A Study to Assess the Effectiveness and Safety of Rivaroxaban in Reducing the Risk of Death, Myocardial Infarction or Stroke in Participants With Heart Failure and Coronary Artery Disease Following an Episode of Decompensated Heart Failure: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study to Assess the Effectiveness and Safety of Rivaroxaban in Reducing the Risk of Death, Myocardial Infarction or Stroke in Participants With Heart Failure and Coronary Artery Disease Following an Episode of Decompensated Heart Failure (NCT01877915), a randomized phase 3 study. The primary outcome is Event Rate of All-Cause Mortality, Myocardial Infarction (MI), or Stroke.

2. Clinical Question

The trial asked whether rivaroxaban, added to standard care, was associated with a different rate of all-cause death, myocardial infarction, or stroke than placebo plus standard care among participants with heart failure and significant coronary artery disease after an episode of decompensated heart failure. The primary result is an event rate per 100 patient-year, while the registry also reports a sponsor analysis using a time-to-event comparison.

3. Trial Design

This was a phase 3, randomized, quadruple-masked, parallel-group study sponsored by Janssen Research & Development, LLC. The registry lists enrollment of 5081 participants. Participants were assigned to rivaroxaban or matching placebo, and both groups received standard care for heart failure and coronary artery disease. The study was described as event-driven and multicenter. The registry record does not report the number or locations of participating centers.

The rivaroxaban regimen was 2.5 mg twice daily. The placebo group received matching placebo twice daily. Standard care was prescribed by each participant’s managing physician. The registry describes the study as comparing the efficacy and safety of rivaroxaban with placebo after an episode of decompensated heart failure in people with heart failure and significant coronary artery disease.

The listed start date was 2013-09-10, and primary completion was 2018-04-19. The study status is completed. The primary outcome timeframe was up to the global treatment end date, described as approximately 54 months. This timeframe is not an assertion that each participant was followed for the same length of time; the event-rate measure uses exposure time.

The primary analysis population was intent-to-treat. It included randomized unique participants who signed valid informed consent, with analyses according to the group assigned rather than the treatment actually received. The registry also lists nine outcome measures and three analyses. Details of the original sample-size justification, the randomization procedure, center-level conduct, and any planned interim analyses 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.

Before assignment: A total of 3 participants randomized twice were only counted once in the ITT analysis set (comprised of 5,022 participants).

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Intent-to-Treat analysis set included all randomized unique participants who signed a valid informed consent. Participants were analyzed according to the treatment group assigned, irrespective of the actual treatment received.

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 Event Rate of All-Cause Mortality, Myocardial Infarction (MI), or Stroke, measured over Up to Global treatment end date (approximately 54 months).

The primary endpoint is the event rate of all-cause mortality, myocardial infarction, or stroke. The registry describes the rate as based on the time to the first occurrence of one of these events. Its unit, events per 100 patient-year, expresses the number of events in relation to accumulated time at risk rather than simply counting events in groups of equal size.

This rate is useful when participants contribute different amounts of follow-up time. A participant who is followed longer contributes more exposure time than one followed for less time, so the rate connects event counts with the time over which events could occur. It is distinct from a simple proportion of participants who experienced an event and distinct from a hazard ratio, which compares event timing over follow-up under a time-to-event framework.

The composite brings death, myocardial infarction, and stroke into a single outcome measure. A composite can summarize the occurrence of any qualifying component, but it does not by itself establish that the components have equal clinical importance or that the group comparison is the same for each component. Component-specific details for the primary composite are not reported in the registry record.

The registry states that the rate was estimated from time to the first event and gives a rate formula based on event count and total risk exposure. The endpoint therefore describes the frequency of first qualifying events relative to exposure, not recurrent-event burden. No dispersion measure for the arm-specific rates is reported in the registry record.

7. Results Posted in the Registry

• Rivaroxaban: 13.44 versus 14.27 (Event rate per 100 patient-year) on control; rate ratio 0.94 (95% CI 0.88–1.01, approximate, Poisson model); p ≈ 0.094.

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.

Event Rate of Cardio Vascular Death or Re-Hospitalization for Worsening of Heart Failure (RHHF)

• Endpoint type: Event rate: events per unit of follow-up, compared as a rate ratio.
• Test: Poisson rate ratio, based on the estimated number of events and follow-up time.
• Result:
  • Rivaroxaban: 23.32 versus 23.46 (Event rate per 100 patient-year) on control; rate ratio 0.99 (95% CI 0.94–1.05, approximate, Poisson model); p ≈ 0.828.
• Interpretation:
  • The 95% confidence interval includes no difference (a ratio of 1), so these data do not show a statistically significant difference between Rivaroxaban and the control. The interval gives the range of effects that remain compatible with the data.

Event Rate of Cardio Vascular Death

• Endpoint type: Event rate: events per unit of follow-up, compared as a rate ratio.
• Test: Poisson rate ratio, based on the estimated number of events and follow-up time.
• Result:
  • Rivaroxaban: 9.46 versus 9.96 (Event rate per 100 patient-year) on control; rate ratio 0.95 (95% CI 0.87–1.03, approximate, Poisson model); p ≈ 0.228.
• Interpretation:
  • The 95% confidence interval includes no difference (a ratio of 1), so these data do not show a statistically significant difference between Rivaroxaban and the control. The interval gives the range of effects that remain compatible with the data.

Event Rate of Re-Hospitalization for Worsening of Heart Failure

• Endpoint type: Event rate: events per unit of follow-up, compared as a rate ratio.
• Test: Poisson rate ratio, based on the estimated number of events and follow-up time.
• Result:
  • Rivaroxaban: 17.24 versus 17.45 (Event rate per 100 patient-year) on control; rate ratio 0.99 (95% CI 0.93–1.05, approximate, Poisson model); p ≈ 0.705.
• Interpretation:
  • The 95% confidence interval includes no difference (a ratio of 1), so these data do not show a statistically significant difference between Rivaroxaban and the control. The interval gives the range of effects that remain compatible with the data.

The posted secondary outcomes cover cardiovascular death, hospitalization for worsening heart failure, combinations of these events, re-hospitalization for cardiovascular events, all-cause mortality combined with worsening-heart-failure hospitalization, and bleeding outcomes. Across the secondary rate comparisons with results available in the record, the arm-specific rates are close, and the unadjusted comparisons do not establish clear differences. The pattern is broadly consistent with the primary result: the posted summaries do not provide strong statistical evidence of separation between the assigned groups.

Secondary endpoints address clinically distinct questions and should not be treated as interchangeable with the primary composite. Outcomes that combine death and hospitalization, for example, have different event definitions from the primary combination of death, myocardial infarction, or stroke. The registry lists seven secondary outcomes, while only some have detailed rate comparisons in the posted results summarized for this analysis. Results for secondary outcomes without a posted comparison are not reported in the registry record.

Unadjusted secondary p-values carry less weight when multiple outcomes are examined. In general, testing many outcomes increases the chance that one or more results may appear noteworthy by chance, even if no true differences exist. A multiplicity adjustment or a prespecified hierarchy can address that issue, but whether either was used for the secondary outcomes is not reported in the registry record. The secondary results are therefore best interpreted as supportive descriptive information rather than independent confirmation of an effect.

The secondary endpoints also do not resolve uncertainty in the primary result. Similar patterns across related outcomes can provide context, but they do not replace the primary analysis or establish what would have been found under a different definition, population, or follow-up period. Component-level details and adjustment procedures are not reported in the registry record.

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

• Event Rate of Cardio Vascular Death or Re-Hospitalization for Worsening of Heart Failure (RHHF)
• Event Rate of Cardio Vascular Death
• Event Rate of Re-Hospitalization for Worsening of Heart Failure
• Event Rate of Re-Hospitalization for Cardio Vascular Events (RHCV)
• Event Rate of All-Cause Mortality (ACM) or Re-Hospitalization for Worsening Heart Failure
• Event Rate of Bleeding Events That Requiring Hospitalization
• Event Rate of International Society on Thrombosis and Haemostasis (ISTH) Major Bleeding Event

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is reported as an event rate per 100 patient-year. In general, an event rate divides an event count by the amount of time participants contribute while at risk, then expresses that frequency on a chosen exposure scale. Here the registry defines the rate using the number of events and total risk exposure. This is appropriate as a descriptive measure when follow-up exposure can differ among participants. It is not the same as the fraction of participants who ever experienced the event.

What does the rate ratio compare?

A rate ratio compares the event rate in one group with the rate in the other. A value of 1 represents equal rates; values below 1 indicate a lower rate in the numerator group, and values above 1 indicate a higher rate. The recalculated primary comparison is an unadjusted rate ratio for rivaroxaban relative to placebo. Its interval includes 1, so the results do not distinguish the group rates clearly at the interval’s stated level of uncertainty. The rate ratio is a summary of event frequency per exposure, rather than a direct statement about the probability that an individual participant will experience an event.

How should the sponsor analysis be read?

The registry reports the sponsor’s primary analysis as a log-rank test with a hazard ratio. A hazard ratio compares the instantaneous event rates over time among participants who have not yet experienced the event, under the time-to-event model and assumptions used for the analysis. A hazard ratio of 1 corresponds to no relative difference in the event hazard. The sponsor reports a hazard ratio of 0.94, a 95% confidence interval from 0.84 to 1.05, and p = 0.270. The interval spans 1, so it is compatible with no difference in hazard. The p-value describes how unusual the data would be under a specified null model; it is not the probability that the null hypothesis is true.

The event-rate comparison and hazard-ratio analysis should not be treated as identical. Event rates compress exposure and events into a frequency measure, whereas time-to-event methods use the timing of events and account for censoring through the analysis framework. Their estimates can differ without indicating a data error, because they answer related but not identical statistical questions. The registry record does not report the detailed model assumptions, censoring rules, or covariate adjustments for the sponsor analysis.

What does a confidence interval convey?

In general, a confidence interval describes the range of values reasonably compatible with the data under the method used. It conveys uncertainty about an estimate, not a range containing a known fixed proportion of individual participant outcomes. An interval that includes the no-difference value means the data do not rule out that value at the interval’s confidence level. It may also include values suggesting differences in either direction. The width of the interval reflects precision: narrower intervals are more precise, while wider intervals leave more uncertainty.

What does the p-value add?

A p-value measures compatibility between the observed data and a null hypothesis, assuming the statistical model and analysis conditions hold. It is not an effect size and does not express clinical importance. The recalculated primary p-value is unadjusted, meaning it is not adjusted for multiple comparisons in the posted calculation. The registry does not report whether the sponsor’s p-value was adjusted for any multiplicity procedure. A p-value above a conventional significance threshold should not be read as evidence that the groups are exactly alike.

The recalculation uses the posted arm-level summaries and is unadjusted. It should be understood as a descriptive statistical reconstruction, not as a replacement for the prespecified trial analysis. The sponsor’s reported analysis is based on its stated log-rank approach, and its estimate should be interpreted according to that analysis rather than assumed to equal the rate-ratio estimate.

Why can rates be useful, and what do they leave out?

Exposure-based rates allow a comparison when total follow-up time is not identical across participants. However, a rate alone does not show the distribution of follow-up, when events occurred, or whether event rates changed over time. A single rate can also conceal differences among the composite components. The registry does not provide a separate primary analysis for death, myocardial infarction, and stroke in the results. In general, additional component results and time-to-event displays can help explain a composite, but no such details should be inferred when they are not reported.

The primary outcome description states that the rate was estimated from time to first occurrence. Thus, the reported rate does not represent a count of every recurrent event after the first qualifying event. The registry’s description and posted analysis should be read together: one gives the event-rate summary, while the sponsor’s log-rank analysis supplies a time-to-event comparison.

11. Confidence Intervals and Precision

The sponsor-reported hazard ratio is 0.94, with a 95% confidence interval from 0.84 to 1.05. The interval includes 1, the value corresponding to no relative difference in hazard. Its lower and upper limits span estimates on both sides of that value, so the reported result leaves uncertainty about the direction and size of any difference. The associated p-value is 0.270.

The recalculated primary comparison also has an interval that includes the no-difference value for a rate ratio. The arm-specific event rates are 13.44 and 14.27 per 100 patient-year, and the unadjusted p-value is 0.094. The registry does not report an interval for each arm-specific rate. An interval for the comparison is more informative about the relative contrast than comparing the two point estimates alone, because it reflects uncertainty in that contrast.

Confidence intervals are not guarantees about the true effect, nor do they show the range of outcomes for individual participants. In general, they indicate which effect values are compatible with the data under the method and assumptions used. A wide interval indicates limited precision; a narrow interval indicates greater precision. Even when an interval excludes the no-difference value, that statistical feature alone would not establish clinical importance.

The study enrolled 5081 participants, but sample size by itself does not determine precision. Event frequency, follow-up exposure, variability, censoring, and the analysis method also matter. The registry record does not report all details needed to independently assess these contributors. The primary completion timeframe is described as up to the global treatment end date, approximately 54 months, but participant-specific exposure is not stated in the results summary.

12. 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 a report of the trial’s original design calculation. They use a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is set at 100%, 75%, or 50% of the observed effect. These assumptions are known for the planning illustration; the original sample-size assumptions for the trial are not reported in the registry record.

For the primary rate comparison, the figures are participants per arm. For a planning scenario targeting 100% of the observed effect, the estimate is 7,018 participants per arm for 80% power and 9,395 participants per arm for 90% power. At 75% of the observed effect, the estimates are 12,381 and 16,575 participants per arm, respectively. At 50% of the observed effect, they are 27,645 and 37,009 participants per arm.

In plain terms, 80% power means that, under the planning assumptions and if the specified effect were the true effect, the design would have an 80% chance of detecting it at the stated significance level. The 90% column describes a higher chance of detecting that same specified effect under those assumptions. Smaller effects are harder to distinguish from random variation, so they require more participants. These values are not observed power: they are prospective planning illustrations calculated from the posted effect, not a measure of what the completed trial achieved.

The original trial design was event-driven, but its original sample-size justification and target event count are not reported in the registry record. The planning estimates should not be interpreted as a redesign recommendation without considering the trial question, endpoint, follow-up, analysis plan, and practical context.

13. 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, λ₀ and λ₁ are event rates per patient-year, and T is follow-up per participant in years. This assumes Poisson counts and equal follow-up.

Worked calculation for this trial

1. Control rate λ₀ = 0.1427; treatment rate λ₁ = 0.1344 events per patient-year; follow-up T = 4.50 years.
2. n = (1.96 + 0.842)² × (1/0.1427 + 1/0.1344) / (4.50 × [ln(0.942)]²) = 7,017.9, rounded up to 7,018 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
lambda0 <- 0.1427                      # control rate, events per patient-year
lambda1 <- 0.1344                      # treatment rate
tt      <- 4.5                       # follow-up per participant, years
n_per_arm <- function(power, frac = 1) {
  l1 <- lambda0 * (lambda1 / lambda0)^frac   # smaller effect = fraction of the observed log rate ratio
  ceiling(round((z_a + qnorm(power))^2 * (1 / lambda0 + 1 / l1) / (tt * log(l1 / lambda0)^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 = ",")))

# Simulation 1: Poisson counts. Simulation 2: over-dispersed counts (variance = 1.5 x mean), analysed with the matching variance.
set.seed(1877915)
n    <- n_per_arm(0.80)
reps <- 20000
m0 <- lambda0 * n * tt; m1 <- lambda1 * n * tt
x0 <- rpois(reps, m0); x1 <- rpois(reps, m1)
z  <- log(x1 / x0) / sqrt(1 / x0 + 1 / x1)
cat(sprintf("Simulated power, Poisson counts, n = %s per arm: %.3f\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE)))
phi <- 1.5
y0 <- rnbinom(reps, mu = m0, size = m0 / (phi - 1)); y1 <- rnbinom(reps, mu = m1, size = m1 / (phi - 1))
z2 <- log(y1 / y0) / sqrt(phi * (1 / y0 + 1 / y1))
cat(sprintf("Simulated power, over-dispersed counts (phi = 1.5), same n: %.3f\n", mean(abs(z2) > z_a, na.rm = TRUE)))

Output from running the code

100% of observed effect: n per arm = 7,018 (80% power), 9,395 (90% power)
 75% of observed effect: n per arm = 12,381 (80% power), 16,575 (90% power)
 50% of observed effect: n per arm = 27,645 (80% power), 37,009 (90% power)
Simulated power, Poisson counts, n = 7,018 per arm: 0.797
Simulated power, over-dispersed counts (phi = 1.5), same n: 0.627

The first simulation draws Poisson counts and should land close to the target power. The second allows extra variability (variance 1.5 times the mean, a purely illustrative value), which lowers the power at the same sample size, so rate trials usually need more participants than the Poisson formula suggests.

14. Analyses Reported by the Sponsor

The registry’s reported primary analysis is a log-rank comparison expressed as a hazard ratio of 0.94, with a 95% confidence interval from 0.84 to 1.05 and p = 0.270. The recalculated comparison uses posted arm-level event rates and is an unadjusted rate-ratio analysis. These are different statistical summaries: the sponsor’s analysis concerns event timing, while the recalculation compares events per unit of exposure.

Both results are compatible with no clear group difference: the sponsor’s interval includes 1 for the hazard ratio, and the recalculated rate-ratio interval also includes its no-difference value. The rate comparison’s unadjusted p-value is 0.094. The numerical estimates need not match because the measures and analysis approaches differ. They should not be combined as though they were duplicate estimates of a single quantity.

The sponsor’s analysis is the trial’s reported analysis and should be read as such. The recalculation is a simple unadjusted analysis of the posted summaries, not a reconstruction of the full trial analysis plan. The registry record does not report enough detail about adjustments, censoring, or other modeling choices to reproduce the sponsor’s result independently. No discrepancy between the recalculation and sponsor analysis is flagged in the registry record.

15. Multiplicity, Missing Data and Analysis Populations

The registry identifies a primary outcome and seven secondary outcomes. It does not report a multiplicity adjustment strategy, a testing hierarchy, or whether particular secondary outcomes were designated as confirmatory. Without that information, the secondary p-values should be treated as unadjusted exploratory comparisons unless the record states otherwise. When several outcomes are tested, the chance of at least one apparently notable result can increase; this is why prespecified procedures for controlling multiplicity matter.

The primary recalculation is explicitly unadjusted. The registry reports a sponsor log-rank analysis for the primary outcome, but whether that reported p-value was adjusted for any interim or multiplicity procedure is not reported in the registry record. The record also does not report whether any interim analyses occurred or whether they influenced the analysis thresholds. No assumptions about such procedures can be made from the event-driven description alone.

The primary analysis population is described as intent-to-treat: randomized unique participants who signed valid informed consent were analyzed according to assigned group, irrespective of treatment actually received. This is an important definition because it preserves the randomized comparison. The record does not report detailed exclusions, the amount or pattern of missing outcome data, handling of loss to follow-up, censoring rules, or sensitivity analyses for missing data. The extent to which these issues could influence estimates therefore cannot be assessed from the registry record.

The registry’s primary outcome description refers to time to first occurrence, and the sponsor analysis is identified as a log-rank method. However, the record does not report the detailed statistical analysis plan, model diagnostics, covariate adjustments, or how competing events were handled. These omissions limit independent evaluation of the analysis choices. They do not establish that the methods were inappropriate; they mean only that the relevant details are not reported in the registry record.

16. 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 2% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: pneumonia (2.9%), acute kidney injury (1.1%), sepsis (0.6%), chronic obstructive pulmonary disease (0.6%), urinary tract infection (0.6%).

The registry lists serious adverse events in 479 participants in the rivaroxaban group, with 2507 at risk, and 451 participants in the placebo group, with 2515 at risk. It separately lists other adverse events affecting 53 participants assigned rivaroxaban and 60 assigned placebo. Deaths are listed for 565 and 571 participants, respectively. These are descriptive counts; they do not by themselves show whether an event was caused by study treatment or establish a comparative safety effect.

The most frequently listed serious adverse event terms include pneumonia at 2.9%, acute kidney injury at 1.1%, and sepsis, chronic obstructive pulmonary disease, urinary tract infection, and chronic kidney disease at 0.6% each. The registry record does not provide enough detail in these figures to assess event timing, duration, severity within each term, or attribution. Adverse-event counts should be interpreted alongside the study’s definitions and follow-up, details that are not fully reported in the registry record.

Comparisons of safety counts can be affected by exposure time and reporting practices. The group totals alone are not exposure-adjusted rates, and the record does not report a formal inferential comparison for these adverse-event figures. These results describe what was posted and should not be read as proof of either harm or safety.

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

The posted primary results show event rates of 13.44 per 100 patient-year with rivaroxaban and 14.27 per 100 patient-year with placebo. The sponsor-reported time-to-event estimate is a hazard ratio of 0.94, with a 95% confidence interval from 0.84 to 1.05 and p = 0.270. The recalculated unadjusted rate comparison and the sponsor’s reported analysis both leave uncertainty that includes no difference between groups.

These data do not establish that the treatments are equivalent. Failure to demonstrate a statistically clear difference is not the same as evidence that the true difference is exactly zero. Nor does the numerical direction of a point estimate establish that a treatment effect exists. Precision, analysis method, endpoint definition, and the population studied all shape what can be concluded.

The primary endpoint combines all-cause death, myocardial infarction, and stroke and is based on the first occurrence. It does not, by itself, describe each component separately or recurrent events. The registry also reports secondary outcomes and adverse-event counts, but their presence does not change the primary result’s uncertainty. Secondary findings are less definitive when multiplicity handling is not reported.

The interpretation is limited to the randomized comparison and participant population described in the registry: people with heart failure and significant coronary artery disease after decompensated heart failure, assigned to rivaroxaban 2.5 mg twice daily or matching placebo, each alongside standard care. The results do not automatically apply to other doses, populations, treatment settings, or outcomes. They are statistical findings and do not constitute medical advice or a recommendation for an individual patient.

18. Why This Trial Matters Statistically

This trial illustrates why statistical interpretation should distinguish an event rate from a time-to-event estimate. The event-rate comparison relates events to accumulated exposure, while the sponsor’s log-rank analysis uses event timing and reports a hazard ratio. Both are legitimate kinds of summaries for time-dependent outcomes, but they are not identical estimands. A reader who compares the numbers without recognizing this distinction may mistake different summaries for conflicting results.

The results also show why a point estimate should be interpreted alongside uncertainty. The arm rates point numerically in one direction, but the comparison’s uncertainty includes the no-difference value. The sponsor’s hazard-ratio interval does as well. A p-value alone cannot communicate the plausible range of effects or clinical importance, and a non-significant result is not proof of equality.

The event-driven design and exposure-based primary endpoint make follow-up and event ascertainment central to interpretation. The endpoint is a composite, so its overall rate cannot reveal which component contributes most to the result. Component-specific primary results and details of missing-data handling are not reported in the registry record, limiting further assessment.

Finally, the registry includes multiple secondary endpoints. Without reported multiplicity procedures, those analyses should be interpreted cautiously. Together, these features make the study a useful example of reading the analysis method, effect estimate, interval, outcome structure, and reporting limitations as a connected set rather than treating any single statistic as a complete answer.

19. 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.
• Rate comparisons use a Poisson approximation unless the sponsor posted confidence limits; real event counts are often over-dispersed.

Several trial-specific details needed for a fuller assessment are not reported in the registry record. The primary outcome is a composite, but component-specific results for all-cause mortality, myocardial infarction, and stroke are not reported with the primary analysis. The relative contributions of those events to the composite therefore cannot be evaluated from these results alone.

The primary event-rate summaries do not include a dispersion measure for each arm, and the registry does not report participant-level exposure distributions. The stated timeframe is up to the global treatment end date, approximately 54 months, but this does not show how follow-up varied among participants. Detailed censoring rules and the handling of competing events are not reported in the registry record.

The sponsor analysis is identified as a log-rank test with a hazard ratio, but its detailed model specifications, covariate adjustments, and diagnostic checks are not reported. The registry also does not report the original sample-size justification, an interim-analysis plan, or whether any interim analyses occurred. The planning calculations based on the observed effect are not a substitute for those original design details.

Multiplicity methods for secondary outcomes, missing-data patterns, treatment adherence, crossover, and sensitivity analyses are also not reported in the registry record. The adverse-event information is presented as counts and selected terms; formal comparative safety analyses and detailed attribution are not reported in the results. These limitations restrict independent assessment and the scope of interpretation, but they do not alone demonstrate that the trial conduct or analysis was flawed.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Study to Assess the Effectiveness and Safety of Rivaroxaban in Reducing the Risk of Death, Myocardial Infarction or Stroke in Participants With Heart Failure and Coronary Artery Disease Following an Episode of Decompensated Heart Failure — NCT01877915.
• Registry record last updated 2019-05-10; status last verified 2019-04.
• Results first posted 2019-05-10.
• 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.

23. Record Summary

NCT01877915 is a completed phase 3 randomized, quadruple-masked, parallel-group study of rivaroxaban versus placebo, both alongside standard care, in participants with heart failure and coronary artery disease following decompensated heart failure. Enrollment was 5081. The primary outcome was the event rate of all-cause mortality, myocardial infarction, or stroke per 100 patient-year, assessed through time to first occurrence.

The posted primary event rates were 13.44 with rivaroxaban and 14.27 with placebo per 100 patient-year. The sponsor reported a log-rank hazard ratio of 0.94, with a 95% confidence interval from 0.84 to 1.05 and p = 0.270. The recalculated unadjusted rate comparison also remains uncertain and does not establish a clear group difference. Details of multiplicity handling, missing data, and the original sample-size justification are not reported in the registry record.