High Dose Lurasidone for Patients With Treatment Resistant Schizophrenia: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for High Dose Lurasidone for Patients With Treatment Resistant Schizophrenia (NCT01569659), a randomized phase 4 study. The primary outcome is Change in Positive Symptoms of Schizophrenia.

2. Clinical Question

This is a question about a measured symptom-scale outcome, not a complete assessment of every outcome mentioned in the study summary. That summary also lists negative symptoms, general psychopathology, anxiety, depression, cognitive function, global function, severity of illness, and tolerability. The registry record posts one outcome measure and no secondary outcome results.

The comparison is between two lurasidone dosing strategies, rather than between lurasidone and an inactive comparator or a different drug. The key estimate is the posted difference in mean change between the high-dose and standard-dose groups. Its confidence interval and unadjusted p-value help describe the uncertainty in that comparison; neither alone establishes what the result means for an individual participant.

3. Trial Design

This completed phase IV study was registered as a randomized, quadruple-masked, parallel-group trial. Northwestern University is listed as the sponsor. The record gives an enrollment of 101 and study dates from 2011-10 to primary completion in 2015-07.

The study summary describes a population with treatment-resistant schizophrenia or schizoaffective disorder, with treatment resistance qualifying after two or more antipsychotic drug trials. The primary outcome population is more specifically described as participants with baseline persistent moderate to severe positive symptoms. The arm descriptions indicate that participants proceeded to randomization only if those symptoms persisted, including after some improvement at 80 mg/day. The registry record does not report further details about the screening or run-in process.

Both randomized groups received lurasidone. The standard-dose arm description says participants continued 80 mg. The high-dose description says participants began titration at 120 mg at week 7 and could increase by 40 mg/day each week to 240 mg/day if adequately tolerated. The available high-dose description ends mid-sentence, so the full criteria used to assess tolerability are not reported in the registry record.

The primary outcome timeframe is baseline to the end of the randomized phase at 24 weeks. The posted endpoint values include 34 participants in the standard-dose group and 33 in the high-dose group. These are the arm sizes reported alongside the primary outcome; they should not be confused with the overall enrollment or with the arm counts described as planned in the study summary. The registry record does not report the original sample-size justification or explain the difference between the enrollment and the posted endpoint group sizes.

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: 101 subject were enrolled in phase 1 for a period of 6 weeks to determine Treatment Resistant Status (TRS). 34 determined NOT TRS so were discontinued. The remaining 67 moved to phase 2. 34 remained in standard dose of Lurasidone and 44 were on High dose Lurasidone.

Overall Study

Analysis population for the primary outcome, as defined in the registry: Participants with baseline persistent moderate to severe positive symptoms

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 in Positive Symptoms of Schizophrenia, measured over Baseline to end of randomized phase (24 weeks).

The primary endpoint is change in positive symptoms of schizophrenia, measured in units on the Positive and Negative Symptom Scale (PANSS), from baseline to the end of the randomized phase at 24 weeks. The reported outcome is a mean with standard deviation. The registry describes a clinical interview and ratings of symptoms, informed by the interview and reports from family members or primary care hospital workers.

The PANSS description says symptoms are rated from 1 to 7, with higher item scores indicating more severe symptoms. The positive-symptom subscale contains 7 items and ranges from 7 to 49. The full scale and its other symptom categories are described in the record, but the posted primary outcome is specifically the positive-symptom measure. The record does not state how the change score was calculated or which sign convention was used for change. Therefore, the posted mean changes can be compared between groups, but the direction of change should not be interpreted as improvement or worsening without that convention.

A symptom scale is used to represent the measured severity of positive symptoms in a consistent way across participants and groups. Comparing mean change summarizes how the groups differed on that measure over the stated timeframe. It does not by itself describe every participant’s experience, all aspects of psychopathology, or tolerability. The registry record posts no secondary outcome results, so the endpoint interpretation is limited to the reported primary measure.

7. Results Posted in the Registry

• High Dose of Lurasidone: mean 24.20 versus 22.60 on control; difference 1.60 (95% CI -0.24 to 3.44), standardized effect d = 0.42; z-test p = 0.088.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is continuous: it is a numerical change score, summarized by a mean and standard deviation. In general, a group mean gives the average of the measured values, while the standard deviation describes how dispersed individual values are around that average. The posted values are mean changes of 22.6 for standard-dose lurasidone and 24.2 for high-dose lurasidone, with standard deviations of 3.3 and 4.3, respectively. These summaries do not show the distribution of individual scores or how many participants had changes in a particular direction.

The comparison is expressed as a mean difference, with the high-dose group compared against the standard-dose group. The posted estimate is 1.6. In general, a difference in means puts the contrast on the same scale as the outcome, which is units on the PANSS positive-symptom measure. The estimate is a summary of the observed group contrast; it is not a statement that every participant differed by that amount.

A confidence interval describes a range of values compatible with the data under the assumptions of the analysis method. The posted interval extends from -0.239 to 3.439. Because it includes zero, it allows for the possibility of no mean difference, as well as differences on either side of zero. The registry record does not state the confidence level or the exact method used to produce this interval, so those details should not be inferred.

The posted unadjusted p-value is 0.088. In general, a p-value describes how unusual data at least as discrepant as those observed would be under a specified null model and analysis procedure. It is not the probability that the null hypothesis is true, and it does not measure the size or importance of a difference. The record does not report the precise test or model behind the p-value. It is therefore appropriate to report it as an unadjusted value and not assume covariate adjustment, a particular distributional model, or an analysis population that is not described.

In general, random allocation can support a fair comparison of groups by reducing systematic differences at assignment, while masking is intended to limit the influence of knowledge about treatment assignment. This study is listed as randomized and quadruple-masked, but the registry record does not specify which roles were masked or describe the allocation sequence and concealment procedures. Those implementation details are not reported in the registry record.

Variation matters for a continuous outcome because more dispersed observations generally make a group contrast less precise, all else being equal. In general, over-dispersion means that observed variability is greater than assumed by a model; if that is not accounted for, standard errors and uncertainty estimates may be unreliable. The posted group standard deviations describe observed spread, but the registry record does not report an over-dispersion assessment or a model-based variance diagnostic. There is no basis to say that over-dispersion was present or how it was handled.

There is one posted outcome measure and no posted analysis count. The registry record does not report the detailed statistical analysis plan, any interim analyses, handling of missing outcome values, or whether the primary analysis used an intention-to-treat or another population. It also does not report adjustment for baseline scores or other covariates. These omissions mean that the posted estimate can be described, but the analysis cannot be fully reconstructed from the registry record.

The record’s planning illustration uses the observed effect as its starting point and specifies a two-sided significance level of 0.05, equal allocation, and a normal approximation. It considers detecting the stated fractions of the observed effect. These are planning assumptions for the illustration, not evidence about the original design. The trial’s own sample-size justification and original analysis assumptions are not reported in the registry record. The planning counts are not observed power and should not be used to reinterpret the trial’s result as if additional participants had been enrolled.

10. Confidence Intervals and Precision

The mean difference is reported as 1.6, with a confidence interval from -0.239 to 3.439. The interval spans zero and extends to values on either side. In practical statistical terms, the posted estimate does not sharply distinguish a zero difference from a range of possible group differences. It indicates uncertainty about the size and direction of the contrast on the PANSS positive-symptom change scale.

Precision is also informed by the endpoint group sizes and variation. The reported sizes are 34 for standard-dose lurasidone and 33 for high-dose lurasidone; the standard deviations are 3.3 and 4.3. A confidence interval combines information about the estimated contrast and its uncertainty. It should not be confused with the range containing most individual participants’ scores, which would require a different summary.

The confidence level and calculation method for the interval are not reported in the registry record. The p-value is given as unadjusted, but the record does not describe the precise test or model. Those gaps limit how specifically the uncertainty can be characterized. In particular, the interval should be read as posted rather than as proof of a particular underlying analysis method.

The width of an interval is not itself a measure of clinical importance. Interpretation also depends on the meaning of the scale, the change-score convention, and what magnitude of difference would matter in context; those contextual thresholds are not reported in the registry record. The interval therefore communicates statistical uncertainty about the group contrast, not a treatment recommendation or a definitive judgment about the importance of the observed difference.

11. 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 sample-size table and curve are planning illustrations based on the observed mean difference. The counts are participants per arm, not total events. The assumptions listed for this illustration 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 to the stated fraction of the observed effect. These assumptions describe the illustration only; the trial’s original design assumptions and sample-size justification are not reported in the registry record.

The table gives separate columns for 80% and 90% power. In plain language, those columns show the participants per arm that the illustration estimates would be needed for an 80% or 90% chance of detecting the specified effect under its assumptions, if that effect were present. The rows consider planning for 100%, 75%, and 50% of the observed effect. For example, planning for 75% of the observed effect gives 160 participants per arm for 80% power and 214 participants per arm for 90% power. The corresponding values for the other rows are shown in the table.

Smaller effects are harder to distinguish from random variation, so the illustration requires more participants per arm when it plans for a smaller fraction of the observed effect. The curve visualizes this relationship; it does not add a separate observed result. These figures are not observed power, and they do not say how much power the completed trial actually had. They are hypothetical planning values, not the trial’s actual enrollment target or an explanation of its design.

12. 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 = 1.600; pooled standard deviation σ = 3.825.
2. n = 2 × (1.96 + 0.842)² × 3.825² / 1.600² = 89.7, rounded up to 90 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 <- 1.6                      # observed difference in means
sigma <- 3.82511940487333                     # 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(1569659)
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 = 90 (80% power), 121 (90% power)
 75% of observed effect: n per arm = 160 (80% power), 214 (90% power)
 50% of observed effect: n per arm = 359 (80% power), 481 (90% power)
power.t.test: 90.7 
Simulated power at n = 90 per arm: 0.798 (20000 replicates)

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

13. Multiplicity, Missing Data and Analysis Populations

The registry record posts one outcome measure, the primary change in positive symptoms, and reports no secondary outcome results. The study summary mentions additional symptom, function, cognitive, and tolerability domains, but results for those outcomes are not posted in this record. Their absence limits the range of questions that can be answered from the available results.

Multiplicity arises when analyses test several outcomes, time points, subgroups, or contrasts: the chance of at least one apparently notable result can increase as more comparisons are made. In general, methods such as adjusted significance thresholds or hierarchical testing can address this issue. The registry record does not report whether the trial’s analysis plan included multiplicity adjustment, a hierarchy of outcomes, or a strategy for interpreting multiple tests. No adjustment should be assumed from the single unadjusted p-value.

The record lists no analysis count and does not describe any interim analyses. It also does not report how missing endpoint data were handled, how many participants lacked outcome measurements, or whether any imputation or sensitivity analyses were used. Those details matter because analysis results can depend on who is included and how unavailable measurements are treated.

The registry record does not specify whether the primary analysis followed an intention-to-treat principle, used a per-protocol population, or applied another definition. The endpoint values are accompanied by group sizes, but those counts alone do not establish which participants were included in the analysis or whether all randomized participants contributed data. The analysis population, missing-data procedures, and multiplicity strategy are therefore not reported in the registry record. These limitations affect interpretation of the posted estimate and its uncertainty, but they do not justify guessing what procedures the investigators used.

14. 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: death (3.0%).

The registry record lists serious adverse events affecting 0 participants in the standard-dose group and 1 participant in the high-dose group. The group sizes at risk are reported as 34 and 33, respectively. It also lists other adverse events affecting 28 participants in the standard-dose group and 30 in the high-dose group. These are counts, not a full comparison of event types, severity, timing, or causal relationship to treatment.

The record’s top serious-event terms include “Death” at 3.0%. However, the group-specific death fields are not populated with counts, and the registry record does not provide the context needed to reconcile the term and percentage with the group-level fields. The number of deaths in each group is not reported in the registry record. The term and percentage should therefore be repeated as listed, not converted into an assumed count or group-specific rate.

These summaries do not establish whether an event was caused by treatment, and the record does not provide the detail needed to assess comparative safety comprehensively. They describe what was posted, not a conclusion about the safety of either dosing strategy.

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

The result is a registry-reported comparison of mean change in PANSS positive-symptom scores over the randomized phase. The high-dose group’s posted mean change was 24.2, compared with 22.6 for the standard-dose group, and the reported high-dose-minus-standard-dose mean difference was 1.6. The confidence interval, from -0.239 to 3.439, includes zero; the unadjusted p-value is 0.088. Together, these figures show an uncertain estimate of the between-group difference in this endpoint.

The registry record does not explain the sign convention for change, so the values should not be described as symptom improvement or worsening. Although the PANSS description says that higher ratings indicate more severe symptoms, interpreting a change score also requires knowing whether positive change represents an increase or decrease in the score. That convention is not reported in the registry record.

The result does not show that the groups were identical. Rather, the interval leaves room for different underlying contrasts, including no difference. Nor does the result establish that a clinically meaningful difference was or was not present: the record does not report a threshold for clinical importance. The observed group averages are not individual-level response descriptions.

The findings also do not answer every question raised by the study summary. Secondary outcomes are not posted, details of missing-data handling and analysis populations are not reported, and the serious adverse event summary is limited. The trial compared two lurasidone dosing strategies in the stated population; it did not provide a comparison with another medication or an inactive comparator in the results. These boundaries matter when summarizing what the registry results can support.

16. Why This Trial Matters Statistically

This trial is statistically interesting because it compares two dosing strategies for the same drug in a selected population with persistent positive symptoms. The randomized, quadruple-masked, parallel design is important context for understanding the intended comparison, while the posted outcome is a continuous symptom-scale change rather than a simple event count.

The result illustrates why an estimate, confidence interval, and p-value should be read together. The estimate summarizes the observed group contrast, while the interval shows that the contrast is imprecisely determined in the posted results. The unadjusted p-value is another summary under a statistical testing framework, not a measure of effect size or clinical importance. None of these quantities replaces the need to understand the scale and its change convention.

The record also makes clear the distinction between overall enrollment and the number of participants represented in the primary outcome summaries. It reports enrollment of 101, alongside endpoint group sizes of 34 and 33. Without registry entries on participant flow and analysis populations, readers cannot reconstruct how those figures relate to randomization and outcome availability.

Finally, the sample-size illustration can help readers see how planning depends on the effect size one wishes to detect, but it is not a retrospective measure of trial success or failure. The original design justification, analysis plan, and several reporting details are not available in the registry record. For statistical readers, the value of this example lies both in the posted comparison and in recognizing where the record does not permit stronger conclusions.

17. 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.

Several limits are specific to the information posted for this trial. The primary endpoint is described as change from baseline, but the direction convention for that change is not stated. As a result, the mean values and difference should not be translated into improvement or worsening. The endpoint is also a positive-symptom subscale, so it does not represent all the clinical domains listed in the study summary.

The outcome summaries include 34 participants in the standard-dose group and 33 in the high-dose group, while the overall enrollment is 101. The record does not give a participant-flow account explaining these figures or state which analysis population produced the endpoint summaries. Missing endpoint data and the methods used to address them are not reported in the registry record.

The description of high-dose titration is incomplete: it ends mid-sentence after referring to an assessment of psychopathology. The full titration and tolerability criteria are therefore not available. The registry record also does not report the original sample-size justification, detailed statistical analysis plan, interim analyses, or multiplicity procedures.

Only one outcome measure is posted, with no secondary results. The confidence interval’s level and calculation method are not specified, and the unadjusted p-value is not accompanied by a detailed test description. Finally, the serious adverse event fields provide limited counts, while the serious-event term “Death” at 3.0% is not paired with group-specific death counts. These gaps constrain interpretation of both efficacy-related outcomes and safety reporting.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: High Dose Lurasidone for Patients With Treatment Resistant Schizophrenia — NCT01569659.
• Registry record last updated 2025-02-28; status last verified 2025-02.
• Results first posted 2025-02-28.
• 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.

21. Record Summary

This completed phase IV randomized, quadruple-masked, parallel-group study compared standard-dose and high-dose lurasidone in participants with treatment-resistant schizophrenia or schizoaffective disorder and persistent moderate to severe positive symptoms. The primary endpoint was mean change in PANSS positive symptoms from baseline to the end of the randomized phase at 24 weeks.

The posted mean changes were 22.6 for standard dose and 24.2 for high dose. The reported mean difference was 1.6, with a confidence interval from -0.239 to 3.439 and an unadjusted p-value of 0.088. The record posts no secondary outcome results. Important details, including the change-score convention, missing-data handling, analysis population, and original sample-size justification, are not reported in the registry record.