A Study Exploring Two Strategies of Rivaroxaban (JNJ39039039; BAY-59-7939) and One of Oral Vitamin K Antagonist in Patients With Atrial Fibrillation Who Undergo Percutaneous Coronary Intervention: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study Exploring Two Strategies of Rivaroxaban (JNJ39039039; BAY-59-7939) and One of Oral Vitamin K Antagonist in Patients With Atrial Fibrillation Who Undergo Percutaneous Coronary Intervention (NCT01830543), a randomized phase 3 study. The primary outcome is Percentage of Participants With Clinically Significant Bleeding.

2. Clinical Question

3. Trial Design

This was a randomized, controlled, multicenter, open-label, parallel-group Phase 3 study. The registry record lists enrollment of 2124 participants with atrial fibrillation who underwent percutaneous coronary intervention. Masking was none, so the study was open-label. The registry lists the study as completed.

There were three treatment strategies: rivaroxaban 15 mg once daily; an initial regimen of rivaroxaban 2.5 mg twice daily followed by rivaroxaban 15 mg once daily; and dose-adjusted vitamin K antagonist treatment. For participants with moderate renal impairment, the rivaroxaban once-daily dose was described as 10 mg. The strategies also differed in use of aspirin: the initial rivaroxaban 2.5 mg strategy and the vitamin K antagonist strategy included low-dose aspirin along with clopidogrel or another listed P2Y12 inhibitor, while the rivaroxaban 15 mg strategy included a P2Y12 inhibitor without aspirin in its description.

Clopidogrel was specified in the treatment descriptions, with prasugrel or ticagrelor listed as alternatives. The vitamin K antagonist strategy was dose-adjusted to a target International Normalized Ratio of 2.0 to 3.0. The rivaroxaban strategies therefore represent treatment combinations rather than comparisons of isolated drugs. Differences in aspirin use and regimen sequence are part of the strategies being compared.

The primary outcome was a safety endpoint assessed through Month 12. The primary analysis population was defined as randomized participants who received at least 1 dose of study drug. The registry record reports arm-level outcome percentages and two log-rank hazard-ratio analyses comparing each rivaroxaban strategy with the vitamin K antagonist group. Further details of the original sample-size justification, randomization procedures, and analysis plan 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

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The safety analysis set is defined as all randomized participants who received at least 1 dose of study drug.

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: The intent-to-treat analysis set included all randomized participants.

6. Primary Endpoint

The registry lists the primary outcome as Percentage of Participants With Clinically Significant Bleeding, measured over Up to Month 12.

The primary endpoint was the percentage of participants with clinically significant bleeding up to Month 12. It was a composite of TIMI major bleeding, TIMI minor bleeding, and bleeding requiring medical attention. A participant could meet the endpoint through any of those defined bleeding categories; the composite gathers events of different severity or clinical response into one outcome for the primary safety assessment.

TIMI major bleeding included symptomatic intracranial hemorrhage or clinically overt bleeding associated with a hemoglobin drop of at least 5 g/dL, or an absolute hematocrit drop of at least 15% when hemoglobin was unavailable. TIMI minor bleeding included clinically overt bleeding associated with a hemoglobin fall of 3 to less than 5 g/dL, or a hematocrit fall of 9% to less than 15% when hemoglobin was unavailable. Bleeding requiring medical attention was an event requiring medical treatment, surgical treatment, or laboratory evaluation that did not meet the major or minor criteria.

As a general statistical point, a composite endpoint can capture more events than a narrowly defined component endpoint and can provide a single summary for a safety question. Its interpretation depends on the component definitions: the composite does not mean that all components have the same severity or importance. The registry record identifies the components, but does not report separate primary-outcome component results in the posted arm summaries.

7. Results Posted in the Registry

• Rivaroxaban 2.5 mg Twice Daily (BID)/15 mg Once Daily(QD): 16.6% (117/706) versus 15.7% (109/696) on control; risk difference 0.9 percentage points (95% CI -2.9 to 4.8), risk ratio 1.06 (0.83–1.34); unadjusted two-proportion z-test p = 0.643.
• Vitamin K Antagonist (VKA): 24.0% (167/697) versus 15.7% (109/696) on control; risk difference 8.3 percentage points (95% CI 4.1 to 12.5), risk ratio 1.53 (1.23–1.90); unadjusted two-proportion z-test p = <0.001.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is binary at the participant level: a participant either met the composite bleeding definition during the stated follow-up or did not. The registry presents arm-level percentages for this outcome. A percentage is a descriptive summary of how common the event was in each group, while a comparison estimate quantifies the contrast between groups. These are related but different summaries.

What does an unadjusted comparison mean?

The recalculated comparisons use the posted event counts and group sizes to make simple, unadjusted comparisons. They do not account for covariates, stratification, or other features unless those are already reflected in the posted counts, and the registry record does not establish that they are. An unadjusted estimate is useful as a transparent check of the arm summaries, but it need not match an analysis using participant-level data, a time-to-event model, adjustment, or a different analysis population. When the record indicates a discrepancy, the sponsor's own analysis takes precedence over the simple recalculation.

The risk difference is a comparison of event probabilities between groups. Its sign depends on which group is subtracted from which, so interpretation requires checking the comparison order. The risk ratio compares probabilities as a ratio, and the odds ratio compares odds. The registry includes several recalculated effect measures, but those do not replace the reported sponsor analysis. The p-value summarizes how compatible the data are with a specified null hypothesis under the calculation's assumptions; it is not the probability that the null hypothesis is true and does not measure the size or clinical importance of an effect.

How does a hazard ratio differ?

The registry's reported analyses use the log-rank method and give hazard ratios. A hazard ratio is a relative comparison of event rates over time, rather than a direct comparison of the final percentages. Time-to-event methods can incorporate the timing of events and account for participants whose event time is not observed through the end of follow-up, under method-specific assumptions. The registry record does not provide enough detail to determine how censoring, covariate adjustment, or other modeling choices were handled. The reported hazard ratios should therefore be read as the record's stated analyses, not as equivalent to the simple unadjusted percentage comparisons.

For a hazard ratio, a value below 1 indicates a lower estimated event hazard for the numerator group relative to the reference group as coded in that analysis. The confidence interval describes uncertainty around this relative estimate. The reported 95% confidence intervals and p-values are attached to those analyses; they should not be substituted for intervals or p-values calculated from the binary arm summaries.

Why can analyses differ?

Different analysis methods can produce different estimates because they may target different quantities or use information differently. An analysis of the proportion experiencing an event by a specified time is not identical to an analysis of event timing. Adjustment, censoring rules, endpoint definitions, and analysis populations can also affect results. These are general statistical explanations; the specific choices behind the registry's sponsor analyses are not reported in the registry record. The record does identify the safety analysis set for the primary endpoint as randomized participants who received at least 1 dose of study drug, but it does not report every detail needed to reproduce the sponsor's analysis.

Confidence intervals are ranges of estimates compatible with the data and the model's assumptions at the stated confidence level. They express statistical uncertainty, not a range containing a guaranteed fraction of individual outcomes. Wider intervals indicate less precision, while narrower intervals indicate more precision under the same framework. A p-value and a confidence interval are connected through the underlying test and estimate, but neither describes the likelihood that a treatment will help or harm a particular person.

What is the role of the endpoint definition?

The primary endpoint combines major bleeding, minor bleeding, and bleeding requiring medical attention. Composite outcomes can make a safety assessment more comprehensive, but the total can be influenced by components that occur with different frequency and have different clinical consequences. Component-specific counts or estimates would help show how the overall pattern is distributed. The registry record lists separate secondary bleeding outcomes, but the posted primary results summarized in the record do not provide component-specific primary composite results by treatment group.

The study was open-label. As a general matter, knowledge of treatment assignment can influence reporting or assessment of outcomes that involve judgment or medical attention. The objective criteria in the endpoint definitions may constrain some judgments, but they do not eliminate every possible influence. The registry record does not report blinded endpoint adjudication, so that detail is not reported in the registry record.

Finally, statistical significance is not a complete interpretation. It depends on the estimate, its uncertainty, the analysis method, and the question being tested. A small p-value does not itself indicate the magnitude or practical importance of a difference, and a non-small p-value does not demonstrate equivalence. Establishing equivalence or non-inferiority would require a suitable prespecified framework and margin; these are not reported in the registry record.

10. Confidence Intervals and Precision

The sponsor-reported log-rank estimate for rivaroxaban 15 mg versus vitamin K antagonist treatment is a hazard ratio of 0.59, with a 95% confidence interval from 0.47 to 0.76. For the rivaroxaban 2.5 mg twice-daily/15 mg once-daily strategy versus vitamin K antagonist treatment, the reported hazard ratio is 0.63, with a 95% confidence interval from 0.5 to 0.8. Both reported p-values are <0.001. These intervals indicate the range of relative hazard estimates compatible with each analysis at its stated confidence level, under its assumptions.

The intervals do not describe uncertainty in the same way as an interval around the final event percentages. The record's arm-level percentages are 15.7%, 16.6%, and 24.0%; no confidence intervals are posted alongside those percentages. The simple recalculations have their own uncertainty measures, but the recalculated comparison is not interchangeable with the sponsor-reported log-rank analysis. In particular, the two approaches use different summary measures.

As a general statistical point, precision depends on the amount of information and the variability of the data, as well as on the method used. A confidence interval gives a more informative account of estimation uncertainty than a p-value alone, but it does not address bias, missing data, or whether the chosen analysis answers the most relevant question. The registry record does not report enough methodological detail to evaluate all those issues. Precision should therefore be interpreted within the stated endpoint, analysis population, and method rather than as a complete measure of certainty about the treatment strategies.

11. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning figures are illustrations based on the observed effect, not the study's original design calculation. The original sample-size assumptions and justification are not reported in the registry record. The illustrations use a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. Each scenario targets a stated fraction of the observed effect. The figures for these binary-outcome comparisons are participants per arm, not total events.

The 80% power column gives the number of participants per arm that the illustration estimates would be needed to have an 80% chance of detecting the specified effect under its assumptions, if that effect and those assumptions hold. The 90% power column expresses the corresponding planning target for a 90% chance. These are planning probabilities, not guarantees about what a future study will find. The number needed depends on the effect chosen: planning for 75% of the observed effect requires more participants than planning for 100%, and planning for 50% requires still more.

For the comparison of the rivaroxaban 2.5 mg twice-daily/15 mg once-daily strategy with rivaroxaban 15 mg, the illustration lists 25,553 participants per arm for 80% power and 34,207 for 90% power when planning for 100% of the observed effect. For 75%, the corresponding figures are 45,168 and 60,466; for 50%, they are 101,041 and 135,265.

For vitamin K antagonist treatment versus rivaroxaban 15 mg, the 100% scenario lists 361 participants per arm for 80% power and 483 for 90% power. The 75% scenario lists 617 and 826; the 50% scenario lists 1,330 and 1,779. These figures are not observed power. They are illustrative planning outputs based on the posted results and specified assumptions, and should not be interpreted as the trial's prespecified design requirements.

12. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.157; treatment proportion p₁ = 0.166; average p̄ = 0.161.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.5200 = 1.0191.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.5199 = 0.4376.
4. n = (1.0191 + 0.4376)² / (0.0091)² = 25,552.1, rounded up to 25,553 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
p0 <- 0.156609195402299                          # control proportion
p1 <- 0.165722379603399                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(1830543)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 25,553 (80% power), 34,207 (90% power)
 75% of observed effect: n per arm = 45,168 (80% power), 60,466 (90% power)
 50% of observed effect: n per arm = 101,041 (80% power), 135,265 (90% power)
power.prop.test: 25552.1 
Simulated power at n = 25,553 per arm: 0.799 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

13. Analyses Reported by the Sponsor

The record contains two kinds of results: simple unadjusted recalculations from the posted event counts and sponsor-reported log-rank analyses. The recalculations compare the binary outcome summaries directly. The sponsor analyses report hazard ratios, confidence intervals, and p-values using the log-rank method. These are not identical measures, and they may not target the same statistical quantity.

For rivaroxaban 15 mg compared with vitamin K antagonist treatment, the record reports a log-rank hazard ratio of 0.59 (95% CI 0.47 to 0.76; p < 0.001). For the other rivaroxaban strategy compared with vitamin K antagonist treatment, it reports a hazard ratio of 0.63 (95% CI 0.5 to 0.8; p < 0.001). The recalculated comparisons instead use the posted event counts and arm sizes, without adjustment. Since the record flags a discrepancy, the sponsor's own analysis takes precedence over the recalculation.

Precedence does not mean that every detail of the sponsor's analysis can be independently assessed. The registry record does not report a full statistical analysis plan or all modeling and endpoint-handling details. The recalculation remains a descriptive check on the posted summaries, while the sponsor-reported result is the record's designated analysis. Neither should be read as a medical recommendation.

14. Multiplicity, Missing Data and Analysis Populations

The registry record lists 9 outcome measures, including the primary clinically significant bleeding outcome and secondary bleeding and cardiovascular outcomes. It reports 8 secondary outcomes as posted and lists 18 analyses. The secondary outcomes include TIMI major bleeding, TIMI minor bleeding, bleeding requiring medical attention, a composite of cardiovascular death, myocardial infarction, and stroke, the individual cardiovascular components, and stent thrombosis.

When many outcomes or comparisons are assessed, the chance of at least one apparently unusual result can increase. Statistical methods can address multiplicity, for example by controlling a family-wise error rate or by defining a hierarchy of tests. Whether such methods were used, which outcomes were included in any testing family, and whether any testing hierarchy was prespecified are not reported in the registry record. The reported p-values should therefore be interpreted in light of this missing information rather than assumed to be multiplicity-adjusted.

The primary outcome's safety analysis set is defined as randomized participants who received at least 1 dose of study drug. The registry record does not provide a complete account of missing outcome data, exclusions from each analysis, or the handling of participants with incomplete follow-up. It also does not report whether analyses followed an intention-to-treat approach for every outcome. These details matter because the population included in an analysis and the handling of missing observations can affect estimates.

The record identifies the trial as randomized and open-label, but does not report details of allocation concealment, endpoint adjudication, or an analysis plan sufficient to assess all risk-of-bias and missing-data questions. These items are not reported in the registry record. Their absence from the posted record should not be treated as evidence that a particular procedure was or was not used.

15. Safety Results

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

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 2% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: cardiac failure (4.4%), angina unstable (3.3%), atrial fibrillation (3.3%), angina pectoris (2.2%), pneumonia (2.0%).

The adverse-event summary lists serious adverse events affecting 237 of 696 participants in the rivaroxaban 15 mg group, 225 of 706 in the rivaroxaban 2.5 mg twice-daily/15 mg once-daily group, and 271 of 697 in the vitamin K antagonist group. The record also lists other adverse events affecting 174 of 696, 210 of 706, and 234 of 697 participants, respectively. These are participant counts and group sizes, not adjusted comparisons. The registry does not provide confidence intervals or comparative tests for these totals in the posted summary.

The listed serious-event terms include cardiac failure at 4.4%, unstable angina at 3.3%, atrial fibrillation at 3.3%, angina pectoris at 2.2%, pneumonia at 2.0%, and congestive cardiac failure at 1.6%. These terms and percentages are presented as registry-record adverse-event information; they do not establish that an event was caused by a study drug. The death counts are not reported in the registry record.

Serious adverse-event summaries and the primary bleeding endpoint are different outcome summaries. They should not be treated as interchangeable measures of overall safety. The registry does not report enough detail to determine event overlap, exposure-adjusted rates, or the full criteria used for each adverse-event summary.

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

The registry record describes a randomized comparison of three antithrombotic strategies in participants with atrial fibrillation undergoing percutaneous coronary intervention. The primary safety measure was a composite of clinically significant bleeding through Month 12. The posted arm percentages were 15.7% for rivaroxaban 15 mg, 16.6% for the rivaroxaban 2.5 mg twice-daily/15 mg once-daily strategy, and 24.0% for vitamin K antagonist treatment.

The sponsor-reported log-rank analyses compare each rivaroxaban strategy with vitamin K antagonist treatment and report hazard ratios below 1, with 95% confidence intervals and p-values as. The simple unadjusted recalculation for the two rivaroxaban groups has a p-value of 0.643, while the recalculated vitamin K antagonist comparison with rivaroxaban 15 mg has a p-value of <0.001. Because the record flags a discrepancy, the sponsor's own analysis takes precedence; the recalculations are not replacements for that analysis.

These results do not establish that a treatment works or does not work. They describe outcomes observed in the trial and statistical comparisons reported in the registry. They also do not directly predict an individual's outcome. The treatment groups involved different combinations and sequences of antithrombotic drugs, so the results should be understood as comparisons of the strategies described, not as isolated comparisons of rivaroxaban, aspirin, or a P2Y12 inhibitor.

The interpretation is constrained by missing information about the sponsor's detailed statistical plan, multiplicity handling, missing outcomes, and some analysis procedures. The primary endpoint is composite, and the record does not give component-specific primary outcome estimates in its arm summaries. Consequently, the overall percentages alone do not show how much each bleeding category contributed. The posted findings are best read as a statistical account of the registered comparisons, with uncertainty and analysis differences kept in view.

17. Why This Trial Matters Statistically

This trial presents a useful statistical case because it compares treatment strategies that include different drug combinations, rather than simply contrasting two isolated medications. That distinction affects what the estimates mean. The primary endpoint is also a composite: it combines major bleeding, minor bleeding, and bleeding requiring medical attention. A single percentage can be concise, but it can mask variation among components with different definitions and clinical seriousness.

The record also illustrates why the analysis method matters. The arm percentages and simple unadjusted recalculations describe binary outcomes, while the reported log-rank hazard ratios summarize time-to-event comparisons. Those measures are not interchangeable. A reader who sees a hazard ratio should not interpret it as a direct percentage-point difference, and a comparison of final proportions does not capture event timing in the same way.

Another statistical lesson is the value of examining uncertainty and the analysis population alongside a point estimate. The confidence intervals reported for the hazard ratios provide information beyond the p-values, while the safety population definition clarifies who was included in the primary analysis. At the same time, absent details about multiplicity, missing data, and the full analysis plan limit independent appraisal.

The sample-size illustrations further show that planning depends strongly on the effect assumed for detection. Smaller target effects require more participants under the stated assumptions. Those illustrations are not the trial's original design calculation, and they are not evidence of observed power. Together, these features make the record a practical example of separating descriptive outcomes, inferential analyses, planning calculations, and unanswered methodological questions.

18. 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.
• Participant counts were reconstructed from posted percentages and arm sizes, so small rounding differences are possible.

The registry record does not report the original sample-size justification, so the planning illustrations based on observed effects cannot be used to reconstruct the trial's intended design assumptions. The detailed statistical analysis plan is also not reported. In particular, the record does not explain the sponsor's modeling choices, covariate adjustment, handling of censoring, or how the log-rank analyses relate exactly to the primary endpoint summary.

The primary endpoint is a composite with components of different definitions and potential severity. Although secondary bleeding outcomes are listed, the posted primary arm summaries do not show component-specific contributions. This limits interpretation of which types of bleeding account for the composite percentages. The study was open-label, and blinded outcome adjudication is not reported in the registry record.

The record reports the primary safety analysis set as randomized participants who received at least 1 dose of study drug, but does not give complete details of missing outcome data or the populations used for each reported analysis. Multiplicity procedures for the many outcomes and analyses are not reported. The number of secondary outcomes posted is 8, while the record lists 9 total outcome measures, so the posted record does not include results for every listed outcome.

The three strategies differ in drug combinations and sequence, including the described use of low-dose aspirin. Therefore, estimates compare the strategies as implemented and described, not the independent effect of one medication. The serious adverse-event summary provides counts and selected event terms, but the record does not report comparative uncertainty for those counts, exposure-adjusted rates, or death counts. These limitations constrain how fully the posted results can be independently interpreted.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: A Study Exploring Two Strategies of Rivaroxaban (JNJ39039039; BAY-59-7939) and One of Oral Vitamin K Antagonist in Patients With Atrial Fibrillation Who Undergo Percutaneous Coronary Intervention — NCT01830543.
• Registry record last updated 2017-09-19; status last verified 2017-08.
• Results first posted 2017-07-31.
• 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.

22. Record Summary

This completed randomized, open-label, parallel-group Phase 3 study enrolled 2124 participants with atrial fibrillation undergoing percutaneous coronary intervention. It compared rivaroxaban 15 mg once daily, a rivaroxaban 2.5 mg twice-daily strategy followed by 15 mg once daily, and dose-adjusted vitamin K antagonist treatment.

The primary endpoint was clinically significant bleeding up to Month 12, defined as a composite of TIMI major bleeding, TIMI minor bleeding, and bleeding requiring medical attention. Posted percentages were 15.7%, 16.6%, and 24.0% across the three groups, respectively. The registry also reports log-rank hazard ratios for each rivaroxaban strategy versus vitamin K antagonist treatment. A simple unadjusted recalculation is included as a separate statistical summary, and the sponsor's own analysis takes precedence because the record flags a discrepancy.

Important details, including the original sample-size justification, multiplicity procedures, missing-data handling, and the full analysis plan, are not reported in the registry record.