Brexpiprazole as an Adjunctive Treatment to Paroxetine or Sertraline in Adult Patients Suffering From Post-traumatic Stress Disorder (PTSD): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Brexpiprazole as an Adjunctive Treatment to Paroxetine or Sertraline in Adult Patients Suffering From Post-traumatic Stress Disorder (PTSD) (NCT01987960), a randomized phase 3 study. The primary outcome is PTSD Symptoms Using CAPS-2 Total Score.

2. Clinical Question

3. Trial Design

This phase 3 study was randomized, double-blind, parallel-group, and placebo-controlled. It evaluated brexpiprazole or placebo as an adjunct to open-label paroxetine or sertraline in adults suffering from PTSD. The listed enrollment is 417. The registry describes brexpiprazole dosing as 1mg/day for one week, followed by 2mg/day for 3 weeks, after which dosing was flexible from 1 to 3 mg/day. The placebo arm also received open-label PAR/SER.

The registry lists the study as terminated. It states that few enrolled patients were eligible for randomization and that the sponsor terminated the study early. As a result, the posted efficacy data are descriptive, and the primary and key secondary efficacy analyses were not done. The registry does not report the reason for termination beyond this description, nor does it report the trial's original sample-size justification or planned statistical analysis details.

The primary outcome concerns CAPS-2 total score from baseline to Week 12 of the randomized period. The displayed summaries include baseline and Week 12 means and standard deviations, with different numbers of participants contributing at each assessment. For the primary outcome, the baseline summaries include 17 participants in the placebo plus PAR/SER group and 23 in the brexpiprazole plus PAR/SER group. The Week 12 summaries include 15 and 17 participants, respectively. These counts describe the posted outcome observations; they should not be confused with total enrollment.

One secondary outcome is listed: Global Clinical Impression Severity of Illness (CGI-S) Score. The registry record lists two outcome measures and zero reported analyses. No allocation ratio, detailed recruitment plan, duration of follow-up beyond the stated outcome timeframe, or original protocol sample-size assumptions are reported in the registry record.

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.

Period 1

Reasons for not completing this period:

Period 2

Reasons for not completing this period:

Period 3

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Due to the low number of enrolled patients eligible for randomization and the sponsor's early termination of the study, the data presented are descriptive i.e. the primary and key secondary efficacy analyses were not done

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: The all-patients-treated set (APTS) comprises all randomized patients who took at least one dose of brexpiprazole or placebo.

6. Primary Endpoint

The registry lists the primary outcome as PTSD Symptoms Using CAPS-2 Total Score, measured over Period 2: Baseline to Week 12 (of randomized period).

The primary endpoint is PTSD symptoms measured by the Clinician-Administered PTSD Scale Part 2 (CAPS-2) total score, with the stated timeframe from baseline to Week 12 of the randomized period. The registry describes 17 items in criteria B, C, and D. Each item is rated for frequency and intensity on scales from 0 to 4. The sum has a stated range from 0 to 136, with higher scores indicating greater symptom severity.

A total score combines responses across the listed symptom items into one continuous measure. In general, a continuous outcome can preserve gradations in a symptom measure rather than reducing scores to categories; this can help describe the extent and variability of symptoms. The registry reports means and standard deviations at baseline and Week 12. Those summaries describe the scores among participants with posted observations at each time, but they do not by themselves establish a treatment effect.

The registry specifically states that the primary and key secondary efficacy analyses were not done because few enrolled patients were eligible for randomization and the sponsor terminated the study early. Thus, the endpoint's intended role in a formal efficacy comparison should be distinguished from the descriptive scores that were posted. The registry does not report a prespecified threshold for a clinically important CAPS-2 change or a formal analysis of change from baseline.

7. Results Posted in the Registry

• Period 2 Absolute Mean at Baseline; Brexpiprazole and PAR/SER: mean 83.43 versus 82.82 on control; difference 0.61 (95% CI -7.55 to 8.77), standardized effect d = 0.05; z-test p = 0.884.
• Period 2 Absolute Mean at Week 12; Placebo and PAR/SER: mean 69.67 versus 82.82 on control; difference -13.15 (95% CI -25.11 to -1.19), standardized effect d = -0.78; z-test p = 0.031.
• Period 2 Absolute Mean at Week 12; Brexpiprazole and PAR/SER: mean 69.18 versus 82.82 on control; difference -13.64 (95% CI -24.24 to -3.04), standardized effect d = -0.86; z-test p = 0.012.

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.

Global Clinical Impression Severity of Illness (CGI-S) Score

• 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 deviations.
• Result:
  • Period 2 Absolute Mean at Baseline; Brexpiprazole and PAR/SER: mean 4.54 versus 4.24 on control; difference 0.30 (95% CI -0.19 to 0.79), standardized effect d = 0.39; z-test p = 0.235.
  • Period 2 Absolute Mean at Week 12; Placebo and PAR/SER: mean 3.73 versus 4.24 on control; difference -0.51 (95% CI -1.19 to 0.17), standardized effect d = -0.53; z-test p = 0.143.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Period 2 Absolute Mean at Baseline; Brexpiprazole and PAR/SER and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Period 2 Absolute Mean at Week 12; Placebo and PAR/SER and the control. The interval gives the range of effects that remain compatible with the data.

The registry lists one secondary outcome, CGI-S Score, measured over the same baseline-to-Week 12 timeframe as the primary endpoint. Its posted means move in the same broad descriptive direction as the CAPS-2 summaries: the Week 12 scores are lower than the corresponding baseline scores in both treatment groups. That pattern is not a formal estimate of the randomized treatment contrast and cannot show whether the score changes differ between groups.

The recalculated comparisons for the secondary outcome are unadjusted. The baseline contrast is 0.3, with a 95% confidence interval from -0.195 to 0.795 and p = 0.235. The Week 12 placebo comparison against the placebo baseline mean is -0.51, with a 95% confidence interval from -1.192 to 0.172 and p = 0.143. These are descriptive recalculations, not sponsor-reported efficacy analyses. The registry does not report the primary and key secondary efficacy analyses.

The secondary outcome is relevant as a separate measure of illness severity, but it does not replace the primary CAPS-2 endpoint. Agreement in broad score patterns across measures may be descriptively informative, yet both outcomes are affected by the same early termination, small numbers, and unreported analysis procedures. Since no multiplicity plan is reported, unadjusted secondary results carry less weight than prespecified, appropriately controlled analyses would. They should be treated as exploratory descriptions rather than independent confirmation of a treatment effect.

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

• Global Clinical Impression Severity of Illness (CGI-S) Score

9. Statistical Interpretation

10. Statistical Methods Explained

What do the means describe?

A mean is the arithmetic average of the observed scores in a group. The reported CAPS-2 means describe the participants with scores at baseline or Week 12. A standard deviation describes how spread out individual scores are around the group mean. It is not the same as a confidence interval: the standard deviation concerns variation among participants, whereas a confidence interval expresses uncertainty around an estimated quantity.

The posted means at baseline and Week 12 are absolute scores at each assessment. They are not reported as individual changes, nor as a model-based comparison of change between randomized groups. The number of observations differs between baseline and Week 12, so the summaries need not represent precisely the same set of participants at both times. Without the analysis population definition and missing-data procedures, the relation between the assessment samples cannot be determined from the registry record.

What is a mean difference?

A mean difference expresses one group mean relative to a reference mean, using the ordering stated for the comparison. The baseline comparison contrasts the brexpiprazole group's baseline mean with the placebo group's baseline mean. The Week 12 recalculations contrast each Week 12 group mean with the placebo group's baseline mean. They are not a direct Week 12 brexpiprazole-versus-placebo contrast and do not represent a difference-in-change estimate.

These values are labeled recalculated because they are simple unadjusted re-calculations from the posted summary statistics. The registry states that the sponsor's primary and key secondary efficacy analyses were not done. The recalculated results should be read as descriptive statistical comparisons, not as a substitute for a planned efficacy analysis. No discrepancy between a sponsor analysis and the recalculations is indicated in the registry record.

How should a confidence interval be read?

A confidence interval gives a range of estimates compatible with the data and the calculation method, reflecting sampling uncertainty. A wide interval indicates limited precision: the data are compatible with a broad range of values. An interval that includes zero for a mean difference does not provide clear evidence of a nonzero difference under that calculation, while an interval excluding zero is not by itself proof of a treatment effect. The comparison, design, analysis plan, and multiplicity context still matter.

The intervals posted for the recalculations arise from unadjusted comparisons based on summary statistics. They do not account for an undisclosed model or covariate adjustment. They also do not remedy the fact that the comparisons use baseline as the reference for Week 12 outcomes rather than a treatment-versus-control comparison at Week 12.

What does an unadjusted p-value express?

A p-value describes how surprising data at least as incompatible with a specified null hypothesis would be if that null hypothesis and the calculation assumptions were true. It is not the probability that the null hypothesis is true, nor a measure of clinical importance. Unadjusted p-values do not account for testing multiple outcomes or comparisons. Since the registry does not report the primary efficacy analysis and does not report a multiplicity plan, the recalculated p-values cannot establish confirmatory evidence.

General statistical principle: when many outcomes or comparisons are examined, the chance of obtaining a small p-value by chance across the set can increase. Planned testing procedures can address that issue, but the registry record does not report such procedures. The number of analyses reported is zero, and the key efficacy analyses were not done.

Why do small samples matter?

With relatively few observations, estimates can be unstable and confidence intervals can be broad. A small sample also makes it harder to distinguish a true group difference from random variation. The posted primary outcome counts at Week 12 are 15 and 17, and the baseline counts are 17 and 23. These outcome counts are much more relevant to the precision of the displayed means than the total enrollment figure alone. The record notes low eligibility for randomization, which helps explain why the efficacy data are descriptive, but it does not report detailed reasons for individual participants' missing observations.

General statistical principle: a randomized design can reduce systematic differences between groups on average, but randomization cannot guarantee balance in every small sample. A baseline comparison can describe observed balance; it is not an efficacy test and cannot substitute for a properly specified outcome analysis. Similarly, a within-group change or comparison against a baseline value does not isolate the effect of assignment because time-related changes and other influences are not separated from treatment.

What is not documented?

The registry record does not report the original sample-size calculation, allocation ratio, statistical model, handling of missing outcome data, analysis population definition, planned covariates, interim analyses, or a multiplicity adjustment. It also does not report the sponsor's primary or key secondary efficacy estimates because those analyses were not done. These omissions limit assessment of whether the posted descriptive summaries align with a prespecified plan and how they were intended to support formal inference.

11. Confidence Intervals and Precision

The recalculated CAPS-2 baseline mean difference is 0.61, with a 95% confidence interval from -7.55 to 8.77. The interval includes zero and extends across values on either side of it. This indicates that the unadjusted baseline contrast is imprecise and does not establish a baseline difference.

For the Week 12 placebo plus PAR/SER comparison against the placebo baseline mean, the recalculated mean difference is -13.15, with a 95% confidence interval from -25.11 to -1.194. For the Week 12 brexpiprazole plus PAR/SER comparison against the placebo baseline mean, it is -13.64, with a 95% confidence interval from -24.24 to -3.036. These intervals quantify uncertainty for the stated recalculated contrasts, not for a direct treatment-versus-placebo comparison at Week 12 or a difference in change.

The Week 12 groups have 15 and 17 observations, and the baseline groups have 17 and 23. With these small counts, estimates can be sensitive to the observed participants, and the intervals should be read in light of the descriptive status of the results. The registry does not report formal efficacy analyses, a model-based treatment contrast, or an analysis accounting for missing outcome data. Consequently, the recalculated intervals should not be treated as precision estimates for the trial's intended primary treatment effect. The registry does not report confidence intervals for the individual group means.

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 planning figures are participants per arm, not total events. They are planning illustrations based on the observed effect, not the trial's original design assumptions. The illustration uses a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. For each displayed fraction of the observed effect, the columns describe the participants per arm for 80% power and participants per arm for 90% power.

In plain terms, 80% power means that, under the stated assumptions and if the effect being targeted were the assumed effect, a study of that planned size would have an 80% chance of detecting it at the specified significance level. The 90% column uses a higher target chance of detecting the assumed effect and therefore generally calls for more participants. These are planning concepts, not probabilities that a treatment works or that a particular observed result is true.

The rows that plan for 75% or 50% of the observed effect use smaller target effects than the row based on 100% of the observed effect. Smaller effects are harder to distinguish from random variation and therefore require more participants per arm for the same power target. The illustrations are derived from the observed comparisons, which are themselves unstable because the study ended early and the outcome counts are small.

These figures are not observed power. They do not describe the study's achieved probability of detecting an effect, and they should not be mistaken for the original trial sample-size justification. The original design assumptions are not reported in the registry record.

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 = 0.610; pooled standard deviation σ = 13.045.
2. n = 2 × (1.96 + 0.842)² × 13.045² / 0.610² = 7,178.5, rounded up to 7,179 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 <- 0.610000000000014                      # observed difference in means
sigma <- 13.0445132648094                     # 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(1987960)
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 = 7,179 (80% power), 9,610 (90% power)
 75% of observed effect: n per arm = 12,762 (80% power), 17,085 (90% power)
 50% of observed effect: n per arm = 28,715 (80% power), 38,440 (90% power)
power.t.test: 7179.5 
Simulated power at n = 7,179 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. Multiplicity, Missing Data and Analysis Populations

The registry record lists two outcome measures: the primary CAPS-2 total score and the secondary CGI-S Score. It reports zero analyses and states that the primary and key secondary efficacy analyses were not done. No multiplicity strategy is reported. This matters because interpretation of multiple outcome assessments or comparisons depends on whether the analysis plan controlled the overall risk of chance-positive findings. The displayed unadjusted recalculations do not indicate any such adjustment.

The record also does not describe the analysis populations, such as whether the intended primary analysis would include all randomized participants or a defined subset. It does not report the statistical model, covariate adjustment, treatment of measurements after discontinuation, or procedures for missing outcome data. The differences between baseline and Week 12 observation counts make missingness especially relevant, but the reasons and timing of missing observations are not reported in the registry record.

General statistical principle: missing data can affect an estimate if the likelihood of having an observed measurement is related to participant characteristics or outcomes. Complete-case summaries describe those with observed scores; without a specified missing-data approach and assumptions, they cannot establish how the results would differ for all randomized participants. Small samples can make these concerns more consequential because each observation contributes a larger share of the available information.

The registry does not report interim analyses or stopping criteria. It does state that the sponsor terminated the study early because of the low number of enrolled patients eligible for randomization, and that the data are descriptive. No protocol details beyond those statements should be inferred. Together, the absence of formal analyses, analysis-population definitions, missing-data methods, and multiplicity procedures prevents a complete audit of the intended inferential plan.

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 5% reporting threshold.

Most frequent non-serious adverse event terms

For the randomized period, the registry reports zero participants with serious adverse events in both groups: 0 of 23 for brexpiprazole plus PAR/SER and 0 of 17 for placebo plus PAR/SER. These are the counts recorded as affected among the participants at risk in each group. The registry lists no top serious adverse event terms, so it does not provide term-level detail in this record.

The record also lists 3 participants affected by other adverse events among 23 at risk in the brexpiprazole plus PAR/SER group, and 8 among 17 at risk in the placebo plus PAR/SER group. These counts are distinct from the serious adverse event counts and should not be combined with them. The registry does not report deaths for either group; the death values are not reported in the registry record.

These safety counts are descriptive and involve small groups. They do not establish that serious adverse events could not occur, nor do they support a comparative safety conclusion. The registry record does not give additional event terms, exposure-adjusted rates, or a detailed safety analysis.

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

The posted data describe CAPS-2 and CGI-S scores observed in a small number of participants at baseline and Week 12. Both groups have lower mean scores at Week 12 than at baseline, but the registry says that the primary and key secondary efficacy analyses were not done. The recalculated Week 12 comparisons use the placebo group's baseline mean as their reference; they are not direct comparisons between the randomized groups at Week 12 and do not estimate a difference in change.

Accordingly, these descriptive data do not establish whether adding brexpiprazole changed PTSD symptoms relative to placebo added to PAR/SER. Nor do they establish that either adjunctive treatment caused the observed score patterns. The recalculated confidence intervals and p-values quantify uncertainty for simple unadjusted summary-statistic comparisons, not for the formal treatment effect the study set out to evaluate.

The record's termination and low randomization eligibility are central to interpretation. Only a small number of participants contribute to the Week 12 outcome summaries, and the registry does not report how missing observations were handled, which analysis population was intended, or the planned model. These limitations restrict both precision and the ability to distinguish treatment effects from chance, time-related changes, or differences in who remained observed.

The registry does not provide the original sample-size justification, a multiplicity plan, or formal efficacy results. The planning illustrations based on observed effects are not evidence that the trial had a particular probability of finding a treatment effect. Likewise, the descriptive adverse-event counts cannot establish comparative safety. The results should therefore be read as a limited account of observed scores and events, not as a definitive evaluation of efficacy or safety.

17. Why This Trial Matters Statistically

This record illustrates an important distinction between a trial's intended question and the evidence available after early termination. Randomization and blinding describe the design, but they do not ensure that the study will produce a precise or interpretable efficacy estimate when few participants are eligible for randomization and formal analyses are not completed. The registry explicitly characterizes the efficacy data as descriptive.

The example also shows why the comparator used in a calculation matters. A Week 12 score compared with a baseline score can describe the observed difference from that reference, but it is not the same as comparing randomized groups at Week 12 or comparing their changes from baseline. Treating these contrasts as interchangeable would misstate the statistical question.

Small outcome counts and missing analysis details matter alongside numerical estimates. Confidence intervals and p-values can look formally precise while answering a limited, unadjusted question. In this record, the recalculations do not replace the unperformed primary analysis, and the secondary outcome does not independently resolve the uncertainty. The sample-size illustrations likewise describe planning under assumptions based on observed effects, rather than the study's original design or achieved power.

For clinical researchers, this trial highlights the value of distinguishing enrollment from outcome-specific observations, prespecified analyses from descriptive summaries, and within-group patterns from randomized treatment contrasts. Those distinctions help keep conclusions aligned with what the record actually documents.

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 study is listed as terminated, and the registry states that few enrolled patients were eligible for randomization. The displayed outcome counts are small and vary between baseline and Week 12. The registry says that the primary and key secondary efficacy analyses were not done, limiting direct assessment of the study's central efficacy question.

The posted summaries do not provide a formal randomized-group comparison of change in CAPS-2 score. The recalculated Week 12 contrasts use the placebo baseline mean as their reference. They are unadjusted and should not be interpreted as the treatment effect. No sponsor efficacy estimates are reported, and the record lists zero analyses.

The registry does not report the original sample-size justification, statistical model, intended analysis populations, handling of missing data, multiplicity procedures, or interim analysis details. The reasons for missing Week 12 outcomes and the relationship between baseline and Week 12 observation sets are also not reported in the registry record. These gaps prevent a full evaluation of bias and precision.

The record describes flexible dosing for brexpiprazole after the initial dosing period, and open-label PAR/SER background treatment. It does not report the distribution of paroxetine versus sertraline use by group or details that would permit assessment of subgroup patterns. The posted adverse-event summaries are limited, with no serious-event terms listed and deaths not reported in the registry record. These limitations apply to interpretation of the registry record and do not establish what a completed trial might have shown.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Brexpiprazole as an Adjunctive Treatment to Paroxetine or Sertraline in Adult Patients Suffering From Post-traumatic Stress Disorder (PTSD) — NCT01987960.
• Registry record last updated 2017-03-13; status last verified 2017-01.
• Results first posted 2017-03-13.
• 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

NCT01987960 is a phase 3 randomized, double-blind, parallel-group, placebo-controlled study of brexpiprazole added to paroxetine or sertraline in adults with PTSD. The registry lists enrollment of 417 and status as terminated. Brexpiprazole dosing began at 1mg/day for one week, then 2mg/day for 3 weeks, followed by flexible dosing from 1 to 3 mg/day.

The primary outcome was CAPS-2 total score from baseline to Week 12 of the randomized period. CGI-S Score was listed as a secondary outcome. The registry states that low randomization eligibility and early termination meant the data are descriptive and that primary and key secondary efficacy analyses were not done. Outcome counts at Week 12 were 15 for placebo plus PAR/SER and 17 for brexpiprazole plus PAR/SER. The posted figures do not establish a formal treatment effect.