Oxcarbazepine Versus Placebo in Childhood Autism: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Oxcarbazepine Versus Placebo in Childhood Autism (NCT00467753), a randomized na study. The primary outcome is Vineland Adaptive Behavior Scales.

2. Clinical Question

This is a question about a measured functional outcome, not a direct measure of every aspect of autism. The Vineland domains cover communication, daily living skills, socialization, and motor skills. The registry description says that standardized scores were used and that higher scores above 100 indicate better adaptive behavior. However, it does not report group scores, a between-group contrast, or uncertainty around a contrast. The trial’s stated clinical question can therefore be described, but its answer cannot be estimated from the posted primary-outcome information.

3. Trial Design

The registry describes a randomized, quadruple-masked, parallel-group trial of oxcarbazepine versus placebo in childhood autism. Random allocation is intended to create groups that are comparable on average at the start of a trial. Masking is intended to reduce the influence of treatment knowledge on participants’ experiences, caregiver reports, clinical assessments, or other trial procedures. The record labels the study quadruple-masked but does not specify which roles were masked or how masking was maintained.

The registry lists the study status as completed. Its recorded start is 2006-04 and its primary completion is 2011-12. The sponsor is the University of Medicine and Dentistry of New Jersey. The phase is listed as NA. The registry lists enrollment 5, while the brief summary describes a proposed twelve-week study involving twenty subjects between ages five and seventeen. These are different enrollment descriptions in the record; the information available does not explain the difference or establish how many participants were assigned to each arm.

The intervention entries identify oxcarbazepine and placebo/sugar pill. The arm descriptions identify an oxcarbazepine experimental arm and a sugar-pill arm. The registry record does not report treatment dose, titration, adherence, the details of allocation concealment, or the procedures used to preserve masking. Those details should not be inferred from the design labels alone.

Completion status describes the registry’s study status; it does not mean that the primary outcome was analyzed or that a treatment effect was established. In this record, the primary outcome population is described as zero participants analyzed, and the analysis count is zero. Consequently, the design tells readers what comparison was intended, while the posted primary results do not supply an analyzed 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:

Analysis population for the primary outcome, as defined in the registry: Zero participants analyzed

5. Baseline Characteristics

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

Population: 5 patients started in the study, 2 dropped out after 1-2 weeks , 3 were lost to follow up.

6. Primary Endpoint

The registry lists the primary outcome as Vineland Adaptive Behavior Scales, measured over Evaluated during Baseline and Termination.

The primary endpoint is the Vineland Adaptive Behavior Scales. The registry describes it as a semi-structured informant interview administered to a caretaker or family member to assess a subject’s functioning. It includes communication, daily living skills, socialization, and motor skills. Items are scored from 0-2, domain scores are summed, and those domain scores are converted to standardized scores. The registry states that the normative score is 100, with a standard deviation of 15, and that higher scores above 100 indicate better adaptive behavior.

This type of measure is used to describe functioning in everyday adaptive domains, rather than relying only on a symptom label or a laboratory measurement. The registry notes that the scale has been revised and standardized in all populations and has been found to assess social deficits in autism and strengths in daily living skills. These are statements in the record about the instrument, not findings from this trial.

The outcome was scheduled for evaluation during baseline and termination. A comparison would ordinarily need to make clear how the repeated measurements and the scale’s domains were represented in the analysis. The registry does not reduce the endpoint to a single comparison, and the record reports zero participants analyzed. It therefore supplies neither endpoint scores nor an estimate of change or a difference between groups. The minimum and maximum are listed as 0 and infinity, respectively; that range does not replace the missing trial results.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does randomization contribute?

Randomization assigns participants to intervention groups using a chance-based process. In general, this can reduce systematic differences between groups at assignment and supports a causal comparison when allocation, follow-up, outcome measurement, and analysis are handled appropriately. It does not guarantee that groups will be balanced in a small study, nor does the label alone show how randomization was implemented. The registry identifies allocation as randomized but does not report the sequence-generation or allocation-concealment procedures.

The parallel design means the registered comparison was between groups assigned to oxcarbazepine or placebo rather than a design in which each participant necessarily received both interventions. The registry describes quadruple masking. Masking can help limit differences in behavior or assessment that arise when treatment assignment is known, especially when outcomes involve informant reports. The record does not name the masked roles or provide operational details, so the extent and success of masking cannot be assessed from the registry entry.

What would an endpoint comparison estimate?

A treatment-effect analysis would need to define the outcome being compared, the time point or change measure, the group contrast, and the population included. The Vineland outcome is collected through a caretaker or family member interview and converted into standardized scores. Because the outcome description includes multiple adaptive domains and evaluations during baseline and termination, the analysis would need to specify whether it focused on a particular domain, an overall score, a change from baseline, or another prespecified summary. The registry does not provide that single comparison, and it reports zero participants analyzed for the primary outcome.

In general, an estimated difference describes the observed contrast under a defined analysis. A confidence interval expresses the uncertainty associated with that estimate under the statistical model; it does not guarantee that the true effect lies within the interval. A p-value evaluates how compatible the observed data are with a specified null model, given the assumptions used. Neither an interval nor a p-value measures clinical importance by itself. No primary-outcome estimate, confidence interval, or p-value is reported in this record, so these concepts cannot be applied to a trial result.

How should the planning assumptions be read?

The registry’s sample-size assumptions describe planning illustrations based on a stated fraction of the observed effect. They specify a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation. These are assumptions for planning calculations, not evidence that the trial achieved a particular power or detected an effect. A planning illustration asks what sample size might be needed under assumed conditions; it is not an observed result and is not observed power.

The posted primary outcome information reports zero participants analyzed, so it does not establish an observed primary-outcome effect from which a reader can verify those illustrations. The registry does not report the trial’s own sample-size justification. It also does not report whether the investigators used interim analyses, a prespecified statistical model, covariate adjustment, or a particular approach to missing observations. Those choices should not be inferred from the planning assumptions.

Why do model assumptions matter?

Statistical models summarize data under assumptions about the outcome distribution, the relationship between observations, and the way variability changes across groups or over time. When a model’s assumptions do not match the data, estimates and uncertainty measures can be misleading. General examples include skewed outcomes, repeated measurements from the same participant, and over-dispersion, in which variability exceeds what a model expects. Over-dispersion matters because a model that understates variability may report uncertainty too narrowly. These are general considerations, not claims about the distribution of this trial’s outcome: the registry does not provide analyzed participant-level data or identify a fitted model.

For an informant-based adaptive behavior measure, analysts would also need to consider how the standardized scale is summarized and whether the same outcome definition is applied across groups and assessment times. The registry provides the instrument description and its scoring direction, but not the statistical analysis details needed to determine how scores were handled. No adjusted or unadjusted treatment estimate can be reconstructed from the descriptive scale information alone.

What can be concluded from the posted analysis status?

The record indicates that no primary-outcome participants were analyzed and lists no analyses. This is a limitation on inference, not evidence of no effect. A null result would require an analyzed comparison with an estimate and uncertainty; the registry does not report such a result. Likewise, the absence of a reported estimate does not show that the groups were equivalent. The defensible statistical reading is that the primary comparison is not estimable from the posted outcome information.

10. Multiplicity, Missing Data and Analysis Populations

The Vineland description covers four adaptive domains, and the primary outcome timeframe refers to baseline and termination. The registry notes that the outcome has several categories or time points and was not reduced to a single comparison. When a study examines multiple outcomes, domains, time points, or contrasts, the chance of a false-positive finding can increase if results are selected or interpreted without an appropriate plan. General approaches include prespecifying a primary comparison, adjusting for multiple testing, or clearly treating additional analyses as exploratory. The registry record does not report a multiplicity strategy or identify a single analyzed contrast.

The record lists four outcome measures but does not report analyses. The primary outcome population is described as zero participants analyzed. It does not explain whether outcome data were unavailable, whether participants withdrew, or why the primary analysis population is zero. Missing-data procedures, reasons for missingness, and sensitivity analyses are not reported in the registry record. No particular mechanism—such as loss to follow-up or an incomplete assessment—can be assumed.

The analysis population is also not defined beyond the zero-participant statement for the primary endpoint. The record does not say whether the intended analysis was based on all randomized participants, participants who received treatment, or another population. It does not report how adherence, protocol deviations, or post-randomization exclusions would have been handled. Without these details and without analyzed primary data, neither the risk of selective reporting nor the robustness of a treatment estimate can 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 non-serious adverse event terms

The registry reports serious adverse events separately from other adverse events. In the oxcarbazepine group, serious adverse events affected 0 among 3 at risk; in the sugar-pill group, they affected 0 among 2 at risk. For other adverse events, the registry reports 1 affected among 3 at risk in the oxcarbazepine group and 1 among 2 at risk in the sugar-pill group. These are recorded counts, not estimates of comparative risk.

The at-risk groups are small, so the counts provide limited information about how often events might occur in a broader population. A zero serious-event count does not establish that serious events cannot occur, and the similar other-event counts do not establish that the interventions have the same safety profile. No serious adverse-event terms are listed. Deaths are not reported in the registry record; that is different from a reported count of zero deaths.

Safety data should be interpreted separately from the primary adaptive-behavior outcome. The record’s safety counts do not fill the gap left by zero participants analyzed for the primary endpoint, and they do not provide a complete account of adverse-event types, timing, severity, or attribution.

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

The record can establish what comparison was planned: randomized assignment to oxcarbazepine or placebo, with the Vineland Adaptive Behavior Scales identified as the primary outcome. It also documents a completed registry status and reports adverse-event counts for the two named groups. Those details describe the trial record; they do not answer whether oxcarbazepine changed adaptive behavior relative to placebo.

The crucial limitation is that zero participants are reported as analyzed for the primary outcome. There is no reported group score, change estimate, between-group effect, confidence interval, or primary-outcome p-value. Without those quantities, the direction and magnitude of any treatment difference are unknown from the posted results. It is not valid to interpret the lack of an estimate as evidence that the groups had no difference, or that the interventions were equivalent.

The listed Vineland scoring direction explains how higher scores are interpreted, but it does not reveal what scores participants had. Likewise, the normative score and scale properties describe the measure, not an outcome in this trial. The proposed twelve-week duration and intended participant ages describe the brief study summary, not the actual amount of treatment received or follow-up completed by analyzed participants.

The safety counts show a limited set of recorded events among the at-risk groups. They are not enough to establish a comparative safety profile, and deaths are not reported in the registry record. Enrollment 5 and the brief summary’s intended twenty subjects are different descriptions that the record does not reconcile. Neither description should be substituted for the primary outcome’s analyzed population.

The appropriate interpretation is therefore narrow: the registry documents a randomized, masked trial design, but the posted primary results do not contain an analyzable efficacy comparison. Any statement about benefit, lack of benefit, or equivalence would go beyond the results reported in this record.

13. Why This Trial Matters Statistically

This trial is statistically instructive because its registered design and its posted analysis status tell different parts of the story. Randomization and masking describe methods intended to support a fair comparison. They do not ensure that results will be available or that an effect can be estimated. Here, the registry identifies the design and a primary functional outcome, yet reports zero participants analyzed for that endpoint.

The record also illustrates why enrollment, planned sample size, and analyzed population must not be treated as interchangeable. The registry lists enrollment 5, while the brief describes a proposed study involving twenty subjects. The primary outcome population is reported as zero participants analyzed. The record does not explain those differences, so the intended design cannot be used to infer the amount of usable outcome data.

Multiplicity is another important issue. A measure with several domains and assessments can support more than one possible comparison. Without a specified analysis and a reported estimate, readers cannot tell which contrast would have answered the primary question or how uncertainty and multiple testing were handled. A clear analysis plan is necessary for interpreting a result, but that plan is not reported in this registry record.

Finally, safety counts can look concrete while still being highly limited. Counts among small recorded at-risk groups are descriptive and cannot provide a precise account of risk. This distinction matters in biostatistics: a recorded number is not automatically a reliable comparative estimate. The trial demonstrates why study design, analysis status, outcome completeness, and precision all need to be considered before drawing conclusions from a clinical trial record.

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 largest trial-specific limitation is the absence of an analyzed primary outcome. The registry reports zero participants analyzed on the Vineland Adaptive Behavior Scales and lists zero analyses. It therefore offers no outcome estimate or uncertainty measure for the intended oxcarbazepine-versus-placebo comparison. The reason for the absence of analyzed data is not reported in the registry record.

The record contains different descriptions of enrollment: the registry lists enrollment 5, whereas the brief summary describes a proposed study involving twenty subjects. It does not explain whether the proposed enrollment changed, whether the listed enrollment reflects actual enrollment, or how participants were assigned. The group counts shown for adverse events should not be used to resolve this discrepancy or to infer the primary analysis population.

The primary outcome has multiple domains and refers to baseline and termination, but the registry does not specify a single comparison. It does not report the scoring summary selected for analysis, a statistical model, covariate adjustments, missing-data procedures, multiplicity handling, or the analysis population beyond the zero-participant statement. The extent of treatment exposure, adherence, and follow-up is also not reported in the registry record.

The safety record is limited to group counts for serious and other adverse events. Serious-event terms are not listed, and deaths are not reported in the registry record. The listed at-risk groups are small and cannot characterize uncommon harms with precision. These limitations mean the record is useful for identifying the intended study question and its reporting gaps, but not for estimating efficacy or making a robust comparison of safety.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Oxcarbazepine Versus Placebo in Childhood Autism — NCT00467753.
• Registry record last updated 2021-01-13; status last verified 2021-01.
• Results first posted 2013-12-16.
• 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

NCT00467753 is a completed, randomized, quadruple-masked, parallel-group study of oxcarbazepine versus placebo in childhood autism. The sponsor is the University of Medicine and Dentistry of New Jersey. The registry lists enrollment 5; the brief summary describes a proposed twelve-week study involving twenty subjects ages five to seventeen.

The primary outcome is the Vineland Adaptive Behavior Scales, assessed during baseline and termination. The registry reports zero participants analyzed for this outcome and lists zero analyses, so no primary treatment-effect estimate is available. Serious adverse events are reported as 0 among 3 at risk for oxcarbazepine and 0 among 2 for sugar pill. Other adverse events are reported as 1 among 3 and 1 among 2, respectively. Deaths are not reported in the registry record.