This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the results posted in the ClinicalTrials.gov record; calculations added by Clinical Biostats are labelled as such.
1. Trial at a Glance
This page covers the registry results for A Study of LY2624803 in Japanese Participants With Transient Insomnia (NCT01784614), a randomized phase 1 study. The primary outcome is Wake After Sleep Onset (WASO) With LY2624803 Compared to Placebo.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT01784614 |
| Phase | Phase 1 |
| Condition | Sleep Initiation and Maintenance Disorders |
| Allocation | Randomized |
| Design model | Crossover |
| Masking | Double |
| Enrollment | 24 |
| Arms in the primary outcome | 5 |
| Lead sponsor | Eli Lilly and Company |
| Results posted | Yes |
2. Clinical Question
Population
Healthy Japanese participants in a 5-hour phase advance model of transient insomnia. The study enrolled 24 participants. The primary-outcome population was participants who were randomized, received at least 1 dose of study drug, and had polysomnography measurements in Periods 1, 2 and 3.
Intervention
Single oral doses of LY2624803 at 0.1 mg, 1.0 mg, 3.0 mg or 6.0 mg. Participants received treatment across four periods; each dose was given in up to 2 of the 4 periods.
Comparator
Oral placebo, administered in up to 1 of the 4 treatment periods. The study included solution and capsule formulations of LY2624803 and placebo.
Primary question
How did wake after sleep onset (WASO) during the 8-hour recording period compare between each LY2624803 dose and placebo in Periods 1, 2 and 3? The outcome was analyzed as a geometric least-squares mean using a mixed model.
The trial focused on pharmacodynamics and pharmacokinetics after single doses in healthy Japanese subjects, using a 5-hour phase advance model of transient insomnia. It was a Phase 1, randomized, double-blind crossover study sponsored by Eli Lilly and Company. The registry record describes the aim as learning how different doses affect sleep; that aim is not itself a statement about the findings.
The primary endpoint was WASO, measured in minutes. The posted comparisons estimate the ratio of the geometric least-squares mean for each LY2624803 dose to the corresponding placebo mean. A ratio below 1 represents a lower modeled geometric mean WASO for the dose than for placebo; it is a comparison of the recorded outcome in this study, not a standalone conclusion about clinical benefit.
3. Trial Design
This was a completed Phase 1 randomized, double-blind crossover study with an enrollment of 24. The study began in 2009-09 and had primary completion in 2010-03. Its official title describes the evaluation of pharmacodynamics and pharmacokinetics of single doses of LY2624803 in a 5-hour phase advance model of transient insomnia in healthy Japanese subjects.
The design used four treatment periods. Participants received a single oral dose of LY2624803 or placebo in each period. The treatment descriptions list 0.1 mg, 1.0 mg, 3.0 mg and 6.0 mg LY2624803, with each active dose administered in up to 2 of 4 periods, and placebo administered in up to 1 of 4 periods. The listed interventions include solution and capsule formulations for both LY2624803 and placebo. The registry record does not report the treatment sequence details beyond the sequence factor included in the primary mixed model.
Because this was a crossover study, participants could contribute observations under more than one treatment condition. The group sizes attached to the primary endpoint are not all the same: the posted summaries list 8 participants for 0.1 mg, 17 for 1.0 mg, 16 for 3.0 mg, 8 for 6.0 mg and 17 for placebo. These are the reported endpoint-group counts, not separate totals to be added as if each were a distinct set of participants.
The primary endpoint was assessed in Periods 1, 2 and 3. Period 4 was included in the study and is referenced in the listed pharmacokinetic outcomes, but it is not part of the stated primary-endpoint timeframe. The registry record does not report the reason for this division of endpoint periods, a washout schedule, or further details of the treatment allocation sequence.
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: This was a double-blind, randomized, placebo-controlled, 4-period cross-over study. The first 3 of which were used for polysomnography (PSG) measurements, and the fourth period was used to determine plasma concentrations of LY2624803 to estimate pharmacokinetic (PK) parameters.
Only the first 6 of 8 groups are shown.
Period 1
| Stage | Cohort 1 | Cohort 2 | Cohort 3 | Cohort 4 | Cohort 5 | Cohort 6 |
|---|---|---|---|---|---|---|
| STARTED | 3 | 3 | 3 | 3 | 3 | 3 |
| Received at Least 1 Dose of Study Drug | 3 | 3 | 3 | 3 | 3 | 3 |
| COMPLETED | 3 | 3 | 3 | 3 | 3 | 3 |
| NOT COMPLETED | 0 | 0 | 0 | 0 | 0 | 0 |
Period 2
| Stage | Cohort 1 | Cohort 2 | Cohort 3 | Cohort 4 | Cohort 5 | Cohort 6 |
|---|---|---|---|---|---|---|
| STARTED | 3 | 3 | 3 | 3 | 3 | 3 |
| COMPLETED | 3 | 3 | 3 | 3 | 3 | 3 |
| NOT COMPLETED | 0 | 0 | 0 | 0 | 0 | 0 |
Period 3
| Stage | Cohort 1 | Cohort 2 | Cohort 3 | Cohort 4 | Cohort 5 | Cohort 6 |
|---|---|---|---|---|---|---|
| STARTED | 3 | 2 | 3 | 3 | 3 | 3 |
| COMPLETED | 3 | 2 | 3 | 3 | 3 | 3 |
| NOT COMPLETED | 0 | 0 | 0 | 0 | 0 | 0 |
Period 4
| Stage | Cohort 1 | Cohort 2 | Cohort 3 | Cohort 4 | Cohort 5 | Cohort 6 |
|---|---|---|---|---|---|---|
| STARTED | 3 | 2 | 3 | 3 | 3 | 3 |
| COMPLETED | 3 | 2 | 3 | 3 | 3 | 3 |
| NOT COMPLETED | 0 | 0 | 0 | 0 | 0 | 0 |
Analysis population for the primary outcome, as defined in the registry: All randomized participants who received at least 1 dose of study drug and had PSG measurements in Periods 1, 2 and 3.
5. Baseline Characteristics
The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.
Population: All randomized participants.
| Characteristic | Unit | Overall |
|---|---|---|
| Participants analysed | 24 | |
| Age, Continuous | years | 54.1 (SD 14.1) |
| Sex: Female, Male: Female | Participants | 14 |
| Sex: Female, Male: Male | Participants | 10 |
| Race (NIH/OMB): American Indian or Alaska Native | Participants | 0 |
| Race (NIH/OMB): Asian | Participants | 24 |
| Race (NIH/OMB): Native Hawaiian or Other Pacific Islander | Participants | 0 |
| Race (NIH/OMB): Black or African American | Participants | 0 |
| Race (NIH/OMB): White | Participants | 0 |
| Race (NIH/OMB): More than one race | Participants | 0 |
| Race (NIH/OMB): Unknown or Not Reported | Participants | 0 |
| Region of Enrollment: Japan | participants | 24 |
6. Primary Endpoint
The registry lists the primary outcome as Wake After Sleep Onset (WASO) With LY2624803 Compared to Placebo, measured over 8 hours in Periods 1, 2 and 3.
| Element | Registry entry |
|---|---|
| Result type | Geometric Mean |
| Measure of spread | 90% Confidence Interval |
| Unit | minutes (min) |
| Endpoint classification | not reduced to a single comparison |
Wake after sleep onset (WASO) measures time spent awake after sleep onset during the overnight recording. In this trial it was calculated as the total time in awake epochs between sleep onset, defined as the first stage 2 epoch, and the end of the primary recording period, 8 hours after lights-off. The unit is minutes.
WASO describes one aspect of sleep maintenance: it captures wakefulness after sleep has begun, rather than the time taken to fall asleep. It was selected as the primary outcome for this study of transient insomnia in a phase advance model. The registry record also lists latency to persistent sleep (LPS) and total sleep time (TST) among secondary outcomes, which address other aspects of sleep timing and duration.
The posted primary results are geometric least-squares means, estimated using mixed models, with 90% confidence intervals. A geometric mean is a summary on a multiplicative scale; comparing the treatment and placebo summaries as a ratio expresses their relative modeled levels. The model included treatment, treatment sequence, period and participants. This structure reflects the repeated treatment periods and participant contributions in a crossover design.
WASO is an observed study endpoint, not a direct measure of every dimension of sleep or daytime functioning. Its interpretation is limited to the defined recording interval and the participants and treatment periods included in the analysis. The registry record does not report a threshold for a clinically important WASO difference.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The posted geometric least-squares mean WASO was 58.88 minutes for 0.1 mg LY2624803, 40.7 minutes for 1.0 mg, 32.27 minutes for 3.0 mg, and 18.42 minutes for 6.0 mg. The placebo geometric least-squares mean was 61.1 minutes. These are modeled summaries for the primary endpoint in the stated analysis population and timeframe. Their ordering is descriptive; it does not by itself establish a dose-response relationship.
The mixed-model comparison for 0.1 mg versus placebo gave a ratio of geometric least-squares means of 0.96, with a 90% confidence interval from 0.56 to 1.65 and p = 0.908. The interval includes 1, the value representing equal modeled geometric means, and the posted estimate is close to 1. The result is therefore imprecise about the direction and size of the relative difference.
For 1.0 mg versus placebo, the ratio was 0.67, with a 90% confidence interval from 0.45 to 0.98 and p = 0.083. The interval is mostly below 1 but reaches close to it. The estimate is compatible with lower modeled WASO for the active dose, while the uncertainty interval still conveys substantial uncertainty about the magnitude.
For 3.0 mg versus placebo, the ratio was 0.53, with a 90% confidence interval from 0.36 to 0.78 and p = 0.010. For 6.0 mg versus placebo, the ratio was 0.30, with a 90% confidence interval from 0.18 to 0.51 and p < 0.001. Both intervals are below 1, so the posted estimates and intervals indicate lower modeled geometric mean WASO for these doses relative to placebo in this study analysis. They do not establish the size or relevance of an effect outside the study setting.
The reported p-values are attached to the mixed-model comparisons. They quantify how compatible the observed comparison is with the null hypothesis used for each test, under the analysis assumptions; they do not measure the probability that the null hypothesis is true. The registry record does not report a multiplicity adjustment, so the p-values should be read as posted rather than as evidence that a family of dose comparisons was controlled for multiple testing.
The estimates are based on a small study, with endpoint group counts ranging from 8 to 17. The confidence intervals communicate uncertainty, but the registry record does not report a prespecified clinically important difference with which to judge the practical importance of the observed ratios. Nor does it report a confirmatory trial or follow-up evidence. The results describe the posted primary endpoint under the study's particular model and population; they do not alone establish clinical effectiveness or safety.
9. Statistical Methods Explained
The primary outcome was analyzed with mixed models, and the reported summaries are geometric least-squares means. A mixed model can account for modeled factors that structure the observations, while estimating treatment comparisons after taking those factors into consideration. In this study the stated model included treatment, treatment sequence, period and participants. This is relevant in a crossover design, where observations are collected across treatment periods and a participant may contribute data under multiple conditions.
What does the ratio mean?
The reported treatment contrasts are ratios of geometric least-squares means, with placebo as the comparison condition. General statistical interpretation: a ratio of 1 represents equal modeled means, a ratio below 1 represents a lower modeled value for the treatment condition, and a ratio above 1 represents a higher modeled value. Because WASO is measured in minutes but the contrast is a ratio, the ratio expresses relative rather than absolute separation. The posted means give the corresponding endpoint summaries; the contrast is the model-based comparison reported by the sponsor.
For example, the 3.0 mg comparison has a ratio of 0.53, while the 6.0 mg comparison has a ratio of 0.30. These ratios should not be read as differences in minutes. No absolute treatment-placebo difference is stated in the reported analyses, and a new difference should not be inferred from the ratio alone. The ratios also do not describe the experience of every participant; they summarize the modeled group comparison.
How should the confidence intervals be read?
A confidence interval describes a range of values reasonably compatible with the observed data and the statistical model, using the stated confidence level. Here, the posted intervals are 90% confidence intervals. General statistical explanation: narrower intervals indicate greater precision than wider intervals, while an interval spanning a broad range leaves more uncertainty about the effect size. The intervals are reported for the ratio scale, so their relation to 1 is central to interpreting the comparison.
The 0.1 mg interval, 0.56 to 1.65, spans 1 and covers values on both sides of equal modeled means. The 1.0 mg interval, 0.45 to 0.98, lies below 1 but approaches it. The intervals for 3.0 mg, 0.36 to 0.78, and 6.0 mg, 0.18 to 0.51, are also below 1. An interval that does not include 1 is consistent with a difference from the null ratio at the corresponding two-sided level under the model, but it does not establish clinical importance or remove concerns about bias, missing data or multiple comparisons.
What does the p-value add?
A p-value describes the compatibility of the data with a specified null hypothesis under the statistical model. It is not the probability that the null hypothesis is correct, and it does not tell how large or important a difference is. The posted p-values are 0.908 for 0.1 mg, 0.083 for 1.0 mg, 0.010 for 3.0 mg and <0.001 for 6.0 mg. They belong to separate dose-versus-placebo comparisons. The registry record does not report whether the analyses were adjusted for the fact that several doses were compared with placebo.
The p-value and interval provide related but different information. The interval shows a range of plausible ratio values under the model, while the p-value summarizes evidence against the null value of 1. Both are sensitive to the analysis assumptions and the amount of information in the data. Neither is a measure of the probability that a treatment is beneficial or harmful.
Why does over-dispersion matter?
General statistical explanation: over-dispersion occurs when observations vary more than a model's assumed distribution allows. If that extra variation is not represented, standard errors may be too small and intervals or tests may appear more certain than warranted. It is a general consideration in statistical modeling, not a problem identified in this record. The registry record does not report a specific assessment of over-dispersion or a correction for it.
For this endpoint, the reported approach is a mixed model with geometric least-squares means. The registry record does not give the model's distributional details, covariance structure, estimation method, diagnostic checks or handling of departures from assumptions. Those details matter for reproducing the analysis and assessing how robust the uncertainty estimates are.
What do the planning calculations represent?
The sample-size material is described as a planning illustration based on the observed effect, a two-sided significance level of 0.05, equal allocation between two arms and a normal approximation. The calculation uses the stated fraction of the observed effect as the effect to detect. Such figures are hypothetical planning estimates, not a description of the trial's actual design or its observed power. The registry record does not report the trial's own sample-size justification or design assumptions. Any planning values should therefore be distinguished from the enrollment of 24 and from the outcome-group counts in the posted results.
10. Analyses Reported by the Sponsor
| Method | Estimate | 95% CI | P-value |
|---|---|---|---|
| Mixed Models Analysis | Ratio of Geometric LS mean 0.96 | 0.56 to 1.65 | 0.908 |
| Mixed Models Analysis | Ratio of Geometric LS mean 0.67 | 0.45 to 0.98 | 0.083 |
| Mixed Models Analysis | Ratio of Geometric LS mean 0.53 | 0.36 to 0.78 | 0.010 |
| Mixed Models Analysis | Ratio of Geometric LS mean 0.30 | 0.18 to 0.51 | < 0.001 |
The posted dose-versus-placebo analyses are mixed-model comparisons of geometric least-squares means. They report ratios, 90% confidence intervals and p-values for each dose. These are the sponsor's reported analyses in the registry record and should be distinguished from any simple unadjusted recalculation based on summary statistics. The record does not flag a discrepancy between the reported analyses and a recalculation.
The registry results should therefore be read through the stated mixed-model estimates, rather than substituting an unadjusted summary-statistic comparison for the crossover analysis. The model includes treatment, treatment sequence, period and participants, whereas a simple calculation from marginal summaries would not necessarily reproduce that adjustment or account for the design in the same way. The registry record does not provide enough model detail to independently reproduce the mixed-model estimates.
The sponsor's results do not resolve questions that are not documented, such as multiplicity handling, missing-data assumptions or model diagnostics. The p-values and intervals should be interpreted as reported for the individual comparisons, with the limitations of the record kept in view.
11. Multiplicity, Missing Data and Analysis Populations
There are four posted primary-endpoint comparisons: each of the four LY2624803 doses versus placebo. When multiple hypotheses are tested, the chance of at least one apparently unusual result can increase if each test is interpreted independently. General statistical explanation: multiplicity procedures can adjust the inferential threshold or intervals to address a planned family of comparisons. The registry record does not report a multiplicity adjustment, a testing hierarchy or a prespecified primary dose comparison. The reported p-values and 90% intervals should therefore be treated as the individual posted results, not as evidence that family-wise error was controlled.
The primary analysis population is described as all randomized participants who received at least 1 dose of study drug and had polysomnography measurements in Periods 1, 2 and 3. This is a defined population, but it is not identical to all randomized participants if some did not receive treatment or lacked the required measurements. The registry record does not report the number excluded from this population or reasons for any exclusions.
Different endpoint group counts are posted for the dose conditions and placebo. The record does not explain the missingness or why the number of participants contributing to each summary varies. It does not report the amount of missing data by period, the assumptions used for missing observations, an imputation method, or sensitivity analyses. In a crossover trial, incomplete participation across periods may affect which treatment contrasts can be estimated, but the specific pattern and its handling are not reported in the registry record.
The record lists LPS, TST, Cmax after a single oral dose in Period 4, and AUC from time 0 to infinity after a single oral dose in Period 4 as secondary outcomes. Values for these outcomes are not included in the posted outcome details. The registry record does not report a full statistical analysis plan, a sample-size justification, interim analyses, or how missing secondary-outcome data were handled. These omissions limit assessment of whether all planned analyses and outcomes are represented.
12. Safety Results
The registry reports serious adverse events as affected participants divided by participants at risk.
| Arm | Participants with serious adverse events | Participants at risk | Percentage |
|---|---|---|---|
| 0.1 mg LY2624803 - Periods 1-3 | 0 | 9 | 0.0% |
| 1.0 mg LY2624803 - Periods 1-3 | 0 | 17 | 0.0% |
| 3.0 mg LY2624803 - Periods 1-3 | 0 | 17 | 0.0% |
| 6.0 mg LY2624803 - Periods 1-3 | 0 | 8 | 0.0% |
| Placebo - Periods 1-3 | 0 | 18 | 0.0% |
| 0.1 mg LY2624803 - Period 4 | 0 | 6 | 0.0% |
| 1.0 mg LY2624803 - Period 4 | 0 | 5 | 0.0% |
| 3.0 mg LY2624803 - Period 4 | 0 | 6 | 0.0% |
| 6 mg LY2624803 - Period 4 | 0 | 5 | 0.0% |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 5% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| 0.1 mg LY2624803 - Periods 1-3 | 2 of 9 |
| 1.0 mg LY2624803 - Periods 1-3 | 3 of 17 |
| 3.0 mg LY2624803 - Periods 1-3 | 3 of 17 |
| 6.0 mg LY2624803 - Periods 1-3 | 0 of 8 |
| Placebo - Periods 1-3 | 3 of 18 |
| 0.1 mg LY2624803 - Period 4 | 0 of 6 |
Most frequent non-serious adverse event terms
| Event term | Organ system | 0.1 mg LY2624803 - Periods 1-3 | 1.0 mg LY2624803 - Periods 1-3 | 3.0 mg LY2624803 - Periods 1-3 | 6.0 mg LY2624803 - Periods 1-3 | Placebo - Periods 1-3 | 0.1 mg LY2624803 - Period 4 |
|---|---|---|---|---|---|---|---|
| Diarrhoea | Gastrointestinal disorders | 0 of 9 | 1 of 17 | 1 of 17 | 0 of 8 | 1 of 18 | 0 of 6 |
| Poor quality sleep | Nervous system disorders | 0 of 9 | 1 of 17 | 0 of 17 | 0 of 8 | 2 of 18 | 0 of 6 |
| Conjunctivitis | Eye disorders | 0 of 9 | 0 of 17 | 1 of 17 | 0 of 8 | 0 of 18 | 0 of 6 |
| Feeling abnormal | General disorders | 1 of 9 | 0 of 17 | 0 of 17 | 0 of 8 | 0 of 18 | 0 of 6 |
| Contusion | Injury, poisoning and procedural complications | 0 of 9 | 0 of 17 | 1 of 17 | 0 of 8 | 0 of 18 | 0 of 6 |
| Muscle strain | Injury, poisoning and procedural complications | 1 of 9 | 0 of 17 | 0 of 17 | 0 of 8 | 0 of 18 | 0 of 6 |
| Headache | Nervous system disorders | 0 of 9 | 1 of 17 | 0 of 17 | 0 of 8 | 0 of 18 | 0 of 6 |
| Somnolence | Nervous system disorders | 0 of 9 | 0 of 17 | 0 of 17 | 0 of 8 | 1 of 18 | 0 of 6 |
| Cough | Respiratory, thoracic and mediastinal disorders | 1 of 9 | 0 of 17 | 0 of 17 | 0 of 8 | 0 of 18 | 0 of 6 |
| Upper airway obstruction | Respiratory, thoracic and mediastinal disorders | 0 of 9 | 1 of 17 | 0 of 17 | 0 of 8 | 0 of 18 | 0 of 6 |
The adverse-event figures are reported separately for Periods 1-3 and Period 4. For Periods 1-3, serious adverse events were reported as 0 for each listed treatment group: 0.1 mg LY2624803, 1.0 mg, 3.0 mg, 6.0 mg and placebo. The corresponding numbers at risk were 9, 17, 17, 8 and 18. For Period 4, serious adverse events were also reported as 0 for each listed LY2624803 dose; the numbers at risk were 6, 5, 6 and 5.
These are event counts within the listed treatment-period groupings. In a crossover study, participants may contribute to more than one treatment condition, so the groups should not be read as wholly separate sets of people. A zero count in these summaries means no serious adverse events were reported in that listed group; it does not establish that the risk is absent.
Other adverse events are also listed: 2, 3, 3, 0 and 3 participants in the Periods 1-3 groups, in dose order followed by placebo; and 0, 2 and 1 for the 0.1 mg, 1.0 mg and 3.0 mg Period 4 groups, with 3 for 6 mg in Period 4. The record does not provide serious adverse-event terms, and deaths are not reported in the registry record. These counts are descriptive and do not establish comparative safety.
13. What the Result Does, and Does Not, Mean
The results describe modeled WASO during an 8-hour recording period in healthy Japanese participants studied in a 5-hour phase advance model of transient insomnia. The posted mixed-model estimates compare each single dose of LY2624803 with placebo during Periods 1, 2 and 3. The ratios for 3.0 mg and 6.0 mg were below 1, and their 90% confidence intervals were also below 1. The 0.1 mg interval included 1, while the 1.0 mg interval was below 1 but close to it.
These findings concern one sleep measure, in a specific experimental setting, after single doses. They do not by themselves establish how the results relate to everyday sleep, other populations, repeated dosing, longer-term outcomes or other aspects of health. The registry record does not report a clinically important threshold for WASO, so the practical significance of the ratios cannot be determined from the posted comparisons alone.
The results also do not establish that a particular dose is optimal. The comparison estimates do not describe all possible dose-response patterns, and the study was not reported as a confirmatory effectiveness trial. The sample was limited, and the endpoint-group counts varied by condition. The record does not report multiplicity handling, missing-data analyses or detailed model diagnostics, which are relevant when judging the certainty of the inferences.
In general, statistical evidence about a group average does not imply that every participant had the same response. Confidence intervals and p-values characterize uncertainty under an analysis model; they are not guarantees about individual outcomes. The posted findings should be read as estimates from this completed Phase 1 crossover study rather than as a universal or definitive conclusion.
14. Why This Trial Matters Statistically
This study raises useful statistical issues because it combines repeated treatment periods, several dose-versus-placebo comparisons and a modeled sleep endpoint. In a crossover design, each participant may contribute observations under different conditions, so participant and period effects can matter. The stated mixed model includes treatment, treatment sequence, period and participants, making clear that the reported estimates are adjusted model-based comparisons rather than simple raw contrasts.
The reported ratios also illustrate why the scale of an outcome comparison matters. WASO is measured in minutes, but the treatment contrasts are ratios of geometric least-squares means. Readers need to distinguish the endpoint summaries in minutes from the relative comparisons and avoid treating a ratio as an absolute difference. Confidence intervals add important context: the estimate alone does not show how precisely the comparison is known.
Multiple doses create multiple comparisons. Without a reported adjustment or testing hierarchy, each p-value needs to be interpreted as its own posted test rather than as a fully controlled family of inferences. The variation in group counts and the absence of documented missing-data methods further show why analysis-population definitions and data availability are part of interpreting a trial, not merely administrative details.
Finally, planning calculations based on an observed effect are not evidence of observed power and are not a substitute for the original design rationale. The registry record's lack of a reported sample-size justification, multiplicity plan and diagnostic details limits what can be concluded about the analysis strategy. These issues make the record a useful example of how design, estimands, precision and reporting completeness shape statistical interpretation.
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 primary analysis population excludes randomized participants who did not receive at least 1 dose or did not have polysomnography measurements in Periods 1, 2 and 3. The registry record does not report how many participants were excluded or why. The posted endpoint counts vary by treatment condition, but the record does not explain those differences or describe the completeness of measurements across periods.
The reported analyses use a mixed model, but key implementation details are not reported in the registry record. These include the covariance structure, distributional assumptions, diagnostic checks, missing-data assumptions and sensitivity analyses. The model's inclusion of treatment, sequence, period and participants is stated, but that alone is not sufficient to reproduce the analysis fully.
Four dose-versus-placebo comparisons are posted for the primary endpoint. The registry record does not report multiplicity control, a testing hierarchy or which comparison, if any, was intended to be primary among the dose contrasts. The posted confidence intervals and p-values therefore do not document adjustment for the set of comparisons.
The enrollment was 24, and the endpoint summaries use groups ranging from 8 to 17 participants. This limits the amount of information available for estimating treatment contrasts. The record does not report a sample-size justification or a clinically important WASO threshold. Consequently, the precision of the estimates can be described from the intervals, but the adequacy of the planned sample size and the practical importance of the observed differences cannot be judged from the registry record alone.
The study assessed single doses in healthy Japanese participants under a transient-insomnia phase advance model. The primary outcome covers an 8-hour recording in Periods 1, 2 and 3, while pharmacokinetic outcomes are listed for Period 4. These features limit the scope of interpretation. The registry record does not report longer-term outcomes, nor does it provide posted results for the listed secondary outcomes in the registry record for this analysis.
16. Related Tutorials
17. Related Statistical Calculators
18. Sources
- ClinicalTrials.gov: A Study of LY2624803 in Japanese Participants With Transient Insomnia — NCT01784614.
- Registry record last updated 2016-01-28; status last verified 2015-12.
- Results first posted 2016-01-28.
- 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.
Continue through the Clinical Biostats statistical pathway
Use this trial's endpoint as a starting point for deeper study of effect measures, confidence intervals, and sample-size planning.
19. Record Summary
NCT01784614 is a completed Phase 1 randomized, double-blind crossover study sponsored by Eli Lilly and Company. It enrolled 24 healthy Japanese participants and evaluated single oral doses of LY2624803 in a 5-hour phase advance model of transient insomnia. The treatment conditions were 0.1 mg, 1.0 mg, 3.0 mg and 6.0 mg LY2624803, and placebo.
The primary outcome was WASO in minutes during an 8-hour recording period in Periods 1, 2 and 3. The registry reports geometric least-squares means, 90% confidence intervals and mixed-model ratios for each dose versus placebo. The posted estimates for 3.0 mg and 6.0 mg had confidence intervals below a ratio of 1; the intervals for 0.1 mg and 1.0 mg do not convey the same degree of separation from that reference value.
Secondary outcomes listed include LPS, TST, and pharmacokinetic measures Cmax and AUC in Period 4. The registry record does not report multiplicity handling, detailed missing-data methods, the trial's sample-size justification or full model diagnostics.