Hyperbaric Oxygen and Outcome After Below Knee Amputations: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Hyperbaric Oxygen and Outcome After Below Knee Amputations (NCT04717557), a randomized phase 1, phase 2 study. The primary outcome is Number of Participants With Mortality During the Inpatient Period.

2. Clinical Question

3. Trial Design

The registry describes a randomized, parallel-group study with no masking. Its listed phases are PHASE1 and PHASE2. The planned comparison was hyperbaric oxygen added to regular care versus regular care alone after below-knee amputation. Hyperbaric oxygen was described as a course of 2-hour treatments at 2 atmospheres absolute, given 1-2 times daily for up to 10 treatments; usual care was provided in both groups.

Enrollment is listed as 4, and the study status is TERMINATED. The recorded start date is 2021-01-01 and the primary completion date is 2023-07-31. These record details describe the registered study, not a complete account of how recruitment or treatment was carried out. Reasons for termination, recruitment progress, protocol deviations, and the timing or circumstances of treatment are not reported in the registry record.

The posted primary outcome is a count of participants who died during the inpatient stay immediately after amputation, up to 13 days. The outcome entries list 3 participants in the hyperbaric oxygen plus regular care group and 1 in the regular care group. The record also lists secondary outcomes concerning hospital length of stay, 90-day mortality, complete wound healing at follow-up, and ambulation with a prosthesis. Results for those secondary outcomes are not reported in the registry record. The record identifies 9 outcome measures and 10 analyses, but does not describe how those measures were prioritized or analyzed as a family.

The registry does not report masking of outcome assessors, allocation concealment, the randomization procedure, or details of the analysis population. Those design details should not be inferred from the randomized label alone.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Overall Study

Reasons for not completing this period:

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

6. Primary Endpoint

The registry lists the primary outcome as Number of Participants With Mortality During the Inpatient Period, measured over During inpatient stay immediately after amputation, up to 13 days.

The primary endpoint is the number of participants who died during the inpatient period immediately after amputation, with the period defined as up to 13 days. It is reported as a count in each study group. In statistical terms, this is a binary outcome for each participant—death during the specified period or no death during that period—summarized as group counts.

In general, a time-bounded mortality endpoint records whether an event occurred within a defined follow-up window. The window matters: a result for inpatient mortality does not describe mortality over a longer period. The registry separately lists 90-day mortality as a secondary outcome, so that outcome addresses a different follow-up horizon. The primary result should therefore be interpreted only as an account of deaths during the stated inpatient period, not as a measure of later mortality.

Mortality is a direct, clearly defined outcome, but a count alone does not show the precision of a comparison. An estimate of the difference between groups and a confidence interval would help communicate the range of effects compatible with the data. Such an estimate and interval are not reported in the registry record. With only 4 participants enrolled, the observed counts offer very limited information about how the groups might compare in a larger study.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry reports two comparisons of the primary binary outcome: a chi-squared analysis and a corrected chi-squared analysis. The endpoint is a participant count, so each participant contributes an event status for the defined inpatient period. A chi-squared test is a general method for assessing whether categorical outcomes differ across groups. It compares the observed pattern of counts with the pattern expected under a null hypothesis of no association between group and outcome.

What does a chi-squared test assess?

In general, a chi-squared test turns the discrepancy between observed and expected counts into a test statistic. A p-value then describes how unusual a discrepancy at least this large would be under the null model, subject to the assumptions of the method. The test does not directly estimate the size of a difference, and it does not show how uncertain an effect estimate is. For that, an effect estimate paired with a confidence interval is more informative.

Chi-squared approximations work best when the expected counts are not too small. When event counts are sparse, the approximation used to obtain a p-value can be poor. That concern is relevant to reading these results because the posted primary outcome contains only one recorded death across the groups. The registry includes a corrected chi-squared result as well as the uncorrected test, but does not identify the correction procedure or explain why both analyses were selected.

What does a correction change?

In general, a correction modifies the calculation to account for the discrete nature of data or to adjust an approximation. The exact meaning depends on the method used. Since the registry says only “Chi-squared, Corrected,” it is not possible to specify which correction was applied. The reported p-value for that analysis is >0.5. The uncorrected chi-squared result is p = 1.0. These results should be quoted as registry analyses rather than reinterpreted as a precise estimate of the group effect.

Why are estimates and intervals useful?

A group comparison can be expressed as an effect estimate, such as a difference in event proportions, together with a confidence interval. In general, the estimate describes the observed contrast, while the interval conveys the statistical uncertainty around it under the chosen method. A confidence interval is not a guarantee that the true effect lies inside it; it is a way to show the range of values compatible with the data and the method. The registry does not report an effect estimate or confidence interval for this primary comparison. The posted counts alone do not provide a measure of precision.

The record indicates that values were missing or not estimable for a two-arm recalculation of this binary endpoint. Consequently, no recalculated effect estimate or interval is available for comparison with the sponsor's p-values. This is distinct from the registry's own analyses, which report p-values. The absence of a recalculated estimate should not be filled in with a hand-derived contrast or an assumed analysis method.

What can randomization contribute?

In general, random allocation is intended to make assigned groups comparable, on average, with respect to measured and unmeasured characteristics. That protection depends on the conduct of the allocation process and is not a guarantee that a small trial will have balanced groups. The registry labels the study randomized, but does not report the sequence-generation method, allocation concealment, or participant-level baseline characteristics. With enrollment listed as 4, chance imbalances remain a substantial interpretive concern; the record does not allow those characteristics to be assessed.

Likewise, “masking: none” describes the registered masking status but does not establish whether outcome assessment was performed consistently. The record does not say whether the people determining inpatient mortality were masked to assignment. For an objectively defined outcome such as death, the scope for subjective classification is limited, but the lack of assessor details and the sparse data still matter when evaluating the complete study conduct.

Finally, p-values do not replace a prespecified statistical analysis plan. The registry does not report a detailed analysis plan, the intended handling of missing data, an analysis population, a multiplicity procedure, or an interim-analysis plan. General statistical principles can explain what the methods are intended to do, but they cannot establish which procedures were planned or followed in this trial.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("Hyperbaric Oxygen Plus Regular Care", "Regular Care")
n      <- c(3, 1)
events <- c(1, 0)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  Hyperbaric Oxygen Plus Regular Care  3  1  33.3  0.8  90.6  6.1  79.2  6.8  80.6
ARM  Regular Care  1  0  0.0  0.0  97.5  0.0  79.3  1.3  84.2

11. Analyses Reported by the Sponsor

The registry's own analyses report p = 1.0 for a chi-squared test and p >0.5 for a corrected chi-squared test. The posted record does not provide an effect estimate or confidence interval for the primary comparison. A recalculated two-group estimate is not available because the binary endpoint is identified as missing or not estimable for that comparison.

There is no reported discrepancy between the sponsor analysis and a recalculation. The sponsor's own analysis therefore takes precedence, as there is no conflicting recalculated result to resolve. That does not make either p-value a measure of the size of the effect, nor does it turn a nonsignificant test into evidence that the groups are equivalent. The p-values should be understood as the sponsor-reported test results for a very sparse set of event counts.

The registry does not provide enough detail to reproduce the tests fully or to determine the exact correction used. The analyses are labeled as superiority tests, but the record does not report a complete statistical plan. Accordingly, interpretation should remain tied to what is posted: the observed counts and the two reported p-values, without attributing a particular unreported method or conclusion to the sponsor.

12. Multiplicity, Missing Data and Analysis Populations

The registry lists 9 outcome measures and 10 analyses. It names the primary inpatient mortality outcome and several secondary outcomes, including hospital length of stay, 90-day mortality, wound healing at follow-up, and ambulation with a prosthesis. The brief summary also describes postoperative complications as part of the study objective. The registry does not report how the analyses were organized across outcomes or whether any multiplicity adjustment was planned or used.

In general, testing many outcomes can increase the chance of at least one apparently unusual result even when no real differences exist. Statistical approaches to multiplicity may adjust significance thresholds, define a hierarchy of outcomes, or focus interpretation on a prespecified primary outcome. Which approach is appropriate depends on the study question and analysis plan. Since no such plan is reported, it is not possible to say whether the posted analyses account for multiple testing. The primary endpoint can be described on its own, but the result should not be treated as one isolated test within a fully documented family of analyses.

Outcome results for the listed secondary measures are not reported in the registry record. This means the reader cannot assess the findings across those outcomes, determine whether data were unavailable, or compare secondary results with the primary endpoint. The reasons for absent results are not reported.

The registry also does not report an analysis population, how missing outcome information was handled, or whether all randomized participants were included in each analysis. It does not state whether the primary analysis followed assignment regardless of treatment received. General approaches such as intention-to-treat analysis or complete-case analysis cannot be attributed to this study without that information. No interim-analysis plan or stopping rule is reported either. The TERMINATED status does not by itself establish why the study ended or how that status affected analyses. These omissions limit assessment of the full analysis process; they do not justify guessing how missingness, multiplicity, or stopping decisions were handled.

13. 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's adverse-event entries list 0 participants with serious adverse events in the hyperbaric oxygen plus regular care group and 0 in the regular care group. It also lists 0 participants with other adverse events in each group. The recorded numbers at risk are 3 in the hyperbaric oxygen plus regular care group and 1 in the regular care group. The entries separately list 1 death in the intervention group and 0 in the regular care group.

These entries should be read as counts reported in the registry, not as a comprehensive safety comparison. The death counts appear in a separate field from the serious and other adverse-event counts, so the record does not establish from these entries whether the death was classified as an adverse event or how it related to treatment. The registry does not report event descriptions, timing, severity details beyond the serious-event category, attribution, or follow-up completeness. It also does not provide enough observations to characterize uncommon harms. Zero recorded adverse-event counts do not prove that no harms occurred; they describe the posted counts for this small enrolled sample.

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

The result establishes what the registry reports for the primary inpatient mortality outcome: 1 death in the hyperbaric oxygen plus regular care group and 0 in the regular care group. It also establishes that the registry reports p = 1.0 for a chi-squared analysis and p >0.5 for a corrected chi-squared analysis. These findings are descriptions of the posted data and analyses.

They do not establish the size of a treatment effect. No effect estimate or confidence interval is reported, and a recalculated comparison is not available. A p-value is not an effect measure, and a large p-value does not prove that groups are the same or that a clinically meaningful difference is absent. The sparse event data leave substantial uncertainty about what might be observed in a larger or different population.

The primary outcome is limited to inpatient mortality during the stay immediately after amputation, up to 13 days. It does not answer the separate question of 90-day mortality. Nor does the primary result establish findings for wound healing, length of stay, postoperative complications, or prosthesis use. The registry lists these as other study objectives or outcomes, but their results are not reported.

The study's randomized designation is relevant to its design, but randomization cannot remove the uncertainty created by very limited enrollment. The record does not include baseline characteristics, details of the randomization procedure, a full analysis plan, or a reason for termination. Those omissions further limit how much can be concluded about the treatment comparison and the study's conduct. The appropriate interpretation is therefore narrow: the posted primary counts and tests do not provide a precise estimate of the difference between groups, and they do not settle the broader clinical questions described in the registry.

15. Why This Trial Matters Statistically

This record illustrates why a randomized label does not by itself guarantee an informative estimate. Randomization is a valuable design feature because, in general, it aims to reduce systematic differences between assigned groups. But when enrollment is listed as 4 and the primary outcome contains only one recorded death, the data have little capacity to characterize the range of possible group differences. The result is not made precise simply by applying a statistical test.

It also highlights the difference between a test and an estimate. The registry reports two p-values, but no effect estimate or confidence interval. A p-value addresses compatibility with a null model under assumptions; it does not say how large a group difference is, how precisely it is known, or whether two treatments are equivalent. Reading the p-values alongside the event counts makes the limitations easier to recognize.

The record also shows the importance of complete outcome reporting. The study description includes questions about healing, length of stay, 90-day mortality, and ambulation with a prosthesis, yet results for those listed secondary outcomes are not reported in the registry record. Without those outcomes and a statistical plan, readers cannot evaluate the broader set of study questions or understand how multiple analyses were interpreted.

Finally, the TERMINATED status and the absence of a reported reason for termination underscore the value of transparent reporting about recruitment, analysis populations, and missing data. Those details help readers distinguish an under-informative result from a definitive answer. In this case, the available record supports careful description of what was observed, while leaving the magnitude and broader implications of the comparison unresolved.

16. 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 most direct limitation is the small enrollment recorded in the registry: 4 participants. The outcome counts are correspondingly sparse, with one death in one group and none in the other. That leaves little information for estimating a stable group contrast or assessing the precision of a statistical comparison. No confidence interval or effect estimate is reported.

The study is marked TERMINATED, but the reason is not reported in the registry record. The record also does not state a target enrollment or sample-size justification, so it is not possible to assess whether the observed enrollment matched a planned target. A reason for stopping, recruitment details, and information about how termination affected the analyses are likewise not reported.

Other design and analysis details are absent. The registry does not report eligibility criteria in detail, baseline characteristics, the randomization procedure, allocation concealment, outcome-assessor masking, the analysis population, or handling of missing data. It also does not provide a full statistical analysis plan, a multiplicity approach, or an interim-analysis plan. These gaps constrain assessment of both comparability and analysis choices.

The primary outcome covers only inpatient mortality during the period immediately after amputation, up to 13 days. Results for the listed secondary outcomes are not reported, and reasons for their absence are not reported. The adverse-event entries contain counts but limited detail about event descriptions, timing, attribution, and completeness of follow-up. Together, these limitations mean the record offers a narrow account of a small study rather than a complete basis for judging its broader objectives.

17. Related Tutorials

18. Related Statistical Calculators

19. Sources

• ClinicalTrials.gov: Hyperbaric Oxygen and Outcome After Below Knee Amputations — NCT04717557.
• Registry record last updated 2025-08-11; status last verified 2025-08.
• Results first posted 2025-08-11.
• 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.

20. Record Summary

The registry record describes NCT04717557, a randomized, unmasked, parallel study of hyperbaric oxygen plus regular care versus regular care after below-knee amputation. The listed phases are PHASE1 and PHASE2, the enrollment is 4, and the status is TERMINATED.

The primary outcome is the number of participants who died during the inpatient stay immediately after amputation, up to 13 days. The posted counts are 1 in the hyperbaric oxygen plus regular care group and 0 in the regular care group. The sponsor reports chi-squared p = 1.0 and corrected chi-squared p >0.5. No primary effect estimate or confidence interval is reported.

The registry lists additional outcomes concerning hospital stay, mortality, healing, and prosthesis ambulation, but their results are not reported in the registry record. Reasons for termination and key design and analysis details are also not reported.