Comparative Effectiveness of Unilateral vs. Bilateral Pulmonary Collapse in Cardiac De-airing: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Comparative Effectiveness of Unilateral vs. Bilateral Pulmonary Collapse in Cardiac De-airing (NCT02119871), a randomized na study. The primary outcome is Quantitative Assessment of Air Embolism to the Brain After Completion of Open Left Heart Surgery.

2. Clinical Question

3. Trial Design

The registry describes a randomized, single-masked, parallel-group trial, marked completed, with enrollment of 20. The groups each contain 10 participants in the posted primary-outcome data. The phase is listed as NA. The sponsor is Lund University. The recorded start is 2014-01 and primary completion is 2014-06.

The two procedures are bilateral open pleurae and right pleura open. The bilateral group description includes right pulmonary vein drainage; the right-pleura-open group description includes left ventricular apical drainage. The brief summary frames the comparison as unilateral right-lung collapse versus bilateral pulmonary collapse for cardiac de-airing in open left-sided heart surgery.

There is an important description that the registry does not reconcile: it says groups were statistically compared to a historical control group of 10 patients with bilateral open pleurae and left ventricular apical venting. That wording sits alongside the randomized allocation label and the two listed groups. The record does not explain how the historical control relates to the randomized comparison or to the arm-level results, so those descriptions should not be treated as interchangeable without clarification.

The listed secondary outcome is duration of the de-airing procedure. The record lists 7 outcome measures and 3 analyses. Details about who was masked, the randomization procedure, allocation concealment, and the planned statistical analysis beyond the reported test 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: Groups were statistically compared to a historical control group of 10 patients with bilateral open pleurae and left ventricular apical vent (LVAV) in order to evaluate:1. the impact on de-airing of unilateral open pleura compared to bilateral open pleurae, and 2. the impact on de-airing of a right superior pulmonary vein vent compared to LVAV.

5. Baseline Characteristics

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

Population: Groups were statistically compared to a historical control group of 10 patients with bilateral open pleurae and left ventricular apical vent.

6. Primary Endpoint

The registry lists the primary outcome as Quantitative Assessment of Air Embolism to the Brain After Completion of Open Left Heart Surgery, measured over Time from the release of the aortic crossclamp to cardiac ejection, an average of 5-10 minutes.

The primary endpoint is the quantitative assessment of gaseous cerebral microemboli after completion of open left heart surgery. The registry describes counting microembolic signals online with transcranial Echo-Doppler monitoring of the right and left middle cerebral arteries. The reported measure is the sum of signals registered from those two arteries, expressed in gaseous cerebral microemboli.

The stated timeframe begins at release of the aortic crossclamp and ends at cardiac ejection, an average of 5-10 minutes. The endpoint therefore focuses on detected gas signals during a specified period in the operation, rather than on a later clinical outcome. Its use makes the comparison about measured de-airing-related signals. The registry does not report how the signal counts were adjudicated, whether monitoring was blinded to group, or what threshold or signal-classification rules were used.

The endpoint is summarized by the median and inter-quartile range. In general, the median describes the middle of an ordered set of observations, while the inter-quartile range describes the spread of its central half. These summaries can be informative when data are skewed or have outlying values. They do not, by themselves, show the full distribution or the precision of an estimated group difference.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The statistical methods can be understood by separating the descriptive summaries from the group-comparison test. The arm summaries report medians and inter-quartile ranges, while the reported inferential analysis is a Wilcoxon (Mann-Whitney) test. These methods answer related but distinct questions: summaries describe the observed data, while the test assesses how compatible the between-group ordering is with a null comparison under its assumptions.

What does the median describe?

The median is the central value after observations are ordered. Unlike the mean, it is less affected by unusually high or low observations. That can be useful for signal counts, which may have a skewed distribution. The inter-quartile range gives the lower and upper quartiles and describes the central spread of the data. Here, the registry reports the median and inter-quartile range for each group, but not individual observations or a plot-specific numerical description. These summaries cannot reveal every feature of the distributions, including clustering, gaps, or the presence of influential observations.

What does the Wilcoxon (Mann-Whitney) test assess?

In general, the Mann-Whitney form of the Wilcoxon rank-sum test pools observations from both groups, orders them, and compares their ranks. It tests whether observations from one group tend to rank higher or lower than those from the other, under the null model used for the test. It does not directly test a difference between medians in all circumstances. A median-difference interpretation is most straightforward when the group distributions have similar shapes and differ mainly in location; the registry does not report whether that condition was assessed.

The registry reports the median difference (final values) as 14 and the test p-value as 1.00. A p-value is calculated under a specified null model and summarizes how unusual the observed test statistic, or a more extreme one, would be if that model held. It is not the probability that the null hypothesis is true, and it does not state the probability that a result will replicate. The record does not provide the test statistic, exact-versus-approximate calculation details, or a confidence interval for the difference.

Why does spread matter?

In general, dispersion matters because substantial variability can make group contrasts less precise. Skewness and outlying observations can also affect how different statistical summaries behave. Count outcomes may, in general, show over-dispersion when their variability exceeds what a simple count model expects; methods that account for over-dispersion can then be important. That is a general consideration, not a claim that over-dispersion was present in this trial. The registry reports medians and inter-quartile ranges and names a rank-based test, but does not report a count-regression model or diagnostics.

In general, a confidence interval conveys a range of effect estimates compatible with the data and the interval method. It helps readers assess precision and the direction and magnitude of effects, alongside the point estimate. No confidence interval is reported for the posted median difference. The inter-quartile ranges describe within-group data spread; they cannot substitute for an interval estimating uncertainty in the between-group contrast.

What can be inferred from the design description?

Randomized allocation can support a fair comparison when the allocation process is implemented as intended, outcomes are measured appropriately, and the analysis corresponds to the assigned groups. The registry labels this trial randomized and parallel, but does not report the sequence-generation method, concealment, or analysis population. It also refers to comparison with a historical control group. The record does not explain how this description relates to the randomized groups, so it is not possible to determine from the registry text alone whether every reported comparison follows the randomized design.

The record identifies single masking, but does not identify who was masked. Masking can, in general, reduce the influence of knowledge of group assignment on outcome measurement or other trial procedures. Its likely role depends on who was masked and how the endpoint was collected; those details are not reported in the registry record.

What does the sample-size planning mean?

Any sample-size values associated with planning illustrations are not observed trial results. The stated planning assumptions are a two-sided significance level of 0.05, equal allocation, and a normal approximation; the effect to detect is defined as a stated fraction of the observed effect. In general, such illustrations ask how many participants per arm would be needed under specified assumptions to detect a chosen effect with a target power. They do not establish the trial’s original design rationale, and they are not observed power. The trial’s own sample-size justification is not reported in the registry record.

The registry does not report a prespecified statistical analysis plan, a multiplicity procedure, missing-data methods, or a detailed analysis population. These absences mean the posted p-value cannot be assessed against all possible planned decisions about endpoints, exclusions, or repeated testing. Statistical interpretation should stay close to what is documented: group medians and inter-quartile ranges, a reported median difference, and a Wilcoxon (Mann-Whitney) test result.

10. Analyses Reported by the Sponsor

The registry reports the sponsor’s comparison as a Wilcoxon (Mann-Whitney) analysis of final values, with a median difference of 14 and p = 1.00. It does not report a confidence interval for that difference. The record does not flag a discrepancy between the posted analysis and the unadjusted recalculation. The sponsor’s reported analysis is therefore the primary inferential result to describe; a simple recalculation from summary statistics would not replace the analysis using participant-level data.

The arm medians and inter-quartile ranges are descriptive summaries, while the reported test evaluates ranks across the groups. A recalculation from summary statistics cannot reconstruct the full ordering of individual observations or verify all details of the sponsor’s test. It also cannot resolve the record’s wording about a historical control group. Accordingly, the posted p-value and effect estimate should be attributed to the registry’s reported analysis, with the limits of the available information made clear.

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 7 outcome measures and 3 analyses. It identifies duration of the de-airing procedure as a secondary outcome, while the primary outcome is quantitative cerebral gaseous microemboli. The record does not describe how many statistical tests were planned for each outcome, whether there were multiple time points or analysis choices, or whether any adjustment for multiple testing was used. A multiplicity procedure is not reported in the registry record.

In general, evaluating several outcomes or conducting multiple analyses can increase the chance of finding at least one apparently notable result by chance. Adjustment methods and clearly defined primary analyses help readers understand how such testing was handled. Without the planned analysis details, the reported p-value should be interpreted as the result of the stated comparison, not as evidence that all possible comparisons were controlled for multiplicity.

Missing-data handling is not reported in the registry record. The posted primary-outcome group sizes are 10 and 10, but the record does not say whether these are all randomized participants, whether any measurements were unavailable, or how missing observations would have been treated. It also does not report an intention-to-treat or per-protocol analysis population, exclusions after randomization, or a sensitivity analysis. The group counts alone do not answer those questions.

The primary-outcome description’s reference to a historical control group is another unresolved analysis detail. The registry does not explain its relationship to the randomized group comparison, whether the historical control contributed to the reported result, or what features were used to define comparability. These details matter when considering the target comparison and the assumptions behind the inference.

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.

Most frequent non-serious adverse event terms

The adverse-event figures report no serious adverse events among 10 participants in either group. The record reports one other adverse event among 10 participants in each group and no deaths in either group. These counts describe the recorded experience in this trial; they are not estimates of comparative safety with high precision.

The registry does not provide the terms or descriptions of the other adverse events in the posted figures, and the list of top serious-event terms is empty. It does not report event timing, severity detail for non-serious events, attribution, or a statistical comparison of adverse events. The small group sizes mean that these counts should be read as limited descriptive information. Zero recorded serious events does not establish that serious events cannot occur, and the equal count of other events does not show that the underlying risks are identical.

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

The posted results describe two groups of 10 and their measured gaseous cerebral microemboli after open left heart surgery. The bilateral-open-pleurae group has a median of 46 and inter-quartile range of 21 to 107; the right-pleura-open group has a median of 32 and inter-quartile range of 16 to 68. The reported median difference is 14, with a Wilcoxon (Mann-Whitney) p-value of 1.00.

These results do not establish that one procedure is superior, nor do they demonstrate that the approaches are equivalent. A failure to distinguish groups statistically is not proof of no effect. Equivalence or non-inferiority requires a suitable prespecified framework and margin, neither of which is reported in the registry record. The absence of a confidence interval also leaves the precision of the reported difference unclear.

The endpoint measures signals detected during an operative interval, not a later clinical event. The trial’s listed conditions include brain ischemia and reduction of cerebral air emboli, but the posted primary measure is a signal count. The registry does not report a primary clinical outcome of brain ischemia in the result description. It would therefore be inappropriate to treat this signal comparison as a direct estimate of clinical benefit or harm.

The randomized label is relevant, but the historical-control wording in the primary-outcome description is unresolved. The record does not establish how that historical group relates to the posted arm summaries. This limits how confidently the comparison can be attributed to the two randomized procedures as described.

In general, a small study can produce an uncertain estimate even when a p-value is reported. Readers should consider the group summaries, the missing interval estimate, the documented design, and the unanswered questions about analysis and missing data together. The results are a record of the measured outcomes and reported test, not a definitive answer about clinical effectiveness.

14. Why This Trial Matters Statistically

This trial is statistically notable because it combines a small randomized parallel-group comparison with a count-like physiological endpoint summarized using medians and inter-quartile ranges. That combination illustrates why the choice of summary and test should fit the shape and scale of the outcome. Rank-based methods can compare the ordering of observations without requiring a normal model for the raw values, but they do not automatically provide a direct test of median differences under every distributional pattern.

The registry gives a point estimate and p-value but no confidence interval. This makes a useful teaching example of why statistical significance and estimation are different tasks. The test result describes compatibility with a null comparison under its method; an interval would help convey uncertainty about the size and direction of the difference. A p-value of 1.00 cannot by itself support a claim of equivalence.

The record also highlights the importance of consistent descriptions of design and analysis. Randomized allocation is listed, yet the primary-outcome text refers to a historical control group. A reader needs the relationship between those groups to understand the estimand and the evidence being compared. That relationship is not reported in the registry record.

Finally, a physiological signal count and a patient-centered clinical outcome are not the same measure. The endpoint can quantify detected cerebral gaseous microemboli over the specified operative interval, while the registry does not report a corresponding clinical endpoint result in the primary description. Separating measurement, statistical evidence, and clinical interpretation is central to reading trials responsibly.

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.

Several limitations specific to this record constrain interpretation. The primary-outcome comparison involves 10 participants per group, and the registry reports no confidence interval for the median difference. The summaries therefore do not show the precision of the contrast, and the inter-quartile ranges are not uncertainty intervals for it.

The description of the comparison is not fully clear. The trial is labeled randomized, but the primary-outcome text says groups were statistically compared to a historical control group of 10 patients with bilateral open pleurae and left ventricular apical venting. The arm description for bilateral open pleurae specifies right pulmonary vein drainage. The registry does not explain the relationship between these descriptions or how the historical control relates to the reported arm-level results.

The registry does not report participant characteristics, detailed eligibility criteria, the randomization method, who was masked, or how the endpoint was adjudicated. It also does not report the prespecified analysis plan, the analysis population, missing-data handling, or a multiplicity adjustment. The primary endpoint is a signal count over an operative interval; the record does not establish a link between that count and later clinical outcomes.

The posted safety counts are sparse, and the event descriptions are not available in the record. The registry lists a secondary outcome of de-airing duration, but does not provide its results in the registry record described for this analysis. These gaps limit conclusions to the documented primary summaries, the stated test, and the adverse-event counts.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Comparative Effectiveness of Unilateral vs. Bilateral Pulmonary Collapse in Cardiac De-airing — NCT02119871.
• Registry record last updated 2018-01-29; status last verified 2017-06.
• Results first posted 2018-01-29.
• 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

NCT02119871 is a completed, randomized, single-masked, parallel-group trial sponsored by Lund University, with enrollment of 20. It compares bilateral open pleurae with right pleura open during open left-sided heart surgery. The primary endpoint is the summed count of gaseous microembolic signals from the right and left middle cerebral arteries, measured from release of the aortic crossclamp to cardiac ejection, an average of 5-10 minutes.

The registry reports 10 participants per group, medians of 46 and 32, and inter-quartile ranges of 21 to 107 and 16 to 68. The reported median difference is 14, with a Wilcoxon (Mann-Whitney) p-value of 1.00 and no confidence interval. The historical-control wording, analysis details, and missing-data methods are not reported clearly in the registry record. Serious adverse events and deaths are recorded as zero in each group.