Buspirone Therapy for Localized Epilepsy: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Buspirone Therapy for Localized Epilepsy (NCT01496612), a randomized phase 2 study. The primary outcome is Seizure Frequency in the Buspirone (Active) and Placebo Periods.

2. Clinical Question

The trial was designed to assess seizure frequency among people whose seizures were described as localization-related and who were already taking antiseizure medication. The registry also lists anxiety, depression, seizures, epilepsy, and partial epilepsy among the conditions associated with the record. The eligibility information gives additional context about the intended participants, but the record does not report the characteristics of the participants who contributed to the posted outcome summaries.

The question is clinically focused, but the posted results address it only descriptively. The mean seizure count was 33.2 during the buspirone period and 31.2 during the placebo period. Those means are period-specific summaries; they do not by themselves establish how participants’ seizure counts changed between periods. A crossover comparison is most informative when the within-participant results and a suitable comparative analysis are available. Such an analysis is not reported in the registry record.

3. Trial Design

The registry describes a Phase Two, randomized, quadruple-masked crossover trial sponsored by the National Institute of Neurological Disorders and Stroke. The study began on 2011-11-22, had a primary completion date of 2016-04-19, and is listed as terminated. Enrollment is reported as 9. The reason for termination and the relation between the enrollment figure and the number contributing to each posted outcome summary are not reported in the registry record.

Participants were assigned to study treatment under a crossover model involving buspirone and placebo. The brief description says that participants would start taking a study drug, either buspirone or placebo, twice daily, and keep a seizure calendar. The primary outcome describes a three-month treatment period on each treatment. The adverse-event record also lists washout periods following buspirone and placebo. The duration of washout, treatment sequence details, and any rules governing transition between periods are not reported in the registry record. The arm descriptions are blank, so they do not add detail about sequence or treatment allocation.

In a crossover design, participants can contribute information under more than one treatment condition. A paired analysis can use each participant’s experience under both conditions, which may account for stable differences between participants. That is a general feature of crossover methods, not a claim about the analysis used in this trial. The registry reports a mean and standard deviation for each treatment period but does not report participant-level paired results or an inferential analysis that accounts for the crossover structure.

The brief description also outlines screening procedures, including clinical assessments and imaging, as well as mood and memory testing. It names buspirone and placebo as the study drugs. Details such as the actual treatment sequences received, adherence, and participant flow through treatment periods are not reported in the registry record. The record therefore allows a broad description of the design, but not a full reconstruction of how the trial was conducted or how many participants completed each stage.

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: Nine participants were consented to the protocol. Two participants decided not to participate prior to baseline and randomization.

1st Intervention

Reasons for not completing this period:

2nd Intervention

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

6. Primary Endpoint

The registry lists the primary outcome as Seizure Frequency in the Buspirone (Active) and Placebo Periods, measured over Three months.

The primary endpoint is seizure frequency, expressed as the number of seizures recorded during each three-month treatment period. Participants used a seizure calendar, and the registry says that the total number of seizures during each period was calculated before a mean was calculated across subjects. The reported unit is number of seizures, and the stated summary measure is the mean with standard deviation.

Seizure frequency is a direct measure of the event the trial sought to study: seizures experienced while participants were receiving buspirone or placebo. A count can retain information about how often events occurred during a specified observation period. The three-month timeframe also makes the outcome a period-specific measure rather than a single assessment at a clinic visit. A calendar can support repeated recording, although the registry does not report the completeness of calendar entries or how missing days were handled.

General statistical context: count outcomes often have a skewed distribution, because many participants may have relatively few events while some have substantially more. The mean describes the average count, while the standard deviation describes how spread out participant counts are around that mean. When counts vary more than a simple statistical model expects, over-dispersion can matter: a model that does not allow enough variability can give uncertainty estimates that are too small. Whether the trial used a count model, how it addressed distributional features, and whether it analyzed counts as rates are not reported in the registry record.

The endpoint description does not specify an adjustment for time actually observed, a baseline-adjusted measure, or a rule for handling incomplete periods. Those details are not reported in the registry record. The posted means therefore describe the recorded outcome summaries, but do not reveal the completeness or participant-level pattern behind them.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

General statistical context: a mean is the arithmetic average of observations. For this endpoint, the registry says that the total number of seizures was calculated for each treatment period and that a mean was then calculated across subjects. The standard deviation describes how much individual observations vary around the mean. It is a measure of variability in the observations, not a measure of uncertainty in the estimated treatment difference. The registry reports means and standard deviations, but no confidence interval for either a treatment contrast or a period-specific mean.

What does the crossover design imply for analysis?

In a crossover trial, participants receive more than one treatment condition over time. This makes the participant, rather than an independent group of participants for each treatment, central to the comparison. A paired analysis compares each participant’s outcome under buspirone with that same participant’s outcome under placebo. The pairing can reduce the influence of stable differences between participants, but it also requires a method that respects the relationship between repeated observations from the same person.

The posted summaries do not report paired treatment differences, participant-level data, or a statistical method for handling repeated measures. They also do not identify the allocation sequence or say whether period or sequence effects were included in an analysis. Without those details, it is not possible to reproduce or assess a crossover-specific comparison from the registry record. The fact that the trial is labeled crossover does not itself establish that a paired analysis was performed.

How should the reported spread be read?

For the buspirone period, the mean is 33.2 and the standard deviation is 25.51. For the placebo period, the mean is 31.2 and the standard deviation is 32.57. These are descriptive summaries of counts, not estimates of a treatment effect with uncertainty. The standard deviations show that observations varied appreciably around their respective means. They do not show how individual participants’ outcomes changed between periods, because that requires information about the joint pattern of each participant’s observations.

General statistical context: seizure counts are count data, and count distributions can be asymmetric or contain greater variability than a simple model assumes. Over-dispersion refers to variability beyond what a model expects. When present, it can make standard errors and confidence intervals from an unsuitable model misleading. Count regression, robust methods, or other approaches can sometimes address these features, but the registry does not report which, if any, were used. No model specification, link function, distributional assumption, or treatment-effect scale is reported.

What would a comparative estimate add?

A treatment-effect estimate would quantify the contrast between seizure frequency under buspirone and under placebo according to a specified analysis. A confidence interval would describe the range of estimates compatible with the data and the model’s assumptions, and would help communicate precision. A p-value, if reported, would address compatibility with a specified null hypothesis under the analysis assumptions; it would not measure the probability that a treatment is effective. These are general interpretations. The registry record does not report a comparative estimate, confidence interval, or p-value for the primary outcome.

The summary values alone cannot be used to recover those quantities for a crossover comparison. In particular, the standard deviation for each treatment period does not give the standard deviation of within-participant differences. That paired variability is needed for a conventional paired analysis. Recalculating a contrast from the two means would also fail to account for the crossover structure and would not supply the missing uncertainty. No such recalculation is presented as a trial analysis.

How do planning calculations differ from trial results?

Sample-size planning uses assumptions about the effect to detect, variability, significance level, allocation, and statistical method. A planning illustration is not a result observed in participants. It does not show that the trial had a particular probability of detecting an effect after the results are known. The registry record does not report the trial’s own sample-size justification or a protocol-based power calculation. The stated planning assumptions describe an illustration based on the observed effect, with a two-sided significance level of 0.05, equal allocation, and a normal approximation. Those assumptions do not replace a trial-specific design rationale, and a planning illustration should not be described as observed power.

General statistical context: power is a design-stage probability calculated under specified assumptions. Once trial data have been collected, the uncertainty around an estimated effect is more informative than a retrospective power calculation. Here, there is no posted comparative estimate and no confidence interval from which to assess precision directly. The enrollment figure of 9 is not, by itself, evidence that the study was adequately or inadequately powered. The enrollment target and its justification are not reported in the registry record.

What can be concluded from the registry analysis record?

The record lists no posted analyses for the primary outcome. It reports the endpoint definition and descriptive treatment-period summaries, but not the statistical model, analysis population, handling of missing observations, or treatment of period and sequence. The absence of a posted analysis is distinct from a finding of no difference. It means the registry record does not present the inferential results needed to quantify a comparative effect or its precision.

The primary outcome is one component of a broader set of measures. Six secondary outcomes are listed, including anxiety and depression scales and cognitive tasks. Their names establish what was planned or tracked in the record, but results for those outcomes are not reported in the posted registry information. The registry also does not state whether adjustments were made for testing multiple outcomes or whether any outcome was designated for a particular confirmatory role beyond the named primary endpoint.

10. Confidence Intervals for Each Arm

No two-arm comparison is shown for this trial, but each arm's mean can be given with a t-based 95% confidence interval computed from the posted mean, spread and number of participants.

R code and output

arm <- c("Buspirone", "Placebo")
n   <- c(5, 5)
m   <- c(33.2, 31.2)
sp  <- c(25.51, 32.57)
sd  <- sp      # posted standard deviations
for (i in seq_along(n)) {
  se <- sd[i] / sqrt(n[i]); h <- qt(0.975, n[i] - 1) * se
  cat(sprintf("CARM|%s|%d|%.3f|%.3f|%.3f|%.3f\n", arm[i], n[i], m[i], sd[i], m[i] - h, m[i] + h))
}

CARM  Buspirone  5  33.200  25.510  1.525  64.875
CARM  Placebo  5  31.200  32.570  -9.241  71.641

11. Multiplicity, Missing Data and Analysis Populations

The registry identifies seizure frequency as the primary outcome and lists six secondary outcomes: scores on the Hamilton Anxiety Rating Scale, Hamilton Depression Rating Scale, and Beck Depression Inventory, along with scores on the Cancellation Task, Trail Making Test-A, and Trail Making Test-B. The record does not report results for these secondary measures in the information available for this analysis. It also does not report a multiplicity plan, such as whether the secondary outcomes were treated as exploratory, whether any hierarchy was specified, or whether statistical significance was adjusted for testing more than one outcome.

General statistical context: evaluating multiple outcomes can increase the chance of at least one apparently unusual result, even when there is no real effect. Adjustment procedures or a prespecified hierarchy can help control that risk, but they may affect precision and interpretation. Since the registry record does not describe a multiplicity strategy, the status of any secondary-outcome inference cannot be assessed from this record.

The record does not report how missing seizure-calendar entries, incomplete treatment periods, or withdrawals were handled. It does not identify an intention-to-treat, per-protocol, or other analysis population, and it does not report a participant flow through treatment periods. These omissions matter in a crossover design because the availability of observations under both treatment conditions affects which participants can contribute to a paired comparison. The posted summary sample size is 5.0 for each treatment period, while overall enrollment is 9; the record does not explain the relationship between these counts.

Missing-data assumptions, imputation procedures, and sensitivity analyses are not reported in the registry record. It is therefore not possible to determine whether the posted summaries reflect all enrolled participants, only those with complete treatment-period data, or another analysis set. The lack of these details limits assessment of both possible selection effects and the generalizability of the reported summaries.

12. Safety Results

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

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

Most frequent non-serious adverse event terms

The registry reports serious adverse events by treatment or washout period. The number at risk is 7 during buspirone, 7 during the washout period following buspirone, 5 during placebo, and 5 during the washout period following placebo. The number of participants affected by a serious adverse event is 0 in each of those periods. Deaths are also reported as 0 in each period. These figures are period-specific counts and should not be interpreted as counts of distinct participants across the entire trial.

The record separately reports other, non-serious adverse events: 6 of 7 participants during buspirone, 4 of 7 during the washout period following buspirone, 3 of 5 during placebo, and 3 of 5 during the washout period following placebo. These counts are not the serious-event figures. They also do not establish that an event was caused by a particular treatment. The registry does not report event-level details in the serious-event terms, and its list of top serious-event terms is empty.

General statistical context: zero reported serious events in a small group does not establish that the risk is zero. The number at risk is limited, and the registry does not report a comparative safety analysis or uncertainty interval. In a crossover study, a participant may contribute to more than one period, so period counts should not be treated as independent groups or added to infer a number of unique participants. The record does not explain whether all adverse events were collected using the same observation rules across periods or how attribution was assessed.

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

The posted results show mean seizure counts of 33.2 during the buspirone period and 31.2 during the placebo period, with standard deviations of 25.51 and 32.57, respectively. These descriptive figures indicate the average counts recorded for the treatment periods and the spread of observations around those averages. They do not provide a formal estimate of the within-participant treatment contrast.

The results do not establish whether seizure frequency differed between buspirone and placebo. No comparative estimate, confidence interval, or p-value is reported in the registry record. Without a paired analysis, the treatment-period means cannot show how consistently individual participants’ counts changed or how precisely a treatment contrast is estimated. The record’s crossover design makes that distinction important: participants can contribute outcomes under both conditions, and the relation between their outcomes is relevant to the analysis.

The trial is listed as terminated and has enrollment of 9, while the primary outcome summaries report 5.0 observations for each treatment period. The record does not explain the termination, the relationship between enrollment and the outcome counts, or the number of participants who completed both periods. Those gaps prevent a full assessment of participant flow and the extent to which the posted summaries represent the enrolled group.

The record also does not establish how missing calendar data were addressed, what statistical method was used, or how period and sequence effects were handled. The absence of those details does not mean the trial found no effect; it means the registry record does not report enough inferential information to judge the comparative result. No treatment recommendation follows from these descriptive summaries.

General statistical context: a difference between sample means can reflect a treatment effect, random variation, design features, or a combination of these. Precision is assessed using the uncertainty around an appropriate effect estimate, not by comparing means alone. The reported standard deviations describe variation in the period-specific observations but do not supply the uncertainty for a paired treatment comparison. The results should therefore be read as limited descriptive evidence rather than a definitive statistical answer to the trial’s primary question.

14. Why This Trial Matters Statistically

This trial is statistically notable because its crossover structure makes the participant-level pairing central to the treatment comparison. A crossover design can be useful when each participant can be observed under different treatments, but valid interpretation depends on the sequence, the timing of periods, the handling of possible carryover, and the analysis of within-participant outcomes. The registry identifies the design as crossover and lists washout periods, but does not report enough detail to evaluate those features.

The endpoint is a seizure count collected across treatment periods. Count outcomes can be highly variable, and the reported standard deviations are sizeable relative to the period means. That spread makes the distinction between descriptive averages and inferential comparisons especially important. Period-specific means do not show whether participants generally had similar counts under both conditions or whether a subset of observations drove the averages.

The record also illustrates why enrollment and analysis sample sizes should be interpreted in context. Enrollment is 9, while each outcome-period summary reports 5.0. In a crossover study, treatment-period counts are not necessarily counts of separate groups. Without a participant flow or an explanation of the outcome sample, the reader cannot determine who contributed to each period or whether paired observations were available.

Finally, the posted record is statistically incomplete for a comparative interpretation: it contains no reported primary-outcome analysis, uncertainty interval, or description of how missing outcomes and multiple measures were handled. This does not resolve the research question in either direction. It shows why statistical reporting must include the analysis approach and its precision, alongside treatment-period summaries, for readers to evaluate what the data can support.

15. 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 main limitation is that the registry record provides descriptive primary-outcome summaries without a reported comparative analysis. It does not report a treatment-effect estimate, confidence interval, p-value, or method for analyzing the crossover data. Consequently, the precision and statistical compatibility of a buspirone-placebo contrast cannot be assessed from the posted results.

Enrollment is reported as 9, but the primary outcome summaries report 5.0 for each treatment period. The record does not explain the participant flow, the number completing each period, or whether the same participants contributed to both summaries. This makes it difficult to understand the analysis population and the extent of missing outcome data. The registry does not report rules for incomplete seizure calendars, missing periods, or withdrawals.

The crossover details are also limited. Although the record identifies the design and lists washout periods, it does not report treatment sequence, washout duration, period effects, or how possible carryover was considered. It does not state whether the analysis used paired observations or another method. These details are particularly relevant when outcomes are measured repeatedly in the same participants.

The trial is listed as terminated, but the reason is not reported in the registry record. The record also does not report a trial-specific sample-size rationale or a multiplicity plan for the six listed secondary outcomes. The brief summary describes eligibility and procedures, but the registry information does not provide participant characteristics or enough detail about adherence and outcome completeness to assess representativeness. These limitations constrain interpretation of the posted summaries; they do not establish why the trial ended or what its unreported analyses might have shown.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Buspirone Therapy for Localized Epilepsy — NCT01496612.
• Registry record last updated 2021-10-05; status last verified 2016-04-19.
• Results first posted 2021-09-05.
• 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.

19. Record Summary

This Phase Two trial of buspirone for localization-related epilepsy is listed as randomized, quadruple-masked, and crossover. The sponsor is the National Institute of Neurological Disorders and Stroke. Enrollment is reported as 9, and the registry lists the study as terminated. The registry describes seizure frequency over three-month buspirone and placebo periods as the primary outcome.

The posted primary-outcome summaries report a mean of 33.2 seizures with standard deviation 25.51 during buspirone and a mean of 31.2 with standard deviation 32.57 during placebo. Each period summary reports 5.0 participants. No formal comparative analysis or uncertainty estimate is reported in the registry record. The record lists six secondary outcomes, but their results are not reported in the information available for this analysis.

Serious adverse events are reported as 0 in each listed treatment or washout period, with deaths reported as 0. The record does not explain the termination, participant flow, missing-data handling, analysis population, or the statistical method used for the primary outcome.