Efficacy of GXR as Adjunctive Therapy With Psycho-stimulant on Executive Function in Children With ADHD: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Efficacy of GXR as Adjunctive Therapy With Psycho-stimulant on Executive Function in Children With ADHD (NCT01985581), a randomized phase 4 study. The primary outcome is Effect of Adjunctive INTUNIV Extended Release Treatment on Executive Function as Assessed by Change From Baseline on the BRIEF-parent Questionnaires.

2. Clinical Question

The registry describes a crossover evaluation: participants received both adjunctive treatments in different sequences, with one sequence starting with placebo and the other starting with GXR. The central statistical question is therefore a comparison of the reported changes under the two treatment conditions, rather than a comparison of two unrelated groups. The registry record gives arm-level summaries and a sponsor-reported analysis, but does not report enough detail to reconstruct the full crossover model or its assumptions. The primary outcome concerns parent assessments of everyday executive function; quality of life, ADHD symptom ratings, and safety outcomes were also included among the secondary outcomes.

3. Trial Design

NCT01985581 is recorded as a completed Phase 4, randomized, quadruple-masked, placebo-controlled, single-center crossover study. The registry lists enrollment of 50 participants. The study began in 2013-10, and primary completion was recorded as 2015-03. The official title describes an evaluation of GXR as adjunctive treatment with stimulant therapy in children with ADHD, focusing on executive function and quality of life at home and school.

Participants continued their stable usual stimulant therapy while receiving adjunctive GXR and placebo in separate intervention periods. One sequence was placebo first and GXR second; the other was GXR first and placebo second. Each intervention period is described as 12 weeks. The GXR dose was optimized to between 1 and 4mg. The stimulant medicines listed in the arm descriptions include Ritalin, Ritalin SR, Biphentin, Concerta, Vyvanse, Adderall, or Dexedrine.

For a crossover design, each participant can contribute observations under both treatment conditions. In general, this can make a within-participant comparison informative because stable differences between participants may be less influential than in a parallel-group comparison. However, the registry record does not report a washout period, the timing of any washout, or the detailed statistical handling of period and sequence effects. It also does not report the trial’s original sample-size justification or the full analysis plan. The enrollment figure is the planned or recorded enrollment, whereas the outcome-specific counts differ by measure; the primary outcome summaries include 48 participants in the placebo-and-stimulant condition and 47 in the GXR-and-stimulant condition.

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.

First Intervention

Second Intervention

Analysis population for the primary outcome, as defined in the registry: ITT population - consisting of subjects who took at least one dose of treatment and completed at least one non-baseline BRIEF questionnaire during either period 1 or period 2.

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 Effect of Adjunctive INTUNIV Extended Release Treatment on Executive Function as Assessed by Change From Baseline on the BRIEF-parent Questionnaires, measured over measured at baseline and end of each 12 week treament arm.

The primary endpoint is change from baseline in the parent-completed Behavioural Rating Inventory of Executive Function (BRIEF-parent). The registry describes the questionnaire as an 86 item measure of real-world expressions of executive function at home. Executive functions include processes such as emotional control, planning, organization, working memory, behavioral inhibition, and managing time and space. These are assessed through a parent’s report rather than by a direct performance test in the endpoint described.

The questionnaire score is converted to a t-score. The registry states that a score less than 65 is considered within the normal range and that higher scores indicate worsening function. Because the endpoint is change from baseline, the treatment comparison concerns how the score changed over each treatment period, not simply the score at one time point. A negative change indicates a reduction in the score; given the stated score direction, that is a change toward lower reported difficulty. The registry does not provide the baseline or end-of-period mean score levels in the posted outcome summaries, so the change estimates alone do not establish the participants’ absolute level of functioning.

Parent ratings are relevant to the stated focus on everyday functioning at home, but the outcome is still a questionnaire-based assessment. The registry also describes teacher ratings in its study summary, while the posted primary endpoint is specifically the BRIEF-parent measure. The endpoint was measured at baseline and at the end of each 12 week treatment arm.

7. Results Posted in the Registry

• GXR and Stimulant: mean -6.90 versus -5.30 on control; difference -1.60 (95% CI -5.74 to 2.54), standardized effect d = -0.16; z-test p = 0.449.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Effect of Adjunctive INTUNIV Extended Release Treatment on Change in Quality of Life as Assessed by the KINDL®-Child Questionnaire.

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard errors.
• Result:
  • GXR and Stimulant: mean 1.20 versus 0.80 on control; difference 0.40 (95% CI -4.04 to 4.84), standardized effect d = 0.04; z-test p = 0.860.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between GXR and Stimulant and the control. The interval gives the range of effects that remain compatible with the data.

Effect of Adjunct Therapy on ADHD Symptom Control as Assessed by the Change in the ADHD Rating Scale (ADHD-RS-IV)

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard errors.
• Result:
  • GXR and Stimulant: mean -11.10 versus -5.50 on control; difference -5.60 (95% CI -9.95 to -1.25), standardized effect d = -0.52; z-test p = 0.012.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for GXR and Stimulant than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Evaluate the Effect of Adjunctive INTUNIV Extended Release Treatment on Change in Quality of Life as Assessed by the KINDL®-Parent Questionnaire.

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard errors.
• Result:
  • GXR and Stimulant: mean 1.40 versus 3.40 on control; difference -2.00 (95% CI -6.03 to 2.03), standardized effect d = -0.20; z-test p = 0.331.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between GXR and Stimulant and the control. The interval gives the range of effects that remain compatible with the data.

The posted secondary outcomes cover quality of life, ADHD symptoms, and safety. Across the continuous outcomes with posted numerical comparisons, the broad pattern is not uniform: the symptom-rating outcome has more consistent evidence of a difference across the sponsor analysis and the simple recalculation, while quality-of-life outcomes do not show similarly clear evidence. One quality-of-life comparison also illustrates that model-based and unadjusted results need not align exactly. The detailed outcome estimates are best read with their respective outcome definitions and analysis methods, rather than combined into one overall claim.

This pattern provides context for the primary BRIEF-parent result, but it does not independently settle the discrepancy between the sponsor’s primary model and the primary recalculation. The outcomes measure different constructs, use different questionnaires or ratings, and may have different analysis populations. In addition, the registry lists six secondary outcomes, while numerical summaries are posted for only some of them. The registry record does not explain the absence of numerical results for the remaining secondary outcomes.

Secondary outcomes generally carry less confirmatory weight than the primary endpoint because they are not the main stated question and because multiple outcomes create more opportunities for apparently notable findings. No multiplicity adjustment is reported in the registry record. Therefore, unadjusted secondary p-values should be regarded as exploratory evidence rather than treated as a set of independent confirmations. The secondary pattern may help identify areas for further study, but it does not establish an overall treatment effect across executive function, symptoms, and quality of life.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Effect of Adjunctive INTUNIV Extended Release Treatment on Change in Quality of Life as Assessed by the KINDL®-Child Questionnaire.
• Effect of Adjunct Therapy on ADHD Symptom Control as Assessed by the Change in the ADHD Rating Scale (ADHD-RS-IV)
• Subjects Experiencing Suicidal Ideation, Suicidal Behaviour and Self-injurious Behaviour Without Suicidal Intent and Incident of Serious Adverse Events in Each Treatment Arm
• Evaluate the Effect of Adjunctive INTUNIV Extended Release Treatment on Change in Quality of Life as Assessed by the KINDL®-Parent Questionnaire.
• Effect of Adjunct Therapy on ADHD Symptom Control as Assessed by the Change on the Clinical Global Impression of Severity (CGI-S) Scale
• Evaluate the Effect of Adjunct Therapy on ADHD Symptom Control as Assessed by the Change in Clinical Global Impression of Improvement (CGI-I) Scale

9. Statistical Interpretation

10. Statistical Methods Explained

The primary measure is continuous: it is a score on a scale, and the registry reports means and standard errors. A mean describes the average observed change among participants contributing data to a treatment condition. A standard error describes the precision of an estimated mean, rather than the spread of individual participants’ scores. The registry’s primary summaries report 48 participants for placebo plus stimulant and 47 for GXR plus stimulant.

What does the unadjusted recalculation compare?

The recalculation subtracts the posted mean change for placebo plus stimulant from the posted mean change for GXR plus stimulant. The resulting mean difference is -1.6. It is an unadjusted comparison derived from the arm-level summary statistics; it is not the sponsor’s crossover-model estimate. In general, an unadjusted comparison of separate condition summaries does not necessarily reproduce a model that uses the crossover structure or other adjustments. The registry record does not provide the detailed model specification needed to identify exactly which features account for the difference between the two estimates.

The recalculation also gives a standardized effect estimate of -0.155 using the posted pooled standard deviation of 10.3. A standardized effect expresses a difference relative to a measure of variability, so it can describe the estimated difference without retaining the original scale units. It does not replace the scale-based mean difference, and it does not establish clinical importance. The registry record does not report a threshold for a clinically important BRIEF-parent change.

How should the confidence interval and p-value be read?

A confidence interval expresses the uncertainty around an estimated difference under the assumptions of the analysis. The recalculated 95% confidence interval extends from -5.745 to 2.545. Since it includes zero, the posted summary statistics are compatible with no difference as well as with differences in either direction under this calculation. A p-value describes how unusual data at least as inconsistent with a specified no-difference model would be, if that model and its assumptions were correct. The recalculated p-value is 0.449. It is not the probability that the treatments are the same, nor does it measure the size or importance of an effect.

The sponsor’s reported primary analysis is an ANOVA with an LS mean difference of -3.0, a 95% confidence interval from -5.9 to -0.2, and p = 0.039. An LS mean is a model-based mean adjusted according to the model’s terms; the record does not state the full set of terms or the precise contrast coding. The reported interval excludes zero, unlike the recalculated interval. Because the record flags a discrepancy, the sponsor’s analysis takes precedence. The reported and recalculated results should be interpreted as distinct analyses rather than as two estimates of a fully specified common model.

Why does the crossover design matter?

In general, a crossover study compares treatments received by the same participants in separate periods. That can account for stable participant characteristics when the analysis explicitly uses within-participant information. Period effects, order effects, and carryover from one treatment period to another can also matter. The registry record does not report whether or how these effects were assessed, the details of any washout, or the model’s handling of repeated observations. Therefore, the full appropriateness and assumptions of the sponsor’s analysis cannot be evaluated from the posted description alone.

For general context, when repeated observations from one participant are related, an analysis that treats them as independent can misrepresent uncertainty. A model designed for the study structure may account for that relationship, but the details matter. The available record identifies an ANOVA and reports LS mean differences; it does not give enough detail to explain the treatment of within-participant correlation or missing observations.

What can the secondary analyses add?

The secondary outcomes include continuous questionnaire and rating-scale changes, along with safety outcomes. The registry reports sponsor ANOVA results for the posted continuous secondary outcomes and also supplies summary-statistic recalculations. These analyses are not all equivalent: the ANOVA reports model-based LS mean differences, while the recalculations are unadjusted comparisons based on posted means and standard errors. The additional analyses can show whether outcome patterns are broadly consistent, but should be read with attention to their secondary status and the number of outcomes considered.

In general, testing multiple outcomes can increase the chance that at least one result appears statistically notable even when there is no true effect across the outcomes. No multiplicity adjustment is reported in the registry record. That absence does not invalidate an individual result, but it means unadjusted secondary p-values are less persuasive as confirmatory evidence. The record also does not report a complete statistical analysis plan, missing-data method, or interim-analysis plan. Those omissions limit assessment of whether the reported analyses were prespecified and how incomplete outcome data were handled.

11. Confidence Intervals and Precision

The primary unadjusted recalculation estimates a mean difference of -1.6, with a 95% confidence interval from -5.745 to 2.545. The interval is wide enough to include a range of plausible differences, including zero and values on either side of zero. It therefore does not locate the primary effect precisely. The recalculated p-value of 0.449 is consistent with that uncertainty, but the interval is more informative about the range of estimates compatible with the calculation than a single significance decision.

The sponsor-reported 95% confidence interval is from -5.9 to -0.2 around an LS mean difference of -3.0. This interval excludes zero, but it also spans more than one possible magnitude of difference. Its interpretation depends on the sponsor model and the contrast used. The registry record does not provide enough detail to reproduce the model or resolve why this interval differs from the unadjusted recalculation’s interval. Given the flagged discrepancy, the sponsor analysis takes precedence, while the recalculation remains a distinct descriptive check.

In general, a confidence interval indicates statistical precision under the analysis assumptions; it does not show the probability that a particular value is true. A narrow interval indicates greater precision than a wide interval, but even a precise estimate may not be clinically important. Conversely, an interval that includes zero does not prove there is no difference. The registry record does not provide a clinically important difference threshold for BRIEF-parent scores, so the clinical importance of the values within either interval cannot be determined from these results alone.

The outcome-specific sample counts are close to the enrollment figure, but the record does not explain missingness by period or provide the reasons for participants not contributing to each analysis. Without that information and the detailed crossover model, precision should be understood as specific to each posted analysis rather than as a complete account of uncertainty arising from study design and missing data.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The sample-size figures are planning illustrations based on the observed effect, not the trial’s original design justification. They give participants per arm for the listed power targets. They are not totals of events: the outcomes in this record are continuous, and the displayed quantities refer to participants per arm. The planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is the stated fraction of the observed effect.

The 80% power column describes the number of participants per arm in a hypothetical study intended to have an 80% chance of detecting the specified effect under those planning assumptions, if that effect and the assumptions were correct. The 90% power column uses a higher target chance of detection and consequently requires more participants per arm for the same specified effect. These are design calculations, not a summary of what happened in this trial; they are not observed power.

The rows that plan for 100%, 75%, and 50% of the observed effect illustrate how the assumed target effect changes the required sample size. A smaller effect is harder to distinguish from random variation, so the planning calculation requires more participants per arm to target the same power. The numbers can be useful for understanding the scale of a future study under the stated assumptions, but they inherit uncertainty from the observed effect used as their starting point.

The trial’s own sample-size assumptions and original justification are not reported in the registry record. These planning illustrations should therefore not be read as evidence that the completed study was designed for a particular power target, nor as proof that the study was adequately or inadequately sized.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = -1.600; pooled standard deviation σ = 10.302.
2. n = 2 × (1.96 + 0.842)² × 10.302² / -1.600² = 650.8, rounded up to 651 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
delta <- -1.6                      # observed difference in means
sigma <- 10.3018504444903                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(1985581)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 651 (80% power), 872 (90% power)
 75% of observed effect: n per arm = 1,157 (80% power), 1,549 (90% power)
 50% of observed effect: n per arm = 2,604 (80% power), 3,485 (90% power)
power.t.test: 651.7 
Simulated power at n = 651 per arm: 0.798 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Analyses Reported by the Sponsor

The sponsor’s primary analysis is an ANOVA reporting an LS mean difference of -3.0, a 95% confidence interval from -5.9 to -0.2, and p = 0.039. The registry also permits an unadjusted recalculation from the posted means and standard errors: that estimate is -1.6, with a 95% confidence interval from -5.745 to 2.545 and p = 0.449. These results differ in estimate and in whether the interval includes zero.

The recalculation is a simple comparison of posted summary statistics. The sponsor result is model-based, and the registry record does not provide enough detail about the model or contrast to reproduce it. A crossover analysis may use information that is not represented in a simple arm-summary calculation, but the record does not establish which model features explain this discrepancy. It would be inappropriate to choose between the analyses by treating the recalculation as a replacement for the sponsor model.

Because the discrepancy is flagged, the sponsor’s own analysis takes precedence. That means the sponsor-reported estimate is the primary inferential result to emphasize, while the recalculation is useful as a check showing that a simple unadjusted analysis leads to a different conclusion. The available record does not resolve why they differ, so readers should keep the analysis method attached to each estimate rather than presenting either as a universal summary of the primary outcome.

15. Multiplicity, Missing Data and Analysis Populations

The registry identifies one primary outcome and six secondary outcomes. Testing multiple outcomes raises a general statistical concern: when each is assessed separately, the chance of obtaining at least one apparently notable result can rise. This is why confirmatory trials may define a multiplicity strategy, such as a hierarchy or adjustment to the testing thresholds. The registry record does not report a multiplicity adjustment or a testing hierarchy. Accordingly, secondary p-values should be read as unadjusted unless the record states otherwise, and they carry less weight as confirmatory findings.

The registry record also does not report a full statistical analysis plan, the original sample-size justification, or whether interim analyses were conducted. The analysis population for the primary endpoint is described as participants who took at least one dose and completed at least one non-baseline BRIEF questionnaire during either period. That is the posted population definition; the record does not provide enough detail to determine how every randomized participant was handled across all outcomes and periods.

Missing data can affect both the estimated treatment difference and its precision, particularly in a crossover study where observations may be incomplete for one period or one treatment. The registry record does not report reasons for missing outcome data, the number missing by sequence or period, or the method used to address missing observations. It also does not state whether sensitivity analyses were conducted. These omissions limit assessment of whether missingness could have influenced the reported results.

The study’s crossover structure makes the analysis plan especially relevant because treatment, period, sequence, and within-participant relationships can be important to interpretation. The record reports an ANOVA and LS mean differences but does not describe its terms or explain how these design features were handled. This does not show that the analysis was inappropriate; it means that the posted information is insufficient to evaluate those details fully.

16. 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 zero serious adverse events among 47 participants at risk in the stimulant-and-GXR condition and zero among 48 participants at risk in the stimulant-and-placebo condition. For other adverse events, 41 participants are reported as affected in each condition, with the corresponding at-risk counts of 47 and 48. These are counts of participants with events, not a description of how many separate events each participant experienced.

The record does not list serious adverse event terms, and no deaths figure is reported in the posted safety data. The zero serious-event counts describe the registry entry; they do not show that no other safety events occurred, since other adverse events are separately recorded. Nor do the counts alone establish that the conditions have identical safety profiles. The registry does not provide event-level timing, severity details for other adverse events, or a causal assessment for each event.

These safety summaries should be interpreted within the recorded numbers at risk and the trial’s follow-up structure. The registry record does not report a detailed safety analysis plan or enough information to make a broader comparative safety assessment from these counts alone.

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

The posted primary outcome is a parent-rated questionnaire score reflecting everyday executive function at home. The sponsor-reported model estimates a difference between treatment conditions, while the simple recalculation from arm summaries yields a different estimate and a confidence interval that includes zero. Because the discrepancy is flagged, the sponsor’s analysis takes precedence. The results do not support treating the unadjusted calculation as a substitute for that model, and the registry record does not explain the difference in enough detail to reconcile it.

The BRIEF-parent score direction is important: higher scores indicate worsening function, and a score less than 65 is described as within the normal range. The posted outcome is change from baseline, not a report of absolute score levels. No clinically important difference threshold is given in the registry record. Therefore, neither statistical significance in the sponsor model nor the direction of a mean change alone establishes the practical importance of the estimated difference.

The secondary results assess other outcomes, including ADHD symptoms and quality of life, but they are not equivalent to the primary endpoint and do not resolve its analysis discrepancy. Their unadjusted status and the number of outcomes assessed limit how strongly they can be treated as confirmatory. The safety counts likewise describe recorded events without establishing causation or a comprehensive comparative safety profile.

These findings are specific to the population, treatment conditions, measures, and analyses recorded for this trial. They do not establish what would happen in other populations or settings, and they are not a basis for individual treatment decisions. The registry record does not provide a complete analysis plan, missing-data account, or full crossover model description, all of which would help readers assess the limits of the reported comparisons.

18. Why This Trial Matters Statistically

This trial illustrates why the analysis method matters as much as the point estimate. A simple recalculation from posted means and standard errors gives a primary result with a confidence interval that includes zero, while the sponsor-reported ANOVA gives a different estimate and an interval that excludes zero. The discrepancy is material to statistical interpretation. Keeping the methods attached to their estimates prevents a recalculation from being mistaken for the sponsor’s analysis or vice versa.

The crossover design is also statistically important. In general, crossover trials can use within-participant comparisons, but those comparisons require attention to period, sequence, and possible carryover effects. The registry record identifies the design and reports an ANOVA, yet does not give enough detail to assess how those features were modeled. The limitation is not merely technical: without those details, the sponsor estimate cannot be fully reconstructed from the registry description.

The record also distinguishes a statistically notable model result from evidence of practical importance. No threshold for a clinically important BRIEF-parent change is reported, and the endpoint is a parent questionnaire rather than a direct measure of all aspects of functioning. Finally, the secondary outcomes and safety counts broaden the set of questions considered, but their interpretation depends on multiplicity, missing-data reporting, and the analysis population. This trial is therefore a useful example of reading estimates alongside design, model, precision, and reporting limitations.

19. 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 registry record provides limited detail about the crossover implementation and analysis. It does not report a washout period, period or sequence effects, carryover assessment, or the full ANOVA specification. These details would help determine how the treatment contrast was estimated and whether the sponsor analysis can be independently reproduced. The record also does not report the original sample-size justification or the statistical analysis plan.

The primary outcome is parent-reported, and the posted result concerns change in a questionnaire score. Although the registry describes the scale direction and a score threshold for the normal range, it does not report a clinically important difference threshold or absolute baseline and end-of-period scores. The practical meaning of the estimated change is therefore difficult to establish from the posted summaries alone.

The primary analysis and unadjusted recalculation differ, and the registry does not explain the source of that discrepancy. The sponsor analysis takes precedence, but the missing model details limit independent appraisal. The record also does not explain missing outcome data, provide counts by period or sequence for all measures, or describe sensitivity analyses. These gaps matter in a crossover study because incomplete observations can affect the within-participant comparison.

Finally, the registry lists six secondary outcomes, but numerical summaries are not posted for all of them, and no multiplicity adjustment is reported. The safety information is limited to participant counts for serious and other adverse events; event details and a deaths figure are not reported in the posted safety data. These limitations constrain the breadth and certainty of conclusions that can be drawn from the registry record.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Efficacy of GXR as Adjunctive Therapy With Psycho-stimulant on Executive Function in Children With ADHD — NCT01985581.
• Registry record last updated 2016-05-23; status last verified 2016-05.
• Results first posted 2016-05-23.
• 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.

23. Record Summary

NCT01985581 is a completed randomized, quadruple-masked, single-center Phase 4 crossover trial with recorded enrollment of 50. Children with ADHD continued stable stimulant therapy and received adjunctive GXR and placebo in separate 12-week periods. The primary outcome was change from baseline in parent-rated BRIEF executive-function score.

The sponsor-reported ANOVA estimated an LS mean difference of -3.0, with a 95% confidence interval from -5.9 to -0.2 and p = 0.039. The unadjusted recalculation differs, with an estimate of -1.6, a 95% confidence interval from -5.745 to 2.545, and p = 0.449. The sponsor analysis takes precedence because the discrepancy is flagged. The record does not report enough detail to reconcile the analyses or fully assess missing data, multiplicity, and crossover-model assumptions.