Safety and Tolerability of OXC XR as Adjunctive Therapy in Subjects With Refractory Partial Epilepsy: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Safety and Tolerability of OXC XR as Adjunctive Therapy in Subjects With Refractory Partial Epilepsy (NCT00908349), a na phase 3 study. The primary outcome is Percent Change in Seizure Rate.

2. Clinical Question

The question is descriptive rather than a randomized treatment comparison. The study was an extension examining long-term safety and tolerability of OXC XR in people with refractory partial epilepsy. Its primary outcome was percent change in seizure rate. The registry defines the summary as a median and the dispersion as the full range, but it does not report the median, the range, or an analysis of that outcome.

Without a concurrent control group, the record cannot establish how seizure-rate changes compare with what would have occurred under another treatment or without OXC XR. Nor does the endpoint alone summarize safety or tolerability: those are separate aspects of the study, and the record includes adverse-event counts. The clinical question should therefore be read narrowly: it concerns the distribution of observed seizure-rate change in the study group, while the posted record leaves the primary numerical result unavailable.

3. Trial Design

NCT00908349 is titled “Safety and Tolerability of OXC XR as Adjunctive Therapy in Subjects With Refractory Partial Epilepsy.” Its official title describes a multicenter, open-label extension study evaluating long-term safety and tolerability of OXC XR as adjunctive therapy. The registry lists the study as phase 3 and completed, with a single-group model, no masking, and allocation not applicable. Enrollment is listed as 214.

The intervention is oxcarbazepine XR, and the arm description identifies an open-label study. Because there is one study group and no concurrent comparator, this is not a randomized comparison. Any change reported for the group would be an observation during the study, not an estimate of a comparative treatment effect. The registry record does not describe a control arm or a masked assessment.

The recorded study start is 2009-06 and primary completion is 2011-11. The primary endpoint timeframe is one year, and its description specifies change from baseline to the end of the study. The record’s endpoint timeframe should not be read as a full account of each participant’s follow-up; individual follow-up details are not reported in the registry record.

The sponsor is Supernus Pharmaceuticals, Inc. The brief summary describes the study as an open-label extension evaluating long-term safety and tolerability of oxcarbazepine extended-release. These design features frame interpretation: the record is useful for describing the study’s registered question and posted safety counts, but it does not provide the primary outcome estimate or an analysis that would support a controlled comparison.

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

Reasons for not completing this period:

5. Baseline Characteristics

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

6. Primary Endpoint

The registry lists the primary outcome as Percent Change in Seizure Rate, measured over one year.

The primary endpoint is percent change in seizure rate, measured as change from baseline to the end of the study. Its timeframe is one year. The registry specifies the median as the summary measure and the full range as the measure of dispersion. It does not state the seizure-counting window used for baseline, the method for defining a seizure rate, or how the endpoint was handled for participants without complete follow-up.

In general, a change-from-baseline endpoint describes how a measured outcome differs between an initial reference and a later assessment. A percent-change scale expresses that change relative to the baseline value, while the median describes the middle of the observed values rather than their arithmetic average. The full range gives the lowest and highest values, so it can convey the span of observations, although it is sensitive to extreme values.

Seizure rate is relevant to the study’s stated population and objective, and following it over time can describe the observed course during an extension study. However, the registry record does not report the median percent change or its full range. It also does not report a confidence interval, an endpoint analysis population, or an inferential test. Consequently, the endpoint is defined, but its numerical result and precision cannot be assessed from the posted record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry record describes a phase 3, single-group, open-label extension study. Its primary outcome is percent change in seizure rate, with the median as the summary measure and the full range as the dispersion measure. The record does not include a numerical primary result or an analysis. The concepts below explain how such methods are generally interpreted; they do not imply that a particular analysis was performed in this study.

What does a median describe?

In general, the median is the central value after observations are ordered. It is less influenced by unusually high or low values than the arithmetic mean, which can be useful when an outcome has a skewed distribution. The median does not describe every participant’s change, and by itself it does not indicate how concentrated or widely spread the values are. The registry pairs it with the full range, which reports the span from the smallest to the largest observed value. The range is easy to understand but depends on the extreme observations and can be unstable when the sample changes.

Here, the record specifies a median and full range for percent change in seizure rate, but supplies neither statistic. No direction or magnitude of the primary outcome can therefore be described from the registry result. Nor can readers assess whether the distribution was tightly grouped or included extreme observations.

Why does a single group matter?

In general, a treatment comparison requires a contrast between outcomes under different assigned conditions. A one-group study instead describes observations in one cohort. A change from baseline can be informative as a description of that cohort’s course, but it is not the same as a randomized between-group difference. Without a concurrent control, other influences on changes over time cannot be distinguished from the intervention’s effects using the design alone.

The registry classifies this study as single-group, with allocation not applicable and masking absent. It identifies the study as an open-label extension. These characteristics are central to interpreting any within-group result; they are not repaired by calculating a change score. The record does not describe an alternative comparator or a masked outcome assessment.

What would an uncertainty interval add?

In general, a confidence interval gives a range of values compatible with an estimate under the assumptions of the method used. It communicates statistical precision and is not a guarantee that the true value lies within that range. For a median, suitable interval methods depend on the data and analysis choices. The registry does not report a confidence interval or the method used to estimate uncertainty for the primary endpoint. As a result, the precision of the endpoint cannot be evaluated from the posted record.

How should planning calculations be read?

The registry’s stated sample-size assumptions describe a planning illustration based on an observed effect, with a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The stated target effect is a fraction of the observed effect. Such calculations are hypothetical design exercises, not observed power. The registry record does not report the study’s own sample-size justification, and the single-group design is not a like-for-like two-arm comparison. The assumptions therefore should not be mistaken for the analysis plan or for evidence that the study had a particular probability of detecting an effect.

The stated illustration distinguishes time-to-event calculations, which express the total number of events, from other calculations, which express participants per arm. This study’s listed primary endpoint is percent change in seizure rate, and the record identifies no analysis for it. No numerical planning outputs are reported in the registry record for this study, so no participant target from such an illustration can be assessed.

What cannot be inferred from the registry entry?

A statistical method is not established merely because an outcome is named. The record does not report an inferential test, adjustment strategy, confidence interval method, covariate adjustment, or analysis population for the primary endpoint. It also does not describe how incomplete seizure-rate observations were handled. These are not details that can be reconstructed from the choice of median and full range. The registry’s zero listed analyses is consistent with the absence of posted analytical results, but it does not explain whether analyses were performed outside the record.

In general, over-dispersion means that variability in count data exceeds what a simple model assumes. It can matter when analyzing event counts or rates because standard errors and inferences may be misleading if the model understates variability. The record does not specify a count model or report an over-dispersion assessment, so no such method should be attributed to this study. More broadly, general statistical concepts help explain what analysis choices can mean, but they do not substitute for the unreported study-specific methods or results.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists one outcome measure, identifying percent change in seizure rate as the primary endpoint, and zero analyses. It does not report an inferential analysis for that endpoint. The record also does not describe whether additional outcomes were tested outside the registered measure, whether any multiplicity adjustment was planned, or how the primary outcome was prioritized in analysis. A multiplicity strategy is therefore not reported in the registry record.

In general, multiplicity arises when multiple statistical hypotheses are assessed; repeated testing can increase the chance of a chance-positive finding if no suitable strategy is used. That general concern should not be taken as evidence that multiple primary-endpoint tests occurred in this study. The posted record does not establish that they did.

Missing-data handling is also not reported. The record gives enrollment and a one-year endpoint timeframe, but does not state how many participants contributed endpoint data, how incomplete follow-up was addressed, or whether any imputation or sensitivity analysis was used. The endpoint’s analysis population is not reported in the registry record. Enrollment should not automatically be interpreted as the number included in the primary analysis.

The registry does not specify an analysis set, such as a population defined by treatment exposure or endpoint availability, and gives no information about exclusions from the outcome analysis. Without these details, readers cannot assess whether the endpoint summary would represent everyone enrolled or a subset with recorded outcome data. These omissions do not establish that missingness or multiplicity affected results; they mean the posted record does not permit those questions to be evaluated.

11. Safety Results

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

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: grand mal convulsion (0.9%), non-cardiac chest pain (0.9%), epilepsy (0.9%), hypothermia (0.5%), spinal fracture (0.5%).

The registry lists one oxcarbazepine XR group, with 214 at risk for serious adverse events and 18 affected, and 214 at risk for other adverse events and 124 affected. These are separate categories in the record and should be read as posted, not added together. The record does not report deaths.

Listed serious adverse-event terms include Grand Mal Convulsion (0.9%), Non-Cardiac Chest Pain (0.9%), Epilepsy (0.9%), Hypothermia (0.5%), Spinal Fracture (0.5%), and Ataxia (0.5%). These are the serious terms and percentages reported in the registry. They describe recorded safety events in the study group; the registry does not provide a concurrent group for comparison.

Adverse-event counts do not by themselves establish causation. The record does not specify event-level detail, timing, duration, severity beyond the serious classification, or how events were judged related to treatment. A count of participants affected is also distinct from a count of total events, since one participant may experience more than one event. Interpret the listed counts as safety reporting, not as a measure of seizure-rate change or a controlled estimate of treatment effect.

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

The registry record establishes that NCT00908349 was an open-label, single-group extension study of OXC XR as adjunctive therapy in subjects with refractory partial epilepsy. It identifies the primary endpoint as median percent change in seizure rate from baseline to the end of the study, with a one-year timeframe and full range as the dispersion measure. It also lists safety counts for the study group.

Those details do not provide the primary endpoint result. The median and range are not reported, and the record lists zero analyses. Without a numerical estimate, there is no basis in the posted record to describe the direction or size of the observed seizure-rate change. Without a confidence interval or another uncertainty measure, its precision cannot be assessed either.

The study design also limits what any within-group result could mean. Since there is no concurrent comparator, a change observed during the study would not by itself show how outcomes differed from another treatment or from no treatment. The registry identifies the study as open label and reports no masking. These features do not make the registered observations meaningless, but they constrain causal and comparative interpretation.

The adverse-event information is a separate descriptive component. The record reports 18 participants with serious adverse events and 124 with other adverse events, with 214 at risk in each category. It lists selected serious-event terms and does not report deaths. These entries do not establish that an event was caused by the intervention, and they cannot be compared with another group within this study record.

Overall, the record supports an account of the study’s design, endpoint definition, and posted safety reporting. It does not support a quantitative claim about the primary seizure-rate outcome, its precision, or a comparative treatment effect. Unreported analyses, missing-data handling, and analysis populations should remain unknown rather than being inferred from the study title or endpoint definition.

13. Why This Trial Matters Statistically

This study is statistically instructive because its registered question and its available results are not the same thing. The registry defines a specific primary endpoint—median percent change in seizure rate from baseline to the end of the study—and specifies the timeframe and summary approach. Yet it does not report the endpoint value or an analysis. Clear outcome definitions are useful, but they cannot substitute for the estimate needed to judge magnitude and precision.

The design also illustrates the distinction between describing a cohort and estimating a comparative effect. An open-label, single-group extension can document observations over time and report safety events in the enrolled group. It cannot make the same controlled contrast as a randomized study with a concurrent comparator. Any change from baseline would need to be interpreted as a within-group observation, not as a randomized difference between treatment strategies.

The record’s safety counts reinforce another statistical distinction: participant counts and event terms are descriptive summaries, not causal attribution. Without a concurrent control group and additional event details, the listed counts do not quantify excess risk or establish that the intervention caused an event. The registry’s lack of a reported primary analysis also means that the safety data cannot fill the gap in efficacy-outcome reporting.

Finally, the record shows why analysis details matter. The endpoint’s analysis population, missing-data handling, inferential method, and multiplicity approach are not reported. Those details shape how estimates should be interpreted and how much uncertainty they carry. Recognizing what the registry does and does not document is an essential part of statistical reading: it prevents a study label, endpoint definition, or planning illustration from being mistaken for a reported result.

14. 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 limitations specific to this registry record restrict what can be concluded. Most importantly, the primary endpoint is defined but its numerical summary is absent. Neither the median percent change in seizure rate nor the full range is reported, and the record lists zero analyses. The endpoint’s precision cannot be assessed because no confidence interval or other uncertainty measure is reported.

The single-group, open-label extension design has no concurrent comparator. It therefore cannot provide a randomized between-group estimate, and any within-group change would not distinguish intervention effects from other influences over time. Masking is listed as none. The registry does not report whether outcome assessment involved any additional safeguards against ascertainment bias.

Enrollment is reported as 214, but the record does not say how many participants contributed primary endpoint data or identify the analysis population. Follow-up completeness and missing-data handling are not reported. Although the endpoint timeframe is one year, individual observation periods and the details of baseline measurement are not specified in the registry record.

Safety reporting is also limited. The record gives counts of participants affected and selected serious-event terms, but does not provide event-level timing, duration, treatment relationship, or a concurrent comparison. Deaths are not reported. These limitations mean the safety summaries are descriptive rather than comparative or causal.

The registry does not report the study’s own sample-size justification, inferential methods, multiplicity strategy, or detailed statistical analysis plan. The planning assumptions described for illustrative calculations should not be mistaken for the trial’s design rationale or analysis. These gaps leave key questions about estimation and uncertainty unanswered; they do not establish what the unreported analyses or underlying data would have shown.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Safety and Tolerability of OXC XR as Adjunctive Therapy in Subjects With Refractory Partial Epilepsy — NCT00908349.
• Registry record last updated 2017-05-31; status last verified 2017-05.
• Results first posted 2017-05-31.
• 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.

18. Record Summary

NCT00908349 is a completed, phase 3, multicenter, open-label extension study of oxcarbazepine XR as adjunctive therapy in subjects with refractory partial epilepsy. The registry describes a single-group design, allocation not applicable, no masking, and enrollment of 214. The recorded start is 2009-06, with primary completion in 2011-11.

The primary endpoint is median percent change in seizure rate from baseline to the end of the study, with a one-year timeframe and full range as the dispersion measure. Its numerical result and analysis are not reported in the registry record. The record lists 18 participants with serious adverse events and 124 with other adverse events, each in a group with 214 at risk; deaths are not reported. These safety data are descriptive and have no concurrent comparator.

The registry lists one outcome measure and zero analyses. It does not report endpoint precision, analysis populations, missing-data handling, or detailed statistical methods. The posted record therefore describes the design and selected safety information, but does not permit assessment of the primary seizure-rate result or a comparative treatment effect.