Efficacy, Safety and Tolerability of Cariprazine as an Adjunctive Treatment to Antidepressant Therapy (ADT) in Patients With Major Depressive Disorder (MDD) Who Have Had an Inadequate Response to Antidepressants Alone: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Efficacy, Safety and Tolerability of Cariprazine as an Adjunctive Treatment to Antidepressant Therapy (ADT) in Patients With Major Depressive Disorder (MDD) Who Have Had an Inadequate Response to Antidepressants Alone (NCT03738215), a randomized phase 3 study. The primary outcome is Total Score Change From Baseline to Week 6 in the MADRS (Montgomery-Åsberg Depression Rating Scale).

2. Clinical Question

3. Trial Design

This was a randomized, double-blind, parallel-group phase 3 study sponsored by AbbVie. It enrolled 759 participants with major depressive disorder who had an inadequate response to antidepressants alone. The registry record lists the study as completed, with a start date of 2018-11-09 and primary completion on 2021-09-30. Those dates describe the registry record and do not establish when results were analyzed or reported.

Participants were assigned to placebo plus ADT, cariprazine 1.5 mg/day plus ADT, or cariprazine 3 mg/day plus ADT. All groups continued their ongoing antidepressant treatment at the same dose they were taking at baseline. The active treatment and placebo were taken orally during the 6-week double-blind period. The description of the 3 mg regimen indicates an initial 2-week period at 1.5 mg before titration; the registry's arm description is truncated after that point, so further dosing details are not reported in the registry record.

The primary outcome was change from baseline to Week 6 in MADRS total score. The primary analysis population was defined as randomized participants with at least one postbaseline MADRS assessment; the outcome description calls this the mITT population. The registry lists the number analyzed in each arm as the participants with data available for analyses. It reports 231 participants for placebo plus ADT, 231 for cariprazine 1.5 mg/day plus ADT, and 223 for cariprazine 3 mg/day plus ADT.

The registry identifies MMRM as the method used for the sponsor-reported primary comparisons. Details of the original sample-size justification, randomization procedure, allocation concealment, and prespecified analysis plan are not reported in the registry record. The posted information does not establish how participants were distributed among sites or how adherence and treatment discontinuation were handled.

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: A total of 759 participants were randomized to double-blind treatment, 757 participants received at least 1 dose of double-blind investigational product (safety population), and 751 treated participants had at least 1 postbaseline assessment of MADRS total score (modified intent-to-treat \[mITT\] population ). At the end of treatment in the Double-blind Period, participants continued on their background antidepressant therapy (ADT) and entered the Safety Follow-up Period.

Double-blind Treatment Period (6 Weeks)

Reasons for not completing this period:

Safety Follow Up Period (4 Weeks)

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: mITT population includes all randomized participants who had ≥ 1 postbaseline assessment of the MADRS total score.

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: Safety Population included all participants in the randomized population who took ≥ 1 dose of double-blind investigational product.

6. Primary Endpoint

The registry lists the primary outcome as Total Score Change From Baseline to Week 6 in the MADRS (Montgomery-Åsberg Depression Rating Scale), measured over Baseline and Week 6.

The primary endpoint was total score change from baseline to Week 6 on the Montgomery-Åsberg Depression Rating Scale, or MADRS. This is a clinician-rated measure of depressive symptomatology during the past week. It covers feelings of sadness, lassitude, pessimism, inner tension, suicidality, reduced sleep or appetite, difficulty concentrating, and lack of interest. Each item is scored from 0, indicating no symptoms, to 6, indicating symptoms of maximum severity.

A change score compares the Week 6 score with the baseline score. Negative changes indicate that the recorded score decreased; a more negative mean change represents a larger decrease on the scale. The difference between an active-treatment group and the placebo group is therefore a comparison of average change, not a count of participants who improved and not a direct measure of an individual's response.

MADRS is used in this trial to quantify depressive symptom severity in a structured way. A standardized scale makes it possible to compare average changes between randomized groups, while clinician ratings incorporate assessment across several symptom domains. The scale does not, on its own, explain which symptoms drove a group's average change or whether the size of a difference is important to participants. The registry record does not report a threshold for clinically meaningful change or a responder definition for the primary endpoint.

The primary analysis included randomized participants with at least one postbaseline MADRS assessment. Consequently, the posted outcome pertains to that defined analysis population, rather than necessarily to every participant enrolled in the trial.

7. Results Posted in the Registry

• Cariprazine 1.5 mg/Day + ADT: mean -14.10 versus -11.50 on control; difference -2.60 (95% CI -4.54 to -0.66), standardized effect d = -0.24; z-test p = 0.009.
• Cariprazine 3 mg/Day + ADT: mean -13.10 versus -11.50 on control; difference -1.60 (95% CI -3.54 to 0.34), standardized effect d = -0.15; z-test p = 0.106.

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 to Week 6 in the Clinical Global Impressions-Severity (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:
  • Cariprazine 1.5 mg/Day + ADT: mean -1.40 versus -1.10 on control; difference -0.30 (95% CI -0.55 to -0.05), standardized effect d = -0.22; z-test p = 0.018.
  • Cariprazine 3 mg/Day + ADT: mean -1.30 versus -1.10 on control; difference -0.20 (95% CI -0.45 to 0.05), standardized effect d = -0.15; z-test p = 0.116.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Cariprazine 1.5 mg/Day + ADT than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Cariprazine 3 mg/Day + ADT and the control. The interval gives the range of effects that remain compatible with the data.

The registry posts one secondary endpoint: change from baseline to Week 6 in the Clinical Global Impressions-Severity score. It provides a related measure of severity, rather than an independent measure of a different clinical domain. Considering a secondary measure can help describe whether the overall pattern of reported outcomes is coherent, but it does not replace the primary endpoint or resolve uncertainty in it.

The posted arm summaries show decreases in the severity score across groups. The recalculated comparisons and sponsor analyses do not align perfectly, so the secondary outcome should be read with the same distinction between a simple unadjusted recalculation and the sponsor's reported MMRM analysis. The secondary result does not establish a separate, definitive conclusion about the primary outcome.

Secondary endpoints generally carry less inferential weight when they are not protected by a prespecified multiplicity procedure. Testing several outcomes or comparisons increases the chance of at least one apparently unusual result arising by chance. The registry record does not report a multiplicity adjustment or a hierarchy governing the secondary endpoint. Accordingly, secondary p-values should not be treated as if they were an independently confirmed primary finding. The original statistical plan and the status of any prespecified testing hierarchy are not reported in the registry record.

There is only one secondary endpoint posted, so the registry does not support a broad pattern across multiple secondary outcomes. The available secondary measure adds context about severity but cannot establish durability, explain which symptoms changed, or show that a difference is important to participants. Its interpretation remains limited by the analysis and reporting details available in the registry record.

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

• Change From Baseline to Week 6 in the Clinical Global Impressions-Severity (CGI-S) Score

9. Statistical Interpretation

10. Statistical Methods Explained

The outcome is continuous: each participant has a numerical scale score, and the reported arm results are least-squares means of change. A mean summarizes the center of a group distribution. It is useful for comparing average change, but it does not show the full spread of individual outcomes or imply that every participant changed by the group average.

What does the MMRM analysis represent?

The sponsor-reported primary comparison used MMRM, a mixed model for repeated measures. In general, this type of model analyzes repeated measurements while accounting for the fact that measurements from the same participant are related. It can estimate adjusted mean differences using information across scheduled measurement times, rather than relying only on a comparison of two raw arm means at the final visit. The registry reports the method and the least-squares mean differences, but the model specification, covariates, covariance structure, and missing-data assumptions are not reported in the registry record. Those details matter for reproducing or fully assessing the model.

The term “least-squares mean” refers to a model-based group estimate. It need not be identical to the arithmetic mean calculated directly from the observed values. The sponsor's MMRM estimates should therefore be distinguished from the simple recalculation made from the posted arm summaries. That recalculation is unadjusted and does not reconstruct the repeated-measures model.

Why do the recalculated and sponsor results differ?

The posted arm summaries give mean changes and standard errors. A simple contrast of those means, with uncertainty approximated from the reported summaries, produces a convenient unadjusted comparison. It cannot account for all the modeling choices in an MMRM analysis. The recalculated standard deviations are approximate, and the recalculated confidence intervals and p-values are not substitutes for the trial's sponsor analysis. The registry flags a discrepancy; the sponsor's own analysis takes precedence for the reported primary inference.

A p-value describes how incompatible the observed data are with a specified null hypothesis under the assumptions of the statistical test. It is not the probability that the null hypothesis is true, and it does not measure the size or importance of an effect. The unadjusted p-values from the recalculation also do not account for any multiplicity procedure unless such a procedure is explicitly applied; one is not reported for those recalculations.

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 communicates precision as well as direction. An interval that includes zero is compatible with no average difference as well as with differences in either direction covered by the interval. An interval that excludes zero indicates that zero is not among the values in that interval, but it does not establish that the estimated difference is important in practice. Interpretation should consider the interval's width and the scale's meaning, not only a thresholded p-value.

For a continuous outcome, uncertainty around a mean difference is influenced by the variability of participant outcomes and by the number of observations contributing to the estimate. The posted arm-level spread is a standard error, not a standard deviation. A standard error reflects uncertainty in an estimated mean; it is not a description of how widely individual scores vary. Confusing these quantities can lead to incorrect conclusions about the distribution of participant outcomes.

Why can over-dispersion matter?

Over-dispersion is a general issue in models for count or binary data when the observed variability exceeds what a model assumes. It can make standard errors too small if it is ignored. The primary MADRS endpoint is continuous, so over-dispersion is not the central issue for interpreting its mean-change comparison. More generally, statistical models should match the outcome type and its variability. The registry record does not report an over-dispersion assessment for this trial.

What is the analysis population?

The primary outcome description defines the mITT population as all randomized participants with at least one postbaseline MADRS total-score assessment. This definition excludes randomized participants without such an assessment from the specified primary analysis population. The registry says the number analyzed is the number with data available for analyses. It does not report the count or reasons for missing postbaseline assessments, the amount of missing data at each visit, or the methods used to address missing outcomes. Although MMRM is named as the sponsor's method, the registry record does not give enough detail to establish the precise missing-data assumptions used in that model.

Randomization supports a comparison between assigned groups, but the interpretation still depends on the analysis population, adherence, follow-up, and handling of missing information. The registry record does not report enough detail to assess all of these features. The posted estimates should be understood as results for the analyzed participants and the reported methods, not as a complete account of every possible analysis of the trial.

11. Confidence Intervals and Precision

Precision concerns how much uncertainty surrounds an estimated difference. The sponsor's 95% confidence interval for cariprazine 1.5 mg/day versus placebo plus ADT runs from -4.17 to -0.89. Its endpoints are both below zero, so the reported interval is consistent with a decrease in MADRS score that is larger on average in the active group. The interval still covers a range of possible magnitudes; it does not identify one exact effect.

For cariprazine 3 mg/day versus placebo plus ADT, the sponsor's 95% confidence interval runs from -3.16 to 0.12. Because that interval includes zero, it is compatible with no difference as well as with a range of negative differences. The interval also reflects uncertainty in the direction and magnitude of the comparison. It does not establish that the groups are equivalent.

The recalculated intervals are approximate and unadjusted, while the sponsor intervals come from the reported MMRM analyses. Since the registry flags a discrepancy, the sponsor's own analyses take precedence. The intervals should not be interchanged: they come from different approaches and reflect different calculations.

A confidence interval is not a range that contains every participant's change. It describes uncertainty around an estimated group contrast. Individual outcomes may vary more widely. The registry does not provide a clinically meaningful difference threshold, so the intervals cannot be classified as clinically important or unimportant from these data alone. Their width indicates that the estimates are not exact, and their location relative to zero communicates statistical uncertainty under the stated analysis.

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 for the continuous outcomes in this trial. They are not total event counts. The planning illustration uses a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. It sets the effect to detect as a stated fraction of the observed effect. For time-to-event outcomes, the general convention would instead be to report the total number of events, but that does not apply to these continuous outcomes.

The 80% power column gives the estimated number of participants per arm for a design with an 80% chance of detecting the specified effect under the planning assumptions, if that effect and the assumptions were correct. The 90% power column gives the corresponding per-arm planning number for a higher chance of detection. These are design calculations, not descriptions of the actual trial's analysis population.

Rows based on 100% of the observed effect use the full observed effect as the planning target; rows based on 75% or 50% use smaller targets. Smaller effects are harder to distinguish from random variation and therefore require more participants under the same assumptions. The figures are illustrations calculated from the observed effect, not the trial's original design calculations. The trial's own original sample-size assumptions are not reported in the registry record. 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 = -2.600; pooled standard deviation σ = 10.639.
2. n = 2 × (1.96 + 0.842)² × 10.639² / -2.600² = 262.8, rounded up to 263 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 <- -2.6                      # observed difference in means
sigma <- 10.6390789074995                     # 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(3738215)
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 = 263 (80% power), 352 (90% power)
 75% of observed effect: n per arm = 468 (80% power), 626 (90% power)
 50% of observed effect: n per arm = 1,052 (80% power), 1,408 (90% power)
power.t.test: 263.8 
Simulated power at n = 263 per arm: 0.794 (20000 replicates)

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

14. Analyses Reported by the Sponsor

The sponsor's reported primary results come from MMRM and are presented as least-squares mean differences with 95% confidence intervals and p-values. The simple recalculation uses the posted arm-level means and standard errors to form unadjusted comparisons. It is useful as a check on the direction and rough scale of the contrast, but it is not a reconstruction of the sponsor's model.

For the primary outcome, the recalculated estimate for the 1.5 mg/day comparison is -2.6, whereas the sponsor reports -2.5. For the 3 mg/day comparison, the recalculated estimate is -1.6, whereas the sponsor reports -1.5. The recalculated p-values also differ from the reported MMRM p-values. These differences are expected to matter because the approaches are not identical; the registry explicitly flags a discrepancy. The sponsor's own analysis takes precedence when interpreting the trial's reported result.

The reported sponsor method does not by itself reveal every modeling choice. Covariates, repeated-measures covariance structure, and detailed missing-data assumptions are not reported in the registry record. The sponsor results should be described as the posted MMRM results, without claiming that the simple recalculation validates or replaces them. The recalculation is supplementary context, not a competing primary analysis.

15. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when several hypotheses, outcomes, or comparisons are examined. As the number of statistical tests grows, the chance of finding at least one result that appears notable by chance can also grow. Common approaches include adjusting the significance threshold, controlling a family-wise error rate, or specifying an ordered testing hierarchy. These are general statistical concepts; the registry record does not report which, if any, approach was prespecified for this trial.

The primary endpoint has two active-versus-placebo comparisons, and the registry also posts a secondary endpoint. The record does not describe whether the two primary comparisons were handled as a family, whether the secondary endpoint was included in a testing hierarchy, or whether any adjustment was applied to the reported p-values. The p-values should therefore be interpreted as reported for their respective analyses, not assumed to be multiplicity-adjusted. The original statistical plan is not reported in the registry record.

The primary outcome population is described as randomized participants with at least one postbaseline MADRS assessment. This is an analysis-population definition, not a complete account of missing-data handling. The record does not report how many randomized participants lacked postbaseline assessments, why observations were missing, or how missing measurements at later visits were handled. MMRM is named as the sponsor's primary method, but the assumptions and implementation details needed to evaluate its treatment of missing data are not reported in the registry record.

These omissions limit independent appraisal of the inferential framework. They do not invalidate the posted findings, but they make it harder to determine whether the reported analyses follow a prespecified plan, how uncertainty from multiple testing was addressed, and how sensitive results might be to missing-data assumptions. The sponsor's reported analysis remains the principal registry result, while the unadjusted recalculation should not be interpreted as correcting for these unknowns.

16. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

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: appendicitis (0.4%), kidney infection (0.4%), fall (0.4%), fibula fracture (0.4%), tibia fracture (0.4%).

The registry lists serious adverse events for each treatment group and gives the number of participants at risk and the number affected. The reported affected counts are 2 of 253 for placebo plus ADT, 3 of 252 for cariprazine 1.5 mg/day plus ADT, and 3 of 252 for cariprazine 3 mg/day plus ADT. No deaths are listed in any group.

The registry also lists non-serious adverse events in an “other” category: 39 participants in the placebo group, 68 in the 1.5 mg/day group, and 62 in the 3 mg/day group were affected. These are descriptive counts. They do not establish that an event was caused by the assigned treatment, and they do not alone account for differences in exposure or follow-up.

Serious-event terms listed include appendicitis, kidney infection, fall, fibula fracture, tibia fracture, and spontaneous abortion, each reported at 0.4%. The registry record does not provide enough detail in these summaries to assess event timing, recurrence, relatedness, or the clinical context for each event. Safety counts should therefore be read as reported summaries rather than as a complete comparative safety analysis.

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

The posted results describe average changes in symptom-severity scores during a 6-week period among participants included in the defined primary analysis population. The sponsor's MMRM analysis reports a difference for the 1.5 mg/day comparison whose confidence interval excludes zero, while the 3 mg/day interval includes zero. Those results characterize statistical uncertainty under the reported model; they do not alone tell readers whether an observed difference matters to participants.

The findings do not establish that every participant experienced the group-average change. A mean can conceal considerable variation in individual outcomes, and the registry does not provide the distribution of individual changes or a responder analysis for the primary endpoint. The registry also does not report a threshold that would define a clinically meaningful difference in MADRS score.

The comparisons concern cariprazine added to ongoing antidepressant therapy versus placebo added to ongoing antidepressant therapy. They do not compare cariprazine with antidepressant therapy alone, and they do not answer whether the same estimates apply to other treatment settings or patient groups. The background antidepressant was maintained at its baseline dose, but specific antidepressant regimens are not reported in the outcome description.

The results are limited to the study's posted follow-up and outcome definitions. They do not establish longer-term effects, durability after the treatment period, or the balance of benefits and harms beyond the adverse-event summaries in the registry. The counts do not establish causation, and the record does not give detailed information about event relatedness or exposure time.

Finally, the recalculated results are not a substitute for the sponsor's analysis. The registry flags a discrepancy, and the sponsor's own analysis takes precedence. The registry record does not report all details needed to independently reproduce the MMRM or assess the original multiplicity and missing-data plans. The appropriate interpretation is therefore specific: these are the posted trial estimates and their reported uncertainty, with limits on what can be concluded from them.

18. Why This Trial Matters Statistically

This trial illustrates why a treatment comparison should be read through both its estimated effect and its uncertainty. A mean difference gives the direction and scale of the average contrast, while a confidence interval shows which values remain compatible with the analysis. The primary sponsor intervals differ in whether they include zero, but neither the point estimate nor the p-value alone establishes clinical importance.

It also demonstrates why adjusted model-based analyses and simple recalculations can differ. Posted arm means and standard errors allow an approximate unadjusted comparison, but they do not reproduce a repeated-measures model. When the registry flags a discrepancy, the sponsor's own analysis takes precedence rather than treating a reanalysis from summary values as equivalent evidence.

The trial also highlights the role of endpoint and population definitions. The primary result concerns MADRS change through Week 6 in participants with at least one postbaseline assessment, not necessarily every randomized participant. Knowing who contributed data and how missing measurements were handled is essential to understanding the reach of an estimate.

Finally, the record separates statistical evidence from broader interpretation. A score change is one way to quantify severity, while the registry's safety summaries are descriptive counts. Without a reported clinically meaningful threshold, detailed missing-data plan, or multiplicity strategy, readers should avoid extending the numerical findings beyond the comparisons and analyses that are actually posted.

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 record has important trial-specific reporting limits. It gives the sponsor's MMRM method but does not report the full model specification, including covariates, covariance structure, or detailed missing-data assumptions. It also does not report the original sample-size justification or the complete prespecified statistical plan. These omissions constrain independent reproduction and appraisal.

The primary analysis population includes randomized participants with at least one postbaseline MADRS assessment. The record does not state how many randomized participants were excluded from that population or why postbaseline information was unavailable. It also does not report adherence, treatment discontinuation, or sensitivity analyses addressing missing data. The estimate should therefore be understood as applying to the stated analysis and available data, not automatically as an all-randomized-participant result.

The follow-up for the primary endpoint is Week 6. The posted outcome does not address longer-term symptom changes. A clinically meaningful difference threshold and a responder definition are not reported in the registry record, limiting interpretation of the practical importance of the mean differences.

The registry flags a discrepancy between recalculated unadjusted results and sponsor analyses. The recalculation uses posted summaries and is not a reproduction of the MMRM; the sponsor's own analysis takes precedence. Multiplicity handling for the primary comparisons and secondary endpoint is not reported in the registry record. The secondary endpoint therefore adds context but does not provide an independently protected confirmatory result.

Safety reporting is also limited to summary counts and listed event terms. Detailed event narratives, relatedness assessments, timing, and exposure-adjusted comparisons are not reported in the registry record. These limitations do not negate the reported data, but they narrow the conclusions that can be drawn from the registry summaries.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Efficacy, Safety and Tolerability of Cariprazine as an Adjunctive Treatment to Antidepressant Therapy (ADT) in Patients With Major Depressive Disorder (MDD) Who Have Had an Inadequate Response to Antidepressants Alone — NCT03738215.
• Registry record last updated 2022-10-25; status last verified 2022-09.
• Results first posted 2022-10-25.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

23. Record Summary

The registry record describes a completed, randomized, double-blind, parallel-group phase 3 study of cariprazine added to ongoing antidepressant therapy in participants with major depressive disorder and inadequate response to antidepressants alone. Enrollment was 759. Participants received placebo plus ADT, cariprazine 1.5 mg/day plus ADT, or cariprazine 3 mg/day plus ADT during a 6-week double-blind treatment period.

The primary endpoint was change from baseline to Week 6 in MADRS total score, analyzed in the mITT population defined by at least one postbaseline assessment. The sponsor reports MMRM comparisons against placebo plus ADT. A secondary CGI-S change outcome and adverse-event summaries are also posted. The registry flags differences between recalculated unadjusted estimates and sponsor analyses; the sponsor's own analysis takes precedence. Key details of the original sample-size rationale, multiplicity plan, and missing-data assumptions are not reported in the registry record.