Treatment of Severe Childhood Aggression (The TOSCA Study): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Treatment of Severe Childhood Aggression (The TOSCA Study) (NCT00796302), a randomized phase 4 study. The primary outcome is NCBRF-TIQ D-Total Score.

2. Clinical Question

3. Trial Design

The registry record identifies TOSCA as a phase 4, randomized, quadruple-masked, parallel trial with an enrollment of 168. The study title refers to treatment of severe childhood aggression, and its official title describes stimulant and risperidone use in children with severe physical aggression. The listed condition is Attention Deficit Disorder With Hyperactivity.

The two described treatment groups shared methylphenidate HCl for children and parent management training for parents. In the basic group, children received placebo instead of active risperidone. In the augmented group, children received active risperidone along with methylphenidate HCl. These arm descriptions define the treatment contrast in the record; further details about dosing, treatment duration, or how the behavioral training was delivered are not reported in the registry record.

The record labels the allocation as randomized and the masking as quadruple, but does not identify which participants or study personnel were masked, or describe the randomization and allocation procedures. Such details should not be inferred from the labels alone. The study start is recorded as 2008-08, primary completion as 2012-11, and status as completed. The record lists Michael Aman as sponsor.

Enrollment and the adverse-event risk sets are reported separately. The adverse-event counts list 80 participants at risk in the basic group and 73 in the augmented group. Those event denominators should be read as the risk sets attached to those safety figures, rather than assumed to explain the enrollment figure. The registry record does not provide enough information to reconcile the difference between enrollment and those safety denominators.

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: Between initial contact and randomization, 108 potential participants were lost for the following reasons: 68 subjects failed screen criteria, 4 were ineligible at Baseline, 10 withdrew consent, 4 were lost to follow-up, and 2 were unable to swallow medication.

Overall Study

Analysis population for the primary outcome, as defined in the registry: ADHD \& severe physical aggression

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: ADHD \& severe physical aggression

6. Primary Endpoint

The registry lists the primary outcome as NCBRF-TIQ D-Total Score, measured over Measured at baseline and Weeks 3, 4, 5, 6, 7, 8, 9.

The primary endpoint is the NCBRF-TIQ D-Total Score, a continuous measure based on parent ratings of aggression and hostility. It combines items from the Conduct Problem and Oppositional Behavior subscales of the Nisonger Child Behavior Rating Form-Typical IQ. The record describes the NCBRF as including a prosocial subscale and six problem-behavior subscales. The D-Total uses the six Oppositional subscale scores and the ten Conduct Problem items.

For each component item, the rating runs from 0, meaning that the behavior did not occur or was not a problem, to 3, meaning that it occurred a lot or was a very severe problem. The D-Total is formed by adding the specified ratings and can range from 0-69. Higher scores reflect worse behavior, so the direction of the scale matters when interpreting any group means or changes.

Parent ratings offer a structured way to quantify behavior across assessment visits. The record notes that the measure has excellent internal consistency and distinguishes between controls and subjects with DBDs. It also states that the Conduct Problem and Oppositional Behavior subscales map closely to DSM-IV-TR symptoms of CD and ODD. These properties explain the measure's relevance to the study question, but they do not substitute for the trial's numerical results.

The score was scheduled for measurement at baseline and Weeks 3, 4, 5, 6, 7, 8, and 9. The registry record does not reduce those repeated assessments to one reported primary comparison.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does randomization establish?

In general, randomization assigns participants to treatment groups by chance. Its purpose is to reduce systematic differences between groups at the point of assignment, so that a later group comparison can more credibly be associated with the assigned treatments. Randomization does not guarantee that groups will be identical in every characteristic, and it does not by itself prevent problems caused by missing outcome data, departures from assigned treatment, or selective reporting.

The registry record labels this trial randomized and parallel. A parallel design means the groups are compared as separate treatment groups over the study rather than as a planned sequence in which each participant receives each treatment. The record does not describe the random sequence method, allocation concealment, or the handling of participants who did not follow their assigned treatment. Those details are not reported in the registry record.

What does the masking label tell readers?

The masking field is recorded as quadruple. In general, masking is intended to limit the influence of knowledge about assignment on behavior, care, assessment, or interpretation. The registry record does not state which people were included in that masking label, how masking was maintained, or whether it was assessed. The label alone is not enough to infer those operational details.

How should a continuous outcome be summarized?

The primary outcome is identified as a mean on a continuous score, with standard deviation as the dispersion measure. In general, a mean describes the central value of a set of observations, while a standard deviation describes how spread out individual observations are around that mean. The mean and standard deviation can help readers understand score levels and variability, but neither alone describes the uncertainty in a between-group contrast. An estimate of the difference and an uncertainty interval would be needed to assess the size and precision of that contrast.

The D-Total has an explicitly stated direction: higher scores reflect worse behavior. When interpreting means, the direction needs to be kept in view. A lower mean corresponds to a more favorable score on this scale, but that interpretation must be applied to actual reported values. The record does not give the numerical group means or a primary between-group estimate.

Why do repeated assessments matter?

The score was measured at baseline and Weeks 3, 4, 5, 6, 7, 8, and 9. Repeated measurements can describe how an outcome develops during follow-up, but they also create choices about which time point or pattern should answer the primary question. In general, a longitudinal analysis may use information from multiple visits, whereas a single-visit comparison focuses on one specified assessment. The appropriate approach depends on the protocol and analysis plan. Those details are not reported in the registry record.

Multiple assessments also raise a multiplicity concern. In general, examining several time points can create more opportunities to find an apparently notable result by chance if each is tested separately. A prespecified overall test, a designated time point, or an adjustment strategy can address different aspects of that issue. The record does not report which approach was used. The primary outcome is specifically flagged as not reduced to a single comparison because it has several categories or time points.

What does a confidence interval contribute?

In general, a confidence interval expresses the precision of an estimated effect under the assumptions of the statistical method. A narrow interval indicates greater precision than a wide interval, while an interval spanning effects in different directions may leave the direction uncertain. A confidence interval is not a guarantee that the true effect lies within its endpoints, and its interpretation depends on the analysis and data. No confidence interval for the primary group contrast is reported in the registry record.

A p-value addresses how compatible the observed data are with a specified null hypothesis under the analysis assumptions. It is not a measure of clinical importance, the probability that a treatment is effective, or the size of an effect. The record gives no p-value for a primary comparison, so statistical significance cannot be assessed from the registry outcome description.

What do planning calculations mean?

In general, sample-size planning asks how many participants would be needed under specified assumptions to detect a chosen effect with a chosen error rate and power. The stated planning illustration uses a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation; it frames the effect to detect as a stated fraction of the observed effect. Such calculations are planning illustrations, not observed power and not evidence that the study was designed using those assumptions.

The registry record does not report the trial's own sample-size justification. An illustration based on an observed effect should not be mistaken for the original design rationale, nor should it be read as an additional outcome result. Its interpretation depends on its assumptions, including the effect size used. The record does not report a single primary effect estimate that could be interpreted from the primary outcome entry.

Why does over-dispersion matter in general?

Over-dispersion occurs when observed variability exceeds what a statistical model expects. In general, if a model understates variability, its uncertainty estimates may be too small and statistical evidence may appear stronger than it should. The primary measure in this trial is described as continuous, with a mean and standard deviation, and the registry record does not report an over-dispersion analysis. This concept is therefore general statistical context, not a stated feature of the trial's reported analysis.

10. Multiplicity, Missing Data and Analysis Populations

The primary outcome has assessments at baseline and Weeks 3, 4, 5, 6, 7, 8, and 9. The registry record flags the outcome as having several categories or time points and not being reduced to one comparison. Because no single primary comparison is reported, the record does not establish which visit or longitudinal summary was intended to carry the main inferential conclusion.

In general, testing multiple visits or outcomes can increase the chance of a false-positive finding if the analyses are treated as independent opportunities to claim significance. A prespecified hierarchy or multiplicity adjustment may address that risk, but the analysis plan and any multiplicity procedure are not reported in the registry record. The record lists four outcome measures and zero analyses; it does not explain the absence of analysis results.

Missing-data handling is also not reported in the registry record. The information does not state how many primary endpoint observations were available at each visit, how missing scores were handled, or whether sensitivity analyses were performed. Those details matter because estimates based only on observed visits can be affected if the participants with available data differ systematically from those without them.

The analysis population is not reported in the registry record. It is therefore not possible to determine whether the intended analysis included all randomized participants, only participants with observed outcomes, or another defined group. The record also does not describe interim analyses or stopping rules. These omissions should not be interpreted as evidence that a particular method was or was not used; they indicate that the registry information does not document it.

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 record lists serious adverse events as 0 among 80 participants at risk in the basic group and 0 among 73 participants at risk in the augmented group. It lists other adverse events as 69 among 80 at risk in the basic group and 62 among 73 at risk in the augmented group. These figures are counts within the recorded risk sets. They should not be read as a full description of the types, timing, duration, or clinical consequences of events.

A zero serious-event count means that no serious events are listed in these registry figures for the stated risk sets. It does not establish that no adverse effects of any kind occurred: the record separately lists other adverse events in both groups. The term “other” is not accompanied by event descriptions or a breakdown by event. The record does not report whether events were considered related to treatment, how severity was classified beyond the serious category, or whether participants experienced more than one event.

Death counts are not reported in the registry record. The safety figures also do not, on their own, provide a comparative risk estimate or an uncertainty interval. They are descriptive and should be interpreted alongside the limited event detail and the stated denominators, rather than as a complete safety assessment.

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

The study record describes a randomized comparison of two treatment regimens in children with ADHD and severe physical aggression. Methylphenidate HCl and parent management training were included in both arm descriptions; the augmented regimen added active risperidone, whereas the basic regimen used placebo instead of active risperidone. The primary measure was a parent-rated score for aggression and hostility, with higher scores reflecting worse behavior.

What the record does not provide is the numerical primary outcome comparison. There are no reported group means, between-group effect estimate, confidence interval, or primary p-value in the registry entry. The assessment schedule spans baseline and several follow-up weeks, and the record does not reduce those assessments to a single primary comparison. As a result, the record does not show how the primary scores differed between groups or how precisely any difference was estimated.

The safety entry reports serious-event counts of 0 in each arm, alongside counts for other adverse events. Those figures do not describe the event types or establish whether events were caused by treatment. They are not enough to characterize overall safety in detail, and death counts are not reported in the registry record.

These limits do not mean that the study had no findings, that no analysis was conducted outside the registry, or that any unreported result had a particular direction. They mean that the registry record itself cannot answer the primary comparative question numerically. A treatment effect should not be inferred from the trial title, the stated objective, the endpoint definition, or the presence of a randomized design. The design explains how the intended comparison was structured; the outcome results would be needed to estimate what the comparison showed.

The appropriate reading is therefore bounded: the record documents a trial design, a defined behavioral endpoint, its scheduled assessments, and selected adverse-event counts, but it does not present a single primary outcome estimate. Any conclusion about the size, direction, or precision of an effect would require results not reported in the registry record.

13. Why This Trial Matters Statistically

This trial is statistically notable because its treatment contrast is clearly described while the primary outcome result is not reduced to one reported comparison. The two arms shared methylphenidate HCl and parent management training, with active risperidone in one arm and placebo in the other. That structure makes the intended added-treatment contrast understandable from the arm descriptions. The primary endpoint, however, was assessed repeatedly, and the registry record does not identify a single summarized comparison across those assessments.

This distinction illustrates a core principle of trial interpretation: a well-defined design and endpoint are necessary for evaluating a question, but they do not replace an effect estimate and its uncertainty. The NCBRF-TIQ D-Total is described in detail, including its scoring range and direction, yet the registry entry does not report the group values or a numerical contrast. Readers therefore cannot assess the magnitude or precision of the primary outcome from this record.

The multiple scheduled visits also make prespecification important. In general, a clear plan for selecting or combining repeated measurements helps distinguish a primary analysis from exploratory looks across time. Without a reported analysis plan or a single primary comparison, the registry information does not show how repeated measurements were incorporated, or how multiplicity was addressed.

The safety counts provide another useful statistical lesson. Counts without event descriptions, uncertainty estimates, or treatment attribution are limited summaries. Zero listed serious events in both risk sets is relevant as a descriptive finding, but it cannot replace a fuller account of all adverse events or settle comparative safety questions.

Finally, a planning illustration based on an observed effect answers a different question from a trial's design justification or its observed treatment result. Separating those ideas helps prevent planning assumptions from being mistaken for evidence about effectiveness. The record's main statistical value is in documenting the intended comparison and the gaps that remain for evaluating its outcome.

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 the absence of a single reported primary outcome comparison. The primary measure has multiple scheduled assessments, but no numerical group results, effect estimate, confidence interval, or p-value for a primary contrast are reported. This prevents assessment of the direction, size, and precision of the comparative outcome using the registry entry alone.

The outcome information is also limited to the definition and schedule of the measure. Although the NCBRF-TIQ D-Total Score is described in detail, the record does not give the number of participants with scores at each assessment or explain missing-data handling. It does not describe the analysis population, the statistical model, or sensitivity analyses. These details are not reported in the registry record and should not be inferred.

Further details about the trial procedures are absent from the record. The randomized and quadruple-masked labels do not specify the randomization method, allocation concealment, who was masked, or how masking was maintained. Dose, treatment adherence, and the delivery and fidelity of PMT are also not reported in the registry record. The record does not give eligibility details beyond the stated condition and population description.

The safety figures are limited to counts in serious and other adverse-event categories. Event names, timing, severity details, attribution, and death counts are not reported in the registry record. The safety denominators also do not provide a reconciliation with the enrollment figure. These limitations leave the event counts as a narrow descriptive account rather than a comprehensive safety analysis.

Finally, although the study status is completed, the registry entry lists four outcome measures and zero analyses. It does not explain why outcome analyses are absent or whether results were reported elsewhere. No assumption about the existence, quality, or direction of unreported analyses is warranted from that absence.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Treatment of Severe Childhood Aggression (The TOSCA Study) — NCT00796302.
• Registry record last updated 2017-07-26; status last verified 2017-06.
• Results first posted 2016-04-20.
• 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

The registry record describes a completed, randomized, quadruple-masked, parallel phase 4 trial with an enrollment of 168. It compares methylphenidate HCl and parent management training plus active risperidone with the same background regimen plus placebo instead of active risperidone.

The primary outcome is the parent-rated NCBRF-TIQ D-Total Score, measured at baseline and Weeks 3, 4, 5, 6, 7, 8, and 9. Higher scores reflect worse behavior, and the score can range from 0-69. The record does not provide a single primary comparison or numerical outcome results.

Serious adverse-event counts are 0 among 80 at risk in the basic group and 0 among 73 at risk in the augmented group. Other adverse-event counts are 69 and 62, respectively. Death counts are not reported in the registry record. Further analysis, missing-data, and safety details are not reported in the registry record.