's population, timeframe, and study conditions. The registry record does not report enough of the full statistical analysis plan to determine how all planned outcomes and contrasts were handled.

11. Confidence Intervals and Precision

The sponsor reports a 95% confidence interval from 1.8 to 44.8 for the difference in percentage in each double-blind dose-versus-placebo comparison. The point estimate is 23.3. The interval excludes no difference, but it spans a broad range of positive differences, so the data are more informative about the direction of the observed contrast than about its exact size.

The recalculated unadjusted risk difference for each contrast is 0.233, with interval limits of 0.018 and 0.448. These are the same contrasts expressed on the risk-difference scale used in the recalculation. The sponsor's reported analysis gives the difference in percentage and its 95% interval; both presentations are consistent at the precision stated in the registry.

General statistical context: confidence intervals reflect sampling uncertainty under the analysis method. A wide interval indicates less precision than a narrow interval, and intervals may remain wide even when a p-value is below a conventional threshold. Precision is influenced by the amount of information in the comparison, including the number of participants and the frequency of the outcome. The denominators are 40 in placebo and 38 in each double-blind dose group.

The open-label group has a denominator of 8 and a posted endpoint percentage of 62.5, but the registry does not report a confidence interval for that group. The primary sponsor comparisons are the double-blind groups against placebo; no interval should be inferred for the open-label result from the posted point estimate alone. In addition, the confidence intervals do not address uncertainty from potential bias, missing data, multiplicity, or differences between the study population and other populations.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The sample-size illustrations translate an assumed treatment contrast into the number of participants needed in each of two equally allocated groups under the stated calculation assumptions. These figures are participants per arm, not total participants and not total events. They are planning illustrations based on the observed primary effect, not the trial's original design justification. The registry record does not report the trial's own original sample-size assumptions or justification.

The listed assumptions for the illustration 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 rows for planning for 100% of the observed effect, 75% of the observed effect, and 50% of the observed effect show how the required group size changes as the assumed target contrast becomes smaller. In the primary comparison, the listed estimates for 80% power are 71 participants per arm for the full observed effect, 128 for 75% of the observed effect, and 288 for 50% of the observed effect. The corresponding estimates for 90% power are 95, 171, and 385 participants per arm.

In plain terms, the 80% power column is the estimated group size under which the calculation would have an 80% chance of detecting the specified effect if that effect were the true one and the assumptions held. The 90% power column makes the same calculation for a higher chance of detection. Reaching a higher target power generally requires a larger planned sample. Likewise, smaller effects are harder to distinguish from random variation and require more participants under these assumptions.

These values are not observed power. They are not a retrospective measure of what the completed trial achieved, nor evidence that the trial was designed around the observed effect. A curve based on the same planning assumptions should be read as a display of how required sample size varies with the assumed effect, not as a forecast of clinical outcomes. The registry does not report the original design assumptions that would be needed to judge the trial's planned power.

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 = 0.425; treatment proportion p₁ = 0.658; average p̄ = 0.541.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.7047 = 1.3811.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6852 = 0.5766.
4. n = (1.3811 + 0.5766)² / (0.2329)² = 70.7, rounded up to 71 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 <- 0.425                          # control proportion
p1 <- 0.657894736842105                          # 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(3478982)
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 = 71 (80% power), 95 (90% power)
 75% of observed effect: n per arm = 128 (80% power), 171 (90% power)
 50% of observed effect: n per arm = 288 (80% power), 385 (90% power)
power.prop.test: 70.7 
Simulated power at n = 71 per arm: 0.809 (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 sponsor's posted analysis compares placebo with each double-blind Staccato alprazolam dose using a chi-squared method. For both comparisons, it reports a difference in percentage of 23.3, a 95% confidence interval from 1.8 to 44.8, and p = 0.039. The reported group percentages are 42.5 for placebo and 65.8 for each double-blind dose group.

The simple unadjusted recalculation from the posted counts and denominators gives a risk difference of 0.233, interval limits of 0.018 and 0.448, and p = 0.039 for each dose comparison. These results correspond to the sponsor's reported contrast at the rounding precision shown. The discrepancy indicator is false, so no conflict between the recalculation and the sponsor's results is flagged.

The sponsor's analysis remains the trial's reported analysis and takes precedence over a summary-statistic recalculation. The recalculation is not an adjusted analysis and cannot establish that every detail of the sponsor's method has been reproduced. The registry record does not report further model adjustments or the full analysis plan. The unadjusted result should therefore be treated as a transparent check on the reported contrast, rather than as a competing analysis.

15. Multiplicity, Missing Data and Analysis Populations

The registry reports two primary comparisons: placebo versus Staccato alprazolam 1.0 mg and placebo versus Staccato alprazolam 2.0 mg. It also identifies multiple secondary outcomes. The record does not report whether the primary comparisons were treated as a family of tests, whether a hierarchy or other multiplicity procedure was planned, or whether the secondary outcomes were adjusted for multiple testing. The reported primary p-values should therefore be understood as reported, unadjusted results unless the registry explicitly indicates otherwise; no such adjustment is described.

The primary analysis population is modified intent-to-treat, defined as participants who had a seizure event, received study drug, and had at least one evaluation after administration. This definition narrows the analysis relative to all enrolled participants. The registry record does not report the number or reasons for any exclusions from the primary analysis, the handling of participants without a usable endpoint assessment, or whether analyses were also conducted in other populations.

The record does not report methods for missing outcome data, such as imputation, nor sensitivity analyses examining different assumptions about missingness. It also does not give detailed information on protocol deviations, treatment adherence, or how participants were handled if the observed seizure event did not meet assessment requirements. These details matter because estimates can be affected if endpoint availability or analysis inclusion differs in a way related to treatment or outcome.

The reported analyses name a chi-squared method for the primary comparisons, but the registry record does not report a complete statistical analysis plan or additional model specifications. It is not possible from the posted information to determine whether further planned analyses were conducted but not entered, or how all outcomes were prioritized. These missing details counsel against treating the reported inferential results as a complete description of every planned analysis.

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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: generalised tonic-clonic seizure (2.6%), seizure cluster (2.5%), status epilepticus (2.5%), pneumonia (2.5%).

The registry lists serious adverse event counts by group, with the number at risk and the number affected. The double-blind placebo group lists 40 at risk and 2 affected; the double-blind 1.0 mg group lists 38 at risk and 1 affected; and the double-blind 2.0 mg group lists 38 at risk and 0 affected. The open-label 1.0 mg group lists 8 at risk and 0 affected. Deaths are listed as 0 in each group.

These are group-level counts, not a comparison adjusted for differences in exposure or follow-up. The record also lists several serious-event terms with percentages, but those terms are not presented as a treatment-specific comparison in the information summarized. The counts alone do not establish whether an event was caused by study drug, and they do not replace the trial's safety assessment. The registry record does not report further detail on event timing, duration, severity grading, or causal attribution for these figures.

General statistical context: rare events are difficult to compare precisely in groups of this size. Zero reported events in a group does not establish that the underlying risk is zero. The posted safety counts should be read as descriptive observations from the study, not as proof of comparative safety.

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

The posted primary results show that a higher percentage of participants in each double-blind Staccato alprazolam group met the combined endpoint than in the double-blind placebo group during the specified 2-hour post-dosing period. The sponsor reports the same difference in percentage for the two dose comparisons, with a 95% confidence interval that excludes no difference. The simple unadjusted recalculation agrees with the sponsor's result at the stated precision.

This finding describes the participants included in the modified intent-to-treat analysis and the outcome definition used in the registry. It does not establish the result for people with epilepsy generally, for people without predictable seizure patterns, or for settings other than the inpatient study environment. It also does not describe outcomes after the 2-hour observation window or establish a dose-response pattern: the two double-blind dose groups have the same posted primary percentage, and the registry's reported comparisons do not show a distinction between those doses.

The result is not a guarantee that an individual participant will meet the endpoint. Group percentages summarize observed outcomes, while the confidence interval reflects uncertainty in the comparative estimate. The interval allows a range of possible positive differences, so the result does not identify a precise magnitude. Statistical evidence against no difference is not the same as evidence about clinical importance, which cannot be determined from the p-value alone.

The registry record does not report the full multiplicity plan, missing-data methods, or original sample-size justification. It also does not report numerical results for every listed secondary outcome. Those gaps limit interpretation of the primary result in the context of the complete planned analysis and limit conclusions about other outcomes. The findings should be understood as a report of this trial's endpoint and analysis, not as advice or a general treatment recommendation.

18. Why This Trial Matters Statistically

This trial illustrates why a binary endpoint needs to be read with its definition, analysis population, and observation window. The primary outcome is not simply seizure cessation: it also requires no recurrence during the specified period. That combined definition determines what the percentages mean and makes clear that the result applies to a particular short-term outcome rather than to seizure control in general.

The study also demonstrates how the same estimated contrast can be presented through a group percentage, a difference in percentage, a confidence interval, and a p-value. Each presentation answers a different question. The group percentages describe observed frequencies; the contrast summarizes separation between groups; the interval communicates uncertainty about that contrast; and the p-value relates the data to a no-difference model. None alone supplies the full interpretation.

There are two dose-versus-placebo primary comparisons, and both have the same posted estimate and reported inferential result. That makes the question of multiplicity relevant, while the registry record does not report the plan for handling it. The example also shows why a recalculation from group summaries should be distinguished from the sponsor's analysis: it can check consistency but cannot substitute for a complete analysis plan.

Finally, planning illustrations based on an observed effect are not evidence of the trial's original power or design rationale. Treating them as such would confuse a retrospective calculation with a prospective design decision. Keeping these distinctions explicit helps readers separate observed findings, statistical uncertainty, planning scenarios, and information that the registry record does not report.

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.

The primary analysis is restricted to the modified intent-to-treat population, which the registry defines by seizure occurrence, receipt of study drug, and at least one post-administration evaluation. The record does not report the number excluded from that population or provide detail on missing endpoint assessments. This limits assessment of how representative the analyzed participants were of all enrolled participants.

The primary observation window is limited to 2 hours after dosing on the dosing day. The posted endpoint therefore cannot address recurrence beyond that interval or longer-term outcomes. The inpatient setting and the requirement for predictable seizure episodes also define a specific study context; the registry results do not directly establish what would be observed in other settings or populations.

The primary comparisons are reported for each dose versus placebo, but the registry does not describe a multiplicity strategy. The secondary-outcome record is incomplete for interpretation: numerical results for seizure episode severity and secondary generalization are not reported in the registry record. The available secondary information cannot resolve whether the primary pattern is accompanied by a consistent pattern in those other outcomes.

The registry does not report the original sample-size justification, full statistical analysis plan, missing-data methods, or sensitivity analyses. The sponsor reports the primary analysis using a chi-squared method, while the recalculation is a simple unadjusted comparison based on posted summaries. The latter does not recover any unreported features of the sponsor's analysis. The confidence interval also allows a broad range of positive differences, limiting precision about the magnitude of the observed contrast.

The safety data are presented as group-level counts, with limited detail on event timing and attribution. Counts, especially when few events are recorded, do not provide a precise comparative safety estimate. These limitations concern the interpretation of the posted registry information and should not be mistaken for evidence that an unreported analysis or outcome had a particular result.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Inpatient, Dose-Ranging Study of Staccato Alprazolam in Epilepsy With Predictable Seizure Pattern — NCT03478982.
• Registry record last updated 2021-02-03; status last verified 2021-02.
• Results first posted 2021-02-03.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

23. Record Summary

The registry record describes a completed phase 2 randomized, quadruple-masked, parallel-group inpatient study sponsored by Engage Therapeutics, Inc. It enrolled 156 and evaluated single inhaled doses of Staccato alprazolam 1.0 mg and 2.0 mg against placebo, with an additional open-label 1.0 mg group listed.

The primary endpoint was the percentage meeting seizure activity cessation within 2 minutes and no recurrence within 2 hours after dosing. In the modified intent-to-treat population, the posted percentages were 65.8 for each double-blind dose group and 42.5 for placebo. The sponsor reported a difference in percentage of 23.3 for each dose comparison, with a 95% confidence interval from 1.8 to 44.8 and p = 0.039. The simple unadjusted recalculation is consistent with that result at the stated precision.

The registry record does not report the original sample-size justification, multiplicity strategy, or missing-data methods. Results for some named secondary outcomes are not reported in the registry record, and the safety counts are descriptive. These considerations constrain how broadly and precisely the findings can be interpreted.