The Role of Hyperoxia in Acute Ischemic Stroke: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for The Role of Hyperoxia in Acute Ischemic Stroke (NCT03904017), a randomized na study. The primary outcome is Mean Time to Randomization and Initiation of Intervention.

2. Clinical Question

The study was intended to evaluate hyperoxia in the immediate ischemic period following stroke in a controlled emergency department setting. Its stated broader aims included examining markers of free radical damage and inflammatory markers associated with hyperoxic lung injury. The registry also lists the mean ratio of initial hypoperfused tissue volume to final infarct volume and change in mean NIHSS as secondary outcomes.

The posted primary result therefore addresses how long it took to begin the assigned intervention after emergency department arrival. It does not, by itself, answer whether the groups differed in neurological outcomes, tissue injury, or biological markers. Although two secondary outcomes are listed, their numerical results are not reported in the registry record.

3. Trial Design

This was a randomized, double-masked, parallel-group trial of hyperoxia versus placebo in acute ischemic stroke. Participants in the hyperoxia arm received supplemental oxygen; participants in the placebo arm received medical air. Both were delivered at 15liters per minute through a partial non-rebreather facemask. The registry lists the study phase as NA.

The registry reports enrollment of 13 and a status of terminated. The primary endpoint description gives group counts of 3 for hyperoxia and 10 for placebo. The reason for termination is not reported in the registry record. The start date is 2019-06-28 and the primary completion date is 2023-02-01.

Random allocation is intended to assign participants to groups by chance, while masking is intended to limit the influence of knowledge about assignment on study conduct or assessment. A parallel design means participants are assigned to one group rather than crossing between the two treatments. These design features do not, on their own, establish how the trial was implemented or whether all potential sources of bias were controlled. Details such as the randomization procedure, who was masked, and how masking was maintained are not reported in the registry record.

The primary endpoint is a process measure: the elapsed time from emergency department presentation to initiation of the assigned intervention. The posted results describe this measure using a mean and standard deviation for each arm. No inferential comparison or analysis result is posted. The registry lists three outcome measures in total, including two secondary outcomes, but numerical results for those secondary outcomes 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

Analysis population for the primary outcome, as defined in the registry: The time from ED arrival to initiation of treatment arm was measured in minutes for 3 patients receiving hyperoxia and 10 patients receiving 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 Mean Time to Randomization and Initiation of Intervention, measured over baseline.

The primary endpoint is the mean time from emergency department arrival to randomization and initiation of the intervention, measured in minutes. The endpoint description says that arrival was documented, participants were screened, and intervention initiation was recorded after consent and randomization. The reported summary is the time from emergency department presentation to initiation of the assigned treatment arm.

This timing measure is useful for describing how quickly study treatment began in the emergency department. In a study focused on an intervention intended for the immediate ischemic period, the interval between arrival and intervention is relevant to understanding when exposure started. However, this endpoint is not a measure of stroke recovery, infarct size, or biological response. It describes the timing of study procedures, rather than the effect of the assigned intervention on those outcomes.

The hyperoxia group had a mean of 96.7 minutes, with a standard deviation of 27.3, among 3 participants. The placebo group had a mean of 75.1 minutes, with a standard deviation of 14.8, among 10 participants. These descriptive summaries do not establish why the times differed or whether the difference is due to chance, operational factors, or other influences. A confidence interval and a formal statistical comparison are not reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry identifies the primary measure as continuous and reports a mean with a standard deviation for each group. A mean is the arithmetic average of the observed measurements. A standard deviation describes how dispersed the observations are around that average. These summaries can describe the recorded times, but neither one alone communicates the precision of an estimated group average or the uncertainty in a between-group comparison.

What would an interval estimate add?

In general, a confidence interval gives a range of values compatible with the data and the statistical procedure used, and helps convey the precision of an estimate. It does not mean that a particular interval contains the true value with a specified probability after it has been calculated. For a comparison, an interval can show both the direction and the range of effects reasonably compatible with the data. No confidence interval for this endpoint is reported in the registry record, so the precision of a between-group effect cannot be assessed from the posted summaries alone.

Why are the group summaries not a test?

A formal comparison uses a statistical model or test to evaluate the observed group difference relative to its uncertainty under specified assumptions. The observed means and standard deviations are inputs that may inform such an analysis, but they are not themselves a test result. The registry reports that values were missing or not estimable for a two-arm comparison of this continuous endpoint. No test statistic or p-value is reported. It would therefore be inappropriate to treat the visible difference in the means as a statistically established treatment effect.

How should randomization and masking be understood?

In general, randomization aims to make treatment assignment independent of participants' measured and unmeasured characteristics on average. In a small trial, however, randomization cannot ensure that the groups are similar in every characteristic. Masking aims to reduce the influence of treatment knowledge on care, measurement, or interpretation. The registry labels the study randomized and double-masked, but further operational details about allocation and masking are not reported in the registry record. The labels describe the stated design, not proof that bias was eliminated.

What does the standard deviation tell readers?

In general, a standard deviation describes how variable individual observations are around the group mean. It is not the same as a confidence interval and does not directly describe uncertainty in the mean. When a group has few observations, the estimated mean and standard deviation can be strongly influenced by the individual values. Here the hyperoxia summary is based on 3 participants, whereas the placebo summary is based on 10. The posted summaries therefore warrant particular caution when used to generalize beyond these participants.

For some continuous outcomes, statistical comparisons may rely on assumptions about the outcome distribution and the adequacy of an approximation used to calculate uncertainty. If data are markedly skewed or contain influential values, a simple model may not describe them well. Over-dispersion, in general, means that the observed variability is greater than a model assumes; if ignored, uncertainty can be understated. The registry does not report distributional checks, a model, or a plan for handling unusual values for this endpoint, so it is not possible to assess those choices.

What are the planning illustrations?

General sample-size planning asks how many participants would be needed to detect a specified effect under stated assumptions about the significance level, allocation, and statistical method. The planning illustrations associated with these results use a two-sided significance level of 0.05, equal allocation, and a normal approximation, with the target effect expressed as a fraction of the observed effect. They are planning illustrations based on the observed result, not the original trial's design justification and not evidence about the trial's actual ability to detect an effect. The registry does not report the trial's own sample-size justification, so its planned enrollment rationale cannot be assessed from the record.

Planning from an observed effect can be unstable when the observed estimate is imprecise, particularly in a small study. It should not be interpreted as a retrospective probability that the trial detected a true effect. The registry record does not report an interim analysis plan, stopping rule, or other details that would clarify how study size and termination were handled. Those matters should not be inferred from the posted outcome summaries or status.

10. Confidence Intervals for Each Arm

No two-arm comparison is shown for this trial, but each arm's mean can be given with a t-based 95% confidence interval computed from the posted mean, spread and number of participants.

R code and output

arm <- c("Hyperoxia", "Placebo")
n   <- c(3, 10)
m   <- c(96.7, 75.1)
sp  <- c(27.3, 14.8)
sd  <- sp      # posted standard deviations
for (i in seq_along(n)) {
  se <- sd[i] / sqrt(n[i]); h <- qt(0.975, n[i] - 1) * se
  cat(sprintf("CARM|%s|%d|%.3f|%.3f|%.3f|%.3f\n", arm[i], n[i], m[i], sd[i], m[i] - h, m[i] + h))
}

CARM  Hyperoxia  3  96.700  27.300  28.883  164.517
CARM  Placebo  10  75.100  14.800  64.513  85.687

11. Multiplicity, Missing Data and Analysis Populations

The registry lists three outcome measures: the primary mean time to randomization and intervention initiation, plus the mean ratio of initial hypoperfused tissue volume to final infarct volume and change in mean NIHSS as secondary outcomes. It reports zero analyses. Numerical results for the secondary outcomes are not reported in the registry record, and no formal analysis result for the primary endpoint is posted.

When a study examines multiple outcomes or performs multiple statistical tests, the chance of a false-positive finding can increase if no adjustment or hierarchy is used. In general, multiplicity methods address this issue by controlling how results are interpreted across a family of analyses. The registry record does not report a multiplicity strategy, a hierarchy among outcomes, or a prespecified approach to testing. No adjustment should be assumed.

Missing-data handling is also not reported in the registry record. The primary endpoint description gives counts of 3 and 10, but does not explain whether those counts represent everyone randomized, everyone with an observed endpoint, or a defined analysis population. The analysis populations, including any intention-to-treat or per-protocol population, are not reported in the registry record. There is no posted account of missing observations, exclusions, or sensitivity analyses.

These omissions matter because the meaning of an estimate depends not only on its numerical value but also on who contributed data and how incomplete information was handled. With no reported analysis plan or analysis results, readers cannot determine whether the posted summaries follow the intended analysis approach or how missingness might affect the results.

12. Safety Results

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

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

The registry reports zero serious adverse events among 3 participants at risk in the hyperoxia group and zero among 10 participants at risk in the placebo group. It reports zero other adverse events among 3 participants at risk in the hyperoxia group and zero among 10 at risk in the placebo group. These are counts in small groups, not estimates that rule out events or quantify comparative safety with precision.

The death count is 0 in the hyperoxia group and 1 in the placebo group. The posted record does not establish the cause of death or whether it was related to the assigned intervention. A difference in counts alone cannot establish causation, particularly in a trial with few participants. The registry lists no top serious adverse event terms.

These figures describe the adverse-event information posted for the trial. The registry record does not report further detail about event definitions, timing, severity categories beyond the serious and other groupings, or follow-up completeness. The counts should therefore be read as a limited safety description, not as a comprehensive comparison of risk.

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

The primary result describes the time between emergency department arrival and initiation of the assigned intervention among the participants whose data are summarized. The hyperoxia group mean was 96.7 minutes, with a standard deviation of 27.3, based on 3 participants. The placebo group mean was 75.1 minutes, with a standard deviation of 14.8, based on 10 participants. Those are descriptive statistics, not a formal estimate of a treatment effect on stroke outcomes.

The result does not show whether hyperoxia altered infarct volume, neurological status, free radical damage, or inflammatory markers. Although the registry lists two secondary clinical outcome measures and describes broader biomarker aims, corresponding numerical results are not reported in the registry record. The primary timing endpoint cannot substitute for those outcomes.

The record also does not report a confidence interval, p-value, or formal primary-endpoint comparison. Thus, it does not provide a quantified measure of uncertainty for a between-group effect. The trial's randomized, double-masked, parallel design is relevant context, but the small group counts and lack of posted inferential results constrain interpretation. The study status is terminated, and the reason is not reported in the registry record.

In general, absence of a reported result is not evidence that an outcome was unchanged, and a descriptive difference is not by itself proof of a real group effect. These data can be used to describe intervention timing and the adverse-event counts recorded in the registry. They do not establish that one intervention improves or worsens clinical outcomes, and they do not provide a definitive comparison of safety.

14. Why This Trial Matters Statistically

This trial raises a statistically important distinction between measuring a process and measuring a clinical outcome. Its primary endpoint records how long it took to begin the assigned intervention after emergency department arrival. That information characterizes treatment delivery, but it does not directly measure the effects of treatment on neurological status, tissue injury, or biological markers.

The group summaries are based on 3 hyperoxia participants and 10 placebo participants. With such limited numbers, descriptive estimates can be sensitive to individual observations, and the registry does not report an interval estimate or formal comparison that would quantify uncertainty. This makes the record useful for understanding what was measured and what values were posted, while also illustrating why summary values should not be mistaken for conclusive evidence.

The registry lists secondary outcomes relevant to stroke injury and neurological change, but their results are not reported in the registry record. It also records a terminated status without a stated reason. These gaps demonstrate why complete outcome reporting and clear explanations of analysis populations, missing data, and study completion are important for interpreting a clinical trial.

More broadly, the record shows the limits of drawing conclusions from a small randomized study when the primary result is descriptive and the intended clinical outcomes are unavailable. Statistical interpretation depends on the endpoint, the amount of information, and the uncertainty around estimates—not only on the direction of group averages.

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 endpoint is treatment timing rather than a direct measure of stroke outcome. It can describe when intervention began but cannot answer the study's broader questions about tissue injury, neurological change, or biological markers. Results for the two listed secondary outcomes are not reported in the registry record.

Only 3 participants are included in the hyperoxia summary and 10 in the placebo summary. The record provides means and standard deviations but no confidence interval or formal comparison for the primary endpoint. The reason values were not estimable for a two-arm comparison is not reported in the registry record. The effect of the small and unequal group counts on precision therefore cannot be quantified from the posted results.

The trial is marked terminated, but the reason for termination is not reported in the registry record. The original enrollment target, sample-size justification, interim analyses, stopping criteria, and statistical analysis plan are not reported in the registry record. The posted record also does not describe the analysis populations, missing-data handling, or whether all randomized participants contributed to each endpoint.

Additional operational details about the randomization and masking procedures are not reported in the registry record. The adverse-event counts are limited, and the cause or treatment relationship of the one recorded death is not stated. These limitations constrain both efficacy and safety interpretation; they do not establish that an outcome was absent or that either intervention caused a particular result.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: The Role of Hyperoxia in Acute Ischemic Stroke — NCT03904017.
• Registry record last updated 2024-05-16; status last verified 2024-05.
• Results first posted 2024-05-16.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

19. Record Summary

The registry record describes NCT03904017, a randomized, double-masked, parallel trial of hyperoxia versus placebo in acute ischemic stroke. The hyperoxia intervention was supplemental oxygen at 15liters per minute; the placebo intervention was medical air at 15liters per minute, both through a partial non-rebreather facemask. The study is listed as terminated, with enrollment of 13.

The primary endpoint was mean time from emergency department arrival to initiation of the assigned intervention. The hyperoxia group mean was 96.7 minutes, with a standard deviation of 27.3, among 3 participants; the placebo group mean was 75.1 minutes, with a standard deviation of 14.8, among 10 participants. No formal comparison is reported. Two secondary outcomes are listed, but their numerical results are not reported in the registry record. The reason for termination and the trial's statistical analysis plan are not reported in the registry record.