Study of the Effectiveness of Three Different Doses of OPC-34712 in the Treatment of Adults With Acute Schizophrenia: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Study of the Effectiveness of Three Different Doses of OPC-34712 in the Treatment of Adults With Acute Schizophrenia (NCT01396421), a randomized phase 3 study. The primary outcome is Mean Change From Baseline to Week 6 Positive and Negative Syndrome Scale (PANSS) Total Score..

2. Clinical Question

3. Trial Design

This was a Phase 3, multicenter, randomized, double-blind, placebo-controlled, parallel-group trial of three fixed doses of OPC-34712 in adults with acute schizophrenia. The registry identifies the active drug as brexpiprazole and lists once-daily tablets at 4mg, 2 mg and 0.25 mg, each administered for six weeks. The comparator group received placebo once daily for the same duration. The record describes masking as quadruple but does not identify which roles were masked.

Enrollment was 636, and the trial status is completed. The registry lists the start as 2011-07 and primary completion as 2014-01. It states that the study aimed to compare effectiveness, safety and tolerability across the three doses and placebo. The original sample-size justification, planned allocation ratio, and detailed operational definitions for masking are not reported in the registry record.

The primary efficacy population included participants who received at least one dose and had baseline and at least one post-baseline efficacy evaluation. For the Week 6 PANSS total score, data were reported for 134 participants in the 4mg group, 132 in the 2 mg group, 62 in the 0.25 mg group and 124 in the placebo group. These endpoint counts differ from total enrollment because they describe participants with Week 6 data in the stated efficacy population, not everyone enrolled.

The registry lists 16 outcome measures and 74 analyses. The primary endpoint was mean change in PANSS total score from baseline to Week 6. Secondary measures included other symptom scales, personal and social performance, CGI-I and response rate. The record does not report a complete statistical analysis plan or explain how the multiple outcomes and analyses were prioritized.

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: Adults with schizophrenia as defined by Diagnostic and Statistical Manual of Mental Health Disorders 4th Edition Text Revision criteria and confirmed by the Mini International Neuropsychiatric Interview for schizophrenia and psychotic disorders studies.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Efficacy: consists of all participants who received at least one dose of study medication and have baseline and at least one post-baseline efficacy evaluation. At Week 6, 134, 132, 62 and 124 participants in the Brex 4mg, Brex 2mg, Brex 0.25 mg and placebo group, respectively had data.

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 Mean Change From Baseline to Week 6 Positive and Negative Syndrome Scale (PANSS) Total Score., measured over Baseline to Week 6.

The primary endpoint was mean change from baseline to Week 6 in the Positive and Negative Syndrome Scale (PANSS) total score. The registry describes PANSS as containing 30 symptom constructs across positive, negative and general psychopathology subscales. Each construct is rated on a 7-point scale, from 1 for absence of symptoms to 7 for extremely severe symptoms. The total score ranges from 30, the best possible outcome, to 210, the worst possible outcome.

Change scores describe how participants’ measured symptom scores shifted over the trial period. A negative mean change indicates a lower Week 6 score than baseline; under the registry’s scale description, lower PANSS total scores correspond to less severe symptoms. Comparing mean changes between a dose group and placebo asks whether the average shift differed between those groups. It does not show how every participant changed, and a group average alone does not describe the full distribution of individual outcomes.

The reported group values are least-squares means with standard errors. A least-squares mean is a model-based estimate of a group average, rather than necessarily the raw arithmetic average of observed values. The registry identifies the outcome timeframe as baseline to Week 6 and reports the efficacy population used for the primary outcome. The precise model specification, covariates and missing-data handling are not reported in the registry record.

PANSS is used as a structured measure of schizophrenia symptom severity, allowing changes to be compared across randomized groups. Its total score captures several symptom domains in one outcome, but it does not by itself represent functioning, tolerability or all aspects of the study’s broader effectiveness and safety aims.

7. Results Posted in the Registry

• Brexpiprazole 4mg: mean -19.65 versus -12.01 on control; difference -7.64 (95% CI -11.99 to -3.29), standardized effect d = -0.43; z-test p = <0.001.
• Brexpiprazole 2 mg: mean -20.73 versus -12.01 on control; difference -8.72 (95% CI -13.09 to -4.35), standardized effect d = -0.49; z-test p = <0.001.
• Brexpiprazole 0.25 mg: mean -14.90 versus -12.01 on control; difference -2.89 (95% CI -8.27 to 2.49), standardized effect d = -0.16; z-test p = 0.292.

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.

Mean Change From Baseline to Week 6 in Clinical Global Impression - Severity of Illness Scale (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 errors.
• Result:
  • Brexpiprazole 4 mg: mean -1.20 versus -0.82 on control; difference -0.38 (95% CI -0.62 to -0.14), standardized effect d = -0.42; z-test p = 0.002.
  • Brexpiprazole 2 mg: mean -1.15 versus -0.82 on control; difference -0.33 (95% CI -0.57 to -0.09), standardized effect d = -0.36; z-test p = 0.006.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Brexpiprazole 4 mg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Brexpiprazole 2 mg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Mean Change From Baseline to Week 6 in Personal and Social Performance Scale (PSP)

• 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:
  • Brexpiprazole 4mg: mean 12.72 versus 10.26 on control; difference 2.46 (95% CI -0.19 to 5.11), standardized effect d = 0.24; z-test p = 0.069.
  • Brexpiprazole 2mg: mean 13.15 versus 10.26 on control; difference 2.89 (95% CI 0.24 to 5.54), standardized effect d = 0.28; z-test p = 0.032.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Brexpiprazole 4mg and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Brexpiprazole 2mg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Mean Change From Baseline to Week 6 in PANSS Positive Subscale 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 errors.
• Result:
  • Brexpiprazole 4mg: mean -6.78 versus -4.35 on control; difference -2.43 (95% CI -3.89 to -0.97), standardized effect d = -0.43; z-test p = 0.001.
  • Brexpiprazole 2mg: mean -6.57 versus -4.35 on control; difference -2.22 (95% CI -3.69 to -0.75), standardized effect d = -0.39; z-test p = 0.003.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Brexpiprazole 4mg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Brexpiprazole 2mg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The registry lists 15 secondary outcomes in addition to the primary outcome, covering other PANSS assessments, CGI measures, personal and social performance, and response rate. The posted numerical secondary results include Week 6 changes in CGI-S, PSP and the PANSS positive subscale, with some sponsor-reported comparisons. Taken together, the available results show that the secondary outcomes examine related but distinct dimensions rather than simply repeating the primary PANSS total-score measure.

The general pattern in the posted Week 6 results is that some comparisons for the 4mg and 2 mg groups differ from placebo, while other comparisons are less conclusive. This is broadly compatible with the primary endpoint’s pattern for the higher doses, but the secondary outcomes do not independently settle the primary question. Different scales have different meanings, and a change in one measure cannot be assumed to imply the same change in another. The registry does not provide numerical results for every secondary outcome named in the outcome list, so the full pattern across all listed measures cannot be assessed from the record.

Secondary findings carry less weight when their status in the testing plan and any multiplicity adjustment are unclear. Testing many outcomes can increase the chance of finding a nominally notable result even when no true difference exists for some measures. The record reports 74 analyses but does not report a multiplicity strategy, hierarchy of outcomes, or whether secondary analyses were confirmatory. Unadjusted secondary p-values should therefore be read as exploratory evidence, interpreted alongside effect estimates and confidence intervals rather than as a collection of definitive conclusions.

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

• Mean Change From Baseline to Week 6 in Clinical Global Impression - Severity of Illness Scale (CGI-S) Score.
• Mean Change From Baseline to Week 1, 2, 3, 4 and 5 Positive and Negative Syndrome Scale (PANSS) Total Score.
• Mean Change From Baseline to Week 1, 2, 3, 4 and 5 in Clinical Global Impression - Severity of Illness Scale (CGI-S) Score.
• Mean Change From Baseline to Week 6 in Personal and Social Performance Scale (PSP)
• Mean Change From Baseline to Week 6 in PANSS Positive Subscale Score
• Mean Change From Baseline to Week 6 in PANSS Negative Subscale Score
• Clinical Global Impression- Improvement Scale (CGI-I) Score at Week 6
• Response Rate at Week 6

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is continuous: it is a score that can take many values across a scale. The registry reports least-squares mean changes and standard errors for each group. In general, a least-squares mean is estimated from a statistical model, which may account for features specified by the analysis. The model details and included covariates are not reported in the registry record, so the exact adjustment behind these estimates cannot be reconstructed from the posted information.

What does a mean difference represent?

A mean difference compares the estimated change in one group with the estimated change in another. For the primary outcome, the comparisons use placebo as the reference. A negative difference means that the active-group mean change was more negative than the placebo mean change. Since the registry describes lower PANSS scores as better, this direction corresponds to a greater average reduction in PANSS total score. This remains a comparison of group averages; it does not mean that every participant had the same change.

How should the standard error be understood?

A standard error describes the statistical uncertainty around an estimated mean under the model. It is not the same as the variability of individual scores. In general, a smaller standard error indicates a more precisely estimated mean, while a larger one indicates greater sampling uncertainty. The standard errors for the reported group means help convey precision, but they do not replace a confidence interval for the between-group comparison.

What does a confidence interval add?

A confidence interval gives a range of values compatible with the data and the analysis assumptions at its stated confidence level. It conveys both the estimated direction and the precision of a comparison. If the interval is relatively wide, a range of possible differences remains compatible with the result. If it includes zero, the data do not rule out no difference under that interval procedure. A confidence interval is not a guarantee that the true effect lies in the interval, nor does it describe the range of individual participant responses.

What do the p-values say?

A p-value measures how unusual the observed data, or more extreme data, would be under a specified null hypothesis and analysis assumptions. A small p-value is evidence against that null comparison within the analysis. It does not give the probability that the null hypothesis is true, and it does not quantify clinical importance. The sponsor’s reported p-values and the unadjusted recalculated p-values are not identical in every comparison. The sponsor analysis takes precedence when the two differ.

Why does over-dispersion matter in general?

In general, over-dispersion means that the observed variation is greater than a statistical model assumes. When a model understates variability, standard errors can be too small and intervals too narrow, making results appear more precise than the data warrant. The primary PANSS outcome is continuous, and the registry does not report an over-dispersion assessment or identify this as a concern. This general concept is relevant when evaluating whether a model’s uncertainty estimates adequately reflect the data, but it should not be read as a finding about this trial.

What can be reconstructed from the posted summaries?

The recalculated comparisons use posted group summaries to produce simple, unadjusted mean differences and approximate uncertainty estimates. The pooled standard deviations used in those calculations are marked as approximate because they are inferred from reported standard errors. Such calculations do not reproduce the sponsor’s mixed-model analysis, its handling of repeated measurements or any adjustments not included in the available summaries. They are useful as a transparent cross-check, but not as a replacement for the prespecified or sponsor-reported analysis.

The registry lists 74 analyses across its outcome measures. Multiple outcomes or comparisons create a multiplicity issue: the more statistical tests considered, the greater the chance of at least one apparently notable result by chance if results are interpreted without adjustment. The record does not report whether the secondary analyses were adjusted for multiplicity, whether they were confirmatory, or how any testing sequence was defined. Therefore, secondary p-values should be treated as less definitive than a clearly prespecified primary analysis.

Randomization supports comparison of assigned groups, while masking can reduce some forms of bias in measurement and care. However, the registry’s labels alone do not establish that all potential biases were avoided. The precise randomization procedure, allocation concealment, model assumptions, handling of missing observations and sensitivity analyses are not reported in the registry record. Those omissions limit how fully an outside reader can assess the analysis beyond the posted estimates.

11. Confidence Intervals and Precision

The sponsor’s primary comparisons are reported with 95% confidence intervals. For the combined 4mg and 2 mg comparison with placebo, the estimated treatment difference is -8.18, with an interval from -12.0 to -4.40. For 4mg alone, the estimate is -7.64, with an interval from -12.0 to -3.30; for 2 mg, it is -8.72, with an interval from -13.1 to -4.37. Each of these reported intervals lies below zero. The 0.25 mg estimate is -2.89, with an interval from -8.27 to 2.49, which includes zero.

In general, a confidence interval expresses the uncertainty around an estimate under the analysis assumptions. The intervals for the 4mg and 2 mg comparisons still cover a range of possible differences, so the point estimates should not be read as exact values. The 0.25 mg interval spans negative and positive differences, leaving more uncertainty about both the direction and magnitude of the comparison. That uncertainty is also consistent with its larger standard error and smaller Week 6 group count than the higher-dose groups, although the registry does not establish that one factor alone explains the interval width.

The sponsor-reported confidence intervals and the simple recalculated intervals are not identical. The recalculated intervals use unadjusted summary-statistic methods and approximate pooled standard deviations, whereas the posted sponsor results come from mixed-model analyses. The sponsor’s own analysis takes precedence. Neither interval type quantifies individual response variation, guarantees coverage of the true effect in a particular study, or resolves the unreported details of the model and missing-data approach.

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 describe hypothetical sample sizes for a two-arm comparison, not the number of participants actually enrolled in each trial arm. For the continuous outcomes in this trial, the figures are participants per arm; they are not total event counts. The assumptions listed for these illustrations are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is set as the stated fraction of the observed effect. The registry’s original sample-size assumptions and justification are not reported in the registry record.

The 80% power column gives the illustrative number of participants per arm for a design that would have 80% power under those assumptions to detect the specified effect. The 90% power column gives the corresponding number per arm for 90% power. Power is a design property calculated under assumptions; it is not a measure of whether the completed trial found a result, nor does it describe the probability that a particular conclusion is correct.

The rows use different fractions of the observed effect. Planning for 75% or 50% of the observed effect means targeting a smaller difference than planning for 100% of it. Smaller effects are harder to distinguish from random variation, so they require more participants under otherwise unchanged assumptions. The listed figures are planning illustrations based on an effect observed in this trial, and may be especially uncertain because an observed effect is not known in advance when planning a future study. This is not observed power.

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 = -7.640; pooled standard deviation σ = 17.822.
2. n = 2 × (1.96 + 0.842)² × 17.822² / -7.640² = 85.4, rounded up to 86 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 <- -7.64                      # observed difference in means
sigma <- 17.822012284672                     # 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(1396421)
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 = 86 (80% power), 115 (90% power)
 75% of observed effect: n per arm = 152 (80% power), 204 (90% power)
 50% of observed effect: n per arm = 342 (80% power), 458 (90% power)
power.t.test: 86.4 
Simulated power at n = 86 per arm: 0.802 (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 registry reports mixed-model analyses for the primary PANSS outcome. Those analyses include a combined comparison of Brexpiprazole 4mg and 2 mg with placebo, as well as separate comparisons of each active dose with placebo. The reported estimates, confidence intervals and p-values are the sponsor’s own analysis results and are the appropriate results to prioritize when they differ from summary-statistic recalculations.

The recalculated comparisons are simple, unadjusted estimates based on the posted group means, standard errors and sample sizes. Their pooled standard deviations are approximate, and the calculations do not reproduce the sponsor’s mixed model. For the 4mg and 2 mg comparisons, the recalculated point estimates match the sponsor’s individual estimates to the displayed precision, but intervals and p-values differ. For the 0.25 mg comparison, the point estimate is also -2.89 while the interval and p-value differ slightly. The combined 4mg and 2 mg sponsor contrast has no matching combined recalculation listed.

These differences do not mean the registry’s sponsor analysis is erroneous. They reflect that an unadjusted calculation from summary statistics and a model-based analysis need not use identical assumptions or methods. The registry does not report enough model detail to reconcile every difference. Accordingly, the recalculations are best regarded as a check on the broad numerical pattern, not as an alternative primary inference.

15. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when a study tests several outcomes, time points, doses or contrasts. Even if each test uses a conventional threshold, the chance of at least one apparently strong result can increase as the number of tests grows. This matters for this trial because the registry lists 16 outcome measures and 74 analyses, including multiple doses, secondary outcomes and time points. The record does not report whether the sponsor controlled the overall type I error rate, adjusted secondary p-values, or specified a testing hierarchy. The reported secondary p-values should not be treated as if they were all independently prespecified confirmatory tests.

The primary efficacy population is described as participants who received at least one dose and had baseline and at least one post-baseline efficacy evaluation. The record also gives Week 6 counts for the primary PANSS outcome. It does not report the amount or reasons for missing data, the method used to handle missing observations, or sensitivity analyses assessing the effect of missingness. Without those details, the effect of incomplete outcome data on the estimates cannot be fully evaluated.

The sponsor reports mixed-model analyses, but the model specification, assumptions, covariates, covariance structure and handling of participants with incomplete repeated measurements are not reported in the registry record. The summary statistics used for recalculation do not provide a way to recover those details. The primary population description also does not by itself explain the analysis set used for every secondary outcome.

The record does not report interim analyses, stopping rules, the original sample-size justification, or whether any protocol changes affected the analysis plan. Those details should not be inferred from the number of analyses or from the posted results. Their absence limits assessment of how fully the reported analyses followed a prospective plan.

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 serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: schizophrenia (2.2%), acute myocardial infarction (1.1%), psychotic disorder (1.1%), rhabdomyolysis (0.5%), gastric ulcer (0.5%).

The registry reports serious adverse events by group with both the number at risk and the number affected. Serious events affected 2 of 180 participants in the Brexpiprazole 4mg group, 4 of 182 in the 2 mg group, 4 of 90 in the 0.25 mg group and 7 of 184 in the placebo group. These counts describe participants recorded as affected; they do not, by themselves, establish causation or permit a complete comparison of safety across the groups.

The listed serious-event terms include Schizophrenia at 2.2%, Acute myocardial infarction at 1.1%, Psychotic disorder at 1.1%, Rhabdomyolysis at 0.5%, Gastric ulcer at 0.5% and Grand mal convulsion at 0.5%. The registry does not report a death count in the posted safety summary. The record’s adverse-event figures should be interpreted within its six-week treatment period and alongside the limited information about event definitions, timing and attribution. No conclusion about longer-term safety can be drawn from these figures alone.

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

The posted primary results describe average change in PANSS total score from baseline to Week 6. The sponsor’s mixed-model estimates for the 4mg and 2 mg comparisons with placebo are negative, and their reported 95% confidence intervals lie below zero. The sponsor’s 0.25 mg comparison has an interval that includes zero. These statements describe the registry’s estimates and their statistical uncertainty; they are not a guarantee of benefit for an individual participant or a complete account of treatment effects.

A statistically distinguishable group average does not automatically establish that the difference is important to participants. The registry record does not report a threshold for a clinically meaningful PANSS difference, and the estimates do not show how many individuals experienced improvement, no change or worsening. The primary outcome also does not encompass all aspects of functioning, safety, tolerability or quality of life.

The study lasted six weeks, so the posted endpoint does not establish what happens after that period. The results apply directly to the trial’s enrolled population and stated analysis population; how well they generalize to other populations or settings is not established by the registry summaries. The record also leaves some statistical details unreported, including the full model specification and missing-data approach.

Safety counts are descriptive and do not establish that an event was caused by the assigned intervention. The serious adverse-event summary, with no death count reported, is not sufficient to resolve the full balance of benefits and harms. Overall, the record supports a careful reading of measured group differences at the specified time point, while leaving clinical importance, longer-term outcomes and several analysis details uncertain.

18. Why This Trial Matters Statistically

This trial offers a useful statistical example of why a primary outcome, a set of dose comparisons and uncertainty estimates need to be read together. The primary measure is a continuous symptom score, and the registry reports both group mean changes and model-based between-group contrasts. The comparison intervals show not only the direction of the estimates but also how much uncertainty remains, particularly for the lower-dose comparison.

It also illustrates the distinction between a sponsor-reported analysis and an unadjusted recalculation from summary statistics. The latter can provide a transparent numerical cross-check, but it cannot reproduce a model-based analysis when the full model details are not available. In this record, the recalculations and the sponsor’s results are not identical in all details, so the sponsor’s analysis takes precedence.

The number of outcomes and analyses makes multiplicity important. A set of secondary findings can add context, but without information about outcome hierarchy and adjustment, nominal p-values across many comparisons should not be treated as a series of independent confirmations. The record also shows why analysis-population counts and missing-data methods matter: the number with Week 6 endpoint data is not the same as total enrollment, and the treatment of absent measurements affects interpretation.

Finally, planning estimates based on observed effects are illustrations rather than a record of the original design process. They help explain how target effect size and desired power relate to sample size, while the trial’s own sample-size justification remains unreported. Together, these features make the study relevant for learning to separate estimates, precision, planning assumptions and confirmatory claims.

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 posted primary endpoint results are specific to the baseline-to-Week 6 period. They do not provide evidence about persistence of changes or outcomes beyond the trial’s six-week treatment duration. The primary efficacy population excludes participants without the specified baseline and post-baseline efficacy assessments, and the registry does not report the reasons or extent of missing Week 6 data.

The Week 6 primary outcome counts were 134 for Brexpiprazole 4mg, 132 for 2 mg, 62 for 0.25 mg and 124 for placebo. The smaller observed count in the 0.25 mg group is relevant when considering its precision, but the registry summaries do not establish the reason for that count or the consequences of missingness. The original allocation plan and sample-size justification are not reported.

The record reports model-based sponsor analyses but does not provide the complete model specification, covariates, missing-data method or sensitivity analyses. It also does not report multiplicity handling despite listing 74 analyses. These gaps constrain independent assessment of the inferential procedures and the weight to give secondary results.

The secondary outcome list is broader than the numerical results described in the posted summaries. Consequently, the pattern across every listed outcome cannot be fully evaluated from the registry record. The serious adverse-event summary supplies counts and selected terms, but a death count is not reported in the registry record, and the summary does not establish event causality or longer-term safety. These limitations are specific to the available record and should not be filled in by assumptions about the protocol or analysis.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Study of the Effectiveness of Three Different Doses of OPC-34712 in the Treatment of Adults With Acute Schizophrenia — NCT01396421.
• Registry record last updated 2015-10-29; status last verified 2015-09.
• Results first posted 2015-10-29.
• 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

NCT01396421 is a completed Phase 3, multicenter, randomized, double-blind, placebo-controlled, parallel-group trial in adults with acute schizophrenia. It enrolled 636 participants and compared once-daily brexpiprazole doses of 4mg, 2 mg and 0.25 mg with placebo over six weeks.

The primary endpoint was mean change in PANSS total score from baseline to Week 6. Sponsor-reported mixed-model comparisons for 4mg and 2 mg versus placebo had 95% confidence intervals below zero; the 0.25 mg interval included zero. Because simple recalculations differ from the sponsor’s reported analyses, the sponsor’s results take precedence. The registry lists multiple secondary analyses and serious adverse-event counts, while multiplicity handling, detailed missing-data methods and the original sample-size justification are not reported in the registry record.