Levetiracetam Versus Valproate in Idiopathic Generalized Tonic-clonic Seizures: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Levetiracetam Versus Valproate in Idiopathic Generalized Tonic-clonic Seizures (NCT03940326), a non-randomized phase 4 study. The primary outcome is Time to First Seizure.

2. Clinical Question

Thus, the registry describes the question the study set out to address, but does not provide the primary outcome estimates needed to judge the observed between-arm difference or its precision. The reported design also matters to interpretation: because allocation was non-randomized, the arms cannot be treated as if they were formed by random assignment. Any observed difference could reflect differences between the groups as well as differences associated with treatment. The record does not report how allocation was determined or whether measured participant characteristics were adjusted for in an analysis.

3. Trial Design

The registry identifies this as a Phase 4, parallel-group study with non-randomized allocation and no masking. It is marked completed, with 103 participants enrolled. The listed interventions are levetiracetam and valproate, and the study summary describes an open-label, active-controlled non-inferiority comparison in idiopathic generalized tonic-clonic epilepsy. The sponsor is Mazandaran University of Medical Sciences. The recorded study start was 2018-04-01, and the primary completion date was 2019-10-01.

These design details set important limits on what a comparison can establish. In a randomized study, random assignment is intended to balance, on average, both known and unknown participant characteristics between treatment groups. This study is explicitly non-randomized, and the record does not explain how participants entered either arm. Therefore, the groups may have differed in ways that influence seizure outcomes. The registry record does not report an adjustment strategy that could address such differences.

Open-label means masking was absent, according to the record. Knowledge of treatment assignment can affect behavior, clinical decisions, reporting, or assessment, depending on the outcome and study procedures. The registry does not describe procedures used to limit these possibilities. Parallel-group means the record lists separate levetiracetam and valproate arms; it does not describe a crossover comparison.

The listed secondary outcomes are withdrawal rate and time to withdrawal. The record lists four outcome measures overall, but reports zero analyses. Although the primary endpoint and secondary outcomes are named, no outcome estimates or inferential results are reported in the registry record. The study summary’s non-inferiority description does not by itself establish that a non-inferiority analysis was performed or met its criteria. The margin, analysis population, and statistical decision rule are not reported in the registry record.

4. Participant Flow and Accounting

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

Overall Study

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 Time to First Seizure, measured over 6 months.

The primary endpoint is time to first seizure, expressed in days. The registry defines it as the time interval from the beginning of the study until the first seizure and gives a timeframe of 6 months. It identifies the endpoint as a continuous measure summarized by a mean and standard deviation. That choice describes the intended form of reporting, but the registry record does not report either arm’s mean or standard deviation, or an estimated comparison between arms.

Time to first seizure can be useful because it captures when an event occurs, rather than only whether an event occurred during follow-up. As a general statistical point, timing information may distinguish participants who experience an event earlier from those who experience it later. However, analyzing such data can require decisions about participants who do not experience the event during observation, or whose observation ends for another reason. The registry record does not report how such cases were handled, whether any observations were censored, or which analysis method was used.

The endpoint’s stated mean-and-standard-deviation format should not be mistaken for a reported result. Without the arm-level estimates, an uncertainty interval, or an analysis result, the direction and precision of the primary comparison cannot be assessed from the registry record. Nor does the endpoint description alone establish how the study’s stated non-inferiority objective was evaluated.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry identifies the primary measure as a continuous outcome, with a mean and standard deviation, and describes it in days. It also states that the endpoint is time from study start to the first seizure. These labels describe the intended outcome and summary format; they do not provide the observed values. No arm-level mean, standard deviation, estimated difference, confidence interval, or p-value is reported in the registry record.

What a mean and standard deviation describe

As a general statistical explanation, a mean is a summary of the central value of a set of observations. A standard deviation describes how spread out individual observations are around that mean. Neither quantity alone gives the full distribution of the outcome, and neither indicates how precisely a treatment-group mean or between-group contrast has been estimated. Precision also depends on the amount of information and the analysis method. Because the registry record does not report the primary endpoint summaries, those features cannot be evaluated for this comparison.

What an uncertainty interval would add

As a general explanation, a confidence interval gives a range of estimates compatible with the data under the assumptions of the statistical procedure. It helps readers assess both the direction and precision of an estimated effect. A narrow interval indicates greater precision than a wide interval, while an interval that includes values on both sides of a relevant decision threshold may leave the conclusion uncertain. No confidence interval for time to first seizure is reported in the registry record. In a non-inferiority study, interpretation would also require knowing the prespecified margin and seeing how the interval relates to that margin; neither is reported.

Non-inferiority and the comparison design

The study summary describes a non-inferiority objective. In general, non-inferiority testing asks whether the evidence rules out an unacceptable loss in the effect of one intervention relative to a comparator, using a margin defined in advance. That margin is essential to interpretation: a finding cannot be classified as non-inferior from a p-value or a study label alone. The registry record does not report the margin, the direction of the comparison, or the analysis used to assess the objective. It also reports non-randomized allocation. As a general principle, non-randomized groups may differ before treatment, so a crude contrast can mix treatment-associated differences with pre-existing group differences. Whether adjustment was performed is not reported in the registry record.

Time-to-event outcomes

Although the registry describes the endpoint using a mean and standard deviation, the outcome is defined by the time until an event. As a general statistical point, time-to-event methods can incorporate both event timing and incomplete observation. Participants without a recorded event by the end of follow-up may contribute information up to the time they were observed, rather than being treated automatically as if an event occurred at the end. This approach requires clear rules for defining follow-up, events, and incomplete observation. The registry record does not report those rules, whether any participants lacked an observed seizure event, or whether a time-to-event method was used. A mean-based summary and a time-to-event analysis answer related but not identical questions, so the actual analytic approach matters.

Planning illustrations

The record describes a planning illustration based on a fraction of the observed effect, with a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation. It specifies that time-to-event illustrations use the total number of events under the Schoenfeld approach, while other illustrations use participants per arm. These are planning assumptions, not observed trial results and not evidence of the study’s original design justification. The registry does not report a sample-size justification for the trial itself. No planning values are stated in the record described for this analysis, so no required sample size can be interpreted from those assumptions alone. Such illustrations should not be called observed power: they describe hypothetical planning under stated assumptions, not the probability that the completed study detected an effect.

What is unavailable for evaluating the analysis

The registry lists four outcome measures and zero analyses. It names withdrawal rate and time to withdrawal as secondary outcomes, but does not report their estimates or analysis details. The record also does not report a statistical analysis plan, handling of missing outcome data, analysis populations, covariate adjustment, or interim analyses. These absences matter because analytic choices can affect estimates and their uncertainty. They should not be filled in by assuming customary methods. The available design and endpoint descriptions identify what was intended to be studied; they do not substitute for a reported analysis.

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("Levetiracetam", "Valproate")
n   <- c(45, 58)
m   <- c(169, 178)
sp  <- c(6.1, 2.2)
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  Levetiracetam  45  169.000  6.100  167.167  170.833
CARM  Valproate  58  178.000  2.200  177.422  178.578

11. Multiplicity, Missing Data and Analysis Populations

The registry lists four outcome measures, including time to first seizure as the primary endpoint and withdrawal rate and time to withdrawal as secondary outcomes. It reports zero analyses. No results for the secondary outcomes or other outcome measures are reported in the registry record.

When a study evaluates multiple endpoints or makes multiple statistical comparisons, the chance of at least one apparently notable finding can increase. As a general statistical principle, multiplicity procedures may be used to control that risk, or researchers may clearly distinguish primary from exploratory analyses. The registry record does not report a multiplicity strategy, the number of planned comparisons, or how any secondary outcomes would be interpreted in relation to the primary endpoint. It is therefore not possible to determine whether multiplicity was addressed.

The registry record also does not report the number of participants contributing to the primary endpoint analysis, missing outcome counts, reasons for missingness, or the method used to handle missing data. It does not identify an intention-to-treat, per-protocol, or other analysis population. These details are particularly relevant to a non-inferiority objective, because exclusions and missing observations can affect the interpretation of a comparison. This is a general methodological concern, not evidence that a particular problem occurred in this study.

The record’s enrollment total does not by itself reveal how many participants were included in each endpoint analysis or whether follow-up was complete. With no analysis results or population definitions reported, the robustness of any efficacy conclusion cannot be assessed. The absence of these details in the registry record should be read as unreported information, not as proof that the investigators did or did not use a particular method.

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 serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: rise in liver enzymes (3.4%).

The registry reports serious adverse events separately from other adverse events. In the levetiracetam arm, 0 of 45 participants at risk are listed as having a serious adverse event; in the valproate arm, 2 of 58 are listed. Deaths are recorded as 0 in both arms. For other adverse events, the record lists 17 of 45 participants at risk in the levetiracetam arm and 30 of 58 in the valproate arm. These figures describe the registry entries and should not be conflated: the “other” adverse-event counts are not the serious-event counts.

The registry lists “Rise in liver enzymes” as a top serious term with 3.4%. It does not provide enough context to determine the arm associated with this term, its event count, or the denominator used for that percentage. The record’s listed at-risk counts are relevant context for the arm-level summaries, but they should not be assumed to be the denominator for every term-level entry.

As a general statistical point, adverse-event counts are descriptive unless a comparative analysis and its uncertainty are reported. Differences in counts do not, by themselves, show that an intervention caused an event or changed its risk. Interpretation is further limited by the non-randomized, open-label design. The registry does not report a safety analysis method, event ascertainment details, or adjusted comparisons. The counts therefore describe what was recorded, not a causal conclusion about either intervention.

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

The record establishes that a completed Phase 4, parallel-group study compared levetiracetam with valproate in a population described as having idiopathic generalized tonic-clonic seizures. It also identifies time to first seizure as the primary endpoint and names withdrawal rate and time to withdrawal as secondary outcomes. The reported safety entries include serious and other adverse-event counts, along with deaths recorded as zero in both arms.

Those details do not establish the comparative efficacy result. The registry record does not report the primary endpoint means, standard deviations, an effect estimate, a confidence interval, or a p-value. It lists zero analyses. As a result, readers cannot determine the observed difference in time to first seizure or how precisely it was estimated. The registry’s use of the term non-inferiority describes the stated objective, not a demonstrated result. The non-inferiority margin and analysis are not reported in the registry record.

The study design also limits causal interpretation. Allocation was non-randomized, and masking was absent. In general, non-randomized arms may differ for reasons other than the intervention, and open-label knowledge can influence aspects of care or outcome reporting. The registry does not report how allocation was handled or whether potential group differences were adjusted for. These limitations do not show that bias occurred, but they prevent assuming that a crude arm difference would represent a treatment effect.

The safety counts are likewise not a causal comparison. They report events among participants at risk, but no comparative safety analysis or uncertainty interval is given. The term-level 3.4% entry lacks sufficient detail in the registry record for arm-specific interpretation.

Accordingly, the record is informative about the study’s stated question, broad design, endpoint definitions, enrollment, and selected adverse-event entries. It does not permit a conclusion about whether one intervention performed better, whether the stated non-inferiority objective was met, or whether the observed safety counts reflect a treatment-related difference.

14. Why This Trial Matters Statistically

This trial is statistically notable because its stated objective was non-inferiority, while its recorded allocation was non-randomized. Those features must be considered together. Non-inferiority is not established merely by observing similar group outcomes or by finding no statistically significant difference. As a general principle, it requires a prespecified margin and an estimate precise enough to rule out a difference considered unacceptably unfavorable. The margin, analysis result, and uncertainty interval are not reported in the registry record, so the intended claim cannot be evaluated from the posted information.

The design also illustrates why group comparability matters. Random assignment is commonly used to reduce systematic differences between arms, but this study is identified as non-randomized. The record does not describe how treatment assignment occurred or whether baseline differences were addressed analytically. Therefore, even if an outcome contrast had been reported, its interpretation would need to account for possible differences between the groups that preceded treatment.

The primary endpoint is defined as a time interval, yet the registry identifies its summary measure as a mean with a standard deviation. That makes reporting of the actual analysis method important: a mean summary and a time-to-event analysis may use timing and incomplete follow-up differently. The registry does not report how participants without a recorded first seizure during follow-up were handled or what model was used.

Finally, the record provides design and selected safety information but no outcome analyses. This distinction is central to responsible statistical reading. A study’s objective, completed status, and enrolled population do not substitute for estimates and uncertainty. The record is useful for identifying what was intended and how the study was broadly structured, while also showing where the evidence needed for statistical conclusions is not reported.

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.

Several trial-specific limitations can be identified from the registry record. First, allocation is explicitly non-randomized, and the record does not explain how participants were assigned to levetiracetam or valproate. That leaves the possibility of differences between the groups that could influence the endpoint. No baseline characteristics or adjusted outcome analyses are reported in the registry record.

Second, masking is listed as none. The study summary describes an open-label comparison. The record does not state whether outcome assessment was independent or whether other steps were used to limit the influence of treatment knowledge. Such details are not reported, so their use cannot be assumed.

Third, the primary endpoint is described as time to first seizure and summarized as a mean and standard deviation, but the registry record does not report outcome values or an analysis method. It also does not describe follow-up completeness, handling of participants without an observed first seizure, or missing-data procedures. These omissions make it impossible to assess how the endpoint was analyzed or how much information contributed to the result.

Fourth, although the study summary states a non-inferiority objective, the non-inferiority margin, analysis population, decision rule, and results are not reported in the registry record. The record lists zero analyses, so neither the primary objective nor the secondary outcomes can be evaluated statistically from the posted results.

Finally, the safety entries are descriptive and the term-level serious-event entry lacks arm-specific context. No comparative safety analysis is reported. These limitations concern what the registry record reports; they do not establish that omitted procedures were not conducted.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Levetiracetam Versus Valproate in Idiopathic Generalized Tonic-clonic Seizures — NCT03940326.
• Registry record last updated 2022-02-17; status last verified 2021-12.
• Results first posted 2022-02-17.
• 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

The registry describes a completed Phase 4, open-label, non-randomized, parallel-group comparison of levetiracetam and valproate for idiopathic generalized tonic-clonic seizures. Enrollment was 103. The primary endpoint was time to first seizure, measured in days over a 6 months timeframe and described as a mean with standard deviation. Withdrawal rate and time to withdrawal are listed as secondary outcomes.

The registry reports 0 serious adverse events among 45 at risk in the levetiracetam arm and 2 among 58 at risk in the valproate arm; deaths are recorded as 0 in each arm. Other adverse events are listed for 17 of 45 and 30 of 58, respectively. “Rise in liver enzymes” is listed as a top serious term with 3.4%.

The record lists four outcome measures and zero analyses. Primary outcome estimates, a non-inferiority margin, analysis methods, and uncertainty measures are not reported in the registry record. The posted information therefore describes the trial and selected safety counts but does not establish an efficacy or non-inferiority result.