Lamictal as Add-on Treatment for Bipolar I Disorder in Pediatric Patients: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Lamictal as Add-on Treatment for Bipolar I Disorder in Pediatric Patients (NCT00723450), a randomized phase 3 study. The primary outcome is Time From Randomization to the Occurrence of a Bipolar Event (TOBE).

2. Clinical Question

3. Trial Design

The study is registered as a completed phase 3 randomized, double-blind, parallel-group trial sponsored by GlaxoSmithKline. Its title identifies lamictal as add-on treatment for Bipolar I disorder in pediatric patients. The official title specifies children and adolescents aged 10 to 17 years. The registry reports enrollment of 301 and lists Bipolar Disorder as the condition.

The brief summary describes a multicenter, placebo-controlled trial of lamictal as add-on maintenance treatment in pediatric outpatients. It describes a sequence of screening, open-label treatment, randomization, and taper and follow-up. The reported planned durations were approximately 2 weeks for screening, up to 18 weeks for the open-label phase, up to 36 weeks for the randomized phase, and up to 4 weeks for taper and follow-up. These are phase descriptions, not observed participant-level durations.

The randomized comparison used LTG with flexible dosing and placebo. The record identifies the study as double-blind and parallel-group; it does not report which individuals were masked or the operational procedures used to maintain masking. It also does not report details of the randomization procedure, allocation concealment, stratification for treatment assignment, or the planned enrollment target. The reported enrollment should therefore not be treated as evidence of the trial’s original sample-size target.

The primary endpoint population is defined as randomized participants who received at least one dose of investigational product. Safety information is separated into open-label and randomized phase groups, with taper phases included in the group labels. The open-label LTG safety group is not the placebo-controlled randomized comparison. The registry record is marked completed, but completion status alone does not establish how complete each outcome dataset was or whether every planned analysis was reported.

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: The study consisted of a 2-week Screening Phase, an 18-week Open-Label Phase, a 36-week Double-Blind Randomized Phase and a Taper and Follow-up Phase (up to 4 weeks depending on the dose the participant was taking at the last Open-Label or Randomized Phase visit), which was either open-label or double-blind depending on the phase of the study.

Open-Label Phase

Reasons for not completing this period:

Randomized Phase

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Randomized Intent-to-Treat (ITT) Population: all participants who were randomized to LTG or placebo and received at least one dose of investigational product.

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 From Randomization to the Occurrence of a Bipolar Event (TOBE), measured over From randomization until Week 36.

The primary endpoint was time from randomization to the occurrence of a bipolar event, or TOBE, measured in days from randomization until Week 36. It was defined by a composite of events: the first prescription of additional pharmacotherapy for bipolar symptoms; an increase in the dose of conventional bipolar medication; electroconvulsive therapy; movement to a more restricted environment for observation, safety, or treatment; or withdrawal because of a bipolar-related adverse event, serious adverse event, or lack of efficacy meeting rating-scale thresholds.

A time-to-event endpoint is useful when the timing of an event matters and participants may be followed for different lengths of time. In general, such methods use information about both whether an event occurred and when it occurred, while accounting for participants whose event time is not observed during follow-up. That general rationale does not establish how censoring or missing follow-up was handled in this trial; those details are not reported in the registry record.

The endpoint combines treatment changes, changes in care setting, and specified withdrawals. This can capture clinically consequential events that may not be represented by a single symptom score, but it also means the endpoint reflects decisions and thresholds as well as illness events. The registry describes a log-rank test stratified by index mood state. It labels the outcome as a mean with standard error in its outcome metadata, but gives no numerical group estimates or comparison results. How those metadata relate to the stated log-rank analysis is not explained in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry describes TOBE as a time-to-event outcome and states that it was calculated using a log-rank test stratified by index mood state. The endpoint timeframe runs from randomization until Week 36. The registry also labels the outcome measure as a mean with standard error, but does not provide numerical means, standard errors, or an explanation of how that summary relates to the log-rank test. The distinction is important: a description of an analysis does not reveal its estimate or uncertainty.

What does a time-to-event analysis use?

In general, time-to-event analysis considers both the occurrence of an event and the time at which it occurs. A participant without a recorded event by the end of observation may still contribute information up to the last time their status is known; this kind of incomplete event-time information is commonly described as censoring. The registry record does not report how censoring was defined or applied in this trial, how withdrawals unrelated to the composite endpoint were handled, or how missing follow-up was addressed. Those choices can affect interpretation and should not be assumed.

The composite event itself includes several routes to meeting the endpoint: added treatment, dose escalation, electroconvulsive therapy, a more restricted care environment, or specified withdrawals. In general, a composite can capture multiple ways a condition leads to intervention, but its interpretation depends on which components occur and how they are assessed. The registry record does not report component-specific primary outcome results. It therefore does not allow readers to determine which parts of TOBE contributed most to the observed event pattern.

What does the log-rank test contribute?

In general, the log-rank test compares event-time patterns across groups while using follow-up information over time. It is not a comparison of simple event counts alone. The registry states that the test was stratified for index mood state, with depression, mania/hypomania, and mixed mood as the strata. General stratification can account for group structure in a comparison, but the registry does not give the test statistic, group event-time summaries, estimated effect, uncertainty interval, or p-value. The stated test therefore identifies an intended analytic approach, not the strength or direction of evidence.

The endpoint description and outcome metadata use different statistical language: the description names a log-rank test, whereas the metadata identifies a mean and standard error. A mean summarizes a continuous measure, and a standard error describes uncertainty in an estimated mean under the relevant model. The record does not clarify whether the mean and standard error refer to a separate summary, how they were calculated, or whether they are intended to represent the primary comparison. No reconciliation should be presumed.

Why does uncertainty matter?

In general, an estimate gives a central summary of an effect, while a confidence interval communicates its precision by indicating a range of values compatible with the data and analysis assumptions. A narrow interval generally indicates greater precision than a wide interval, but interval interpretation depends on the method and the data. No confidence interval for TOBE is reported in the registry record. Without an estimate and its uncertainty, readers cannot assess whether the data are compatible with effects of different sizes or directions.

A p-value, in general, describes how unusual data at least as incompatible with a null hypothesis would be if that hypothesis and the test assumptions held. It does not measure the probability that a treatment effect is true, nor does it indicate the clinical importance of a finding. The registry record does not report a primary p-value, so no inference about the primary test’s statistical significance is available.

How should planning illustrations be read?

The stated planning illustration uses the observed effect as its basis, a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation; the effect to detect is described as a stated fraction of the observed effect. For time-to-event outcomes, the illustration reports a total number of events, while other illustrations report participants per arm. Such calculations are planning aids under assumed conditions, not observed trial results and not observed power. They do not establish that the trial was designed around those assumptions. The registry does not report the original sample-size justification or whether the trial’s design used these planning assumptions.

General sample-size calculations depend on assumptions about the target effect, outcome variability or event pattern, analysis, and allocation. Changing an assumed target effect changes the planning requirement. An illustration based on an observed effect is especially not proof that the trial had adequate power: the observed effect is itself uncertain, and planning figures do not replace interpretation of estimates and uncertainty from the actual analysis. No numerical planning values are reported in the registry record.

What remains undocumented?

The record identifies a randomized double-blind parallel design and provides an endpoint analysis description, but does not report the randomization sequence method, allocation concealment, model details beyond the log-rank test, censoring rules, missing-data procedures, sensitivity analyses, or a complete statistical analysis plan. These are not details that can be reconstructed from the design label. Likewise, the record does not report interim analyses or a sample-size justification. Their presence or absence in trial operations should not be guessed.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists 22 outcome measures and one analysis. The secondary outcomes cover time to withdrawal, intervention for mood episodes, relapse or recurrence, and depressive symptom measures. Several outcomes are named for multiple mood states, time points, or follow-up intervals. The primary outcome is not reduced to a single comparison in the registry record. The record does not report a multiplicity adjustment, a hierarchy of testing, or another strategy for controlling the interpretation of multiple outcomes. Those procedures are not reported in the registry record, so the relationship among outcome tests cannot be established.

In general, when many outcomes or comparisons are assessed, the chance of at least one apparently unusual result can increase if each is interpreted independently. Prespecifying a primary outcome, using a testing hierarchy, or adjusting inference are possible approaches, but the registry information does not identify which approach, if any, was used beyond naming the primary outcome. The number of listed outcome measures alone does not show how many were tested or how their results were interpreted.

The primary population is described as randomized participants who received at least one dose of investigational product. That is the population definition in the record; it should not be silently replaced with a different analysis population. The record does not report the number analyzed for TOBE, the number with endpoint events, or how participants with incomplete follow-up were treated. It also does not report whether all randomized participants were included in a sensitivity analysis or how deviations from assigned treatment were handled.

In general, missing outcome information can affect both precision and the validity of a comparison, depending on why it is missing and how the analysis addresses it. The registry record does not report missing-data assumptions, imputation methods, censoring rules, or sensitivity analyses for missingness. Therefore, it is not possible to assess from this record whether missing data could have altered the primary result. The record’s endpoint definition and population description support a clear account of what was intended, but not a full evaluation of multiplicity or missing-data handling.

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: suicidal ideation (1.7%), mania (1.2%), bipolar i disorder (1.2%), infectious mononucleosis (1.2%), urinary tract infection (1.2%).

The registry separates safety information by study phase and group. In the open-label and open-label taper phases, the LTG group lists 298 participants at risk for serious adverse events, with 19 affected, and 298 at risk for other adverse events, with 177 affected. These figures describe the open-label group and should not be interpreted as a randomized comparison with placebo.

For the randomized and double-blind taper phases, the placebo group lists 86 at risk for serious adverse events, with 5 affected, and 86 at risk for other adverse events, with 41 affected. The LTG group lists 87 at risk for serious adverse events, with 1 affected, and 87 at risk for other adverse events, with 45 affected. These are counts of participants affected as reported, not counts of event episodes. The registry record does not explain whether a participant could contribute to more than one event term.

The record lists serious adverse event terms with percentages: suicidal ideation, 1.7%; mania, 1.2%; Bipolar I disorder, 1.2%; infectious mononucleosis, 1.2%; urinary tract infection, 1.2%; and intentional overdose, 1.2%. The term-specific percentages should be read as the registry reports them; their denominators and any overlap among terms are not specified in this summary. Deaths are not reported in the registry record.

In general, adverse-event counts describe events observed during the stated exposure periods and do not alone establish causation. Differences between groups may be difficult to interpret without aligned exposure duration, event definitions, and analysis details. The open-label and randomized safety summaries have different contexts, so they should not be combined into a single comparative estimate.

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

The registry record documents a randomized, double-blind comparison of flexible-dose LTG and placebo as add-on maintenance treatment in pediatric Bipolar I disorder. It also defines a time-to-event primary endpoint and names a stratified log-rank test. These details clarify the intended comparison and the kind of outcome under study. They do not, on their own, show what the comparison found.

The primary effect estimate, its uncertainty, and the primary p-value are not reported in the registry record. Consequently, the posted information cannot establish whether event timing differed between groups, how large any difference was, or how precisely it was estimated. It also cannot establish that the groups were equivalent. A missing result is not evidence of no effect, and design information is not evidence of a favorable or unfavorable outcome.

The endpoint is a composite that includes medication changes, care setting changes, and specified withdrawals. In general, a composite may summarize several clinically meaningful pathways, but it can obscure variation among its components. Since component-specific primary results are not reported, the record does not show whether any particular component was more common or occurred sooner in one group.

The safety summaries describe serious and other adverse events in groups defined by study phase. They are useful for understanding what the registry recorded in those groups, but they do not establish that an event was caused by treatment. The open-label safety group is distinct from the randomized comparison, and the registry does not report the detailed exposure and analytic information needed to make a causal safety assessment.

In general, a result should be interpreted using its estimate, uncertainty, design, outcome definition, and analysis population together. Here, the record supports a description of the trial design and reporting gaps, not a conclusion about treatment benefit, lack of benefit, or comparative safety. It also does not report the statistical analysis plan, missing-data procedures, interim analyses, or original sample-size justification. Those limitations restrict how far interpretation can go, regardless of the study’s completed status.

13. Why This Trial Matters Statistically

This trial is statistically notable for using a time-to-event primary endpoint in a pediatric maintenance-treatment study. Time from randomization to a bipolar event can retain information about when an event occurred, rather than reducing follow-up to a simple event count. The registry’s endpoint definition also illustrates how a composite outcome may combine treatment escalation, changes in care, and specified withdrawal criteria. Those components make the operational definition central to interpretation.

The record names stratification by index mood state for the log-rank test. In general, stratification is a way to account for prespecified groups when comparing event-time patterns. The registry states the strata but does not report the analysis results or details needed to evaluate the implementation. This is a useful distinction for readers: knowing the intended method is different from being able to assess the estimate it produced.

The reporting also shows why outcome definitions, populations, and safety groups must be read together. The primary population includes randomized participants who received at least one dose, whereas safety information is separated into open-label and randomized phases. Enrollment, the primary analysis population, and phase-specific safety groups are not interchangeable quantities. Treating them as if they described the same set of participants would obscure the design.

Finally, the record illustrates the limits of statistical interpretation when the primary estimate and uncertainty are unavailable. A trial may have a clearly stated phase, allocation, masking, endpoint, and test while still lacking the results needed to judge the comparative evidence. Planning illustrations based on an observed effect are also conceptually separate from observed power and from the original design justification. These distinctions matter whenever readers use registry records to understand what was studied and what conclusions the reported statistics can support.

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.

The main trial-specific limitation in the registry record is that the primary outcome comparison is not reported as a numerical result. The endpoint is described, but no group estimate, confidence interval, or p-value is available. The record also does not provide the event counts or component-specific results needed to understand which parts of the composite contributed to the outcome pattern. This prevents assessment of effect direction, magnitude, and precision.

The outcome metadata identify a mean with standard error, while the primary endpoint description specifies a stratified log-rank test. The registry does not explain the relationship between these descriptions or provide results that would resolve the ambiguity. The record also does not report censoring rules, missing-data methods, sensitivity analyses, or details of the statistical model beyond the stated test and stratification. The robustness of the primary analysis cannot therefore be assessed.

The trial summary describes staged treatment, including an open-label phase before randomization. The record does not report how eligibility for randomization was determined, how many participants entered each phase, or how phase transitions were handled analytically. It does not report dosing details beyond flexible dosing, background medication regimens, or treatment adherence. These omissions limit interpretation of treatment exposure and generalizability to the population described by the title and summary.

The registry lists multiple secondary outcomes but does not report a multiplicity strategy, a complete analysis plan, interim analyses, or the original sample-size justification. Safety information is organized by phase, and deaths are not reported in the registry record. The serious adverse event term percentages are listed without denominator details in this summary. Together, these reporting gaps mean the record supports a careful description of design and recorded safety counts, but not a comprehensive assessment of efficacy, missingness, or comparative safety.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Lamictal as Add-on Treatment for Bipolar I Disorder in Pediatric Patients — NCT00723450.
• Registry record last updated 2017-01-09; status last verified 2016-11.
• Results first posted 2014-06-04.
• 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

NCT00723450 is a completed phase 3 randomized, double-blind, parallel-group study sponsored by GlaxoSmithKline. It evaluated lamictal as add-on maintenance treatment for pediatric Bipolar I disorder in outpatients aged 10 to 17 years. The study summary describes screening, open-label treatment, randomization, and taper and follow-up phases. The randomized comparator was placebo, and the LTG arm used flexible dosing.

The primary endpoint was time from randomization to TOBE through Week 36, with a composite definition and a stated stratified log-rank analysis. The primary population was defined as randomized participants who received at least one dose. The registry record reports enrollment of 301 and safety summaries for open-label and randomized phase groups.

The primary group comparison, its estimate, uncertainty, and p-value are not reported in the registry record. Secondary outcomes are listed, but the record does not provide enough information to assess multiplicity, missing-data handling, or the original sample-size rationale. The available record therefore characterizes the design and endpoint, while leaving the main comparative result unresolved.