Abnormal Ventilatory Response to Carbon Dioxide: a Potential Biomarker for Seizure Induced Respiratory Depression & Modification by SSRI: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Abnormal Ventilatory Response to Carbon Dioxide: a Potential Biomarker for Seizure Induced Respiratory Depression & Modification by SSRI (NCT02929667), a randomized phase 2 study. The primary outcome is Study Retention Rate.

2. Clinical Question

3. Trial Design

The registry describes a Phase 2, randomized, triple-masked, parallel-group study. Participants were assigned to fluoxetine or placebo. The record lists an enrollment of 30 and a completed status. The study began on 2017-02-16, and its primary completion date is 2019-03-06. These dates and status describe the trial record; they do not establish when every participant completed follow-up.

The two arms used corresponding titration schedules. Fluoxetine and placebo were each described as starting at 10 mg per day for 1 week, increasing to 20 mg per day for 1 week and then 40 mg per day for 2 weeks, followed by 20 mg per day for 1 week and 10 mg per day for 1 week, then stopping. The registry does not report whether participants adhered to this schedule or whether dose changes occurred.

The primary outcome was study retention, measured as participants completing the study every 3 months. Its stated timeframe runs from enrollment until completion or loss to follow-up, with assessments every 3 months up to 2 years and 3 months. The posted primary outcome gives n=11 for each arm. This differs from the overall enrollment of 30, but the registry record does not explain the difference or identify the disposition of participants who are not represented in those outcome counts.

Three secondary outcomes are listed: change in minute ventilation during hypercapnic ventilatory response testing, change in PHQ-9 score, and change in the slope of the hypercapnic ventilatory response. Although five outcome measures are indicated in total, results for these listed secondary outcomes are not reported in the registry record. The record does not report the randomization method, allocation concealment, masking procedures, sample-size justification, or the statistical analysis plan.

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: A total of 6 subjects with HCVR slope of \> 2.0 L/min/mm Hg were excluded.Two additional subjects did not meet other inclusion criteria for randomization.

Overall Study

Analysis population for the primary outcome, as defined in the registry: Due to being a feasibility study, recruitment is reported for the whole population (drug + placebo).

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 Study Retention Rate, measured over From date of enrollment until either completion of study or lost to follow up every 3 months up to 2 years and 3 months.

The primary endpoint is study retention, described as the number of participants completing the study every 3 months. Its timeframe begins at enrollment and continues until study completion or loss to follow-up, with assessments every 3 months up to 2 years and 3 months. Retention can describe whether participants remain in a study over follow-up and can help readers assess how much outcome information may be available. It is not a direct measure of the proposed respiratory mechanism or of seizure-related breathing.

The registry reports a mean of 1.2 participants per 3 months in the fluoxetine arm, with a full range from 0.0 to 4.0, and a mean of 1.1 in the placebo arm, with a full range from 0.0 to 3.0. Each arm has n=11 for this outcome. The outcome description says that the number completing the study was recorded every 3 months, but the record does not explain how those repeated counts were combined into the reported mean or whether the mean represents a particular follow-up interval or an average across intervals. The dispersion is reported as a full range, not as a standard deviation or confidence interval.

Consequently, these figures should be read as the posted retention summaries, not as a retention percentage or as an estimate of respiratory response. The registry does not report a comparison estimate or uncertainty interval for the difference between arms.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The statistical information available for interpreting the primary outcome is limited to arm-level summaries: the number of participants included, the mean, and the full range. The registry describes the outcome as participants completing the study every 3 months. Understanding what these summaries can and cannot say is important because the outcome is based on repeated follow-up and is not described as a simple one-time measurement.

What does the mean represent?

A mean is an average of recorded values. Here, the posted means are 1.2 participants per 3 months for fluoxetine and 1.1 for placebo. The unit and description indicate that the values concern participant completion counts over follow-up. The registry does not explain the calculation used to turn the repeated completion counts into one mean for each arm. For example, it does not specify whether all scheduled intervals contributed equally, how incomplete follow-up was handled, or whether the mean corresponds to a particular set of intervals. Those details matter because different aggregation choices can produce summaries with different interpretations.

These means should not be converted into completion percentages. A percentage would require a defined denominator and a clearly specified interval or follow-up summary. Although overall enrollment is reported as 30, the primary outcome lists n=11 in each arm, and the record does not explain the relationship between those counts. The outcome means therefore describe the posted measure, not a proportion of all enrolled participants.

What does the full range show?

The full range gives the lowest and highest recorded values for an arm. The reported ranges are 0.0 to 4.0 for fluoxetine and 0.0 to 3.0 for placebo. A range can show that the observations varied, but it does not show how frequently values occurred within that span. It is sensitive to the most extreme observations and does not provide the same information as a standard deviation, standard error, or confidence interval. None of those measures of uncertainty is posted for this outcome.

Why is a confidence interval useful?

In general, a confidence interval expresses uncertainty around an estimated quantity under a specified statistical method. For a treatment comparison, an interval around the estimated effect can show which effect sizes are reasonably compatible with the data and model. It also helps distinguish a point estimate from the precision of that estimate. The registry record does not report a confidence interval for the retention outcome, so the precision of an arm-to-arm effect cannot be assessed from the posted summaries.

What can randomization contribute?

In general, random assignment aims to make treatment groups comparable with respect to both measured and unmeasured characteristics, on average. That design can support a causal comparison when the allocation process is implemented appropriately and the analysis respects the randomized groups. The record identifies the study as randomized, but does not describe the sequence-generation method, allocation concealment, the analysis population, or how missing outcomes were handled. The label alone does not establish that every potential source of bias was prevented.

The record also identifies the design as triple-masked, but does not state which people were masked or how masking was maintained. Masking can reduce the influence of expectations on care, participant reporting, and outcome assessment, depending on which parties are masked and how the outcome is measured. The registry entry does not provide enough detail to assess those mechanisms for this trial.

Why can’t the result establish a treatment difference?

A comparative analysis would ordinarily estimate a contrast between arms and quantify its uncertainty, using a method suited to the outcome and design. No analysis entries are posted for a two-arm comparison, and no effect estimate, confidence interval, or test result for retention appears in the record. The means alone are descriptive. A reader should not infer statistical significance or a treatment effect by comparing the rounded means or ranges.

Repeated completion counts also raise methodological questions that cannot be resolved from the posted summary. Measurements over time from the same participants are related rather than independent. A formal analysis may need to account for that dependence and define how follow-up time is represented. The registry record does not report whether the outcome was analyzed as repeated counts, a time-to-event measure, or another form of summary. It would be speculative to assign a particular method to the trial.

In general, over-dispersion is a concern when count data vary more than a chosen statistical model expects. If a count model were considered, its assumptions about variation would need evaluation. The record does not say that a count model was used, so this is a general methodological consideration, not a description of the trial’s analysis.

The registry lists several secondary outcomes, but their numerical results and analyses are not reported. There is consequently no basis in the record for assessing estimates, uncertainty, or consistency across those outcomes. The primary endpoint summaries should be interpreted as descriptive retention information, not as evidence about the proposed ventilatory mechanism or as a complete statistical evaluation of the trial.

10. Multiplicity, Missing Data and Analysis Populations

The record lists the primary outcome of study retention and three secondary outcomes: change in minute ventilation during hypercapnic ventilatory response testing, change in PHQ-9 score, and change in the slope of the hypercapnic ventilatory response. Five outcome measures are indicated in total. When a study evaluates multiple outcomes, the chance of obtaining at least one apparently notable result can increase if many separate tests are conducted. General approaches to this issue include defining a primary outcome in advance, specifying a testing hierarchy, or adjusting interpretation or inference for multiple comparisons. The registry record does not report a multiplicity plan, testing hierarchy, or adjustment method.

The record contains no posted analysis entries for a two-arm comparison. Results for the listed secondary outcomes are not reported in the registry record. It is therefore not possible to determine which outcomes were analyzed, whether there were additional analyses, or how any multiplicity was handled. The presence of several registered outcomes alone does not establish that all were measured or analyzed as planned.

Missing-data information is also limited. The primary outcome description includes completion or loss to follow-up, and the stated timeframe extends to 2 years and 3 months. However, the record does not report the number or timing of losses to follow-up, withdrawals, protocol deviations, or missing measurements. It does not state whether missing observations were imputed, excluded, or addressed through another method.

The overall enrollment is 30, while the posted primary outcome lists n=11 for each arm. The record does not explain this difference. Without participant flow and an account of outcome availability, it is not possible to assess whether the outcome summaries represent all randomized participants, a subset with observed data, or another analysis population. The record does not report whether the primary analysis followed an intention-to-treat principle or used a different population. These omissions limit interpretation of both the retention summaries and any possible 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 reports serious adverse events for 11 participants at risk in the fluoxetine arm and 11 in the placebo arm. No serious adverse events were reported in either arm, and deaths were reported as zero in both arms. The record also reports one other, non-serious adverse event in the fluoxetine arm and none in the placebo arm. These are counts in the posted safety summary; they are not estimates of comparative risk and are not accompanied by an uncertainty interval or statistical test.

Serious adverse events are distinct from the category of other adverse events in the record. The absence of reported serious events in these groups does not demonstrate that no such events could occur, and the small arm-level counts limit how much can be learned about uncommon events. The registry does not provide event descriptions for the one other adverse event, a fuller account of how events were collected, or a detailed safety analysis. The record lists no top serious adverse event terms because no serious events are reported in the posted group summaries.

These figures should be read as the adverse-event counts recorded for the participants at risk, not as a comprehensive judgment of safety or as evidence for a treatment comparison.

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

The posted results answer a narrow question about retention summaries. They report a mean of 1.2 participants per 3 months for fluoxetine, with a full range of 0.0 to 4.0, and a mean of 1.1 for placebo, with a full range of 0.0 to 3.0. Each arm has n=11 for the primary outcome. These values describe the available registry summary; they do not provide a formal estimate of the difference between arms or show how precise such a difference would be.

The primary endpoint is not the physiological outcome that motivated the trial. The study rationale proposes a relationship between central responsiveness to carbon dioxide and peri-ictal respiratory depression, and proposes that fluoxetine might affect that response. The listed secondary outcomes address ventilatory response measures and PHQ-9 score, but their results are not reported in the registry record. There is also no posted SUDEP outcome comparison. Accordingly, the retention data cannot establish whether fluoxetine altered breathing responses, peri-ictal respiratory depression, or SUDEP risk.

Nor can the arm-level retention summaries establish that the treatments were equivalent. A conclusion of equivalence would require an appropriate design and analysis with a prespecified margin and an estimate precise enough to exclude differences outside that margin. Such information is not reported in the registry record. Likewise, the descriptive values alone do not establish a statistically significant difference.

The registry states an enrollment of 30 but reports n=11 in each arm for the primary outcome. Because participant flow and missing-data handling are not reported, the relationship between these counts cannot be resolved. The summary should not be treated as a complete account of follow-up or as a substitute for a participant-level analysis.

The record does establish that a randomized, triple-masked, parallel-group Phase 2 study was registered as completed, with fluoxetine and placebo arms and a retention endpoint. Its posted results are limited: retention summaries and adverse-event counts are available, while comparative inference and the listed physiological outcomes are not. Interpretation should remain within those boundaries.

13. Why This Trial Matters Statistically

This trial is statistically interesting because its scientific rationale and its posted primary outcome address different questions. The investigators’ stated rationale concerns carbon-dioxide sensing, ventilatory response, and breathing changes around seizures. The primary endpoint reported in the registry is study retention. Retention can help characterize participation over follow-up, but it does not directly measure the proposed biological pathway. Separating the registered primary outcome from the mechanistic rationale is essential when interpreting what the results can support.

The record also illustrates why an outcome’s operational definition matters. A repeated count of participants completing study assessments can be summarized in several ways, and the registry gives an arm-level mean and range without explaining how repeated counts were combined. That ambiguity affects the meaning of the mean and makes it difficult to reproduce or compare the summary. A clearly stated denominator, follow-up interval, and aggregation rule would help readers understand the estimand—the quantity the analysis is intended to estimate.

Randomization and masking are important design features, but they do not replace transparent reporting of participant flow, outcome definitions, analysis populations, and statistical methods. The overall enrollment and primary-outcome counts differ, while the registry does not explain the relationship. Without that context, readers cannot assess completeness of follow-up or the role of missing data.

Finally, descriptive summaries are not substitutes for comparative estimates and uncertainty. The posted means and ranges show the observed retention summaries, but there is no reported treatment-effect estimate or confidence interval. This distinction is central to biostatistical interpretation: values observed in two groups do not by themselves establish whether an underlying treatment contrast is present, absent, or estimated precisely. The unreported secondary outcomes further limit assessment of whether the study addressed its mechanistic questions.

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.

Several trial-specific limitations of interpretation arise from the registry record. The primary endpoint is described as the number of participants completing the study every 3 months, but the record does not explain how those repeated values were combined into the reported arm-level means. It also does not specify the denominator for each interval or whether the summary reflects all scheduled assessments. That makes the scale and practical interpretation of the mean less clear than a fully defined retention rate.

The reported primary-outcome sample size is n=11 in each arm, whereas overall enrollment is 30. Participant disposition, reasons for missing outcome data, and the analysis population are not reported. The record therefore does not show whether the primary summaries include all randomized participants or a subset, or how any missing observations were handled. This constrains interpretation of the outcome even before considering treatment comparisons.

No effect estimate, confidence interval, or inferential test is posted for the primary outcome. The statistical analysis plan, sample-size justification, and randomization and masking procedures are not reported in the registry record. There is not enough information to assess the planned estimand, model assumptions, or whether analysis decisions followed a prespecified plan.

The trial’s mechanistic rationale is not directly resolved by the posted primary results. Although ventilatory and PHQ-9 outcomes are listed, their numerical results are not reported. The registry record also does not provide a SUDEP outcome comparison. The adverse-event summary is limited to group counts and does not include detailed event descriptions or a comparative safety analysis. These limitations concern what can be concluded from the registry record; they do not establish what other information may exist outside it.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Abnormal Ventilatory Response to Carbon Dioxide: a Potential Biomarker for Seizure Induced Respiratory Depression & Modification by SSRI — NCT02929667.
• Registry record last updated 2020-05-04; status last verified 2020-04.
• Results first posted 2020-05-04.
• 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

NCT02929667 is a completed Phase 2 randomized, triple-masked, parallel-group study of fluoxetine versus placebo in people with epilepsy. The record lists an enrollment of 30 and dates the study from 2017-02-16, with primary completion on 2019-03-06. The primary outcome is study retention, described as the number of participants completing the study every 3 months through completion or loss to follow-up, up to 2 years and 3 months.

The posted primary summaries include n=11 in each arm, a mean of 1.2 for fluoxetine and 1.1 for placebo, and full ranges of 0.0 to 4.0 and 0.0 to 3.0, respectively. The relationship between the reported outcome counts and overall enrollment is not explained. Serious adverse events and deaths are reported as zero in both arms; one other adverse event is reported in the fluoxetine arm and none in the placebo arm.

Three secondary outcomes are listed, including ventilatory response measures and PHQ-9 score, but their results are not reported in the registry record. No inferential comparison of the primary outcome is posted.