Safety, Tolerability, Pharmacokinetic, and Efficacy Study of AZD5213 in Adolescents With Tourette's Disorder: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Safety, Tolerability, Pharmacokinetic, and Efficacy Study of AZD5213 in Adolescents With Tourette's Disorder (NCT01904773), a randomized phase 2 study. The primary outcome is Total Tic Severity Score (Part 2 Only) Crossover Analysis Over 6 Periods.

2. Clinical Question

The primary outcome is described as a crossover analysis over six periods, with lower scores indicating less tic severity on the Yale Global Tic Severity Scale. The registry record reports least-squares means and standard errors for the treatment groups, alongside two sponsor analyses. Those analyses do not use the same summary measure: the reported comparison for AZD5213 2.0 mg is a net mean difference, while the reported comparison for AZD5213 0.5 mg is a net median difference.

The question is therefore not answered by treating the group summaries as if they were the full crossover analysis. A crossover design uses repeated treatment periods within participants, and the registry does not give the details needed to reconstruct that analysis. The unadjusted recalculations offer a separate, approximate comparison of posted summaries; they do not replace the sponsor’s analysis.

3. Trial Design

This was a Phase 2, randomized, multicenter, triple-masked, two-part study of AZD5213 in adolescents with Tourette’s disorder. The registry record lists enrollment of 29 and a crossover model. The recorded study dates are 2013-08 for the start and 2015-02 for primary completion; the status is completed.

The study summary describes an eligibility screening period of up to 21 days. Part 1 assessed safety, tolerability, and pharmacokinetics during a 1-week period. Part 2 assessed safety, tolerability, and preliminary efficacy across six consecutive four-week crossover periods. The summary states that participants received AZD5213 and placebo, with two AZD5213 doses depending on tolerability in Part 1. A follow-up visit was planned for 14 (±) 7 days after the last dose.

The primary outcome is specifically labelled as applying to Part 2. Its analysis population is all Part 2 participants. The reported outcome groups contain 23 participants for AZD5213 0.5 mg, 19 for AZD5213 2.0 mg, and 21 for placebo. These group counts do not sum to the overall enrollment, and the registry record does not explain the reasons for the different counts or how participant overlap across crossover treatment groups was handled in each reported analysis.

Randomization and triple masking are recorded, but the method of sequence generation, the identity of the masked parties, allocation concealment, and the planned statistical analysis details are not reported in the registry record. The record describes the crossover schedule but does not provide the sequence assignments or period-specific outcome summaries. Those omissions limit independent reconstruction of the intended repeated-measures 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.

Before assignment: There were 29 screening visits for 28 subjects, 1 subject was initially screen failed for exclusion 32 but rescreened and qualified and entered Part 1 of the study(104-4003 rescreened and qualified as 104-4004). Eligible subjects participated in Part 1 and then continued into a randomized 6 period crossover Part 2 of the study

Only the first 6 of 9 groups are shown.

Screen

Reasons for not completing this period:

Part 1 - Single Dose- AZD5213 0.5 mg

Part 1 - Multiple Dose AZD5213 2.0 mg

Reasons for not completing this period:

First Intervention

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All Part 2 participants

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: All patients enrolled in Part 1 of the study, 23 of whom continued into Part 2

6. Primary Endpoint

The registry lists the primary outcome as Total Tic Severity Score (Part 2 Only) Crossover Analysis Over 6 Periods, measured over 3 week period of treatment.

The primary endpoint is the Total Tic Severity Score from Part 2 of the Yale Global Tic Severity Scale. The registry describes the score as ranging from 0 to 50 and states that lower scores are better. The outcome is reported for a three-week treatment period, with least-squares means and standard errors for each treatment group.

A tic-severity score provides a structured way to quantify the symptom domain targeted by this study. In general, a continuous score can retain gradations in severity rather than reducing each participant’s experience to a binary category. Comparing scores across treatments can therefore describe differences in the measured scale, while the score’s clinical interpretation still depends on the scale’s properties and the study context.

The registry identifies the endpoint as a crossover analysis over six periods, but it does not report period-specific scores, the exact model specification, or details of how repeated observations within a participant were used. The posted group summaries alone do not show the within-participant changes or the variability of those changes. Nor do they establish how missing period data, if any, were treated. As a result, the endpoint definition is clear at a broad level, but the posted information is not sufficient to reproduce the primary crossover analysis.

7. Results Posted in the Registry

• AZD5213 0.5 mg: mean 24.32 versus 22.99 on control; difference 1.33 (95% CI -4.27 to 6.93), standardized effect d = 0.14; z-test p = 0.642.
• AZD5213 2.0 mg: mean 25.41 versus 22.99 on control; difference 2.42 (95% CI -3.23 to 8.07), standardized effect d = 0.27; z-test p = 0.402.

8. Statistical Interpretation

9. Statistical Methods Explained

What does a mean difference describe?

A mean difference expresses the distance between two group means on the outcome’s original scale. For this outcome, the posted unadjusted recalculations compare each AZD5213 dose-group mean with the placebo mean. The resulting estimates are 1.33 for AZD5213 0.5 mg and 2.42 for AZD5213 2.0 mg. Since lower Total Tic Severity Scores are described as better, positive differences indicate that the AZD5213 group mean is higher than the placebo group mean in these unadjusted comparisons.

These are simple comparisons based on the posted summaries, not a reconstruction of the six-period crossover model. They do not account for within-participant correlations or other model features unless those are represented in the summaries, which the registry record does not establish.

What does an ANCOVA do?

Analysis of covariance, or ANCOVA, compares outcomes while adjusting for specified covariates. In general, adjustment can improve precision or account for baseline or design-related differences, depending on the model and data. The sponsor reports ANCOVA for both dose comparisons, but the registry record does not identify the covariates, model terms, or handling of the crossover periods. The sponsor’s estimates therefore cannot be independently reproduced from the information in the record.

The sponsor labels the 2.0 mg result a net mean difference and the 0.5 mg result a net median difference. A mean and a median summarize a distribution differently: the mean uses all observed values in its arithmetic average, while the median identifies the middle of the ordered values. Their differences need not agree. This distinction matters when comparing the sponsor’s 0.5 mg result with the recalculated mean difference.

Why does a crossover design matter?

In a crossover study, participants receive multiple treatments over different periods, so observations from the same participant are generally related. In general, analysis can use those within-participant comparisons to separate treatment effects from some between-participant variation. It must also consider design features such as period and treatment sequence where appropriate. The registry describes six crossover periods and receipt of AZD5213 and placebo, but it does not report the model terms or sequence-level results. The unadjusted comparisons based on treatment-group summaries should consequently be viewed as approximations, not as a substitute for a complete crossover analysis.

What is a standard error?

A standard error describes the estimated variability of a statistic, such as a mean, across repeated samples under a statistical model. It is not the same as the spread of individual participant scores. The registry reports standard errors of 2.01, 2.05, and 2.03 for the least-squares means in the 0.5 mg, 2.0 mg, and placebo groups, respectively. The recalculation uses those values to approximate group standard deviations; the record marks that approximation as such. In general, a standard error can be used to characterize uncertainty in an estimated mean, while a standard deviation describes variation among observations.

What does a confidence interval add?

A confidence interval gives a range of values compatible with the data and the calculation’s assumptions at a stated confidence level. It communicates uncertainty around an estimate, rather than merely indicating whether a test crosses a chosen threshold. The recalculated intervals are 95% confidence intervals and include zero for both unadjusted mean differences. Their breadth indicates that the approximate comparisons do not pin down the difference precisely. These intervals belong to the simple recalculations, not to the sponsor’s ANCOVA results.

Why is an unadjusted p-value not the whole result?

A p-value summarizes how unusual the observed data would be under a specified null model and its assumptions. It is not the probability that the null hypothesis is true, nor does it quantify the clinical importance of a difference. The unadjusted p-values of 0.642 and 0.402 accompany calculations that do not reproduce the sponsor’s crossover ANCOVA. The sponsor’s own reported p-values come from a different analysis and therefore address a different model-based comparison. Interpretation should consider the estimate, uncertainty, outcome scale, study design, and analysis method together.

In general, over-dispersion occurs when observed variability is greater than a model assumes. If present and unaccounted for, standard errors can be too small and inferential results too optimistic. The registry record does not report an over-dispersion assessment or enough model details to determine whether it was relevant to the sponsor’s analysis. This is a general statistical consideration, not a finding about this trial.

10. Confidence Intervals and Precision

The unadjusted 95% confidence interval for the AZD5213 0.5 mg mean difference is -4.269 to 6.929. For AZD5213 2.0 mg, it is -3.235 to 8.075. Each interval crosses zero, so each includes values consistent with a lower score on AZD5213 as well as values consistent with a higher score, under the assumptions of the recalculation. The intervals are fairly broad relative to their point estimates, which means the summary-based comparisons do not determine the size or direction of a difference precisely.

A confidence interval reflects both the estimated effect and its uncertainty under a particular analysis. In general, its width is influenced by variability, the amount of information, and the model used. These intervals rely on approximated standard deviations derived from the posted standard errors and treat the comparison as an unadjusted calculation. The registry’s crossover design involves repeated treatment periods, but the recalculation does not incorporate reported within-participant covariance or a reported crossover model. The intervals should not be interpreted as the sponsor’s model-based uncertainty estimates.

The sponsor’s registry analyses include no confidence limits, so their precision cannot be assessed from sponsor-reported intervals in this record. The sponsor reports p-values, but a p-value alone does not give a range of plausible effect sizes. The record also does not report a clinically important threshold for the score, limiting interpretation of whether any particular difference would be meaningful. In general, precision and clinical importance are distinct: a narrow interval can still include effects of little practical importance, while a broad interval may leave substantial uncertainty about the effect.

11. 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 planning estimates are participants per arm, not total events. They are illustrations based on the observed effect in the unadjusted comparisons, not the trial’s original sample-size design. The stated assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, a normal approximation, and an effect to detect set to the stated fraction of the observed effect. The trial’s own original design assumptions are not reported in the registry record.

The 80% power column means the illustrated number of participants per arm needed to have an 80% chance of detecting the specified effect under those assumptions, if that effect and model were correct. The 90% power column uses the same idea with a higher probability of detecting that effect. These are planning quantities, not a description of what happened in the completed study.

The rows consider planning for 100%, 75%, and 50% of the observed effect. For both dose comparisons, the required participant counts rise as the effect being targeted becomes smaller. This is a general feature of power calculations: smaller differences are harder to distinguish from random variation and typically require more information. The illustrations use an equal-allocation, two-arm framework and do not model the full crossover design or establish what a future crossover trial would require.

These estimates are not observed power. They are prospective planning illustrations calculated from the observed effects, and should not be used to explain the observed results as if they were a post-study probability that the trial succeeded.

12. 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 = 1.330; pooled standard deviation σ = 9.481.
2. n = 2 × (1.96 + 0.842)² × 9.481² / 1.330² = 797.6, rounded up to 798 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 <- 1.33                      # observed difference in means
sigma <- 9.48064268466467                     # 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(1904773)
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 = 798 (80% power), 1,068 (90% power)
 75% of observed effect: n per arm = 1,419 (80% power), 1,899 (90% power)
 50% of observed effect: n per arm = 3,191 (80% power), 4,272 (90% power)
power.t.test: 798.6 
Simulated power at n = 798 per arm: 0.796 (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.

13. Analyses Reported by the Sponsor

The registry reports a sponsor ANCOVA net mean difference of 2.42 for AZD5213 2.0 mg versus placebo and a net median difference of 1.34 for AZD5213 0.5 mg versus placebo. The corresponding sponsor p-values are 0.009 and 0.120. No confidence intervals are reported for these sponsor comparisons.

The unadjusted recalculation for 2.0 mg also gives an estimate of 2.42, but its p-value is 0.402 and its confidence interval includes zero. For 0.5 mg, the recalculation is a mean difference of 1.33 with an unadjusted p-value of 0.642; the sponsor’s result is a median difference, so the two estimates summarize the data differently. The difference in inferential results for the 2.0 mg comparison demonstrates that matching point estimates do not make the analyses equivalent.

The record flags a discrepancy between the sponsor results and the recalculations. The sponsor’s own analysis takes precedence for describing the sponsor-reported findings. The recalculations are useful only as simple checks based on posted summaries; they do not reproduce the sponsor’s ANCOVA or the repeated-measures crossover analysis. The model details needed to explain the discrepancy fully are not reported in the registry record.

14. Multiplicity, Missing Data and Analysis Populations

The registry record reports two sponsor analyses for the primary outcome, one for each AZD5213 dose against placebo, and lists four outcome measures overall. It reports no secondary outcome measures as posted. The record does not state whether the dose comparisons were treated as a family for multiplicity adjustment, whether any adjustment was planned, or how multiple treatment periods or repeated assessments were handled in the inferential plan. Multiplicity methods and decisions are not reported in the registry record.

In general, when several hypotheses are tested, the chance of at least one apparently unusual result can increase if each test is interpreted separately without accounting for the number of comparisons. This is a general statistical issue, not evidence that a particular adjustment was or was not used in this trial. Without the prespecified analysis plan or further model details, the sponsor p-values cannot be interpreted as adjusted or unadjusted for multiplicity beyond what the registry explicitly reports.

The primary outcome description names all Part 2 participants as the analysis population, but the registry record does not give a fuller definition of the analysis set, such as eligibility rules for inclusion in each comparison. It does not report the extent or pattern of missing outcome data, reasons for missingness, imputation methods, or sensitivity analyses. Nor does it state whether participants with incomplete crossover sequences were included and how their available periods were handled.

For a crossover study, the treatment-period structure and within-participant data can affect the analysis. The registry record does not provide the sequence allocation, period-specific summaries, carryover assessment, or details of how period effects were addressed. These are important uncertainties when interpreting the reported analyses, but the absence of those details in the registry record does not establish that the issues were ignored in the study analysis.

15. 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 1% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: testicular torsion (part 2) (4.8%), convulsion (4.2%).

The registry’s overall-study safety group lists 24 participants at risk, with 2 participants affected by serious adverse events. In Part 1, the serious-event counts were 0 for AZD5213 0.5 mg, 1 for AZD5213 2.0 mg, and 0 for placebo, each with 24 participants at risk. In Part 2, the counts were 0 for AZD5213 0.5 mg, 0 for AZD5213 2.0 mg, and 1 for placebo; the respective groups at risk were 23, 19, and 21.

These are group-level counts across the listed safety categories, not a complete description of event timing, severity, or causal attribution. The record also lists Testicular torsion (Part 2) at 4.8% and Convulsion at 4.2% among the top serious event terms. The registry does not provide further detail about those terms in the information.

Deaths are not reported in the registry record. Safety counts should be interpreted in light of the different group sizes and the crossover design; the Part 2 treatment groups are not necessarily independent sets of participants. These figures describe reported adverse events and do not establish that a treatment caused an event.

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

The results record outcome summaries and sponsor analyses from a randomized crossover study of AZD5213 and placebo in adolescents with Tourette’s disorder. The posted least-squares mean scores are 24.32 for AZD5213 0.5 mg, 25.41 for AZD5213 2.0 mg, and 22.99 for placebo. Because lower scores are described as better, the group means alone do not show a lower score for either AZD5213 dose than for placebo.

The sponsor reports an ANCOVA result for each dose, but the 0.5 mg comparison is a net median difference while the 2.0 mg comparison is a net mean difference. The registry record flags a discrepancy between sponsor analyses and simple recalculations. The sponsor’s analysis takes precedence when stating the sponsor-reported findings; the recalculations are approximate, unadjusted comparisons and do not reproduce the crossover model. Sponsor confidence intervals are not reported, and the analysis details needed to independently verify the model are unavailable in the registry record.

These findings do not, on their own, establish a causal effect, an effect’s clinical importance, or a conclusion for adolescents beyond the population studied. The registry provides no clinical-importance threshold for the scale and does not report the details needed to assess missing data, multiplicity, or the full analysis population. The sample-size illustrations are planning calculations based on observed effects, not evidence about the probability that this trial detected an effect.

The safety counts describe reported serious adverse events in the listed groups. They do not establish that AZD5213 or placebo caused those events, and the record does not provide enough detail to make broader safety conclusions. The appropriate interpretation is limited to the recorded outcome summaries, the sponsor’s stated analyses, and the gaps in reporting.

17. Why This Trial Matters Statistically

This trial is statistically notable because its crossover design and its posted summaries invite different kinds of comparison. The primary outcome is described as an analysis over six periods, in which participants received AZD5213 and placebo. In general, crossover designs can make within-participant comparisons informative, but that depends on an analysis that accounts appropriately for repeated observations and the treatment-period structure. The registry does not provide enough model detail to reconstruct that analysis.

The contrast between sponsor analyses and unadjusted recalculations illustrates why a point estimate is not the whole result. For AZD5213 2.0 mg, both display an estimate of 2.42, yet the sponsor’s reported ANCOVA p-value differs from the recalculated unadjusted p-value, and only the recalculated comparison has a posted confidence interval. For AZD5213 0.5 mg, the sponsor reports a median difference whereas the recalculation uses a mean difference. Statistical summaries can look similar numerically while answering different model-based questions.

The record also shows the limits of planning calculations derived from observed effects. The required participant counts increase when planning for smaller fractions of those effects, but the illustrations use a simplified equal-allocation, two-arm framework rather than the reported crossover schedule. They can explain how assumed effect size affects sample-size planning, but they do not reveal the original design rationale.

Finally, the outcome and safety results are reported with important information gaps: sponsor confidence intervals, model specification, missing-data handling, and multiplicity procedures are not reported in the registry record. Recognizing those gaps is part of statistical interpretation, not a reason to substitute an unreproduced analysis for the sponsor’s reported result.

18. 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 does not report the sponsor ANCOVA model specification, including its covariates and treatment of period or sequence effects. It also lacks period-specific outcome summaries, sequence assignments, within-participant correlations, and sponsor confidence intervals. These omissions prevent independent reconstruction of the primary crossover analysis and limit direct assessment of the precision of the sponsor estimates.

The unadjusted recalculations use approximate standard deviations derived from the posted standard errors. They are not fitted crossover analyses and do not incorporate the within-participant structure described for Part 2. Their confidence intervals and p-values therefore answer a simpler comparison than the sponsor’s reported ANCOVA. This distinction is particularly important because the registry flags a discrepancy between the results.

The sponsor reports different summary measures across dose comparisons: a net mean difference for AZD5213 2.0 mg and a net median difference for AZD5213 0.5 mg. The registry record does not explain that difference in reporting. It also does not provide confidence intervals for either sponsor analysis, a clinically important threshold for the tic-severity score, or a detailed account of the analysis populations.

Missing outcome data, reasons for missingness, imputation, sensitivity analyses, multiplicity procedures, and the planned sample-size rationale are not reported in the registry record. The study enrolled 29 participants, while the primary outcome groups list 23, 19, and 21 participants; the record does not explain the differing group counts or participant overlap across crossover treatments. These limitations restrict interpretation of precision, generalizability, and the robustness of the reported results.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Safety, Tolerability, Pharmacokinetic, and Efficacy Study of AZD5213 in Adolescents With Tourette's Disorder — NCT01904773.
• Registry record last updated 2016-09-23; status last verified 2016-09.
• Results first posted 2016-07-25.
• 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.

22. Record Summary

NCT01904773 is a completed Phase 2 randomized, triple-masked crossover study of AZD5213 in adolescents with Tourette’s disorder. The registry lists enrollment of 29 and describes a two-part study, with Part 1 assessing safety, tolerability, and pharmacokinetics and Part 2 assessing preliminary efficacy across six crossover periods.

The primary Part 2 outcome is Total Tic Severity Score on the Yale Global Tic Severity Scale, with lower scores described as better. Posted least-squares means and standard errors are listed for AZD5213 0.5 mg, AZD5213 2.0 mg, and placebo. The sponsor reports ANCOVA comparisons; simple unadjusted recalculations give different inferential results and are not a substitute for the crossover analysis. The record does not report sponsor confidence intervals or enough model detail to reproduce that analysis.

Safety results include serious adverse event counts across the listed study groups. Several analysis and reporting details, including missing-data handling, multiplicity procedures, and the original sample-size rationale, are not reported in the registry record.