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 Evaluating Efficacy and Safety of E2007 (Perampanel) Given as Adjunctive Therapy in Subjects With Refractory Partial Seizures (NCT00700310), a randomized phase 3 study. The primary outcome is Percent Change in the 28-day Seizure Frequency From Baseline to the End of the Double-blind Phase (Titration and Maintenance Phases).
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT00700310 |
| Phase | Phase 3 |
| Condition | Refractory Partial Seizures |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | Double |
| Enrollment | 712 |
| Arms in the primary outcome | 4 |
| Lead sponsor | Eisai Inc. |
| Results posted | Yes |
2. Clinical Question
Population
Participants with refractory partial seizures. The primary analysis population was the Full Intent-to-Treat (ITT) Analysis Set: randomized participants who received study drug and had any seizure-frequency data during the double-blind phase.
Intervention
Perampanel given as adjunctive therapy, with groups listed as 2 mg, 4 mg, and 8 mg.
Comparator
Placebo, also given as adjunctive therapy in the randomized, double-blind study.
Primary question
What was the median percent change in seizure frequency per 28 days from baseline through the end of the double-blind phase, which extended through Week 19?
The trial asked whether seizure frequency changed during the study in participants assigned to perampanel or placebo. The primary measure was based on seizure frequency per 28 days, derived from information in participant diaries. The reported endpoint is a change from baseline, expressed as a percentage, and the summary statistic is the median for each group.
This is a question about the observed distribution of change within each randomized group. The registry record does not post a direct statistical comparison between each perampanel group and placebo, such as an estimated treatment contrast with a confidence interval or a p-value. The group medians can therefore be read descriptively, but they do not by themselves establish the size or precision of a difference between groups.
The record lists responder rate and percent change in complex partial plus secondarily generalized seizure frequency as secondary outcomes. Those outcomes provide additional ways to characterize seizure experience, but their numerical results are not reported in the registry record. The posted primary results consequently address a particular summary of seizure-frequency change rather than every seizure-related outcome measured or planned.
3. Trial Design
This was a randomized, double-blind, placebo-controlled, parallel-group phase 3 study of perampanel as adjunctive therapy in participants with refractory partial seizures. The registry lists enrollment of 712. The study is recorded as completed, with a start date of 2008-08 and primary completion in 2009-08. Eisai Inc. is listed as the sponsor.
Participants were assigned to placebo or one of three perampanel groups: 2 mg, 4 mg, and 8 mg. The intervention list describes the drug assignments, but the arm descriptions are blank. Further details of dose escalation, including how doses were increased and the timing of changes, are not reported in the registry record. The official title describes the study as dose-escalation, but that title does not supply the missing operational details.
Double masking means that the study was designed so that participants and relevant study personnel did not know the assigned treatment during the masked phase. Random allocation is intended to reduce systematic differences between groups at the point of assignment. These design features support comparison, but they do not replace reporting of the analysis used to estimate treatment differences.
The primary outcome timeframe runs from baseline, before randomization, through Week 19, at the end of the double-blind phase. The endpoint description refers to the titration and maintenance phases. The primary population was not simply everyone enrolled: the registry defines it as randomized participants who received study drug and had any seizure-frequency data during the double-blind phase. The reported group sizes for this endpoint are 184 for placebo, 180 for perampanel 2 mg, 172 for perampanel 4 mg, and 169 for perampanel 8 mg.
Additional protocol details, including the planned sample-size justification, allocation procedures, masking procedures, and statistical analysis plan, are not reported in the registry record. The posted information identifies the broad design and groups but does not establish how all design decisions were implemented.
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
| Stage | Placebo | Perampanel 2mg | Perampanel 4mg | Perampanel 8 mg |
|---|---|---|---|---|
| STARTED | 187 | 180 | 174 | 171 |
| COMPLETED | 166 | 154 | 158 | 145 |
| NOT COMPLETED | 21 | 26 | 16 | 26 |
Reasons for not completing this period:
| Reason | Placebo | Perampanel 2mg | Perampanel 4mg | Perampanel 8 mg |
|---|---|---|---|---|
| Adverse Event | 6 | 10 | 5 | 11 |
| Lost to Follow-up | 4 | 1 | 0 | 1 |
| Withdrawal by Subject | 8 | 9 | 8 | 8 |
| Lack of Efficacy | 0 | 3 | 0 | 1 |
| Administrative/Other | 1 | 3 | 1 | 3 |
| Randomized, Not Treated | 2 | 0 | 2 | 2 |
Analysis population for the primary outcome, as defined in the registry: Full Intent-to-Treat (ITT) Analysis Set - group of subjects who were randomized to study drug, received study drug, and had any seizure frequency data during the Double-blind Phase.
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.
| Characteristic | Unit | Placebo | Perampanel 2mg | Perampanel 4mg | Perampanel 8 mg | Total |
|---|---|---|---|---|---|---|
| Participants analysed | 185 | 180 | 172 | 169 | 706 | |
| Age, Customized: <18 Years | Participants | 14 | 21 | 13 | 12 | 60 |
| Age, Customized: 18-64 Years | Participants | 169 | 156 | 158 | 153 | 636 |
| Age, Customized: >64 Years | Participants | 2 | 3 | 1 | 4 | 10 |
| Sex: Female, Male: Female | Participants | 90 | 95 | 84 | 92 | 361 |
| Sex: Female, Male: Male | Participants | 95 | 85 | 88 | 77 | 345 |
| Race/Ethnicity, Customized: White | Participants | 119 | 119 | 105 | 116 | 459 |
| Race/Ethnicity, Customized: Asian | Participants | 34 | 35 | 37 | 28 | 134 |
| Race/Ethnicity, Customized: Chinese | Participants | 31 | 25 | 29 | 25 | 110 |
| Race/Ethnicity, Customized: Other | Participants | 1 | 1 | 1 | 0 | 3 |
6. Primary Endpoint
The registry lists the primary outcome as Percent Change in the 28-day Seizure Frequency From Baseline to the End of the Double-blind Phase (Titration and Maintenance Phases), measured over Baseline (Pre-randomization) through Week 19.
| Element | Registry entry |
|---|---|
| Result type | Median |
| Measure of spread | Full Range |
| Unit | Percent Change |
| Endpoint classification | not reduced to a single comparison |
The primary endpoint is the percent change in seizure frequency per 28 days from baseline through the end of the double-blind phase, at Week 19. Seizure frequency was derived from information recorded in participant diaries. The registry reports the median and the full range for each group.
A percent change expresses a participant’s follow-up seizure frequency in relation to that participant’s baseline frequency. In general, a negative change indicates a lower frequency than at baseline, while a positive change indicates a higher frequency. The group median is the middle value in the distribution of participant-level changes; it is not the average change and does not show how many participants had a particular result.
Seizure frequency is a direct way to summarize the occurrence of the seizures targeted in this study. Expressing frequency per 28 days provides a common time unit for the diary-derived measure. A change-from-baseline endpoint also accounts for each participant’s starting frequency when describing the outcome, although the registry does not report the detailed calculation rules beyond the endpoint description.
The median is useful when participant outcomes are unevenly distributed or include extreme values, because it is less affected by extremes than an arithmetic mean. The full range adds information about the spread and extremes, but it does not show the distribution’s shape or the uncertainty around the median. These summaries describe observed outcomes; they are not, by themselves, estimates of a treatment effect relative to placebo.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The posted primary results are median percent changes in seizure frequency per 28 days. The placebo group’s median was -10.69, with a full range from -100.0 to 420.6. The perampanel 2 mg group’s median was -13.63, with a range from -100.0 to 346.3. For perampanel 4 mg, the median was -23.33 and the range was -100.0 to 416.0. For perampanel 8 mg, the median was -30.8 and the range was -100.0 to 390.6.
Each group median is negative, indicating that the middle participant-level change in each group was a decrease relative to baseline. The medians differ across the listed groups, and the sequence of medians is increasingly negative across the perampanel dose groups. This is a descriptive pattern, not a reported statistical test of a dose trend or of any perampanel group against placebo. No treatment-effect estimates, confidence intervals, or p-values are reported in the registry record.
The ranges show substantial variability in observed percentage change within each group. Every group includes a minimum of -100.0, while the maximum is positive and differs by group. Positive maximum values indicate that some recorded changes were increases relative to baseline. The range alone does not reveal how many participants had increases, where most outcomes fell, or how much precision there is around a group median. It also cannot show whether the distributions in two groups differ reliably.
The primary endpoint analysis population contains 184 participants in the placebo group, 180 in the perampanel 2 mg group, 172 in the 4 mg group, and 169 in the 8 mg group. These are the endpoint group sizes stated for the Full ITT Analysis Set. They should not be confused with the overall enrollment or with the adverse-event denominators, which are listed separately in the safety results.
The registry lists responder rate and percent change in complex partial plus secondarily generalized seizure frequency as secondary outcomes, but numerical results for those outcomes are not reported in the registry record. There are no posted analysis records. Thus, the available results support a description of group medians and ranges, but not a complete account of adjusted comparisons, statistical uncertainty, or the broader outcome pattern.
For the adverse-event results, serious adverse events affected 9 participants in the placebo group and 6 in each perampanel group. Other adverse events affected 48 participants on placebo, 61 on perampanel 2 mg, 66 on perampanel 4 mg, and 79 on perampanel 8 mg. These counts are descriptive. They do not establish that treatment assignment caused an event or that observed differences between groups are statistically distinguishable.
9. Statistical Methods Explained
The registry reports a median and full range for the primary endpoint in each randomized group. It does not report the inferential method used to compare groups. The concepts below are general explanations of what these summaries mean and what additional methods can contribute; they should not be read as a description of an unreported analysis plan.
What does a median summarize?
General: A median is the central value after observations are ordered. It divides the observations so that half are at or below the median and half are at or above it, subject to ties and the conventions used for a particular dataset. Unlike a mean, it does not depend on adding all observations and dividing by their count. This can make it a useful summary when outcomes are skewed or include unusually large values.
For this endpoint, the reported median summarizes participant-level percent changes in seizure frequency. It does not mean that every participant experienced the median change. Nor does it identify the proportion whose seizure frequency decreased. A group-level median also does not show the treatment contrast: a contrast would quantify how outcomes in one group compare with outcomes in another under a specified statistical method.
What does the full range tell the reader?
General: A range reports the smallest and largest observed values. It conveys the span of the observations, including extremes, but leaves out the values between those extremes. Two groups can have the same range and very different distributions, or different ranges while having similar central patterns. The range is not a confidence interval and should not be read as a measure of precision around the median.
The posted ranges extend from -100.0 to a positive maximum in each group. This communicates that the observed outcomes included both a complete reduction relative to baseline at the lower end and increases at the upper end. The ranges do not tell how frequent either extreme was, whether values clustered near the median, or how much the group distributions overlapped.
Why are confidence intervals and tests relevant?
General: A confidence interval describes uncertainty around an estimated quantity under a specified statistical model and procedure. A confidence interval for a between-group effect would help show the range of effects compatible with the data and method. A hypothesis test produces a measure of how compatible the data are with a specified null hypothesis, subject to the assumptions of the test. Neither a confidence interval nor a p-value alone establishes clinical importance.
No confidence intervals, p-values, or between-group effect estimates for the primary endpoint are reported in the registry record. Consequently, readers cannot use the posted primary summaries to assess the statistical precision of a difference between perampanel and placebo. The pattern in medians is descriptive and does not substitute for those quantities.
Why might the choice of summary and model matter?
General: The statistical method should reflect the outcome’s distribution, the study design, and the question being asked. Seizure-frequency outcomes can have uneven distributions, and participant-level changes can vary considerably. A method designed for means may answer a different question from a method focused on medians or ranks. Adjustment for baseline or other prespecified variables can also change the estimate and its precision. Which method is appropriate depends on the analysis plan and data properties.
Over-dispersion is a general feature of count data in which the variability exceeds what a simple model assumes. If present, a model that does not account for it can misstate uncertainty. This is a general statistical consideration, not a claim about the distribution in this trial. The registry record does not report the model, distributional checks, covariate adjustments, or other procedures used for the primary endpoint.
How should planning illustrations be understood?
General: Sample-size calculations estimate how many participants may be needed to detect a specified effect under stated assumptions. They are planning illustrations, not evidence about the probability that a completed study’s result is true. In particular, an observed result should not be described using “observed power.” Any planning figures associated with the posted results use the stated assumptions of a two-sided significance level of 0.05, equal allocation, and a normal approximation, with the effect defined as a fraction of the observed effect. The registry’s own sample-size justification is not reported in the registry record. These illustrations therefore do not establish the trial’s original design rationale.
The primary results give group-specific medians and ranges, while the design is randomized and parallel-group. A complete statistical interpretation would also require the prespecified estimand, analysis method, handling of missing diary data, and multiplicity strategy. Those details are not reported in the registry record. Without them, the posted descriptive summaries can be interpreted on their stated scale, but they cannot be expanded into an unreported inferential conclusion.
10. Multiplicity, Missing Data and Analysis Populations
The registry lists one primary outcome and two secondary outcomes: responder rate, and percent change in 28-day complex partial plus secondarily generalized seizure frequency from baseline to the end of the double-blind phase. Numerical results for the secondary outcomes are not reported in the registry record. The record does not provide a hierarchy of hypotheses, a family of confirmatory tests, or a multiplicity adjustment. It is therefore not possible to determine how the study handled the risk of false-positive findings across multiple outcomes or comparisons.
No analysis records are posted, and the statistical analysis plan is not reported in the registry record. The primary endpoint is presented as group medians and full ranges, without a reported treatment contrast, confidence interval, or p-value. Whether analyses were adjusted for baseline values or other covariates is not reported in the registry record.
The primary analysis population is defined as randomized participants who received study drug and had any seizure-frequency data during the double-blind phase. This definition excludes participants without any such data, but the record does not state how partially missing diary data were handled, how missing periods were treated, or whether any imputation or sensitivity analyses were conducted. Those details are not reported in the registry record.
The outcome group sizes are listed separately from adverse-event denominators. The safety record gives participants at risk for serious and other adverse events; these denominators should be read as the safety populations reported for those outcomes, not assumed to be identical to the primary endpoint population. The reasons for differences in denominators are not reported in the registry record.
Multiplicity and missing-data decisions can affect both the interpretation and precision of results. Without the prespecified strategy and supporting analyses, the record permits description of the posted primary summaries but not a full evaluation of those analytical choices.
11. 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 |
|---|---|---|---|
| Placebo | 9 | 185 | 4.9% |
| Perampanel 2mg | 6 | 180 | 3.3% |
| Perampanel 4mg | 6 | 172 | 3.5% |
| Perampanel 8 mg | 6 | 169 | 3.6% |
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 |
|---|---|
| Placebo | 48 of 185 |
| Perampanel 2mg | 61 of 180 |
| Perampanel 4mg | 66 of 172 |
| Perampanel 8 mg | 79 of 169 |
Most frequent serious adverse event terms
| Event term | Organ system | Placebo | Perampanel 2mg | Perampanel 4mg | Perampanel 8 mg |
|---|---|---|---|---|---|
| Convulsion | Nervous system disorders | 3 of 185 | 0 of 180 | 1 of 172 | 0 of 169 |
| Wound infection | Infections and infestations | 1 of 185 | 0 of 180 | 0 of 172 | 1 of 169 |
| Epilepsy | Nervous system disorders | 1 of 185 | 0 of 180 | 1 of 172 | 0 of 169 |
| Goitre | Endocrine disorders | 0 of 185 | 1 of 180 | 0 of 172 | 0 of 169 |
| Conjunctivitis allergic | Eye disorders | 0 of 185 | 0 of 180 | 1 of 172 | 0 of 169 |
| Iritis | Eye disorders | 0 of 185 | 0 of 180 | 1 of 172 | 0 of 169 |
| Cholelithiasis | Hepatobiliary disorders | 0 of 185 | 0 of 180 | 1 of 172 | 0 of 169 |
| Appendicitis | Infections and infestations | 0 of 185 | 0 of 180 | 0 of 172 | 1 of 169 |
| Bronchitis | Infections and infestations | 1 of 185 | 0 of 180 | 0 of 172 | 0 of 169 |
| Orchitis | Infections and infestations | 1 of 185 | 0 of 180 | 0 of 172 | 0 of 169 |
Most frequent non-serious adverse event terms
| Event term | Organ system | Placebo | Perampanel 2mg | Perampanel 4mg | Perampanel 8 mg |
|---|---|---|---|---|---|
| Dizziness | Nervous system disorders | 18 of 185 | 18 of 180 | 28 of 172 | 45 of 169 |
| Somnolence | Nervous system disorders | 12 of 185 | 22 of 180 | 15 of 172 | 27 of 169 |
| Headache | Nervous system disorders | 16 of 185 | 16 of 180 | 19 of 172 | 18 of 169 |
| Fatigue | General disorders | 5 of 185 | 8 of 180 | 13 of 172 | 9 of 169 |
| Upper respiratory tract infection | Infections and infestations | 5 of 185 | 11 of 180 | 6 of 172 | 3 of 169 |
| Nasopharyngitis | Infections and infestations | 3 of 185 | 7 of 180 | 9 of 172 | 3 of 169 |
| Gait disturbance | General disorders | 2 of 185 | 1 of 180 | 2 of 172 | 9 of 169 |
Serious events with the highest arm-level rate: convulsion (1.6%), appendicitis (0.6%), wound infection (0.6%), contusion (0.6%), head injury (0.6%).
The safety results list participants at risk and participants affected by serious and other adverse events in each group. For serious adverse events, 9 of the 185 participants at risk in the placebo group were affected. The corresponding affected counts were 6 of 180 for perampanel 2 mg, 6 of 172 for perampanel 4 mg, and 6 of 169 for perampanel 8 mg. For other adverse events, the affected counts were 48 of 185 on placebo, 61 of 180 on perampanel 2 mg, 66 of 172 on perampanel 4 mg, and 79 of 169 on perampanel 8 mg.
These are counts of participants, not counts of distinct event episodes, and the registry does not provide event-level detail for all adverse events in these summaries. The serious-event terms listed include convulsion at 1.6%, appendicitis at 0.6%, wound infection at 0.6%, contusion at 0.6%, head injury at 0.6%, and post concussion syndrome at 0.6%. The record does not explain in this summary whether an individual could contribute to more than one listed term.
Adverse-event counts describe what was recorded in each group; they do not establish that an event was caused by study treatment. No comparative safety analysis or uncertainty estimate is reported in the registry record. Death counts are not reported in the registry record, so the missing death values should not be interpreted as zero deaths.
12. What the Result Does, and Does Not, Mean
The primary endpoint results show negative median percent changes in 28-day seizure frequency in all four groups over the period from baseline through Week 19. The medians are -10.69 for placebo, -13.63 for perampanel 2 mg, -23.33 for perampanel 4 mg, and -30.8 for perampanel 8 mg. The reported ranges show that individual participant outcomes varied widely, including positive changes in every group.
These summaries do not establish whether a difference between any perampanel group and placebo is statistically distinguishable from chance. The registry record does not report a treatment-effect estimate, confidence interval, p-value, or the inferential analysis method. The median pattern also does not by itself show how many participants experienced a reduction, how consistent outcomes were, or whether the pattern would persist under a different analysis.
The endpoint concerns change in seizure frequency as captured by participant diaries. It does not represent every aspect of seizure experience, functioning, quality of life, or safety. The registry lists additional seizure-related outcomes, but their numerical results are not reported in the registry record. The adverse-event counts offer a separate descriptive view and should not be treated as proof of causation or as a complete comparative safety analysis.
The trial’s randomized, double-blind, placebo-controlled design is relevant when considering comparisons, but a design description cannot fill gaps in the posted analysis. Important information about missing diary data, the analysis plan, multiplicity, and uncertainty is not reported in the registry record. The results therefore support a careful account of the posted group summaries, not a definitive claim about the magnitude, precision, or clinical importance of a treatment effect.
13. Why This Trial Matters Statistically
This trial illustrates the difference between reporting group outcomes and reporting a comparative treatment effect. The primary endpoint is summarized with a median and full range in each randomized group. Those statistics are informative about the center and span of observed participant-level changes, but they do not quantify the uncertainty around the difference between a perampanel group and placebo.
The endpoint also highlights why the summary statistic matters. Participant-level percentage changes can extend from a complete reduction to substantial increases, as reflected by the posted full ranges. A median offers a central summary that is less influenced by extreme observations than a mean, while the range makes the extremes visible. Neither alone describes the full distribution or establishes how outcomes compare between groups.
The record provides a useful opportunity to distinguish design strength from analytical completeness. Randomization, masking, and a placebo group are important design features. Yet readers also need the analysis method, treatment-effect estimates, uncertainty intervals, and decisions about missing observations and multiple outcomes to evaluate the statistical evidence fully. Those details are not reported in the registry record.
Finally, the safety information demonstrates why denominators and outcome definitions matter. Serious and other adverse-event counts are presented against group-specific at-risk numbers, while death values are missing. Treating an unreported value as zero or comparing raw counts without regard to denominators would go beyond the record. A disciplined reading separates what is numerically posted from what remains uncertain.
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 main limitation of the posted efficacy results is that they are descriptive summaries by group. The registry record does not report a direct comparison of perampanel with placebo for the primary outcome, an effect estimate, a confidence interval, or a p-value. Consequently, the precision and statistical compatibility of between-group differences cannot be assessed from the posted summaries.
The full ranges indicate wide variation, but no additional distributional summaries are reported. The record does not give individual-level data, quantiles beyond the median, or the proportion of participants meeting a responder definition. Although responder rate is listed as a secondary outcome, its numerical results are not reported in the registry record.
Several analytical details are absent. The statistical analysis plan, handling of missing diary entries, model specification, baseline adjustment, and sensitivity analyses are not reported in the registry record. This limits assessment of how the diary-based endpoint was constructed and how robust the summaries are to missing or incomplete observations. Multiplicity handling for the primary and secondary outcomes is also not reported.
The safety results are counts without comparative uncertainty estimates. The registry’s death values are not reported, and the reasons for differences between endpoint and safety denominators are not reported in the registry record. The specific details of dose escalation and arm procedures are also not provided in the arm descriptions. These gaps constrain interpretation of both the observed outcomes and the conduct of the study; they should not be filled by assumptions.
15. Related Tutorials
16. Related Statistical Calculators
17. Sources
- ClinicalTrials.gov: Evaluating Efficacy and Safety of E2007 (Perampanel) Given as Adjunctive Therapy in Subjects With Refractory Partial Seizures — NCT00700310.
- Registry record last updated 2016-01-21; status last verified 2015-11.
- Results first posted 2012-11-22.
- 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.
18. Record Summary
The registry record describes a completed randomized, double-blind, placebo-controlled, parallel-group phase 3 study of perampanel as adjunctive therapy for refractory partial seizures. Enrollment was 712. The primary outcome was median percent change in 28-day seizure frequency from baseline through Week 19, in the Full Intent-to-Treat Analysis Set.
The reported medians were -10.69 for placebo, -13.63 for perampanel 2 mg, -23.33 for perampanel 4 mg, and -30.8 for perampanel 8 mg. Full ranges were reported for each group. The registry lists two secondary outcomes, but their numerical results are not reported in the registry record. Serious and other adverse-event counts are posted by group; death counts are not reported in the registry record.
No inferential analyses are posted. The record does not report a treatment-effect estimate, confidence interval, p-value, statistical analysis plan, or handling of missing diary data. The available results therefore support descriptive interpretation of group summaries, with important uncertainty about comparative effects and analytical decisions.