Suvorexant and Trauma Related Insomnia: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Suvorexant and Trauma Related Insomnia (NCT02704754), a randomized phase 4 study. The primary outcome is Change in Insomnia Severity Index Score From Baseline..

2. Clinical Question

The primary measure is a patient-level score intended to capture the severity of insomnia over the preceding two weeks. The registry describes a seven-item scale, with a total score from 0 to 28 and higher scores indicating greater severity. The key comparison is between the groups’ mean changes, rather than whether either group changed on its own. Comparing groups is important because changes during a trial can also reflect factors that affect participants in both arms.

This is a randomized, placebo-controlled study of a drug for insomnia associated with trauma exposure. The posted results include 18 participants in the suvorexant group and 19 in the placebo group for the primary outcome. They show a lower mean change in the suvorexant group, but the recalculated difference is imprecise and its confidence interval includes values on either side of no difference. The registry also reports a sponsor-conducted two-sided t-test with p <.05. Because the posted sponsor result and the unadjusted recalculation do not tell the same inferential story, both need to be described distinctly rather than treated as interchangeable analyses.

3. Trial Design

The registry describes a completed phase 4, randomized, triple-masked, parallel-group trial sponsored by Howard University. Enrollment is listed as 41. The trial start is recorded as 2016-05-01, and primary completion as 2021-04-30. The summary describes a six week placebo-controlled evaluation of suvorexant for insomnia that developed in relation to trauma exposure. It also says that polysomnography was used to identify biomarkers of response.

Participants were assigned to suvorexant or placebo. The placebo arm description specifies a pill without active ingredients and says randomization was 1:1, with stratification for gender and PTSD status. The suvorexant arm began with 10mg before bedtime during the first week, with an increase to 20 mg before bedtime if well tolerated. The registry does not explain how tolerance was assessed, how many participants had a dose increase, or how treatment adherence was measured.

Triple masking indicates that the registry classifies the trial as masked at three levels, but it does not identify the people or roles covered by that label. The exact masking procedures are not reported in the registry record. Nor does the record give details of sequence generation, allocation concealment, or the intended analysis population. These omissions limit what can be said about implementation without making assumptions beyond the registry description.

The primary endpoint is the change in Insomnia Severity Index score from baseline to six weeks or the last observation. Scores were also obtained at two and four weeks; the registry says the mathematical mean for the last observation was five weeks. The primary outcome counts are 18 for suvorexant and 19 for placebo. A separate polysomnographic outcome is posted for 13 participants per group, and its timeframe is baseline values minus the values at two weeks. The reason for the smaller number contributing to that measure is not reported in the registry record.

The record lists two secondary outcomes: change in Clinician Administered PTSD Scale score and polysomnographically measured Wake After Sleep Onset. It does not describe the study’s original sample-size justification, its planned statistical analysis in detail, or any interim analyses. The listed completion status is completed; that status by itself does not establish how every planned analysis was carried out.

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.

Overall Study

Reasons for not completing this period:

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 Insomnia Severity Index Score From Baseline., measured over Baseline score minus 6 weeks or last observation (measure was also obtained at 2 and 4 weeks, the mathematical mean for the "last observation" was 5 weeks)..

The primary endpoint is change in Insomnia Severity Index score from baseline to six weeks or the last observation. The registry defines the change as baseline score minus the later score. This direction matters: a more positive change represents a greater decrease in the score, while the scale itself assigns higher scores to greater insomnia severity.

The Insomnia Severity Index is described as a seven-item measure of insomnia severity over the preceding two weeks. Its items are scored on a five-point scale and summed to produce a total from 0 to 28. Using a standardized score gives the investigators a way to summarize participants’ reported insomnia symptoms in a common measure and compare the average change between randomized groups.

A change endpoint can account for where participants began, while the randomized group comparison helps distinguish the pattern in the suvorexant arm from changes also seen with placebo. The registry posts group means and standard deviations for the change measure. It does not provide baseline score summaries, individual participant data, or a detailed account of how observations were handled when a six-week score was unavailable.

In the posted primary results, the mean change is 6.8 units in the suvorexant group and 8.6 units in the placebo group. Given the baseline-minus-follow-up coding, the group means describe reductions in severity in both arms, with a smaller average reduction in the suvorexant arm. These summaries do not alone establish whether the between-group difference is distinguishable from chance variation; that requires an inferential comparison and attention to its uncertainty.

7. Results Posted in the Registry

• Suvorexant: mean 6.80 versus 8.60 on control; difference -1.80 (95% CI -5.49 to 1.89), standardized effect d = -0.31; z-test p = 0.339.

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 in Clinician Administered PTSD Scale 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 deviations.
• Result:
  • Suvorexant: mean 4.20 versus 3.80 on control; difference 0.40 (95% CI -3.33 to 4.13), standardized effect d = 0.07; z-test p = 0.834.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Suvorexant and the control. The interval gives the range of effects that remain compatible with the data.

Polysomnographically Measured Wake After Sleep Onset

• 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 deviations.
• Result:
  • Suvorexant: mean 21.00 versus 22.30 on control; difference -1.30 (95% CI -12.31 to 9.71), standardized effect d = -0.09; z-test p = 0.817.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Suvorexant and the control. The interval gives the range of effects that remain compatible with the data.

The two posted secondary outcomes cover different aspects of the study: change in Clinician Administered PTSD Scale score and polysomnographically measured Wake After Sleep Onset. The point estimates and recalculated intervals are available in the results tables. Across these outcomes, the recalculated intervals include no difference, and the registry lists no sponsor-reported analyses for either secondary outcome.

Taken together, the secondary results do not provide a clear, consistent pattern of separation between the randomized groups in the posted unadjusted comparisons. They also do not resolve the difference between the sponsor’s reported primary test and the primary summary-statistic recalculation. Secondary outcomes can offer context about whether findings across measures appear aligned, but they are not substitutes for the prespecified primary question.

Unadjusted secondary results generally carry less confirmatory weight when several outcomes are examined. Looking at multiple outcomes creates more opportunities for apparently notable results to arise by chance. A multiplicity plan can control or otherwise address this issue, but such a plan is not reported in the registry record. It is therefore not possible to establish from the record whether the secondary analyses were intended as confirmatory, how their testing was prioritized, or whether any adjustment was planned.

The secondary measures also differ in their timing and data availability. The PTSD score uses the baseline-to-six-week or last-observation timeframe described in the record, while Wake After Sleep Onset is based on baseline values minus values at two weeks and is posted for 13 participants per group. Differences in construct, measurement method, timing, and available observations mean that the outcomes should not be treated as interchangeable replications of the primary endpoint.

The sensible overall reading is descriptive: the tables show the direction and uncertainty of each posted comparison, but the secondary results do not independently establish a treatment effect. They can help identify questions for further study, while conclusions about the primary endpoint should remain anchored in the primary comparison and the limitations of its reported analyses.

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

• Change in Clinician Administered PTSD Scale Score
• Polysomnographically Measured Wake After Sleep Onset

9. Statistical Interpretation

10. Statistical Methods Explained

The primary measure is continuous: each participant has a numerical change score, and the posted group summaries describe means and standard deviations. A mean gives the arithmetic average of the observed scores in a group. A standard deviation describes how dispersed individual scores are around that group mean. It is not the uncertainty in the mean itself; uncertainty in a group comparison also depends on the number of participants contributing data.

What does a mean difference represent?

The comparison is framed as the mean change in the suvorexant group minus the mean change in the placebo group. The recalculated estimate is -1.8 units. Because the endpoint is baseline minus follow-up, the negative sign indicates that the average change was smaller in the suvorexant group than in the placebo group. This is a contrast between group averages, not a change from baseline within one group and not a statement about every participant.

The posted recalculation is an unadjusted comparison based on summary statistics. It does not adjust for gender, PTSD status, baseline score, or any other characteristic. Although the registry says randomization was stratified for gender and PTSD status, the posted recalculation does not incorporate those factors. Whether the sponsor’s analysis used adjustment or other modeling choices is not reported in the registry record.

Why is a confidence interval useful?

A confidence interval gives a range of estimates compatible with the data under the method used. For the primary recalculation, the interval runs from -5.49 to 1.89. It includes zero and values in both directions. This communicates more than a single point estimate: the point estimate is -1.8, but the data do not pin the difference down closely. The registry does not state the confidence level or all construction details for this interval, so those details should not be inferred.

A confidence interval is not a list of the effects that are equally likely, nor does it describe the range of individual participant outcomes. It concerns uncertainty in the estimated group contrast under a statistical procedure. The individual outcome spread is reflected separately in the group standard deviations. When sample sizes are modest and outcomes vary, the estimated contrast can remain uncertain even when the group means differ.

How should the t-test result be read?

The registry’s reported analysis is described as a two-sided t-test for superiority and gives p <.05. A two-sided test evaluates evidence against a null hypothesis of no mean difference in either direction; it does not restrict attention to a prespecified favorable direction. A p-value describes how unusual data at least as incompatible with that null would be under the test’s assumptions if the null were true. It is not the probability that the null hypothesis is true, nor does it measure the size or clinical importance of an effect.

The sponsor’s posted result lacks a test statistic, estimate, interval, and analysis details. A separate unadjusted recalculation from the posted summary statistics gives p = 0.339. Because those results differ, it is important not to merge them into a single conclusion. The registry record does not explain whether the sponsor’s analysis used different data, a different model, or another procedure. Such explanations would be speculation.

What do the standard deviations and sample sizes contribute?

The primary arm summaries are 6.8 with standard deviation 5.0 for 18 participants, and 8.6 with standard deviation 6.4 for 19 participants. The standard deviations describe variability of individual change scores. The reported standard errors in the comparison summaries quantify uncertainty in the respective group means, not variation among individual participants. They are inputs to the recalculated contrast and its interval.

A t-test is a method for comparing means while accounting for estimated variability and sample size. Its use does not guarantee that all assumptions are appropriate; the registry does not report diagnostic checks, distributional assessments, or a plan for handling influential observations. With small groups, estimates can be sensitive to individual observations, and summary-level data do not allow that sensitivity to be examined.

What is meant by over-dispersion?

For continuous outcomes, greater variability means participant scores are more spread out relative to the size of the group difference. This makes the mean difference less precisely estimated. In other settings, such as counts, over-dispersion means the observed variability exceeds what a simple model assumes; a model that ignores it can understate uncertainty. That general issue illustrates why the outcome’s distribution and analysis model matter. The registry gives standard deviations for these continuous outcomes but does not report a special model for over-dispersion, and one should not assume that such a model was needed or used.

The calculations labelled as recalculated are simple unadjusted re-calculations from the posted summary statistics. They are useful as a transparent check of the descriptive contrast, but they are not a reconstruction of an undocumented sponsor analysis. The posted record does not give enough detail to reproduce the sponsor’s test exactly or reconcile its p-value with the unadjusted result.

11. Confidence Intervals and Precision

The recalculated primary mean difference is -1.8 units, with a confidence interval from -5.49 to 1.89. The interval is wide enough to include zero and estimates on both sides of zero. This means the direction and magnitude of the underlying mean difference are not determined precisely by this recalculation. In particular, the observed point estimate should not be read as if it were the only effect compatible with the data.

The interval’s endpoints are expressed in the same units as the change score. They describe uncertainty in the between-group average contrast, not the spread of participant responses. The standard deviations of 5.0 and 6.4 describe that participant-level spread in the two arms. The registry does not report a confidence level for the posted or recalculated interval, so this page does not attach a conventional coverage percentage to it.

Precision is affected by the amount of outcome variation and the number of participants observed. The primary outcome is reported for 18 participants receiving suvorexant and 19 receiving placebo. The standard errors in the recalculation are 1.179 for the suvorexant mean and 1.468 for the placebo mean. They help convey why the group means do not translate into a narrowly estimated contrast. The posted group summaries do not include individual data, which would be needed to inspect the distribution or evaluate sensitivity to particular observations.

The sponsor’s reported two-sided t-test gives p <.05, whereas the simple unadjusted recalculation gives p = 0.339. Since the sponsor’s estimate and interval are not posted, its result cannot be assessed for precision from the registry record. The reported p-value alone does not show the range of effect sizes compatible with that analysis. The uncertainty displayed by the recalculation should therefore be read as belonging to that calculation, not silently substituted for an unavailable sponsor confidence interval.

Neither a confidence interval including zero nor a non-significant recalculated p-value establishes that there is no effect. Conversely, a sponsor-reported p-value below .05 does not by itself indicate that the effect is large or precisely estimated. Assessing practical importance would also require a clearly stated threshold for meaningful change, which is not reported in the registry record.

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 sample-size table is a planning illustration based on the observed effect, not the trial’s original design calculation. For these continuous outcomes, the entries are participants per arm, not a combined total and not total events. The stated planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is a stated proportion of the observed effect. The trial’s own original sample-size assumptions and justification are not reported in the registry record.

The 80% power column describes the number of participants per arm that the illustration estimates would be needed for an 80% chance of detecting the specified effect under the stated assumptions, if that effect were the true one and the planning model applied. The 90% power column uses a higher target chance of detection, so it requires more participants per arm for the same assumed effect. These are prospective planning quantities, not descriptions of what happened in this trial.

The rows consider planning for 100%, 75%, and 50% of the observed effect. The effect used becomes smaller down those rows, and smaller effects need more participants to detect because they are harder to distinguish from random variation. The effect observed in a small trial can itself be unstable, so planning directly from it is only an illustration; the table does not establish a reliable target effect for a future study.

The figures are not observed power. Observed power is not a useful substitute for interpreting the estimate and its uncertainty after results are known. Here, the planning table helps show how sample requirements depend on an assumed effect size and chosen power target. It does not say that the trial was designed around these numbers, nor does it recover the sponsor’s original protocol rationale.

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 = -1.800; pooled standard deviation σ = 5.763.
2. n = 2 × (1.96 + 0.842)² × 5.763² / -1.800² = 160.9, rounded up to 161 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 <- -1.8                      # observed difference in means
sigma <- 5.76263828467482                     # 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(2704754)
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 = 161 (80% power), 216 (90% power)
 75% of observed effect: n per arm = 287 (80% power), 383 (90% power)
 50% of observed effect: n per arm = 644 (80% power), 862 (90% power)
power.t.test: 161.9 
Simulated power at n = 161 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 that the sponsor used a two-sided t-test for superiority and gives p <.05. It does not report the sponsor’s estimated difference, test statistic, confidence interval, or enough detail about the analysis to reproduce it. The sponsor result therefore cannot be fully evaluated from the posted information alone.

The simple unadjusted recalculation from the posted primary outcome summaries gives a mean difference of -1.8 units, a confidence interval from -5.49 to 1.89, and p = 0.339. This does not match the sponsor’s reported p-value. The record does not explain the difference, and it would be inappropriate to guess whether it reflects a different model, analysis population, data handling, or another choice.

The sponsor’s finding should be described as the reported analysis, while the recalculation should be described as an independent unadjusted calculation from the posted group summaries. They answer related questions but are not interchangeable results. Without the sponsor’s full analysis details, readers cannot determine how the sponsor’s procedure relates to the summary-statistic recalculation. The registry record does not identify a reason for the contrast between them.

15. Multiplicity, Missing Data and Analysis Populations

The registry posts one primary endpoint and two secondary outcomes. It does not report a multiplicity strategy, such as a hierarchy for testing outcomes or an adjustment to account for multiple comparisons. Without this information, readers cannot tell whether the secondary outcomes were intended to support confirmatory claims or to provide exploratory context. Unadjusted p-values across several outcomes should be interpreted cautiously because repeated testing can increase the chance of a nominally notable finding.

The primary outcome timeframe allows a score at six weeks or the last observation, and the record says the measure was also obtained at two and four weeks. It describes the mathematical mean for the last observation as five weeks. However, the detailed procedure for identifying a participant’s last observation, handling missing scores, or analyzing incomplete follow-up is not reported in the registry record. The label last observation does not by itself establish a particular imputation or missing-data method.

The record also does not report how many participants were randomized, treated, or included in each analysis beyond the outcome-specific group counts shown. The registry lists enrollment of 41, while the primary analysis summaries include 18 in the suvorexant arm and 19 in the placebo arm. It does not explain why those counts differ from the total enrollment, or whether the analysis followed an intention-to-treat, per-protocol, or other population definition. No reasons for missing data are reported.

For the polysomnographic outcome, 13 participants per group contribute to the posted comparison. The registry does not state why that outcome has fewer observations than the primary endpoint, or whether missingness could be related to participant characteristics or outcomes. These details matter because analyses based only on observed measurements can be affected if the availability of data is systematically related to the results.

Other analysis details are also absent, including the sponsor’s full statistical model, any covariate adjustment, diagnostics, sensitivity analyses, interim analyses, and the original sample-size justification. These are not evidence that a procedure was or was not used; they are simply not reported in the registry record. The available results support transparent description of the posted summaries and stated test, but not reconstruction of a complete analysis plan.

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 0% reporting threshold.

Most frequent non-serious adverse event terms

The registry lists serious adverse events for 18 participants at risk in the suvorexant group and 19 in the placebo group. It records zero serious adverse events in each group. It also records zero deaths in each group. These are the serious-event counts posted for the trial; they do not provide a complete account of all possible safety outcomes.

For other adverse events, the record lists one affected participant among 18 at risk in the suvorexant group and zero among 19 at risk in the placebo group. The registry does not give a term for the event in the posted summary. The serious adverse event terms list is empty. These counts should be read as what was recorded in this trial, not as a comparison capable of establishing a reliable difference in safety between groups.

With a small number of participants, the absence of a recorded serious event does not show that such events cannot occur. The registry does not report event timing, severity details for the other adverse event, whether events were judged related to study treatment, or a broader safety analysis. The counts are therefore descriptive and limited to the information posted.

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

The results describe a randomized comparison of suvorexant and placebo in a trial of insomnia associated with trauma exposure. The primary outcome is the change in Insomnia Severity Index score, coded as baseline minus follow-up. The posted group means are 6.8 for suvorexant and 8.6 for placebo. The simple recalculation estimates a mean difference of -1.8 units, with an interval that includes zero and spans estimates in both directions.

This means the recalculated result is uncertain: it does not precisely establish the size or direction of the underlying between-group difference. It also does not establish that the treatments are equivalent or that no meaningful difference exists. The registry does not report a threshold for meaningful change, and the recalculated interval cannot be judged against one.

The sponsor’s reported two-sided t-test gives p <.05, while the recalculation from the posted summary statistics gives p = 0.339. Since the sponsor’s estimate, interval, and analysis details are not reported, the record does not allow a full reconciliation. The reported test should not be silently replaced by the recalculation, and the recalculation should not be presented as if it were the sponsor’s analysis. The results should be read with this distinction explicit.

The study’s size and outcome variability limit precision, and the registry leaves important analysis details unspecified. The secondary outcomes offer additional descriptive information but do not resolve these limitations; no sponsor analyses are posted for them, and a multiplicity plan is not reported. Likewise, the adverse-event counts are small and descriptive.

These findings do not establish whether suvorexant should be used by any person, and they are not medical advice. They describe the posted results for this particular trial. The record supports discussion of the measured group changes and the uncertainty surrounding them, but it does not support a broad conclusion beyond the study population and its reported data.

18. Why This Trial Matters Statistically

This trial is statistically instructive because the registry presents two different forms of evidence about the primary comparison: a sponsor-reported t-test and a simple recalculation from group summaries. The sponsor reports p <.05, whereas the unadjusted recalculation gives p = 0.339 and a confidence interval that includes zero. Showing both makes clear why an isolated p-value cannot stand in for an effect estimate, interval, or transparent analysis description.

The study also illustrates the distinction between observed group means and the uncertainty around their difference. The posted means are not themselves evidence that the groups differ reliably. The recalculated interval is broad relative to the point estimate and includes estimates in both directions. Small group sizes and variability in change scores make precision especially important when describing a trial result.

There is also a practical lesson in the secondary outcomes. Multiple endpoints can describe different dimensions of a condition, but unadjusted analyses across them do not automatically provide confirmatory evidence. Without a reported multiplicity plan or analysis hierarchy, readers need to distinguish the primary question from secondary, more exploratory information.

Finally, the planning illustrations show how sample requirements depend on the assumed effect and desired power. They are not the study’s original design rationale and are not observed power. Treating them as planning scenarios rather than retrospective proof of adequacy helps keep the interpretation tied to what the registry actually reports.

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 primary result has limited precision. The posted primary summaries include 18 participants in the suvorexant group and 19 in the placebo group, and the recalculated interval spans values below and above zero. The registry does not provide a threshold for clinically important change, so the observed difference cannot be judged against such a benchmark.

The sponsor’s reported t-test and the simple unadjusted recalculation yield different p-value descriptions. The sponsor’s estimate, interval, test statistic, and detailed model are not reported, and the registry does not explain the contrast. This prevents independent reproduction of the sponsor analysis and limits interpretation of how its result relates to the posted summaries.

Analysis populations and missing-data procedures are not described. The primary outcome counts are lower than overall enrollment, and the polysomnographic outcome has 13 participants per group, but the reasons are not reported. The record does not explain how missing observations were handled or whether the analyses followed an intention-to-treat definition.

The registry also does not report the original sample-size justification, the full statistical analysis plan, multiplicity procedures, interim analyses, or sensitivity analyses. Though randomization was stratified by gender and PTSD status, the posted recalculation is unadjusted, and the sponsor’s handling of those factors is not stated. The trial is described as triple-masked, but the roles masked and procedures are not detailed.

These limitations are specific to the information available for this trial record. They do not establish that a particular method was omitted in practice; they mean that the record does not allow readers to verify it. The secondary outcomes and adverse-event counts are also descriptive and based on limited posted information.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Suvorexant and Trauma Related Insomnia — NCT02704754.
• Registry record last updated 2023-02-15; status last verified 2023-01.
• Results first posted 2023-02-15.
• 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 describes a completed phase 4, randomized, triple-masked, parallel-group trial sponsored by Howard University, with enrollment of 41. It compared suvorexant with a placebo pill for insomnia related to trauma exposure. The primary endpoint was change in Insomnia Severity Index score from baseline to six weeks or the last observation, coded as baseline minus follow-up.

For the primary outcome, the posted mean changes were 6.8 among 18 participants receiving suvorexant and 8.6 among 19 receiving placebo. The simple unadjusted recalculation gives a mean difference of -1.8, a confidence interval from -5.49 to 1.89, and p = 0.339. The registry separately reports a sponsor-conducted two-sided t-test with p <.05, without posting its estimate or interval.

Two secondary outcomes and descriptive adverse-event counts are also posted. The registry does not report the original sample-size justification, a multiplicity plan, detailed missing-data methods, or enough statistical detail to reconcile the sponsor test with the recalculation. The results therefore require careful separation of the sponsor’s reported analysis from the uncertainty shown by the summary-statistic calculation.