Lurasidone - A 24-week Extension Study of Patients With Bipolar I Depression: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Lurasidone - A 24-week Extension Study of Patients With Bipolar I Depression (NCT00868959), a na phase 3 study. The primary outcome is Number of Participants With Serious and Non-serious Treatment-emergent Adverse Events Who Have Completed 24 Weeks of Extension Study Treatment.

2. Clinical Question

The study is described as a Phase 3 extension for patients with bipolar I depression. Its registry summary says the study was designed to test whether lurasidone is effective, tolerable, and safe. However, the listed primary outcome concerns adverse events among participants completing extension treatment. The registry also lists changes in depression severity on the MADRS and the depression component of the CGI-BP-S as secondary outcomes. Results for those changes are not reported in the registry record.

The extension design can describe events and symptom outcomes during treatment in the enrolled group. It cannot, by itself, separate treatment effects from changes that might have occurred over time or under another treatment. The clinical question answered by the posted adverse-event counts is therefore narrower than whether lurasidone caused improvement or harm compared with an alternative.

3. Trial Design

This was a completed Phase 3, open-label, flexible-dose extension study of lurasidone for bipolar I depression. The registry classifies it as a single-group study, with allocation listed as not applicable and masking listed as none. No randomized or concurrent control group is described. The registry gives enrollment as 817 and lists the study start as 2009-04 and primary completion as 2013-02.

The extension was intended for participants who had reached the six-week endpoint of one of three earlier studies. The registry names those studies as NCT00868452, NCT00868699, and NCT01284517. That entry describes the prior-study endpoint as the point of completion for the relevant participants; further details about selection into the extension are not reported in the registry record. The listed primary outcome timeframe is 24 weeks, and the outcome title refers to participants who completed 24 weeks of extension treatment.

Open-label means the study was not masked, as also reflected in the registry classification. Flexible dosing indicates that treatment was not described as a single fixed dose, but the dose range and rules for changing doses are not reported. There is no reported allocation process because the design is single group.

The registry lists changes from the open-label extension baseline to Week 24, using a Month 6/LOCF endpoint, for MADRS total score and CGI-BP-S depression severity as secondary outcomes. The record lists three outcome measures overall, but does not report the analysis results for those symptom measures. No detailed protocol, analysis population, or fuller account of follow-up and discontinuation is 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: 817 represents the total number of subjects who provided informed consent and enrolled, which is different from the total number of subjects who were treated with study drug, which was 813.

Overall Study

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 - received at least one dose of study medication

6. Primary Endpoint

The registry lists the primary outcome as Number of Participants With Serious and Non-serious Treatment-emergent Adverse Events Who Have Completed 24 Weeks of Extension Study Treatment, measured over 24 weeks.

The primary outcome is the number of participants with serious and non-serious treatment-emergent adverse events who completed 24 weeks of extension treatment. Its unit is participants, and the registry describes the parameter as a number rather than a change in a symptom score. The timeframe is 24 weeks.

Counting affected participants is a direct way to describe how many people in a study group experienced events categorized as serious or non-serious. In general, safety monitoring uses such counts to characterize the kinds of events observed during treatment. A count does not, on its own, establish that treatment caused an event. Nor does a count show how the event burden compares with what would have occurred under another treatment or without treatment.

The registry’s outcome description also refers to participants who completed the six-week endpoint of one of the earlier studies. That context matters when considering who could enter the extension. The posted adverse-event information lists 24 participants in the serious-event category and 246 in the other-event category, with 813 listed at risk in each category. The record does not report a single combined count for the primary outcome, and the counts should not be added to create one.

The registry does not report detailed event definitions, a primary-endpoint analysis method, or a comparison group. These omissions limit what can be concluded from the endpoint beyond the descriptive counts recorded for the extension group.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The design determines which statistical questions the data can answer. A single-group extension follows participants receiving lurasidone without a concurrent control group. It can describe outcomes observed in that group, but it does not estimate a randomized between-group treatment contrast. In general, a control group helps distinguish changes associated with treatment from changes that might occur over time, through participant selection, or under other care. No such comparison is reported for this extension.

The registry classifies the study as open label and unmasked. In general, masking can reduce the influence of expectations on treatment decisions, participant reports, and outcome assessment. An unmasked design does not automatically invalidate descriptive event counts, but it can affect how events are noticed, reported, or assessed. The registry record does not report procedures for independent event adjudication or blinded assessment.

The primary outcome is expressed as a number of participants with serious and non-serious treatment-emergent adverse events. This is a participant-count outcome, not a time-to-event estimate or a symptom-score change. In general, a count answers how many participants met an outcome definition within the specified study context. It is not interchangeable with the total number of event occurrences: one participant can potentially experience multiple events, and the registry does not provide event totals or a rule for combining events. The reported figures should be read in the categories and units the record gives.

A denominator is needed to interpret a count as a proportion or rate. The adverse-event entries list 813 at risk for both categories, while enrollment is listed as 817. The record does not explain the relationship between enrollment and the at-risk entries. No proportion or rate should be inferred without clarity about the relevant population, observation period, and calculation. The serious-event terms are presented as percentages, but the denominator and method used for those term-specific percentages are not reported in the registry record.

In general, a confidence interval describes the range of values compatible with an estimate under a statistical model, and its width conveys precision. The registry entries do not include confidence intervals or other uncertainty measures for the adverse-event counts. A count alone does not show how much sampling uncertainty surrounds an underlying event frequency. It also cannot correct for the lack of a control group.

Over-dispersion is a general issue when observed counts vary more than a chosen statistical model expects. It can make uncertainty estimates from an overly restrictive model too narrow. The primary outcome is reported as affected participants, and the registry does not describe a count model, an over-dispersion assessment, or model-based uncertainty estimates. No claim about such an analysis can be made from the record.

The registry lists three outcome measures and zero analyses. The outcome descriptions identify the two symptom measures as changes from extension baseline to Week 24, with a Month 6/LOCF endpoint. LOCF, in general, means carrying a participant’s last observed value forward to a later endpoint when a later observation is missing. This approach can simplify an endpoint calculation, but it relies on assumptions about the meaning of the carried-forward value. The registry titles alone do not establish exactly how LOCF was applied, which participants were included, or whether any sensitivity analyses were performed.

The registry record does not report a statistical analysis plan, hypothesis-testing procedure, confidence intervals, covariate adjustment, interim analyses, or sample-size justification. It also does not report a multiplicity adjustment plan for the listed outcomes. These details matter because choices about analysis populations, missing data, and multiple outcomes can affect estimates and uncertainty. Their absence from the registry record should not be filled in by assumption.

The sample-size planning illustrations use a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation, with the effect to detect specified as a fraction of the observed effect. Such calculations are planning illustrations, not observed power and not evidence that the extension used those design assumptions. The stated two-arm allocation is unlike the single-group extension design, and no sample-size justification for this trial is reported in the registry record. The illustration therefore should not be read as a reconstruction of the trial’s actual design or as proof that its outcomes were adequately powered.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("Lurasidone")
n      <- c(813)
events <- c(529)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  Lurasidone  813  529  65.1  61.7  68.3  61.7  68.3  61.7  68.3

11. Multiplicity, Missing Data and Analysis Populations

The registry lists three outcome measures and zero analyses. The primary outcome is the participant count for serious and non-serious treatment-emergent adverse events among those completing 24 weeks of extension treatment. The listed secondary outcomes are changes from extension baseline to Week 24 in MADRS total score and CGI-BP-S depression severity, both with a Month 6/LOCF endpoint in their titles.

When a study examines multiple outcomes, multiplicity is a general statistical concern: the more opportunities there are to test hypotheses, the greater the chance of finding an apparently notable result by chance if tests are interpreted separately. Whether adjustment is appropriate depends on the study’s goals and prespecified analysis plan. A multiplicity strategy or hierarchy for outcomes is not reported in the registry record. No adjustment should be assumed.

The outcome titles mention LOCF for the secondary symptom measures, but the registry does not report how missing values were handled in the analyses, whether other approaches were used, or what analysis populations were defined. The complete-case population, intention-to-treat population, and any safety population are not described. The primary outcome’s emphasis on participants completing 24 weeks also leaves the handling of participants who did not complete treatment unclear.

Missingness can affect interpretation when participants with unavailable measurements differ from those with recorded measurements. In general, carrying a last observation forward does not remove that concern; it substitutes an earlier value under an assumption about later status. Without analysis details and results, the consequences of missing data for these outcomes cannot be assessed. Further information on outcome handling and analysis populations is not reported in the registry record.

12. 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: depression (0.7%), depression suicidal (0.2%), mania (0.2%), suicide attempt (0.2%), pancreatitis (0.1%).

The adverse-event entries list 813 participants at risk in each category. The serious-event category lists 24 affected participants, and the other-event category lists 246. These are participant counts, not counts of all event episodes. The registry does not provide a combined participant count, so the category counts should not be added; the same participant could potentially appear in both categories.

The serious-event terms and percentages listed are Depression, 0.7%; Depression Suicidal, 0.2%; Mania, 0.2%; Suicide Attempt, 0.2%; Pancreatitis, 0.1%; and Disease Progression, 0.1%. The denominator for these term percentages is not reported in the registry record. The terms identify events recorded, not proof of a causal relationship to lurasidone.

The record does not report a death count. It also does not supply the detailed definitions, follow-up by individual, or comparative event rates needed to make a controlled safety comparison. The figures are best read as descriptive registry counts from an open-label extension group.

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

The posted results describe adverse events recorded during an open-label, single-group extension of lurasidone treatment. They include participant counts for serious and other adverse events and a list of serious-event terms with registry-reported percentages. The study was completed, and the primary outcome was framed around participants completing 24 weeks of extension treatment.

These data do not establish whether lurasidone caused any particular event. Events occurring during treatment may have multiple explanations, and the absence of a comparator prevents a randomized assessment of how event frequency differed under another condition. The counts also do not describe event duration, severity beyond the serious/non-serious classification, recurrence, or timing in detail.

The registry’s study summary describes an aim to evaluate effectiveness, tolerability, and safety. However, the outcome information available in the record does not include numerical results for the listed changes in MADRS or CGI-BP-S scores. It therefore does not support a quantitative conclusion about change in depression severity or its precision. The adverse-event figures should not be treated as an efficacy result.

The interpretation is also limited by incomplete reporting about the analysis population, missing observations, event definitions, and the relationship between enrollment and the at-risk count. No uncertainty intervals are reported for the event counts. The results can characterize what the registry records for this extension group, but they cannot by themselves determine a comparative treatment effect, establish causation, or resolve how the findings would apply beyond the participants represented.

14. Why This Trial Matters Statistically

This trial matters statistically because it illustrates the distinction between an extension study’s descriptive safety record and a comparative treatment-effect estimate. A single-group, open-label design can document adverse events and follow outcomes during continued treatment. Without a concurrent control group, however, it cannot isolate treatment effects from time trends, selection into the extension, or other influences.

The outcome framing also matters. The primary measure is the number of participants with serious and non-serious treatment-emergent adverse events who completed 24 weeks of extension treatment. The registry reports event-category counts, while the trial summary also refers to effectiveness and lists symptom-score changes as secondary outcomes. Those different aims call for different kinds of evidence. Participant counts describe safety observations; comparative symptom effects require interpretable outcome estimates and a suitable contrast.

The record gives enrollment as 817 and adverse-event at-risk counts as 813. Because the relationship between these entries is not explained, readers should avoid assuming that they refer to identical analysis populations. Similarly, term percentages are reported without a stated denominator or calculation method. Clear population definitions, denominators, analysis methods, and uncertainty estimates are central to interpreting clinical statistics.

Finally, the secondary outcome titles specify an LOCF endpoint, but the analysis details and results are not reported. This is a reminder that an outcome name alone does not reveal how missing data, multiple outcomes, or uncertainty were handled. The registry record supports a cautious descriptive reading, while leaving substantial questions about comparative efficacy and analytical precision unanswered.

15. 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 extension has no randomized comparator and is open label. These design features limit causal interpretation of adverse events and symptom changes. The registry’s stated single-group design means the posted counts cannot show how outcomes would compare under another treatment or under a control condition.

The population is shaped by prior-study participation: the description refers to patients who reached a six-week endpoint in one of three earlier studies. The registry does not report enough detail about eligibility, entry into the extension, or representativeness to assess how closely this group reflects other people with bipolar I depression. Findings should be understood in the context of this extension population rather than assumed to apply broadly.

There is an unexplained difference between the enrollment entry of 817 and the at-risk entries of 813 for the adverse-event categories. The record does not describe the relevant safety population or explain exclusions. The primary outcome’s focus on participants completing 24 weeks also raises an interpretive limitation: outcomes among completers may not reflect outcomes among all participants who started the extension. Reasons for non-completion and their handling are not reported.

Further limitations are the absence of numerical symptom-score results, uncertainty intervals, a detailed analysis plan, and information on event definitions and missing-data handling. The term-specific percentages lack a stated denominator. These gaps prevent assessment of precision and weaken interpretation of the adverse-event terms as frequencies. The registry also does not report whether multiple outcomes were adjusted or prioritized in analysis.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Lurasidone - A 24-week Extension Study of Patients With Bipolar I Depression — NCT00868959.
• Registry record last updated 2016-04-14; status last verified 2016-03.
• Results first posted 2014-04-01.
• 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.

19. Record Summary

The registry record describes a completed Phase 3, open-label, flexible-dose, single-group extension study of lurasidone for bipolar I depression. The study lists enrollment as 817, with a 24-week treatment timeframe. Participants entered after reaching the six-week endpoint of one of three earlier studies.

The primary outcome concerns the number of participants with serious and non-serious treatment-emergent adverse events who completed extension treatment. The adverse-event entries list 813 at risk, with 24 participants in the serious-event category and 246 in the other-event category. Serious-event terms are reported with percentages, but their denominator is not stated.

Three outcome measures and zero analyses are listed. Secondary outcomes concern changes in MADRS and CGI-BP-S depression scores, but their numerical results are not reported in the registry record. The record does not provide a comparator, detailed analysis plan, or enough information to assess missing-data handling and precision. The available results are descriptive and do not establish a comparative treatment effect.