A Pilot Study of Suvorexant for Insomnia in Parkinson Disease: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Pilot Study of Suvorexant for Insomnia in Parkinson Disease (NCT02729714), a randomized phase 4 study. The primary outcome is Change in Sleep Efficiency as Measured by Polysomnogram.

2. Clinical Question

3. Trial Design

This was a randomized, double-blind, placebo-controlled pilot study with a crossover model. The registry identifies the study as Phase 4 and reports enrollment of 21. Its brief summary anticipated a total of 20 participants, aged 30 to 80 years, with Parkinson disease and symptoms of insomnia. The study status is completed. The registry lists a start of 2016-04 and primary completion of 2022-04.

Participants were randomized 1:1 to begin with suvorexant or matching placebo. The first treatment period was 4 weeks, followed by a 2 week washout and crossover to the alternate treatment. The suvorexant description gives a bedtime dose of 10 mg with optional up-titration to 15 mg after 2 weeks. The placebo description indicates matching treatment. The arm descriptions are not fully complete: one ends mid-sentence, so all sequence and washout details are not reported in the registry record.

A crossover study exposes participants to both conditions in different periods. This design can support a within-participant comparison, since each participant may serve as their own reference, but period effects, treatment sequence, and carryover need consideration in analysis. The registry identifies the crossover model but does not report a detailed statistical plan for these issues. It reports 19 participants for each primary-outcome treatment-period result; it does not explain the relationship between those counts and the 21 enrolled participants.

Polysomnograms were performed at baseline, at the end of treatment period 1, and at the end of treatment period 2. The primary outcome was change in sleep efficiency, assessed as a difference between change during the suvorexant period and change during the placebo period. Secondary outcomes included sleep measures, insomnia severity, sleepiness, and participant and clinician global impressions of change. The registry lists six secondary outcomes.

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.

Treatment Period 1 (4 Weeks)

Washout Period (2 Weeks)

Treatment Period 2 (4 Weeks)

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 Sleep Efficiency as Measured by Polysomnogram, measured over 4 weeks.

Sleep efficiency describes the share of time in bed spent asleep. In this study, it was defined as total sleep time divided by total time in bed and expressed as a percent. Polysomnography was used to measure it at baseline and at the end of each treatment period. The primary outcome was change in sleep efficiency, with a positive change defined in the registry as improvement. The intended comparison was the change during the suvorexant period against the change during the placebo period.

This endpoint offers an objective sleep measure, complementing the study’s reported secondary outcomes, which include wakefulness after sleep onset, latency to persistent sleep, and questionnaire-based assessments. It is relevant to insomnia because it relates sleep duration to the time allocated for sleep. It does not, by itself, describe every aspect of a participant’s sleep experience or daytime functioning.

The registry identifies the primary summary as a median with inter-quartile range. It reports median changes of 8.9 for the suvorexant period and 9.4 for placebo, with 19 participants listed for each. A spread value is not reported in the posted results, although lower and upper values are listed for each period. The registry does not identify the type of those lower and upper values, so they should not be interpreted as a particular kind of interval. The primary endpoint is a measure of change over the treatment periods, not a direct report of the final sleep-efficiency level.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary summary is a median change in sleep efficiency for each treatment period. A median is the middle value when observations are ordered. It describes a typical observation while being less influenced by extreme values than a mean can be. The registry specifies an inter-quartile range as the dispersion measure, which generally describes the middle portion of the observed values. However, a numerical spread value is not reported alongside these outcome summaries. The registry also lists lower and upper values without identifying their statistical interpretation.

What does the reported test compare?

The registry names a Wilcoxon (Mann-Whitney) test and reports p = 0.72. In general, this rank-based test compares the distributions of observations between two groups without requiring a normal distribution in the same way as a standard two-sample t-test. A p-value describes how compatible the observed data are with a specified null hypothesis, under the assumptions of the test. It is not the probability that the null hypothesis is true, and it does not measure the size or clinical importance of an effect.

The trial is described as a crossover study, in which participants receive both treatment conditions in different periods. Such data can be paired: each participant’s outcomes under the two conditions are linked. The named Mann-Whitney procedure is ordinarily an unpaired comparison, so readers need details about the analysis to know how its use fits the crossover structure. The registry does not report whether the analysis accounted for participant pairing, treatment sequence, period effects, or possible carryover after the washout.

Why does the crossover structure matter?

General statistical consideration: a crossover design can reduce the influence of stable differences between participants because outcomes can be compared within the same person. At the same time, outcomes may change over time for reasons unrelated to treatment. Treatment sequence and period can therefore matter, and residual effects from an earlier period can complicate interpretation if they persist. A washout is intended to separate treatment periods, but the registry entry alone does not establish whether any carryover occurred or how it was assessed.

The primary outcome description states that the comparison is between change during the suvorexant period and change during the placebo period. The reported period-specific median changes are 8.9 and 9.4. Those summaries are not themselves a paired treatment-effect estimate. The registry does not report a numerical estimate for the between-period contrast, its uncertainty, or the participant-level analysis details needed to reconstruct one.

How should the interval information be read?

In general, an interval estimate describes uncertainty or spread, depending on how it is constructed. A confidence interval is designed to express uncertainty around an estimated quantity, whereas an inter-quartile range describes variation among observed values. These concepts are not interchangeable. Here the registry identifies median and inter-quartile range as the intended summary, but the reported outcome lists lower and upper values and does not label their meaning. They should therefore be read as bounds of unspecified type, not assigned a more specific interpretation.

General statistical consideration: a small study may yield imprecise estimates even when a test does not indicate a statistically detectable difference. A non-significant result is not evidence that the treatments are equivalent. Demonstrating equivalence or non-inferiority requires a prespecified margin and a suitable analysis; neither is reported in the registry record.

What do planning calculations represent?

Sample-size calculations are prospective planning illustrations, not observed outcomes or a measure of what the trial achieved. General explanation: calculations based on an observed effect can be sensitive to the effect estimate from a small study. An illustration using a stated fraction of that effect describes hypothetical planning under its assumptions, not a justification for the actual enrollment or observed power. The registry does not report the trial’s own sample-size justification, target effect, or detailed design assumptions. No conclusion about observed power can be drawn from the reported test.

The registry lists seven outcome measures and seven analyses, including the primary outcome and six secondary outcomes. It does not report a comprehensive analysis plan, model specification, handling of missing outcomes, or a multiplicity strategy. These omissions limit assessment of how outcome selection, multiple testing, or incomplete data may have influenced the reported results.

10. Analyses Reported by the Sponsor

The registry reports the sponsor’s primary comparison as a Wilcoxon (Mann-Whitney) test with p = 0.72. It does not report a treatment-effect estimate for the contrast between the suvorexant and placebo periods, nor does it identify the statistical meaning of the listed lower and upper values. The record therefore supports describing the reported test result and the two period-specific median changes, but not constructing a more detailed estimate of the contrast.

The posted period summaries are 8.9 for suvorexant and 9.4 for placebo, each based on 19 participants. They are medians, not an adjusted or paired treatment-effect estimate. No separate recalculated contrast is reported in the registry record. The sponsor’s reported analysis is therefore the available inferential result, rather than a result that can be checked against a posted recalculation.

The registry describes a crossover design, but does not explain how the reported test handled the repeated observations, sequence, or period. That missing detail affects interpretation of the method, not the numerical p-value recorded. The reported p-value should not be read as demonstrating equivalence or as quantifying the size of any treatment difference.

11. Multiplicity, Missing Data and Analysis Populations

The registry lists seven outcome measures and seven analyses. The primary outcome is change in sleep efficiency measured by polysomnography. The six secondary outcomes are wakefulness after sleep onset, latency to persistent sleep, Insomnia Severity Index, Epworth Sleepiness Scale, Subject’s Global Impression of Change, and Clinician’s Global Impression of Change. The primary comparison is reported with p = 0.72.

When multiple outcomes or comparisons are examined, the chance of obtaining at least one apparently notable result can increase. General statistical practice may address this through a prespecified hierarchy, adjustment, or a clearly stated focus on estimation rather than multiple significance tests. The registry does not report a multiplicity plan, whether all listed outcomes were tested, or how the findings should be interpreted across outcomes. No inference about adjustment can be made from the outcome list alone.

The record also does not report the statistical analysis plan, detailed analysis populations, or rules for excluding observations. It gives 21 enrolled participants and lists 19 participants for each primary treatment-period result. It does not explain why those counts differ, whether the same participants contributed to both periods, or how incomplete period data were handled. Missing-data methods and reasons for missing outcomes are not reported in the registry record.

For a crossover study, analysis populations and missing observations can be especially relevant because a participant may have data for one period but not the other. General explanation: a paired comparison may require information from both conditions for a participant, while other approaches can use incomplete records under specified assumptions. The record does not identify which approach was used. It also does not report sensitivity analyses for missing data.

These gaps do not invalidate the reported result, but they limit how fully it can be assessed. Readers can identify the outcomes, sample counts, and named primary test, while the rationale for analysis choices and the handling of multiplicity and incomplete observations remain not reported in the registry record.

12. 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 21 participants at risk in each safety group. It reports zero serious adverse events in the suvorexant group and zero in the placebo group. Other adverse events are listed for 2 participants receiving suvorexant and 3 receiving placebo. Deaths are reported as zero in both groups.

These counts describe what was recorded in this study; they do not establish that no serious events could occur. General statistical explanation: when a study includes a limited number of participants, uncommon events may not be observed, and zero recorded events does not mean the underlying risk is zero. The registry does not provide serious adverse-event terms, as none are listed, or additional detail about event timing or severity beyond the categories reported.

The safety groups list 21 participants at risk for each condition, whereas the primary outcome results list 19 for each treatment period. The registry does not explain this difference. The adverse-event counts should be read as recorded group totals and not compared with the primary-outcome counts as though they necessarily describe the same analysis population. These figures alone do not establish a comparative safety profile.

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

The posted results describe median changes in polysomnographic sleep efficiency during two treatment periods: 8.9 during the suvorexant period and 9.4 during placebo. The registry reports a Wilcoxon (Mann-Whitney) comparison with p = 0.72. The test result does not indicate a statistically detectable difference using the reported method, but it does not establish that the treatments are equivalent or that their effects are identical.

The result is limited to this primary outcome and the participants represented in the reported treatment-period results. It does not show whether the secondary outcomes changed, and the registry does not report their numerical results in this record. Nor does sleep efficiency alone capture all aspects of sleep or daytime experience. The study’s secondary outcomes include sleep measures and participant- and clinician-reported scales, but their presence in the registry is not itself evidence of a particular result.

Because this is a crossover study, the treatment periods and the relation between observations within a participant are important. The registry does not explain whether the analysis accounted for that structure. It also does not identify the type of lower and upper values attached to the period summaries or report a numerical contrast with an uncertainty interval. Consequently, the precision of the treatment comparison cannot be fully evaluated from the primary outcome entry.

The study enrolled 21 participants, and the primary outcome lists 19 for each period. This is a pilot study, and the registry does not report a sample-size justification. The findings should therefore be understood as a limited set of recorded results, not as a definitive determination of effect or equivalence. The safety counts similarly describe recorded events in the study and cannot rule out uncommon harms.

Overall, the registry records period-specific sleep-efficiency changes and a non-significant test result. It does not provide enough detail to infer a precise treatment contrast, verify how crossover features were handled, or make claims about outcomes and analyses not reported in the registry record.

14. Why This Trial Matters Statistically

This trial is statistically instructive because the design and the named test raise different interpretive considerations. A crossover design collects outcomes under both treatment conditions, often allowing within-participant comparisons. The registry names a Wilcoxon (Mann-Whitney) test, which is generally used for an unpaired comparison. Without further analysis details, readers cannot determine how the test was applied to the crossover data or whether pairing and period structure were accounted for.

The study also illustrates why a p-value should not be read in isolation. The reported p = 0.72 does not quantify the treatment difference, establish equivalence, or describe the uncertainty around an effect estimate. The period-specific medians are 8.9 and 9.4, but the registry does not report a numerical between-period estimate. The listed lower and upper values are not identified as confidence limits or as another specific interval type.

Another useful distinction is between a pilot study and a definitive comparison. Enrollment was 21, and 19 participants are listed for each primary treatment-period result. The registry does not provide a sample-size justification or detailed analysis plan. In general, pilot results can help describe feasibility and observed outcomes, but a small dataset may leave substantial uncertainty about the size of a treatment contrast.

Finally, the record shows why missing analysis details matter. The listed secondary outcomes create a multiplicity question, and the difference between enrolled and outcome counts raises questions about analysis populations and missing data. The registry does not report how these issues were handled. Statistical interpretation depends not only on a test name and p-value, but also on how the design, outcomes, and incomplete observations are incorporated into the analysis.

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 registry does not report a sample-size justification, detailed statistical analysis plan, or the rationale for the named Wilcoxon (Mann-Whitney) test in this crossover design. It does not state whether the analysis accounted for pairing, treatment sequence, period effects, or carryover. Those omissions limit assessment of whether the reported test fully reflects the study structure.

The primary results list 19 participants for each treatment period, while enrollment is 21. The record does not explain the difference, whether the same participants contributed to both period results, or how missing observations were handled. It also does not report analysis populations or sensitivity analyses. These details matter for judging the possible influence of incomplete data.

The outcome entry specifies median and inter-quartile range, but no numerical spread is reported. Lower and upper values are listed, but their type is not identified. A numerical treatment-period contrast and a clearly defined uncertainty interval are not reported, so the precision of the primary comparison is difficult to assess.

The registry lists six secondary outcomes but does not report a multiplicity strategy or the numerical results for those outcomes in this record. The treatment descriptions also do not fully document the sequence: one description ends mid-sentence. Although a 2 week washout and crossover are described, the record does not report whether carryover was assessed. Finally, the study’s 21-participant enrollment limits the amount of information available about uncommon safety events. These are limits on interpretation of the registry record, not evidence that a particular bias or problem occurred.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Pilot Study of Suvorexant for Insomnia in Parkinson Disease — NCT02729714.
• Registry record last updated 2024-08-20; status last verified 2024-07.
• Results first posted 2024-08-20.
• 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

NCT02729714 is a completed randomized, double-blind, placebo-controlled Phase 4 pilot study with a crossover design. It enrolled 21 participants and compared bedtime suvorexant with matching placebo. Treatment periods lasted 4 weeks, with a 2 week washout described before crossover. The primary outcome was change in polysomnographic sleep efficiency, measured at baseline and at the end of each treatment period.

The registry reports median changes of 8.9 during the suvorexant period and 9.4 during placebo, with 19 participants listed for each result. The named Wilcoxon (Mann-Whitney) comparison has p = 0.72. The registry does not identify the statistical type of the lower and upper values or report a numerical between-period treatment-effect estimate.

Six secondary outcomes are listed. Safety records show zero serious adverse events and zero deaths in each group, with other adverse events recorded for 2 participants receiving suvorexant and 3 receiving placebo. The registry does not report a sample-size justification, detailed analysis plan, multiplicity strategy, or missing-data methods.