Oxytocin and Social Cognitive Skills Groups: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Oxytocin and Social Cognitive Skills Groups (NCT02918864), a randomized phase 2 study. The primary outcome is Change From Baseline in Social Behavior Impairment (SBI) Composite.

2. Clinical Question

This is a question about change over time between randomized groups, rather than a question about whether either group changed on its own. The registry’s description says that larger negative slopes represent improvement because higher SBI scores indicate relatively more impairment. Without the group-specific slopes, their difference, or uncertainty estimates, the direction and precision of the primary comparison cannot be assessed from the record.

3. Trial Design

The study is registered as a phase 2 randomized trial with a parallel-group model and single masking. It enrolled 40 participants and is marked completed. The registry lists a start date of 2016-06-15 and primary completion in 2021-09. It does not identify who was masked, so the meaning of single masking for participants, clinicians, assessors, or analysts is not reported in the registry record.

Participants were assigned to one of two groups: ION-ASD or facilitated play. ION-ASD combined intranasal oxytocin with social-cognitive skills group training, using the Seaver-NETT curriculum. The comparison group received facilitated play therapy, described as an active comparison condition. The record does not give the allocation ratio, sequence-generation method, concealment procedures, treatment frequency, duration of sessions, or detailed dosing schedule.

The primary endpoint was measured at baseline and Week 12, designated the endpoint assessment. It was a continuous composite score, with results intended to be represented as the change over time or slope for each group and compared through a group-by-time interaction. The registry states that all participants contributed to the intent-to-treat model. It does not describe how missing assessments, deviations from assigned treatment, or participants with incomplete follow-up were handled.

The study’s stated purpose was to evaluate feasibility, safety, and preliminary efficacy. The record lists 27 outcome measures and 15 analyses, including responder status at several follow-up points and social-functioning outcomes. Numerical results for the primary endpoint and the listed secondary outcomes are not reported in the registry record. Thus, the design and planned question can be described, but the primary comparison’s observed magnitude and precision cannot be reconstructed from these registry entries.

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: All participants contributed to the intent to treat (ITT) model.

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 Change From Baseline in Social Behavior Impairment (SBI) Composite, measured over Baseline and Week 12 (Endpoint).

The primary endpoint was change from baseline in the Social Behavior Impairment (SBI) composite, expressed as a z-score. It combines measures from the Social Relations and Nonverbal Communication subscales of the Children’s Communication Checklist-2 and the Griffith Empathy Measure. The registry describes standardizing these measures using the study sample’s baseline means and standard deviations, reversing the GEM total score, and averaging the resulting z-scores. This construction makes higher values indicate more impairment across the included measures.

The SBI was used to represent several aspects of social behavior in one composite outcome. A composite can provide a broad summary when a study is interested in related skills rather than a single isolated measure. It also means that the result depends on how the component measures are standardized and combined. The endpoint is based on the approach attributed in the record to Soorya et al. (2015).

At baseline, the sample average SBI z-score equals 0 by construction; that does not mean every participant begins at zero. Individual scores above or below the average indicate relative position within the baseline sample. The primary question concerns the difference between groups in the rate of change from baseline to Week 12, not simply the endpoint score in one group. The record explains that larger negative slopes represent improvement. The actual group slopes, their difference, and an uncertainty interval are not reported in the registry record, so the observed change cannot be characterized numerically.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary endpoint is continuous: each participant’s SBI is a numerical score, and the study’s stated comparison concerns change in that score over time. The registry describes the outcome as a slope for each group and identifies the primary contrast as the group-by-time interaction. In general, an interaction asks whether the pattern of change over time differs between groups. It is not the same as testing whether one group changes from its own baseline, nor is it simply a comparison of the groups at one follow-up assessment.

What does a slope represent?

In general, a slope describes the direction and rate of change across measurement occasions. For this outcome, the registry states that larger negative slopes represent improvement because higher SBI scores mean more impairment. The slope’s value would convey the rate of change under the fitted model, while the difference between group slopes would address the randomized comparison. The numerical slopes and their difference are not reported in the registry record.

What does the standard error tell a reader?

The primary outcome is specified as a mean with standard error. In general, a standard error describes the estimated uncertainty of a statistic, such as a group mean or model estimate, due to sampling variation under the analysis. It is not the same as the spread of individual participant scores. A confidence interval, when reported, combines an estimate with its uncertainty to show a range of values compatible with the data and model assumptions. Neither the primary estimate nor its standard error or confidence interval is reported in the registry record, so its precision cannot be assessed.

Why use a composite z-score?

The SBI averages standardized scores from the listed CCC-2 subscales and the GEM, with the GEM total reversed. In general, standardization places measures with different original scales onto a common reference scale, which can make averaging more interpretable than averaging raw scores with unlike units. Here, the registry says standardization used the sample means and standard deviations at baseline. The resulting composite is anchored to the sample’s baseline distribution, with a sample average of 0 at baseline. An individual’s score still varies, and the score’s meaning is relative to that baseline reference.

Composite outcomes also involve trade-offs. In general, combining measures can summarize a wider construct, but it can obscure whether changes are concentrated in one component or shared across components. The registry does not report component-level primary results or additional details about the construction beyond the stated measures and averaging procedure. It is therefore not possible to determine how much each component contributed to any observed change.

What does the intent-to-treat description establish?

The registry says all participants contributed to the intent-to-treat model. In general, intent-to-treat analysis aims to compare groups according to randomized assignment, regardless of whether participants fully follow their assigned treatment. This approach can preserve the advantages of randomization. The label alone does not explain the specific model, the treatment of missing measurements, or the handling of protocol deviations. Those details are not reported in the registry record.

How should the sample-size illustrations be read?

The record describes a planning illustration based on the observed effect, using a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The illustration considers effects defined as stated fractions of the observed effect. In general, such calculations estimate the number of participants per arm needed under their assumptions to detect a specified effect at a chosen power. They are planning illustrations, not observed power and not a report of the trial’s actual design justification. The registry record does not report the numerical sample-size outputs or the trial’s own sample-size justification.

The assumptions matter. In general, a calculation based on a fraction of an observed effect changes the effect size the calculation is intended to detect; it does not change the effect observed in the trial. Equal allocation is an assumed planning arrangement in the illustration. The record separately describes the trial as randomized, but does not report its allocation ratio. The normal approximation is a calculation assumption, not evidence that all model assumptions were checked or satisfied. Details of any underlying effect estimate used for the illustrations are not reported in the registry record.

Finally, statistical significance and practical importance are different concepts. In general, a p-value addresses compatibility between data and a specified null model under assumptions; an effect estimate and interval are needed to assess magnitude and precision. Because these primary numerical results are not reported in the registry record, neither statistical evidence for a group difference nor the size of any difference can be evaluated from the posted information.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists 27 outcome measures and 15 analyses. The listed secondary outcomes include CGI-I responder classifications at Week 12, Week 16, and Week 24, along with changes and rates of change in SRS-2 social functioning. Repeated outcomes and follow-up times can generate multiple statistical comparisons. In general, the more comparisons considered, the greater the chance that at least one result appears notable by chance if no adjustment or prespecified interpretation strategy is used.

The registry record does not report whether multiplicity was addressed, which outcomes or time points were treated as confirmatory, whether any hierarchy was used, or whether results were adjusted for multiple comparisons. These details cannot be inferred from the number of outcomes or analyses. Numerical results for the primary and listed secondary outcomes are not reported in the registry record, so the possible influence of multiplicity on interpretation cannot be evaluated.

The primary outcome description states that all participants contributed to the intent-to-treat model. This identifies the stated analysis population, but does not explain whether all participants had observed endpoint measurements or how missing measurements were handled. The registry record does not report the amount or pattern of missing data, reasons for missingness, imputation methods, sensitivity analyses, or exclusions from particular analyses.

In general, missing-data methods matter because an analysis may depend on assumptions about why measurements are unavailable and how unobserved values relate to observed information. Stating an intent-to-treat population does not, by itself, resolve these questions. The record also does not report the full statistical analysis plan, any interim analyses, or whether planned analyses changed. Those aspects are not reported in the registry record and should not be guessed from the completed status or the listed outcomes.

11. 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: impulsivity (5.0%), aggression (5.0%), anxiety (5.0%), irritable (5.0%), mood lability (5.0%).

The registry lists 20 participants at risk for serious adverse events in each group. Serious adverse events affected 0 participants in ION-ASD and 1 participant in facilitated play. Deaths were recorded for 0 participants in both groups. These are participant counts in the registry record; they do not provide a complete description of the events or establish their relationship to treatment.

For other adverse events, the record lists 20 participants at risk and 20 affected in each group. These figures are distinct from the serious adverse-event counts. The record does not provide enough detail in these figures to describe the types, severity, timing, or duration of all other events, or to determine which participants experienced more than one event.

The listed serious-event terms are Impulsivity, Aggression, Anxiety, Irritable, and Mood lability, each reported as 5.0%. The record does not identify the group for each term in the listed information. The terms should not be interpreted as an arm-by-arm comparison, nor assumed to represent distinct participants. In general, adverse-event counts are descriptive: they require context about event definitions, ascertainment, timing, and attribution for a fuller safety interpretation. That context is not reported in the registry record.

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

The trial addresses whether a combined intervention—targeted intranasal oxytocin and social-cognitive skills group training—has a different pattern of change in social behavior impairment than facilitated play therapy. Its primary outcome and the direction of its score are described, as are the randomized parallel-group design and the stated intent-to-treat model. These details define the question, but they are not a substitute for the outcome estimates.

The primary group-by-time result is not reported in the registry record. Consequently, the posted information does not establish the direction, magnitude, or precision of a difference between groups in SBI change. It does not establish that the groups had the same trajectory either. A missing numerical result should be treated as unavailable evidence, rather than interpreted as a null result.

The record also includes secondary measures and follow-up points but does not report their numerical findings. Therefore, no conclusion about responder counts, later follow-up, or SRS-2 change can be drawn from the listed outcome names alone. Nor can the feasibility or preliminary-efficacy aims be evaluated using numerical results absent from the record.

The adverse-event entries provide limited descriptive information: serious adverse events affected 0 participants in ION-ASD and 1 in facilitated play, deaths were recorded for 0 in either group, and other adverse events affected 20 participants in each group. These counts do not establish causation, rule out risk, or fully characterize comparative safety. The registry’s event-term list also lacks the group attribution needed to interpret those terms as an arm comparison.

In general, a clinical trial result is best interpreted using the estimated contrast, its uncertainty, the analysis population, and the methods used to handle missing data and multiple outcomes. Several of these details are not reported in the registry record. The appropriate interpretation is therefore limited: the record describes the design and measurement approach, but it does not supply the numerical primary outcome evidence needed to judge the observed between-group pattern.

13. Why This Trial Matters Statistically

This trial illustrates a statistical challenge common in behavioral and clinical research: translating a multidimensional construct into an outcome that can be compared between randomized groups. The SBI combines standardized social-relations, nonverbal-communication, and empathy measures. Its composite structure offers a broad summary, while making it important to understand how its components are scaled, combined, and interpreted.

The primary question is also framed around trajectories rather than a single endpoint value. Comparing rates of change through a group-by-time interaction focuses the analysis on whether the groups’ patterns differ over the baseline-to-Week-12 period. That distinction matters because within-group change alone does not answer the randomized comparison. The record describes the intended contrast and the direction of the score, but does not report its estimate or uncertainty.

The design includes an active comparison group, facilitated play therapy, rather than describing the comparator as no treatment. This makes the question specifically about the contrast between the integrated ION-ASD approach and that group condition. The record does not provide detailed treatment schedules or other implementation information, so interpretation of the contrast’s exact content remains limited.

The record’s 27 outcome measures and 15 analyses also illustrate why statistical interpretation depends on understanding the analysis plan, outcome hierarchy, and multiplicity approach. Those features are not reported in the registry record. Similarly, an intent-to-treat label identifies the stated population for the primary model but does not answer how missing follow-up data were handled.

Finally, the case shows the difference between having a registered trial description and having enough posted results to assess an effect. A stated endpoint, sample-size assumptions, and adverse-event counts do not replace the primary estimate with its precision. Without those results, the statistical contribution is chiefly the definition of the question and the limits of what can be concluded from the available registry information.

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 for interpretation is that the registry record does not report numerical results for the primary SBI comparison. Group slopes, the group-by-time estimate, its uncertainty, and a p-value are not available in the record. This prevents assessment of the observed difference and its precision, despite the endpoint and analysis population being described.

The endpoint is a composite assembled from multiple measures. The record explains the baseline standardization, reversal of the GEM total score, and averaging, but does not report component-level results or further validation information for this trial. A change in the composite would not, by itself, show whether change was similar across all components. No such change is numerically reported in the registry record.

The trial enrolled 40 participants, but the record does not report the allocation ratio, detailed randomization procedures, who was masked, treatment adherence, or detailed intervention schedules. Single masking is recorded, but the masked role is not reported in the registry record. These omissions limit evaluation of implementation and potential sources of bias.

The record does not describe missing-data amounts or methods, multiplicity handling, or a full statistical analysis plan. Although all participants are said to have contributed to the intent-to-treat model, that statement does not reveal whether data were complete or how incomplete observations were addressed. The listed number of outcomes and analyses makes these unspecified details relevant, but does not establish that any particular method was or was not used.

Safety interpretation is also constrained. The registry provides participant counts for serious and other adverse events and a list of serious-event terms, but limited detail about timing, severity, attribution, repeated events, or group assignment of those terms. Finally, the registry states planning assumptions for sample-size illustrations without reporting their numerical outputs or the trial’s own sample-size justification. These limitations narrow what can be inferred; they do not indicate the direction of unreported results.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Oxytocin and Social Cognitive Skills Groups — NCT02918864.
• Registry record last updated 2024-04-05; status last verified 2024-04.
• Results first posted 2024-04-05.
• 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

NCT02918864, titled “Oxytocin and Social Cognitive Skills Groups,” is a completed randomized phase 2 parallel-group trial with single masking. The registry lists 40 enrolled participants and identifies autism spectrum disorder as the condition. The intervention group combined targeted intranasal oxytocin with Seaver-NETT social-cognitive skills training; the active comparison group received facilitated play therapy.

The primary endpoint was change from baseline to Week 12 in the SBI composite, assessed as a difference between group slopes. Higher scores indicate relatively more impairment, and larger negative slopes represent improvement. All participants are stated to have contributed to the intent-to-treat model. Numerical primary and secondary outcome results are not reported in the registry record.

The registry records serious adverse events in 0 of 20 participants at risk in ION-ASD and 1 of 20 in facilitated play, with no deaths in either group. Other adverse events affected 20 participants in each group. The record does not provide enough detail to make a full comparative safety assessment.