Evaluation of the Hemocompatibility of the Direct Oral Anti-Coagulant Apixaban in Left Ventricular Assist Devices: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Evaluation of the Hemocompatibility of the Direct Oral Anti-Coagulant Apixaban in Left Ventricular Assist Devices (NCT04865978), a randomized phase 2 study. The primary outcome is Freedom From Death or Hemocompatibility Related Adverse Events (Stroke, Device Thrombosis, Bleeding, Aortic Root Thrombus, and Arterial Non-CNS Thromboembolism).

2. Clinical Question

This is a question about a composite clinical outcome in a small randomized comparison, not about any single event type alone. A participant could meet the composite outcome through any of its listed components, so the combined count does not identify which component accounts for an arm-level difference. The registry also lists secondary outcomes addressing specific survival-free outcomes and mortality, but the primary question remains the combined endpoint as defined in the record.

The randomized design makes the between-arm comparison the central statistical contrast. Open-label status means masking was not used; the registry does not report further details about outcome assessment procedures. The record identifies the study as Phase 2 and reports completion, but those design labels do not by themselves determine how precise or conclusive the outcome estimates are. Those questions depend on the observed counts, uncertainty, and details of the analysis.

3. Trial Design

The registry describes a prospective, randomized, controlled, open-label trial with a parallel-group model. Allocation was 1:1. The study enrolled 30 participants and is marked completed. The listed study condition is heart failure, and the intervention set includes apixaban, warfarin, and an LVAD implant. The comparison concerns antithrombotic treatment among LVAD patients; the record does not describe the implant procedure as a randomized treatment arm.

There were 16 participants in the apixaban arm and 14 in the warfarin arm in the posted primary outcome data. The apixaban regimen was 5 mg twice daily. Warfarin was dosed to achieve an INR goal of 2-2.5. Masking was listed as none, so this was an open-label comparison.

The primary outcome was recorded as the number of participants free from death or hemocompatibility-related adverse events through the end of treatment at 24 weeks. Its composite components were stroke, device thrombosis, bleeding, aortic root thrombus, and arterial non-CNS thromboembolism. The registry lists 11 outcome measures and identifies nine posted secondary outcomes. It does not report a primary-outcome analysis model, an analysis population, or an account of missing outcome data.

The registry lists the study start as 2021-12-14 and primary completion as 2023-10-05. These dates identify the reported study timeline; they do not supply details about follow-up completeness or the timing of individual events. The registry record does not report the trial's sample-size justification, a prespecified statistical analysis plan, or interim analyses. The absence of those details limits what can be concluded about the rationale behind the planned enrollment and the extent to which the posted analysis follows a prespecified approach.

The trial's randomized design is relevant to interpreting the comparison, but randomization does not remove uncertainty from a small study. The numerical estimates remain based on the participants and outcomes reported, and their precision must be judged from the confidence interval rather than from the design 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 Freedom From Death or Hemocompatibility Related Adverse Events (Stroke, Device Thrombosis, Bleeding, Aortic Root Thrombus, and Arterial Non-CNS Thromboembolism), measured over From enrollment to end of treatment at 24 weeks.

The primary endpoint counts participants who remained free from death or a hemocompatibility-related adverse event over the stated period, from enrollment to the end of treatment at 24 weeks. The composite includes stroke, device thrombosis, bleeding, aortic root thrombus, and arterial non-CNS thromboembolism. The registry describes the outcome as a count of participants.

A composite endpoint is used to combine several clinically relevant event types into one outcome. In general, this approach can capture whether participants experience any event in a group of related outcomes, rather than relying on one event type alone. It can also make an overall comparison possible when individual component events are uncommon. However, a combined result does not reveal which component contributes most to the observed pattern. Components may also differ in frequency and clinical meaning, so interpretation should preserve the specific definition rather than treating the composite as a single event type.

For this endpoint, a higher count of participants free from the composite corresponds to more participants meeting the stated freedom criterion. The posted counts are 16 of 16 for apixaban and 12 of 14 for warfarin. Those counts should be read together with the confidence interval and the small number of participants. They do not, on their own, establish why participants did or did not meet the endpoint.

The endpoint's 24-week timeframe is important: it describes the period covered by this outcome, not longer-term experience. The registry does not report component-specific primary outcome counts or a detailed definition of how each event was adjudicated. Those details are not reported in the registry record.

7. Results Posted in the Registry

• Warfarin: 85.7% (12/14) versus 100.0% (16/16) on control; risk difference -14.3 percentage points (95% CI -32.6 to 4.0), risk ratio 0.86 (0.69–1.06); unadjusted two-proportion z-test p = 0.118.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Survival Free of Gastrointestinal Bleeding

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Warfarin: 85.7% (12/14) versus 100.0% (16/16) on control; risk difference -14.3 percentage points (95% CI -32.6 to 4.0), risk ratio 0.86 (0.69–1.06); unadjusted two-proportion z-test p = 0.118.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Warfarin and the control. The interval gives the range of effects that remain compatible with the data.

The registry lists nine posted secondary outcomes, including survival-free outcomes for stroke, ischemic stroke, hemorrhagic stroke, device thrombosis, gastrointestinal bleeding, and major non-gastrointestinal bleeding, along with all-cause and cardiovascular mortality. The detailed posted secondary comparison included in the record is survival free of gastrointestinal bleeding, with the same arm counts and comparison statistics as the primary result. The record does not provide analyses for the other listed secondary outcomes in the registry record.

The secondary outcome list gives context for the range of events the trial sought to describe, but it does not establish a consistent pattern across all outcomes. The primary endpoint combines multiple hemocompatibility-related events and death; separate secondary outcomes focus on particular events or mortality. A secondary endpoint may therefore differ in definition and event frequency from the composite, and its results cannot be treated as a direct decomposition of the primary result unless component-specific data are reported.

General explanation: when many outcomes are examined, the chance of observing an unusual result among them can increase. Unadjusted secondary comparisons do not account for this multiplicity, so isolated nominal findings generally carry less weight than a prespecified primary comparison. The registry does not report a multiplicity strategy or a prespecified hierarchy for interpreting these outcomes. It also does not report detailed results for most listed secondary outcomes, so no overall direction across them can be established from the registry record.

The primary comparison itself is uncertain, with a confidence interval that includes zero. Secondary results should be viewed in that context rather than as independent confirmation or contradiction. Without estimates, uncertainty intervals, and analysis details for each endpoint, the secondary-outcome list mainly describes what was assessed, not the strength or consistency of evidence across outcomes.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Survival Free of Any Stroke
• Survival Free of Ischemic Stroke
• Survival Free of Hemorrhagic Stroke
• Survival Free of Device Thrombosis
• Survival Free of Gastrointestinal Bleeding
• Survival Free of Major Non-gastrointestinal Bleeding
• All-cause Mortality
• Cardiovascular Mortality

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is binary at the participant level: each participant is counted as free or not free from the composite during the stated period. The registry reports the number meeting the freedom criterion in each randomized arm. A comparison of these counts can be expressed as a difference between proportions, which describes the absolute separation in observed outcome proportions. The posted risk difference is -0.143 for warfarin relative to apixaban, with the confidence interval reported from -0.326 to 0.040.

General explanation: a risk difference is an absolute contrast, rather than a relative comparison. Its sign depends on the order used to subtract the groups, so the arm labels and reported direction matter. For an outcome framed as freedom from an event, a positive or negative contrast must be interpreted in light of which group is subtracted from which. The registry lists apixaban as the reference arm and warfarin as the comparison arm in the outcome data.

The reported confidence interval gives a range of values reasonably compatible with the data under the statistical procedure used. In repeated applications of the same procedure, intervals constructed this way would cover the underlying parameter at the stated confidence level at the stated frequency. It is not a probability statement that the particular underlying value lies in this one interval. The interval includes zero, indicating that the data do not clearly distinguish a difference from no difference at the usual two-sided significance threshold.

The unadjusted p-value is 0.118. A p-value describes how incompatible the observed data, or data more extreme, would be with a specified null hypothesis under the assumptions of the test. It is not the probability that the null hypothesis is true, and it does not quantify the probability that a result is due to chance. The registry does not report the exact statistical test or full modeling details, so the calculation underlying the posted p-value cannot be independently specified from the record alone.

General explanation: with binary outcomes and small groups, the observed proportions can change substantially when only a few participant outcomes differ. Confidence intervals make that uncertainty visible. A non-significant p-value in a small trial is not evidence of equivalence; equivalence or non-inferiority conclusions require suitable prespecified margins and analyses, which are not reported in the registry record.

The trial uses random allocation, which is intended to make treatment groups comparable on average. Randomization does not guarantee balance on every characteristic in a small sample, and the registry does not provide baseline characteristics in the registry record. The record also does not report adjusted analyses. Thus, the posted summary comparison is best understood as an unadjusted comparison of the displayed group counts, rather than as an estimate adjusted for other participant characteristics.

The outcome is a composite. General explanation: combining outcomes can increase the number of participants with an endpoint event and can provide a broad summary of clinical experience. At the same time, the components may have different frequencies and consequences. When only the combined measure is reported, the overall result cannot establish how each component behaved. The registry lists the component types but does not report primary component-level counts.

Sample-size assumptions, where available, are planning illustrations based on an observed effect and are not observed power. In general, planning calculations describe how many participants per arm might be needed under specified assumptions; they do not show the probability that this completed trial detected an effect. No numerical planning illustrations are reported in the registry record, and the trial's own sample-size justification is not reported in the registry record. The stated planning assumptions describe a two-sided significance level of 0.05, equal allocation, and a normal approximation, with the effect set as a fraction of the observed effect. Those are planning conventions, not evidence that the trial had a particular power.

The registry reports 11 outcome measures and nine posted secondary outcomes. It does not report the prespecified hierarchy of outcomes, any multiplicity adjustment, or the statistical methods for the secondary endpoints. Consequently, secondary findings should be considered exploratory in the absence of those details. The registry also does not report interim analyses, an analysis population, missing-data procedures, or a detailed protocol for outcome adjudication. These omissions limit assessment of the assumptions and robustness of the analysis.

The simple recalculated comparison is unadjusted. Because the registry does not flag a discrepancy, there is no reported conflict between that calculation and the posted summary. A sponsor's own analysis would take precedence if a discrepancy were present; a fuller sponsor analysis is not reported in the registry record. The summary statistics therefore support a description of the posted group counts and their uncertainty, but not reconstruction of every analytical decision.

11. Confidence Intervals and Precision

The estimated risk difference is -0.143, and its 95% confidence interval ranges from -0.326 to 0.040. The interval includes zero and spans values on both sides of no difference. This makes the estimated direction uncertain: the data are compatible with a difference favoring either arm, as well as no difference. The interval's width also shows that the estimate is imprecise.

Precision matters because the displayed counts are based on 16 participants in the apixaban group and 14 in the warfarin group. With relatively small groups, a limited number of participant outcomes can have a substantial effect on the estimated proportions. The point estimate is therefore only one part of the interpretation; the interval conveys how much uncertainty accompanies it.

General explanation: a confidence interval is not a guarantee that the true value lies within its bounds, nor does inclusion of zero prove that the groups are equivalent. It is a range produced by a statistical procedure that reflects sampling uncertainty under its assumptions. A narrow interval would indicate greater precision, while a wider interval would indicate less precision. Here, the interval leaves room for meaningfully different underlying risk differences, so the result does not settle the size or direction of the contrast.

The p-value of 0.118 is consistent with the interval's inclusion of zero for a conventional two-sided comparison. It should not be used alone to judge the result. The registry does not report a prespecified threshold for clinical importance, and the confidence interval should not be translated into a treatment recommendation. Component-specific estimates and longer-term results are also not reported in the registry record.

For these reasons, the most accurate reading is that the observed counts differ, but the available estimate is uncertain. The study's small enrollment and the breadth of the interval restrict the precision of conclusions about the primary endpoint.

12. Multiplicity, Missing Data and Analysis Populations

The registry identifies 11 outcome measures and nine posted secondary outcomes. It does not report whether a hierarchy of primary and secondary outcomes was prespecified, whether secondary comparisons were adjusted for multiplicity, or how the analyses were selected for reporting. With multiple outcomes, unadjusted tests can produce nominally notable results by chance even when no underlying effect is present. General explanation: multiplicity procedures can control or otherwise address the increased chance of false-positive findings, but their use and interpretation depend on the study's prespecified objectives and analysis plan.

The record does not report a statistical analysis plan or interim analyses. It is therefore not possible to determine whether outcomes, timeframes, analysis methods, or decision thresholds were finalized before data were examined. The absence of such information should be treated as an evidence gap, not as proof that these procedures were or were not used.

The analysis population is not reported in the registry record. The posted primary outcome counts include 16 apixaban participants and 14 warfarin participants, but those numbers alone do not establish whether the analysis followed an intention-to-treat approach, a per-protocol approach, or another definition. The record also does not explain whether every randomized participant completed follow-up or how participants with missing outcome information were handled.

General explanation: missing data can affect a comparison if outcome availability differs between groups or is related to the unobserved outcomes. The appropriate response depends on why data are missing and on the analysis plan. Without a participant flow, missing-data description, and analysis-population definition, the robustness of the posted estimates to missing information cannot be assessed.

The registry does not report details of event adjudication or whether assessors were independent of treatment assignment. Because the trial was open-label, knowledge of assigned treatment was not masked, but the record does not specify how this was handled in outcome assessment. These missing details limit evaluation of potential measurement and reporting biases, particularly for outcomes requiring clinical judgment.

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.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: serious ae other (28.6%), serious ae infection (25.0%), serious ae gastrointestinal bleeding (21.4%), serious ae cardiac (12.5%), serious ae device related (6.2%).

The registry's adverse-event summary reports serious adverse events in 7 of 16 participants in the apixaban group and 8 of 14 participants in the warfarin group. It reports other adverse events in 14 of 16 and 12 of 14 participants, respectively, and zero deaths in both groups. These are descriptive counts from the posted safety record; they are not the same measure as the primary endpoint, which counts participants free from a composite outcome.

The serious-event terms listed most frequently include Serious AE Other at 28.6%, Serious AE Infection at 25.0%, Serious AE Gastrointestinal bleeding at 21.4%, Serious AE Cardiac at 12.5%, and Serious AE Device Related at 6.2%. The record does not specify in the registry record whether individual participants could contribute to more than one term or the denominator used for each percentage. These percentages therefore should not be summed or interpreted as mutually exclusive groups.

General explanation: adverse-event counts describe events recorded during the trial and do not by themselves establish that a treatment caused an event. In a small study, differences in counts are uncertain, and the registry does not report a comparative safety model or confidence intervals for these summaries. The absence of reported deaths in both groups is a description of the posted observations, not evidence about mortality risk beyond the study period.

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

The result describes how many participants in each randomized arm met the registry's definition of freedom from death or a specified hemocompatibility-related adverse event through 24 weeks. The posted counts are 16 of 16 for apixaban and 12 of 14 for warfarin. The estimated risk difference is -0.143, with a 95% confidence interval from -0.326 to 0.040 and an unadjusted p-value of 0.118. These statistics describe the observed comparison and its uncertainty; they do not establish a definitive difference between the treatments.

The confidence interval includes no difference and allows a range of possible underlying contrasts. The p-value does not show that there is no effect, and it does not establish equivalence. Conversely, the observed difference in counts does not by itself establish that one treatment caused better or worse outcomes. The small enrolled sample makes the estimate imprecise, and the composite endpoint combines event types that may not occur equally often or have the same clinical meaning.

The result is limited to the specified endpoint and timeframe. It does not provide a detailed account of the separate primary endpoint components, longer-term outcomes, or a complete comparison of all safety outcomes. The registry lists secondary outcomes, but detailed results for most of them are not reported in the registry record.

The trial was randomized, but it was open-label. The registry does not report the analysis population, missing-data approach, statistical analysis plan, multiplicity procedures, or outcome-adjudication details. These gaps constrain assessment of how the comparison was conducted and how robust it is to analytical choices or incomplete information.

Accordingly, the posted data support a cautious description of the observed group counts and the uncertainty around their difference. They do not establish a treatment recommendation, prove that the treatments are equivalent, or resolve the effects on individual event types. Further conclusions would require additional information beyond what is reported in the registry record.

15. Why This Trial Matters Statistically

This trial raises a statistical question that is important in device-related research: how to compare treatments when the endpoint combines several types of adverse events and the randomized sample is small. The composite outcome gives a single measure of freedom from death or specified hemocompatibility-related events, but its overall count cannot show which component drives the result. Component-level reporting is essential for understanding whether an aggregate pattern is shared across outcomes or mainly reflects one type of event.

The confidence interval illustrates why an observed difference and a conclusive estimate are not the same thing. The reported risk difference is -0.143, but the interval from -0.326 to 0.040 includes zero and allows effects in either direction. The p-value of 0.118 likewise does not settle whether a real difference exists. Statistical interpretation should emphasize both the estimated contrast and its precision, rather than treating a thresholded test result as a complete answer.

Randomization supports a fair comparison in principle, but small randomized trials can still have considerable uncertainty. Transparent reporting of participant flow, analysis populations, missing data, endpoint adjudication, and prespecified statistical methods would help readers judge the reliability and scope of the estimates. Those details are not reported in the registry record.

The registry's multiple secondary outcomes make multiplicity another important consideration. Without a reported hierarchy or adjustment strategy, secondary comparisons are less persuasive as confirmatory evidence. This is especially relevant when each endpoint may have few events and results are interpreted separately. The record's detail therefore illustrates a broader statistical lesson: clear outcome definitions and complete analytic reporting are needed to distinguish exploratory signals from robust, precise estimates.

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 total enrollment was 30 participants, with 16 in the apixaban arm and 14 in the warfarin arm in the primary outcome data. The resulting estimate is imprecise, as reflected by the confidence interval from -0.326 to 0.040. Counts based on groups of this size provide limited resolution for a composite endpoint and cannot support precise comparisons of less common component events.

The primary endpoint combines death with stroke, device thrombosis, bleeding, aortic root thrombus, and arterial non-CNS thromboembolism. The registry does not report component-specific primary outcome counts, so it is not possible to tell which events contributed to the arm-level results. A composite can also conceal differing patterns among its components.

The trial was open-label, and the registry does not report details of outcome adjudication or assessment. It also does not report baseline characteristics in the registry record, the analysis population, participant flow, missing-data methods, or a statistical analysis plan. These omissions prevent a full assessment of comparability, data completeness, and analytical choices.

The registry lists multiple secondary outcomes but does not report a multiplicity strategy or detailed results for most of them. This limits conclusions about consistency across outcomes. The record also does not report the trial's sample-size justification or interim analyses, so the intended precision and decision framework cannot be evaluated.

Finally, the primary endpoint covers the period from enrollment to the end of treatment at 24 weeks. The posted outcome does not establish longer-term results. These limitations are specific to the available registry entries and should not be mistaken for evidence that any unreported procedure or analysis was absent.

17. Related Tutorials

18. Related Statistical Calculators

19. Sources

• ClinicalTrials.gov: Evaluation of the Hemocompatibility of the Direct Oral Anti-Coagulant Apixaban in Left Ventricular Assist Devices — NCT04865978.
• Registry record last updated 2024-11-14; status last verified 2024-10.
• Results first posted 2024-11-14.
• 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

NCT04865978 is a completed Phase 2, randomized, open-label, parallel-group trial of apixaban versus warfarin in LVAD patients. The registry reports enrollment of 30 participants, with 16 assigned to apixaban and 14 to warfarin in the primary outcome data. Apixaban was prescribed at 5 mg twice daily; warfarin was dosed to achieve an INR goal of 2-2.5.

The primary outcome was freedom from death or hemocompatibility-related adverse events through 24 weeks. The posted counts were 16 of 16 in the apixaban arm and 12 of 14 in the warfarin arm. The risk difference was -0.143, with a 95% confidence interval from -0.326 to 0.040 and an unadjusted p-value of 0.118.

The confidence interval includes zero and is wide, so the estimated comparison is uncertain. The registry lists multiple secondary outcomes and reports adverse-event counts, but detailed results for most secondary outcomes, analysis-population definitions, missing-data procedures, and the statistical analysis plan are not reported in the registry record.