Study to Evaluate the Efficacy and Safety of Aripiprazole Administered With Lithium or Valproate Over 12 Weeks in the Treatment of Mania in Bipolar I Disorder: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Study to Evaluate the Efficacy and Safety of Aripiprazole Administered With Lithium or Valproate Over 12 Weeks in the Treatment of Mania in Bipolar I Disorder (NCT00665366), a randomized phase 3 study. The primary outcome is Change From Baseline in Total Score on the Young Mania Rating Scale (YMRS) (LOCF Data Set).

2. Clinical Question

3. Trial Design

The registry describes a completed Phase 3 randomized, double-blind, parallel-group, placebo-controlled study. The study title and official title specify a 12-week assessment of adjunctive aripiprazole in people with mania in Bipolar I disorder who were treated with lithium or valproate and needed further clinical improvement. Enrollment is listed as 493.

The treatment contrast was aripiprazole added to ongoing lithium or valproate versus placebo added to ongoing lithium or valproate. Participants were randomly assigned to study treatment, while their background treatment continued. The registry describes the placebo arm as randomized in a 1:1 allocation to study drug. Details about how randomization was generated or concealed are not reported in the registry record.

Double masking means that the study was designed to keep treatment assignment hidden from participants and relevant study personnel. The record does not describe who specifically remained masked, how masking was maintained, or whether masking was assessed. It also does not report stratification factors or other allocation procedures.

The study started in 2008-06, reached primary completion in 2011-10, and is marked completed. The primary outcome is change from baseline in YMRS total score from baseline to Week 12, using a last observation carried forward analysis set. The registry lists 12 outcome measures and 56 analyses. The trial's original sample-size justification, its planned statistical analysis details, and any interim analyses are not 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.

Before assignment: Of 493 participants enrolled, 370 were randomized, and 369 received treatment with double-blind study medication.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All randomized participants who received at least 1 dose of study medication and who had at least 1 efficacy evaluation after the start of study drug.

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

6. Primary Endpoint

The registry lists the primary outcome as Change From Baseline in Total Score on the Young Mania Rating Scale (YMRS) (LOCF Data Set), measured over Baseline to Week 12.

The primary endpoint is change from baseline in total score on the Young Mania Rating Scale, or YMRS, from baseline to Week 12. The YMRS is administered by a clinician and assesses manic symptoms and clinical condition over the preceding 48 hours. Its total score ranges from 0 to 60, with 12 or greater signifying hypomania or mania. Items cover symptoms such as mood, motor activity, sleep, irritability, speech, thought content, behavior, appearance, and insight.

Each item receives a severity rating, and the total score provides a structured measure of the overall severity of manic symptoms. A change score describes movement from a participant's baseline score; negative changes in the posted results indicate a decrease in the score. The endpoint is used because it measures the symptom domain targeted by this study and permits comparison of change between randomized groups.

The registry specifies a last observation carried forward (LOCF) data set for the primary outcome. In general, LOCF uses a participant's last recorded outcome to stand in for a later outcome that is missing. This can retain participants with incomplete follow-up in an analysis, but it relies on assumptions about what the unobserved later measurements would have been. The rationale for choosing LOCF and any sensitivity analyses addressing missing outcomes are not reported in the registry record.

7. Results Posted in the Registry

• Aripiprazole + Valproate or Lithium: mean -13.57 versus -11.54 on control; difference -2.03 (95% CI -4.19 to 0.13), standardized effect d = -0.19; z-test p = 0.066.

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.

Change From Baseline in Score on Clinical Global Impression-Bipolar Version (CGI-BP) Severity of Illness (Overall) Scale at Week 12 (LOCF Data Set)

• 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:
  • Aripiprazole + Valproate or Lithium: mean -1.59 versus -1.51 on control; difference -0.08 (95% CI -0.38 to 0.22), standardized effect d = -0.05; z-test p = 0.607.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Aripiprazole + Valproate or Lithium and the control. The interval gives the range of effects that remain compatible with the data.

The registry lists secondary outcomes spanning clinician-rated illness severity, functioning, response and remission on the YMRS, longitudinal functioning, and patient-rated improvement. These outcomes address related but distinct aspects of illness and functioning. They can offer supporting context for the primary symptom-scale result, but they should not be treated as interchangeable measures of the same endpoint.

Numerical information is available for the change in CGI-BP overall severity at Week 12. The reported group means are -1.51 for placebo added to valproate or lithium and -1.59 for aripiprazole added to valproate or lithium. The recalculated mean difference is -0.08, with an interval from -0.385 to 0.225 and an unadjusted p-value of 0.607. The sponsor-reported analysis gives -0.08, with a 95% confidence interval from -0.37 to 0.20 and p = 0.567. Both intervals span values on either side of no difference.

Detailed numerical findings for the other listed secondary outcomes are not reported in the registry record. Consequently, the overall pattern across the full secondary outcome set cannot be evaluated from the available results. The one numerical secondary comparison does not offer clear evidence of a group difference, and the primary result is also uncertain; this does not allow a broad conclusion about every domain measured.

In general, multiple secondary comparisons increase the chance that some result may appear notable by chance. Unless multiplicity is addressed through a prespecified testing strategy or other appropriate methods, unadjusted secondary p-values carry less confirmatory weight than a prespecified primary analysis. The registry does not report a multiplicity adjustment or testing hierarchy.

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

• Change From Baseline in Score on Clinical Global Impression-Bipolar Version (CGI-BP) Severity of Illness (Mania) Scale (LOCF Data Set)
• Change From Baseline in Score on Clinical Global Impression-Bipolar Version (CGI-BP) Severity of Illness (Depression) Scale (LOCF Data Set)
• Change From Baseline in Score on Clinical Global Impression-Bipolar Version (CGI-BP) Severity of Illness (Overall) Scale at Week 12 (LOCF Data Set)
• Change From Baseline in Total and Subscale Scores on the Functional Assessment Short Test (FAST)(LOCF Data Set)
• Percentage of Participants Showing A Response From Baseline on the Young Mania Rating Scale (YMRS)(OC Data Set)
• Percentage of Participants Showing Remission in the Young Mania Rating Scale (YMRS) Score From Baseline (LOCF Data Set)
• Change From Baseline in Total Score on the Longitudinal Interval Follow-up Evaluation-Rating Impaired Functioning Tool (LIFE-RIFT)(OC Data Set and Week 12 LOCF Data Set)
• Participant Scores on Patient Global Impression Improvement (PGI-I) Scale (OC Data Set)

9. Statistical Interpretation

10. Statistical Methods Explained

What does a mean change represent?

The primary outcome is continuous: each participant has a numerical YMRS score, and the group result is expressed as a mean change. A mean summarizes the average direction and amount of change among the observations included in the analysis. It does not show the distribution of individual changes, how many participants improved, or whether every participant moved in the same direction. In this trial, the reported group means are negative, indicating decreases in score from baseline on average.

The comparison of interest is the difference between the group results, not whether either group's mean change differs from baseline on its own. The recalculated estimate is -2.03 for aripiprazole relative to placebo. Its sign gives the direction of the contrast under that subtraction order. The mean difference alone does not indicate the precision of the estimate or its clinical importance.

Why distinguish the recalculation from the reported analysis?

The recalculated comparison is an unadjusted re-analysis based on posted summary statistics. It does not reproduce the trial's full participant-level model. The registry reports an ANCOVA analysis for the primary outcome, described as a mean difference in final values. ANCOVA is a regression-based approach that can compare groups while accounting for specified covariates. The registry record does not identify the covariates, model specification, or other details needed to reconstruct that analysis.

The recalculated estimate and the sponsor-reported estimate are similar, but they are not identical. That is expected when an unadjusted calculation and a model-based analysis use different information or assumptions. The sponsor's ANCOVA result is the reported trial analysis; the recalculation is useful as a transparent summary-statistic comparison, not as a substitute for the trial's own analysis. The discrepancy flag is false.

What does the p-value say?

In general, a p-value measures how compatible the observed data are with a specified null model, assuming that model and the analysis assumptions hold. It is not the probability that the null hypothesis is true, and it does not measure the size or clinical importance of an effect. The unadjusted p-value is 0.066; the sponsor-reported p-value is 0.058. These values do not meet the conventional 0.05 threshold, but a threshold-based reading should not replace consideration of the estimated difference and its confidence interval.

A p-value also does not demonstrate equivalence when it is not below a chosen threshold. Establishing equivalence or noninferiority requires a prespecified margin and an analysis designed for that question. Such a margin and analysis are not reported in the registry record.

Why consider dispersion and uncertainty?

In general, a mean is more informative when read alongside a measure of uncertainty. The standard errors for the primary group means are 0.78 in each arm. A standard error describes the estimated precision of a mean, rather than the variation among individual participant scores. Individual-level variability is therefore not conveyed by these standard errors alone.

The recalculated comparison uses a pooled standard deviation of 10.5, with an estimated standardized difference of -0.193. These recalculated quantities offer a scale-relative description of the contrast, but they do not replace the original trial analysis. A standardized effect depends on the outcome's variability, and a numerical effect size should not be interpreted as a clinical threshold unless such a threshold has been established.

Over-dispersion is a concern for outcome models when the observed variability is greater than a model's assumed variability. In general, this can make standard errors and confidence intervals too small if not handled appropriately. The primary outcome is continuous and the registry does not report an over-dispersion assessment or correction. The issue should not be presumed to have occurred; the available record simply does not describe that diagnostic.

How does missing-data handling matter?

The primary outcome uses an LOCF data set. In general, carrying a last observation forward treats a previously observed score as a stand-in for a later unobserved score. This makes an assumption about the trajectory after the last measurement and can affect the group comparison if missingness or outcomes differ between groups. LOCF is not the same as observing a participant's Week 12 score.

The primary analysis population includes randomized participants who received at least one dose and had at least one efficacy evaluation after study drug began. Thus, the result does not represent every person randomized regardless of treatment initiation or post-start assessment. The registry record does not report the number excluded from this analysis, the reasons for missing observations, or sensitivity analyses under alternative missing-data assumptions.

The record's outcome listings refer to different analysis sets, including LOCF and observed-case analyses. In general, observed-case analyses use measurements that are available rather than filling in missing values, and their findings can be affected by who has follow-up data. Without detailed participant flow and missingness information, the robustness of the reported results to these choices cannot be assessed from the registry record.

11. Confidence Intervals and Precision

The recalculated primary mean difference is -2.03, with an interval from -4.192 to 0.132. The range includes estimates favoring a greater decrease with aripiprazole, as well as estimates on the other side of no group difference. This means the posted summary is compatible with a modest group separation in the estimated direction, but also with little or no separation. The interval does not support treating the point estimate as exact.

The registry-reported ANCOVA estimate is -2.04, with a 95% confidence interval from -4.14 to 0.07. In general, a confidence interval describes a range of estimates that are reasonably compatible with the data and the method's assumptions. It is not a range containing a known proportion of individual participant outcomes, nor does it assign a probability to the true effect after observing the data. The sponsor-reported interval likewise extends across values on either side of no difference.

Both intervals are consistent with uncertainty about the size of any between-group difference. The point estimates are close, but precision is not determined by closeness between estimates. It is represented by the spread of the interval and depends on the analysis, variability, and information available. The registry does not provide a clinical importance threshold for the YMRS difference, so the interval cannot be classified against such a threshold.

The standard error of each primary group mean is 0.78. That quantity concerns uncertainty in the corresponding mean, not the full range of participant-level scores. The comparison interval is the more direct representation of uncertainty for the treatment contrast. The reported confidence interval and the recalculated interval differ slightly because the recalculation is unadjusted while the registry's analysis uses ANCOVA.

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 illustrations based on the observed effect, rather than the trial's original design calculations. The original sample-size assumptions and justification are not reported in the registry record.

The illustrations assume a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is set to a stated fraction of the observed effect. For planning for 100% of the observed effect, the figures are 420 participants per arm for 80% power and 562 participants per arm for 90% power. For planning for 75% of the observed effect, they are 747 participants per arm for 80% power and 1,000 participants per arm for 90% power. For planning for 50% of the observed effect, they are 1,680 participants per arm for 80% power and 2,248 participants per arm for 90% power.

In general, the 80% power column represents the planning number per arm associated with an 80% chance of detecting the specified effect under the stated assumptions if that effect is real. The 90% power column makes the corresponding planning target more demanding in power. Smaller effects are harder to distinguish from random variation, so planning for a smaller fraction of the observed effect requires more participants per arm. These are planning illustrations, not observed power; they do not measure the probability that the completed trial detected an effect.

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 = -2.030; pooled standard deviation σ = 10.498.
2. n = 2 × (1.96 + 0.842)² × 10.498² / -2.030² = 419.8, rounded up to 420 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 <- -2.03                      # observed difference in means
sigma <- 10.4978521612757                     # 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(665366)
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 = 420 (80% power), 562 (90% power)
 75% of observed effect: n per arm = 747 (80% power), 1,000 (90% power)
 50% of observed effect: n per arm = 1,680 (80% power), 2,248 (90% power)
power.t.test: 420.8 
Simulated power at n = 420 per arm: 0.799 (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 an ANCOVA analysis for the primary outcome: a mean difference in final values of -2.04, a 95% confidence interval from -4.14 to 0.07, and p = 0.058. The posted-summary recalculation is unadjusted and gives a mean difference of -2.03, an interval from -4.192 to 0.132, and an unadjusted p-value of 0.066. The estimates are close in direction and size, while their intervals and p-values differ slightly.

This distinction matters because the recalculation uses summary statistics and does not reproduce the sponsor's model. ANCOVA may account for covariates, but the specific covariates and model details are not reported in the registry record. The sponsor's analysis is the registry-reported inferential analysis; the recalculation provides a separate, unadjusted view of the posted group summaries. The discrepancy flag is false, and the recalculation should not be presented as a correction to the sponsor's result.

Both analyses leave uncertainty about whether there is a group difference: their intervals include values on either side of no difference, and neither reported p-value is below 0.05. Neither result establishes equivalence or proves that the intervention has no effect.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists 12 outcome measures and 56 analyses, but it does not report a multiplicity-control strategy, testing hierarchy, or adjustment for examining multiple outcomes or analyses. When many statistical comparisons are made, the chance of observing at least one seemingly unusual result by chance can increase. In general, this makes isolated unadjusted secondary findings less persuasive as confirmatory evidence. The absence of a reported multiplicity approach does not establish that no such approach was used; its details are not reported in the registry record.

The record describes the primary outcome as change in YMRS total score from baseline to Week 12 using LOCF. It also lists other measures using LOCF and observed-case analysis sets. The choice of analysis set and missing-data method can affect results, particularly if missingness differs across groups or relates to outcomes. Participant-level missingness patterns, reasons for withdrawal, and sensitivity analyses are not reported in the registry record.

The primary analysis population includes randomized participants who received at least one dose of study medication and had at least one efficacy evaluation after study drug began. This population is narrower than all randomized participants, and the record does not report the participant flow needed to assess the consequences of that restriction. The specific handling of participants with incomplete follow-up, additional exclusions, or protocol deviations is not reported in the registry record.

The registry reports an ANCOVA analysis for the primary outcome, but its covariates, model diagnostics, and detailed assumptions are not reported. The full statistical analysis plan, including how the primary comparison was selected from the listed analyses and whether any interim looks were planned or conducted, is also not reported in the registry record. These omissions limit assessment of prespecification and robustness, without implying that the analyses were necessarily inappropriate.

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: mania (3.7%), depression (2.2%), alcohol abuse (1.1%), hyperthermia (0.6%), coma (0.6%).

The registry reports serious adverse events for the aripiprazole and placebo study-drug groups. Serious adverse events affected 11 of 180 participants in the aripiprazole group and 17 of 189 participants in the placebo group. Other adverse events affected 45 of 180 and 29 of 189 participants, respectively. These are group-specific counts as posted; they should not be combined with the efficacy analysis counts, which use different populations.

The registry lists Mania at 3.7%, Depression at 2.2%, Alcohol abuse at 1.1%, Hyperthermia at 0.6%, Coma at 0.6%, and Psychotic disorder at 0.6% among the most frequent serious adverse event terms. The record does not identify group attribution for these listed terms in the available information, so they should not be assigned to one treatment group. Death counts are not reported in the registry record.

These figures describe recorded events during the study; they do not by themselves establish causation or provide a complete comparison of safety. The registry does not provide event timing, exposure-adjusted rates, or details needed to evaluate whether participants had multiple events. The different group counts should be read descriptively rather than as a definitive estimate of comparative risk.

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

The primary outcome summaries show mean decreases in YMRS score in both groups. The estimated mean difference favors a greater decrease in the aripiprazole group, but the recalculated interval and the sponsor-reported ANCOVA interval both include values on either side of no group difference. The reported p-values are 0.066 for the unadjusted recalculation and 0.058 for the sponsor analysis. The data therefore leave uncertainty about the size and direction of the underlying group contrast.

This is not proof that the treatments are equivalent, and it is not proof that adjunctive aripiprazole has no effect. A nonsignificant result can reflect uncertainty as well as a small difference. The registry does not report an equivalence or noninferiority design or margin, so such conclusions cannot be drawn from these results.

The group means do not tell how individual participants changed, whether there were subgroups with different responses, or whether the average difference reaches a clinically important threshold. No such threshold is reported in the registry record. The primary result concerns a clinician-rated mania scale over the stated follow-up, not every symptom, aspect of functioning, or longer-term outcome.

The study compares adding aripiprazole with adding placebo while participants continue lithium or valproate. It does not answer questions about using aripiprazole alone, comparing lithium with valproate, or outcomes in populations outside the enrolled study context. Safety counts are descriptive, and the available information does not establish that recorded events were caused by a particular study treatment.

18. Why This Trial Matters Statistically

This trial illustrates why an estimated treatment difference should be read together with its uncertainty. The primary point estimate is in one direction, while the confidence interval includes values on either side of no difference. The p-values are close to, but do not cross, the stated conventional threshold of 0.05. Treating a result as simply positive or negative would obscure the uncertainty and the range of effects compatible with the analysis.

It also illustrates the difference between a sponsor-reported model and an unadjusted recalculation from summary statistics. Their results are similar but not identical, because they do not represent the same analysis. Reporting each clearly helps readers distinguish the trial's inferential analysis from a transparent summary-statistic comparison.

The outcome itself matters to interpretation. YMRS change captures clinician-rated manic symptoms, while secondary outcomes address global severity, functioning, and patient perspectives. Those additional domains could complement the primary endpoint, but detailed numerical results are not reported for most listed secondary measures. That limits the degree to which findings can be triangulated across outcomes.

Finally, a planning illustration based on an observed effect is not evidence about the power achieved by the completed trial. It shows how participant requirements can grow when planning for smaller effects under stated assumptions. The trial's original design assumptions are not reported, so those illustrations cannot explain why the study enrolled the number it did or establish whether its original design was adequately powered.

19. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The registry record does not report the trial's original sample-size justification or planned design assumptions. Therefore, the planning illustrations based on the observed effect cannot be used to infer the original rationale for enrollment. The detailed statistical analysis plan, model covariates, randomization procedures, allocation concealment, and any interim analyses are also not reported in the registry record.

The primary endpoint uses LOCF, which carries forward a prior measurement when a later value is missing. Without the amount and pattern of missing data or sensitivity analyses, the effect of that method on the result cannot be assessed. The primary population also excludes randomized participants who did not receive study drug or did not have a post-start efficacy evaluation; participant flow and reasons for exclusions are not reported.

The registry lists many outcomes and analyses, but does not report a multiplicity strategy. This makes it difficult to judge how secondary findings were prioritized or whether inferential error from multiple comparisons was addressed. Numerical findings for most listed secondary outcomes are not reported in the registry record, limiting assessment of consistency across domains.

The primary endpoint is a mean change on a symptom scale. Group means and standard errors do not describe the full distribution of individual outcomes, and the registry does not provide a clinically important difference threshold. The registered safety summary also lacks event timing, treatment-group attribution for the listed serious-event terms, and deaths counts. These limitations constrain interpretation of efficacy and safety without resolving the underlying treatment effects.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Study to Evaluate the Efficacy and Safety of Aripiprazole Administered With Lithium or Valproate Over 12 Weeks in the Treatment of Mania in Bipolar I Disorder — NCT00665366.
• Registry record last updated 2013-12-02; status last verified 2013-02.
• Results first posted 2013-02-04.
• 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

NCT00665366 is a completed Phase 3 randomized, double-blind, parallel-group, placebo-controlled study of adjunctive aripiprazole for mania in Bipolar I disorder. Participants continued lithium or valproate and received aripiprazole or placebo for 12 weeks. Enrollment is listed as 493.

The primary outcome was change in YMRS total score from baseline to Week 12 using an LOCF data set. The recalculated unadjusted comparison and the registry-reported ANCOVA analysis both estimate a greater mean decrease with adjunctive aripiprazole, but their confidence intervals include values on either side of no group difference. The record lists multiple secondary outcomes, while numerical results for most are not reported in the registry record.

The registry reports serious and other adverse-event counts by study-drug group, but the available safety information does not establish causation. Key details about the original sample-size justification, full analysis plan, missing-data sensitivity analyses, and multiplicity handling are not reported in the registry record.