Suvorexant and Cocaine: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Suvorexant and Cocaine (NCT03937986), a randomized early/phase 1 study. The primary outcome is Reinforcing Effects of Cocaine.

2. Clinical Question

The scientific aim described in the registry is to examine whether orexin antagonism is associated with motivation for cocaine and other cocaine-associated behaviors in active cocaine users. The primary endpoint operationalizes one aspect of that question through a choice task: more cocaine choices represent a greater number of choices of cocaine over the available money in that task. It is a laboratory measure, not a direct measure of cocaine use outside the study.

The registry record does not report a specific statistical contrast for the primary question. In particular, it does not say which posted conditions should be compared to estimate the effect of suvorexant, or whether the primary contrast is between maintenance conditions, acute administrations, or a combination of both. The results therefore need to be read as condition-specific summaries rather than as an established treatment comparison.

3. Trial Design

The study is titled “Suvorexant and Cocaine,” with the official title “Influence of Orexin Antagonism on Motivation for Cocaine.” It is classified as EARLY PHASE1 and is marked COMPLETED in the registry. The record lists a randomized allocation, triple masking, and a crossover model. Enrollment is 8. The listed start date is 2019-07-11, and primary completion is recorded as 2022-04-30.

The intervention descriptions indicate oral maintenance with placebo or suvorexant dose 1, dose 2, or dose 3. During each maintenance condition, cocaine and placebo are described as acute administrations. This structure distinguishes the continuing maintenance condition from the acute administration occurring during it. The registry does not report the dose amounts, the sequence of conditions, the washout arrangements, or the detailed procedure for administering the acute substances. Those design details should not be inferred from the crossover label alone.

A crossover design generally means that participants receive more than one study condition, so comparisons may use observations from the same participant under different conditions. As a general statistical point, such a design can account for stable differences between participants when analyzing within-person contrasts. It also requires attention to the relationship between repeated observations, the order of conditions, and any carryover. The registry record does not report how these features were handled in the analysis.

The primary outcome is described as the mean number of cocaine choices over money, measured 12 times over approximately 1 month of inpatient admission. The outcome summaries are labeled as means with standard errors. The posted condition labels run from Dose Condition 1 through Dose Condition 12, but the registry does not map those labels to the intervention descriptions. No formal primary analysis, estimated contrast, or analysis model is posted. The purpose and details of any sample-size justification are not 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.

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 Reinforcing Effects of Cocaine, measured over 12 times over approximately 1 month inpatient admission..

The primary endpoint is titled “Reinforcing Effects of Cocaine.” It measures the number of times a participant chooses cocaine rather than money, with a maximum of 10 choices. The registry describes this as a mean of a continuous measure and reports the unit as “Number of Cocaine Choices.” The timeframe is 12 assessments over approximately 1 month of inpatient admission.

The endpoint is used as a behavioral measure of choice under the study conditions. In general, a choice task can quantify how often a participant selects one available option over another, making the outcome a way to examine motivation or reinforcing effects in a controlled setting. A higher mean indicates more cocaine choices within the task; it does not, by itself, establish how a participant behaves outside the study or explain why the choices were made.

The record reports a mean and a spread labeled standard error for each posted dose condition. A mean describes the average number of choices in that condition. A standard error describes the precision of an estimated mean, rather than the spread of individual participants’ values. As a general point, a standard error is affected by both variability in observations and the amount of information contributing to the estimate. The registry does not provide individual-level data or confidence intervals for these means.

The endpoint has an explicit upper limit of 10 choices, but the record does not state how missing assessments were handled or how repeated assessments were combined in the posted summaries. It also does not report a threshold that would define a meaningful difference in this laboratory measure.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry describes the primary outcome as a continuous measure summarized by a mean and standard error. These are descriptive statistics. The mean represents the average number of cocaine choices in a condition, while the standard error describes the precision of that estimated average. Neither statistic is, by itself, a test of a difference between conditions. No inferential analysis is posted for this endpoint.

What does the standard error tell the reader?

As a general statistical explanation, a standard error quantifies uncertainty in an estimated summary, such as a mean. It should not be interpreted as the standard deviation of individual observations, which describes how participant-level values vary. It is also not a confidence interval. A confidence interval would give a range of values compatible with the data under a stated method and assumptions. The registry record does not report confidence intervals for these outcome means or for comparisons.

The reported standard errors vary considerably among the conditions. A mean paired with a larger standard error is less precisely estimated than a mean paired with a smaller standard error, considered in isolation. However, comparing the standard errors of separate means does not supply the uncertainty for a difference between them. In a crossover design, that difference may depend on the paired observations and their covariance: how outcomes from the same participant move together across conditions. The registry does not report those paired data or a model for them.

Why does the crossover structure matter?

In general, a crossover study observes participants in multiple conditions. An analysis can then compare outcomes within the same participant, which can help account for stable participant characteristics. The advantage depends on appropriately analyzing the repeated measurements. Outcomes from the same person are not ordinarily independent, and a model that treats them as unrelated can misrepresent uncertainty. The condition sequence can also matter if responses depend on what was received earlier or if effects persist into a later condition. The registry record identifies the crossover model but does not report the analysis method, sequence, or handling of possible carryover.

The posted record gives means and standard errors by labeled dose condition, but does not state which conditions form the intended contrast. Nor does it identify the labels with the placebo-maintenance and suvorexant-maintenance descriptions. Without those links and a defined comparison, it is not possible to interpret a difference between particular labels as an estimate of a specified intervention effect.

What can be concluded without a comparison analysis?

The displayed condition summaries describe the observed average choice counts in each labeled condition. They do not provide a test of a treatment hypothesis. A p-value would assess compatibility of data with a particular null model and contrast; a confidence interval would describe uncertainty around an estimated contrast. Neither is reported. General statistical interpretation cannot replace the missing contrast or analysis.

The record includes sample-size planning assumptions: a two-sided significance level of 0.05, equal allocation between two arms, a normal approximation, and an effect to detect defined as a stated fraction of the observed effect. These assumptions describe planning illustrations, not the trial’s own design or an observed power calculation. The registry record does not report the sample-size justification for the study or numerical planning results. Planning calculations are not evidence that the study had a particular probability of detecting an effect after the data were collected.

For count outcomes, the distribution and variance can also matter. As a general explanation, over-dispersion occurs when variation is greater than a model’s assumptions would predict; if present, it can make standard errors based on an overly simple model too small. This endpoint is a bounded choice count, with a maximum of 10, so the assumptions of any model would need to be considered in light of its scale and repeated-measure structure. The registry record does not report a model, distributional checks, or an assessment of over-dispersion. It is not possible to determine from the posted summaries whether these issues were evaluated.

The primary outcome’s posted summaries use means and standard errors, but a summary format does not reveal the analysis model used to generate them. The registry reports no outcome analyses, so there is no sponsor-reported inferential estimate to interpret. The record also does not report an adjusted analysis, a prespecified contrast, a confidence interval, or a p-value. Consequently, the sound statistical reading is limited to describing the means, their labeled standard errors, and the variation in participant counts across conditions.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists 1 outcome measure and no posted analyses for it. Although the outcome is described as being measured 12 times over approximately 1 month, the record does not explain whether those assessments were combined, analyzed separately, or used in another way. It also does not report a multiplicity adjustment or a plan for controlling the chance of false-positive findings across repeated assessments or condition comparisons. Because no inferential results are posted, there is no reported family of tests or adjustment to evaluate.

The record does not report how missing outcome data were handled. The condition-level summaries include participant counts of 7 or 8, while enrollment is 8. Those counts alone do not explain why observations are absent from some summaries, whether missingness occurred at the participant or assessment level, or whether participants contributed data to all conditions. No imputation method, complete-case rule, or sensitivity analysis is reported in the registry record.

The analysis population is not fully described for the primary outcome. The record does not state whether summaries include all randomized participants, only those completing particular conditions, or another analysis set. It also does not report exclusions or a protocol for handling incomplete crossover sequences.

These omissions limit interpretation of precision and potential bias. As a general point, missing outcomes can affect an estimate when the likelihood of missingness is related to the outcome or study condition. In repeated-measure crossover data, the pattern of incomplete observations can also influence which within-participant comparisons are available. Without the missing-data and analysis-population details, the posted means should be treated as descriptive summaries of the recorded condition data, not as a complete account of the randomized comparison.

11. Safety Results

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

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

Most frequent non-serious adverse event terms

The registry’s safety summary lists 8 subjects included in the analysis. It reports 0 subjects with a serious adverse event among 8 at risk, 8 subjects with other adverse events among 8 at risk, and 0 deaths. The record does not list any top serious adverse-event terms.

These counts describe what the registry reports for the stated analysis group; they do not establish whether an event was caused by a study intervention. The entry for other adverse events indicates that all 8 subjects in the listed group were recorded as affected, but the registry summary does not specify the event types, severity, timing, or relation to suvorexant, cocaine, or placebo. The meaning and scope of the “other” category should therefore not be expanded beyond the record’s label.

The serious-event count is a count of affected subjects, not necessarily a count of all event episodes. No serious events are listed in the summary, and no deaths are reported. These registry figures should not be interpreted as proving that serious harms could not occur, particularly given the small enrollment and the limited detail posted. They also cannot be used to compare safety between conditions because condition-specific safety counts are not reported.

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

The posted results describe average cocaine-choice counts across 12 labeled dose conditions in a randomized, triple-masked crossover study. The means vary across those labels, and the posted summaries include standard errors and participant counts. This provides descriptive information about the choice-task outcome recorded under the study conditions.

The record does not report inferential analyses for the primary outcome. It supplies no estimated comparison between suvorexant and placebo, no confidence interval for such a contrast, and no p-value. Furthermore, the condition labels are not mapped to the intervention descriptions. The observed summaries therefore cannot establish the direction or size of an effect attributable to suvorexant, nor whether a difference between any two labels reflects an intervention effect.

The endpoint is a behavioral choice measure collected during an inpatient study. It can inform research about choices made in that controlled task, but it is not the same as measuring cocaine use in daily life, treatment retention, recovery, or longer-term outcomes. The registry does not report those outcomes as part of the primary result.

The study’s brief summary states that the research sought initial clinical evidence concerning orexin antagonism and motivation for cocaine. That statement describes the research aim, not a finding. The posted data do not support a conclusion that suvorexant changes cocaine motivation, either favorably or unfavorably, because the registry lacks a specified and analyzed intervention contrast.

As a general statistical principle, an observed mean is subject to sampling variability, and the uncertainty of a treatment comparison depends on the design and analysis as well as the individual condition summaries. The crossover structure makes within-participant information relevant. Since that information and its analysis are not reported, the most defensible conclusion is limited to the recorded condition-specific summaries and the absence of a posted inferential result.

13. Why This Trial Matters Statistically

This trial is statistically notable because it uses a crossover structure to study a choice-based outcome in an inpatient setting. In general, measuring the same participants under multiple conditions can make within-person contrasts informative, since participants may differ in their baseline tendency to choose cocaine. That potential benefit depends on reporting and analyzing the repeated measurements appropriately.

The registry makes clear that the outcome is a mean number of cocaine choices over money, with a maximum of 10, and gives condition-level means, standard errors, and participant counts. Those summaries let readers understand the basic scale of the task and see that the recorded averages are not uniform across the labels. They also show why labeling and analysis details matter: without a mapping from dose-condition labels to actual regimens, a descriptive pattern cannot be assigned to a particular intervention.

The absence of posted inferential analyses is also important. A study can report outcome summaries without reporting a treatment contrast, but the summaries alone do not answer whether observed differences are consistent with a treatment effect. For a crossover design, interpreting a contrast requires information about paired observations and the statistical handling of repeated measurements. Neither is available in the registry record.

The record illustrates the distinction between scientific motivation and statistical evidence. A stated plan to investigate a mechanism or generate initial clinical evidence does not substitute for a reported estimate with uncertainty. The endpoint may be useful for studying behavior in a controlled task, while remaining a limited proxy for behavior beyond that setting. These features make careful separation of descriptive evidence, inferential evidence, and broader clinical meaning especially important.

14. 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 record leaves key details unresolved for interpreting this particular study. Although it identifies a crossover model and randomized, triple-masked allocation, it does not report the sequence of conditions, the procedures for acute cocaine and placebo administration, or any washout arrangements. The listed intervention descriptions do not give the actual suvorexant dose amounts.

The 12 posted labels are generic Dose Condition labels, with no mapping to placebo maintenance, the suvorexant dose levels, or the acute administration conditions. This prevents attributing the condition-level means to the named interventions. The record also does not specify the intended primary contrast, so even a descriptive comparison between labels cannot be aligned confidently with the main scientific question.

Enrollment is 8, and outcome summaries report participant counts of 7 or 8. The reason for the varying counts is not reported in the registry record. The record does not state whether observations were missing, whether participants completed all conditions, or how incomplete data were treated. It also does not identify the primary outcome analysis population.

No inferential analysis is posted for the continuous endpoint. There is no reported analysis model, estimated contrast, confidence interval, or p-value. The record does not explain how the repeated assessments were used, whether within-participant correlation was handled, or whether any order or carryover effects were considered. A sample-size justification for the trial itself is not reported; planning assumptions do not establish the study’s original design rationale.

The outcome is a laboratory choice measure and the record does not report how it relates to behavior outside the inpatient setting. Safety reporting is limited to summary counts, without event terms for the “other” category or condition-specific counts. These limitations make the posted findings descriptive and narrow in scope.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Suvorexant and Cocaine — NCT03937986.
• Registry record last updated 2023-02-09; status last verified 2023-02.
• Results first posted 2023-02-08.
• 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.

18. Record Summary

NCT03937986, “Suvorexant and Cocaine,” is an EARLY PHASE1 randomized, triple-masked crossover study marked COMPLETED. The record lists enrollment of 8, a start date of 2019-07-11, and primary completion on 2022-04-30. The condition is Cocaine Use Disorder.

The intervention descriptions include oral placebo or suvorexant dose 1, dose 2, or dose 3 as maintenance, with acute cocaine and placebo administrations described during maintenance. The primary outcome is the mean number of cocaine choices over money, with a maximum of 10 choices, measured 12 times over approximately 1 month of inpatient admission.

Condition-specific means, standard errors, and participant counts are posted, but the dose-condition labels are not mapped to the intervention descriptions. The registry reports no inferential analyses for the primary endpoint. Its safety summary records 0 serious adverse events among 8 at risk, 8 affected by other adverse events among 8 at risk, and 0 deaths.