← Clinical Trials
Cardiovascular Disease Phase 3 Randomized NCT01594333

CIRT: Complete Statistical Analysis of Methotrexate in Cardiovascular Disease

An independent statistical analysis of the Cardiovascular Inflammation Reduction Trial (CIRT), a randomized phase 3 comparison of methotrexate versus placebo in cardiovascular disease, focusing on its time-to-event endpoints, hazard-ratio estimates, stratified log-rank analyses, multiplicity considerations, and reported safety results.

Trial status: COMPLETED  ·  Enrollment: 4786  ·  Primary completion: April 15, 2019
Scope of this record

This page separates reported trial results from statistical interpretation. Trial-specific numerical results and design facts on this page are taken from the ClinicalTrials.gov record. The registry reports two primary time-to-event analyses and six additional posted statistical analyses.

Registry record: This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record.

1. Trial at a Glance

CIRT was a randomized, parallel-group, triple-masked, phase 3 trial comparing methotrexate with placebo in participants with cardiovascular disease. The trial enrolled 4786 participants and had two registered primary endpoints, both analyzed as time-to-event outcomes.

4786
Enrolled
Randomized trial
2
Study arms
Methotrexate vs placebo
1.01
Primary MACE HR
95% CI 0.82–1.25
0.96
Primary composite HR
95% CI 0.79–1.16
FeatureCIRT
Brief titleCardiovascular Inflammation Reduction Trial
PhasePhase 3
StatusCOMPLETED
Therapeutic areaCardiovascular
ConditionCardiovascular Disease
AllocationRANDOMIZED
Design modelPARALLEL
MaskingTRIPLE
Primary purposeTREATMENT
Enrollment4786
Lead sponsorBrigham and Women's Hospital
Start2013-04
Primary completion2019-04-15
Registered primary endpoints2
Outcome measures posted13
Statistical analyses posted8

2. Clinical Question

The primary statistical question was whether assignment to methotrexate, compared with placebo, changed the time to occurrence of major cardiovascular events in participants with cardiovascular disease. The registry specified two primary time-to-event endpoints and a superiority hypothesis for each.

Population

Participants enrolled in the phase 3 trial with the registry condition of cardiovascular disease.

Intervention

Methotrexate.

Comparator

Placebo.

Primary question

Does methotrexate change the hazard of the prespecified cardiovascular time-to-event outcomes compared with placebo?

3. Trial Design

01
Randomize4786 participants
02
Two armsMethotrexate or placebo
03
FollowTime-to-event outcomes
04
CompareStratified log-rank
05
EstimateCox hazard ratio
Allocation
Randomized allocation to two parallel study arms.
Masking
Triple-masked trial according to the registry record.
Phase
Phase 3 treatment trial.
Hypothesis
Superiority testing was used for the posted statistical analyses.
ARM A · METHOTREXATE

Methotrexate

  • Randomized study intervention.
  • Included in the primary and secondary time-to-event comparisons.
  • Serious adverse events: 569 affected participants among 2391 at risk.
ARM B · PLACEBO

Placebo

  • Randomized comparator.
  • Included in the primary and secondary time-to-event comparisons.
  • Serious adverse events: 549 affected participants among 2395 at risk.
Early termination: The registry states that the trial was terminated early because the pre-specified boundary for futility was crossed for both the original and the final primary endpoints and because of the lack of evidence of a reduction in hsCRP level with methotrexate treatment. This is a design and interpretation feature of the trial and should not be confused with a treatment-effect estimate.

4. Primary Endpoints

The two registered primary endpoints were both defined around the first occurrence of clinically important cardiovascular events. Both were analyzed from randomization to the trial end, April 2, 2018, with a maximum follow-up of 5 years.

Primary endpointRegistered definitionTime frame
Number of Subjects With Major Adverse Cardiovascular Events The first occurrence of Major Adverse Cardiovascular Event, defined as the occurrence of one or more of the following: Cardiovascular Death, Non-Fatal Myocardial Infarction or Non-Fatal Stroke. From randomization to trial end (April 2, 2018) up to a maximum of 5 years
Number of Subjects With Major Adverse Cardiovascular Event or Hospitalization for Unstable Angina That Led to Urgent Coronary Revascularization The first occurrence of Major Adverse Cardiovascular Event or Hospitalization for Unstable Angina that led to Urgent Coronary Revascularization. From randomization to trial end (April 2, 2018) - up to a maximum of 5 years

5. Statistical Methodology

Time-to-event analysis

Both primary endpoints were treated as time-to-event outcomes. The relevant quantity is not simply whether an event occurred, but when the first qualifying event occurred during follow-up. Participants who did not experience the endpoint before the end of their available follow-up contribute information until their censoring time.

This distinction is important because two randomized groups can have similar proportions of observed events at one point in time while having different event timing. Survival-analysis methods use the ordering and timing of events rather than reducing the entire follow-up period to a single binary outcome.

Log-rank test

The registry reports a log-rank test for both primary endpoints. The log-rank procedure compares the observed pattern of events between randomized groups over follow-up, accounting for the changing number of participants at risk.

Conceptual comparison
Observed events − expected events under the null hypothesis

The log-rank test evaluates whether the observed event experience differs systematically between the treatment groups under the specified survival-analysis framework. Its P-value addresses evidence against the null hypothesis; it is not a measure of how large or clinically important an effect is.

Hazard ratio

The effect measure reported for the primary analyses was a Cox proportional-hazard estimate. A hazard ratio compares the estimated instantaneous event rate between groups within the fitted model.

Interpretation of the hazard ratio
HR = estimated hazard in methotrexate group ÷ estimated hazard in placebo group

An HR below 1 indicates a lower estimated hazard in the methotrexate group; an HR above 1 indicates a higher estimated hazard. The HR is not an absolute risk difference and does not state the probability that an individual participant will experience an event.

Stratified analysis

The primary analyses were stratified according to type of qualifying event, time of qualifying event, and risk status. The registry specifies risk status using diabetes or metabolic syndrome alone for the first primary endpoint and metabolic syndrome alone or diabetes at enrollment for the second primary endpoint.

Stratification allows the comparison to account for prespecified categories that can affect the event process. Rather than assuming that all participants contribute to a single completely homogeneous risk set, the analysis compares treatment experience within the relevant strata and combines the resulting information according to the statistical model.

Analysis population

The primary MACE analysis used all randomized subjects. The second primary analysis used all randomized participants. This is consistent with an analysis anchored to randomized assignment rather than restricting the efficacy comparison to participants who completed a particular amount of treatment.

6. Primary Results: Major Adverse Cardiovascular Events

The first primary endpoint was the first occurrence of a Major Adverse Cardiovascular Event, defined in the registry as cardiovascular death, non-fatal myocardial infarction, or non-fatal stroke. The posted analysis compared methotrexate with placebo using a stratified log-rank test and a Cox proportional-hazard estimate.

Hazard ratio for Major Adverse Cardiovascular Events

1.01

95% CI: 0.82–1.25   ·   P = 0.91

Analysis population: all randomized subjects  ·  Two-sided 95% confidence interval

EndpointMethodEffect measureEstimate95% CIP-value
Major Adverse Cardiovascular Events Stratified log-rank test Cox proportional hazard 1.01 0.82–1.25 0.91
Clinical Biostats interpretation

The estimated hazard ratio of 1.01 is very close to 1.00. Within the fitted Cox model, this corresponds to an estimated instantaneous event rate that is approximately the same in the methotrexate and placebo groups.

The estimate does not mean that the two groups had exactly identical event risks for every participant or at every point in time. A hazard ratio is a model-based relative measure of event rates over the analyzed follow-up.

The 95% confidence interval of 0.82–1.25 shows the statistical uncertainty around the estimate. It spans values below and above 1.00, so the interval is compatible with a lower hazard, little difference, or a higher hazard under the statistical model and sampling framework.

The P = 0.91 value addresses the strength of evidence against the specified null hypothesis; it does not measure effect size, clinical importance, or the probability that the treatment has no effect. The size and location of the hazard-ratio estimate and its confidence interval are needed to understand the magnitude and precision of the comparison.

The analysis notes also state that confidence intervals were not adjusted for multiplicity. Therefore, the reported interval should be interpreted in the context of the broader set of analyses rather than as an independently multiplicity-adjusted interval.

7. Primary Results: Major Adverse Cardiovascular Event or Urgent Coronary Revascularization

The second primary endpoint expanded the event definition to the first occurrence of a Major Adverse Cardiovascular Event or hospitalization for unstable angina that led to urgent coronary revascularization.

Hazard ratio for the second primary endpoint

0.96

95% CI: 0.79–1.16   ·   P = 0.67

Analysis population: all randomized participants  ·  Two-sided 95% confidence interval

EndpointMethodEffect measureEstimate95% CIP-value
Major Adverse Cardiovascular Event or Hospitalization for Unstable Angina That Led to Urgent Coronary Revascularization Stratified log-rank test Cox proportional hazard 0.96 0.79–1.16 0.67
Clinical Biostats interpretation

The hazard ratio of 0.96 corresponds to an estimated hazard approximately 4% lower in the methotrexate group than in the placebo group under the fitted model. That small relative difference is an estimate, not a statement that each participant had exactly 4% lower risk.

The 95% confidence interval of 0.79–1.16 includes 1.00. It therefore encompasses both a lower estimated hazard and a higher estimated hazard, illustrating the uncertainty around the point estimate.

The P = 0.67 value is evidence against the null hypothesis at the level represented by that test, but it should not be interpreted as a 67% probability that the treatment works or does not work. P-values do not quantify the magnitude of an effect.

As with the first primary endpoint, the analysis was stratified and the confidence interval was not adjusted for multiplicity. The registry specifically cautions that inferences drawn from these intervals may not be reproducible.

8. Secondary Time-to-Event Results

The registry also posts formal analyses for four secondary endpoints. Each used all randomized participants, compared methotrexate with placebo, and used a log-rank test with a Cox proportional-hazard effect measure. The analyses incorporated stratification and the registry notes that confidence intervals were not adjusted for multiplicity.

Secondary endpointHR95% CIP-valueTime frame
All-cause mortality 1.16 0.87–1.56 0.32 From randomization to trial end (April 2, 2018) up to a maximum of 5 years
Major Adverse Cardiovascular Event or Any Coronary Revascularization 0.95 0.81–1.12 0.57 From randomization to trial end (April 2, 2018) - up to a maximum of 5 years
Hospitalization for Congestive Heart Failure 0.89 0.60–1.31 0.54 From randomization to trial end (April 2, 2018) up to a maximum of 5 years
Major Adverse Cardiovascular Event, Coronary Revascularization, Hospitalization for Congestive Heart Failure or All Cause Mortality 0.98 0.84–1.14 0.80 From randomization to trial end (April 2, 2018) up to a maximum of 5 years

How these secondary estimates should be read

The secondary estimates range from 0.89 to 1.16. Some point estimates are below 1 and one is above 1, but the confidence intervals posted on ClinicalTrials.gov for these analyses all include 1.00. These estimates therefore do not provide a simple basis for treating any one secondary endpoint as an isolated definitive treatment-effect finding.

There is an additional statistical issue: the registry explicitly states that the confidence intervals were not adjusted for multiplicity. When multiple endpoints are examined, the probability of obtaining an apparently unusual result somewhere among the collection of analyses can increase. Consequently, nominal confidence intervals and P-values from individual secondary analyses should be interpreted as part of the overall analysis set rather than as though each were a standalone confirmatory test.

9. Additional Pre-Specified Analyses

Two additional pre-specified time-to-event analyses were posted in the registry. The first compares methotrexate for hospitalization for unstable angina that led to unplanned coronary revascularization; the registry's analysis record lists the compared group as Methotrexate. The second evaluates coronary revascularization and explicitly compares methotrexate with placebo.

EndpointGroups comparedHR95% CIP-value
Hospitalization for Unstable Angina That Led to Unplanned Coronary Revascularization Methotrexate 0.81 0.53–1.22 0.31
Coronary Revascularization Methotrexate vs Placebo 0.92 0.75–1.12 0.38

Both analyses used a log-rank test and Cox proportional-hazard effect measure. Both used two-sided 95% confidence intervals, and both included the registry's stratification and multiplicity caveat. The point estimate of 0.81 for hospitalization for unstable angina that led to unplanned coronary revascularization is compatible with a lower hazard, but its confidence interval of 0.53–1.22 also includes values above 1.00.

10. Statistical Methods Explained

Why was a log-rank test used?

All eight posted statistical analyses use time-to-event methods. A log-rank test is designed for comparing survival or event-time distributions between groups while accounting for the timing of events and censoring. That makes it more informative for these endpoints than simply comparing the proportion of participants who eventually experienced an event.

What does a hazard ratio of 1.01 mean?

An HR of 1.01 means the fitted model estimates the instantaneous event hazard in the methotrexate group to be about 1.01 times that in the placebo group. Because 1.01 is very close to 1, the estimated relative difference is small. It does not mean that the cumulative risks are exactly equal at every time point.

Why does the confidence interval matter?

The point estimate is only one estimate from the observed trial data. The 95% confidence interval describes statistical uncertainty around that estimate under the model and sampling framework. For the first primary endpoint, 0.82–1.25 is considerably wider than the single value 1.01 and includes both sides of the null value 1.00.

Why doesn't the P-value measure the size of the treatment effect?

A P-value evaluates how compatible the observed data are with a specified null hypothesis under the statistical model. It depends on both the estimated effect and the amount of information in the analysis. It therefore cannot be used as a substitute for the hazard ratio or its confidence interval.

Why was stratification used?

The registry states that the analyses were stratified according to type of qualifying event, time of qualifying event, and risk status. Stratification permits the analysis to account for these specified categories when comparing event times, rather than treating all risk sets as if they were identical.

What does the proportional-hazards assumption contribute?

The Cox proportional-hazards framework expresses treatment effect through a hazard ratio. The interpretation of a single HR is most straightforward when the relative hazards are reasonably represented by a proportional-hazards relationship over time. The registry supplies the Cox hazard-ratio estimates, but the ClinicalTrials.gov record does not provide a diagnostic assessment of proportional hazards. Therefore, this page does not infer that the assumption was verified.

Why is multiplicity important here?

The registry reports two primary endpoints plus multiple secondary and additional pre-specified analyses. It also explicitly states that the reported confidence intervals were not adjusted for multiplicity. That means a reader should distinguish the prespecified analysis framework from an interpretation in which every posted endpoint is treated as an independent confirmatory test.

11. Confidence Intervals, Null Values, and Effect Size

Estimate95% CIP-valueWhat the numbers describe
Primary MACE HR 1.010.82–1.250.91Point estimate close to 1, with uncertainty spanning lower and higher hazards
Primary expanded endpoint HR 0.960.79–1.160.67Small estimated relative difference, with uncertainty spanning 1
All-cause mortality HR 1.160.87–1.560.32Point estimate above 1, with substantial uncertainty around the direction and magnitude
MACE or any coronary revascularization HR 0.950.81–1.120.57Point estimate slightly below 1, with interval including 1
Hospitalization for congestive heart failure HR 0.890.60–1.310.54Point estimate below 1, with a relatively wide interval
Expanded cardiovascular composite HR 0.980.84–1.140.80Point estimate close to 1, with interval spanning both sides

A useful discipline in reading these results is to separate three quantities. The hazard ratio describes the estimated relative effect. The confidence interval describes uncertainty around that estimate. The P-value addresses compatibility with a null hypothesis. None of the three alone describes absolute event probability or an individual's treatment response.

A practical reading rule

For a hazard ratio, 1.00 is the null value. Values below 1 indicate a lower estimated hazard in the treatment group, while values above 1 indicate a higher estimated hazard. Whether the estimate is precise enough to support a particular conclusion depends on the confidence interval, the prespecified analysis plan, multiplicity, censoring, and the assumptions underlying the survival model.

12. Multiplicity and the Broader Analysis Set

CIRT's registry record contains 13 posted outcome measures and 8 posted statistical analyses. The two primary analyses were accompanied by several secondary and additional pre-specified analyses.

Multiple endpoints

Testing or estimating several outcomes creates a broader inferential context than evaluating a single endpoint in isolation.

Unadjusted intervals

The registry states that confidence intervals were not adjusted for multiplicity and cautions that inferences from them may not be reproducible.

Nominal P-values

A P-value reported for an individual endpoint should not automatically be interpreted as if it came from a separate confirmatory testing family.

Interpretation hierarchy

Primary endpoints, secondary endpoints, and other pre-specified analyses should remain distinguishable when summarizing the evidence.

This distinction is especially important for clinical-trial pages because a table of many P-values can visually imply that every row carries the same evidentiary status. The CIRT registry data do not support treating the eight posted analyses as eight equivalent confirmatory tests.

13. Early Termination and Futility

The registry reports that CIRT was terminated early because the pre-specified boundary for futility was crossed for both the original and the final primary endpoints. It also reports a lack of evidence of a reduction in hsCRP level with methotrexate treatment.

What the futility statement means statistically

Futility ≠ treatment-effect estimate

A futility boundary is a prespecified decision rule concerning whether continued trial follow-up is sufficiently promising under the trial's design assumptions.

Futility monitoring is different from declaring that a treatment has been proven ineffective. The statistical meaning depends on the prespecified monitoring rule, the information available at the interim assessment, and the assumptions used to define the boundary. The ClinicalTrials.gov record identifies the reason for early termination but do not provide the numerical futility boundary, interim information fraction, or interim estimate. Those quantities therefore are not reproduced here.

14. Safety Results

The ClinicalTrials.gov record reports serious adverse events by randomized arm as affected participants divided by participants at risk.

Safety measureMethotrexatePlacebo
Serious adverse events569 / 2391549 / 2395
Serious adverse events: affected participants / at risk
Methotrexate
569 / 2391
Placebo
549 / 2395

The registry supplies these counts as the serious-adverse-event measure. They should not be converted here into an independently calculated percentage because the page's data rules require reported numerical values to be preserved exactly and the ClinicalTrials.gov recordset does not provide an explicit percentage. The at-risk denominators also differ slightly between the two arms.

Safety versus efficacy: serious adverse events and cardiovascular efficacy endpoints answer different questions. A safety comparison should not be merged into a single numerical "net benefit" measure unless the trial explicitly prespecifies such an analysis.

15. Reading the Primary Results Together

The two primary hazard-ratio estimates are 1.01 for Major Adverse Cardiovascular Events and 0.96 for Major Adverse Cardiovascular Event or hospitalization for unstable angina that led to urgent coronary revascularization. The corresponding 95% confidence intervals are 0.82–1.25 and 0.79–1.16, respectively.

Both estimates are close to the null value of 1.00, and both confidence intervals include 1.00. The corresponding P-values are 0.91 and 0.67. These numbers describe the results of the posted superiority analyses; they do not by themselves establish that every possible cardiovascular outcome is unaffected, nor do they quantify the probability that methotrexate has no biological effect.

The registry's early-termination statement provides an important design context. Because the trial crossed a pre-specified futility boundary for both primary endpoints, the observed results should be understood within a trial that stopped early rather than as if it had necessarily continued to its originally intended endpoint of follow-up.

Clinical Biostats interpretation

The central statistical lesson is that point estimates, uncertainty, hypothesis tests, and trial-monitoring decisions are different pieces of evidence. The HR tells us the direction and relative magnitude estimated by the model. The CI describes uncertainty. The P-value quantifies evidence relative to a null hypothesis. The futility decision reflects a prespecified interim-monitoring rule. None should be substituted for another.

16. Limitations

17. Why This Trial Matters Statistically

CIRT is a useful teaching case because its registry record connects randomized treatment assignment with multiple time-to-event outcomes and illustrates how statistical interpretation changes when a trial has several endpoints, stratified survival analyses, multiplicity considerations, and early stopping for futility.

ConceptHow it appears in CIRT
RandomizationThe trial used randomized allocation to methotrexate or placebo.
Parallel designThe registry identifies a parallel-group design model.
Triple maskingThe registry identifies the trial as triple-masked.
Time-to-event endpointsBoth primary endpoints and the posted secondary analyses were analyzed as time-to-event outcomes.
Log-rank testEach posted statistical analysis used a log-rank comparison.
Hazard ratioCox proportional-hazard estimates were used as the effect measure.
Stratified analysisAnalyses were stratified according to event type, event timing, and risk status.
Confidence intervalsPrimary analyses reported two-sided 95% confidence intervals.
MultiplicityThe registry explicitly states that confidence intervals were not adjusted for multiplicity.
Superiority testingThe posted analyses identify superiority as the hypothesis type.
Futility monitoringEarly termination occurred after crossing a pre-specified futility boundary for both primary endpoints.
Safety analysisSerious adverse events are reported by randomized arm as affected participants among those at risk.

18. Statistical Interpretation: What the Hazard Ratios Do — and Do Not — Mean

Hazard ratio of 1.01

The MACE HR of 1.01 indicates an estimated instantaneous event hazard approximately 1% higher in the methotrexate group than in the placebo group under the fitted Cox model. This is a relative model-based estimate; it does not mean that 1% more participants experienced MACE.

Hazard ratio of 0.96

The second primary HR of 0.96 indicates an estimated instantaneous event hazard approximately 4% lower in the methotrexate group than in the placebo group under the fitted model. It does not mean that the cumulative probability of the composite endpoint was exactly 4% lower.

Confidence intervals

The first primary 95% CI, 0.82–1.25, and the second primary 95% CI, 0.79–1.16, quantify uncertainty around the corresponding point estimates. Both include the null value of 1.00, so the intervals encompass both lower and higher hazards.

P-values

The primary P-values, 0.91 and 0.67, are hypothesis-test quantities. They should not be interpreted as effect sizes or as probabilities that the null hypothesis is true. The hazard ratio and confidence interval provide the essential information about estimated magnitude and precision.

19. Analysis Population, Censoring, and Model Interpretation

The primary MACE analysis used all randomized subjects, and the second primary analysis used all randomized participants. This preserves the randomized comparison as the basis for efficacy inference.

For a time-to-event analysis, participants can contribute information even when they do not experience the endpoint during the observed follow-up. Their follow-up may end through administrative study termination or another censoring mechanism. The statistical model uses their available event-free time rather than simply discarding them.

The Cox proportional-hazard model then summarizes the treatment contrast through a hazard ratio. A single HR is convenient and widely interpretable, but it is not a complete description of two survival distributions. If the relative hazard changes substantially over time, the meaning of one summary HR can become less straightforward.

What cannot be reconstructed from the ClinicalTrials.gov record: the registry summary does not provide participant-level event and censoring times, so a new Kaplan-Meier curve, risk table, or independent Cox-model reproduction cannot be generated from the ClinicalTrials.gov record alone.

20. Early Stopping: A Statistical Perspective

Early stopping for futility is fundamentally a decision rule, not an effect-size metric. A trial can be stopped because the accumulating information crosses a boundary indicating that continuing is unlikely to provide the prespecified level of evidence or benefit under the design assumptions.

What is known

The registry states that a pre-specified futility boundary was crossed for both the original and final primary endpoints.

What is not reported

The ClinicalTrials.gov record do not give the numerical futility boundary, interim hazard-ratio estimate, or information fraction at the stopping decision.

Why it matters

Stopping changes the amount and timing of information available for the final analysis and should therefore remain part of the statistical interpretation.

What it does not mean

A futility decision should not automatically be restated as proof of no treatment effect across every possible endpoint or population.

21. Registry Results Summary

Endpoint roleEndpointHR95% CIP-value
Primary Major Adverse Cardiovascular Events 1.01 0.82–1.25 0.91
Primary Major Adverse Cardiovascular Event or Hospitalization for Unstable Angina That Led to Urgent Coronary Revascularization 0.96 0.79–1.16 0.67
Secondary All-cause Mortality 1.16 0.87–1.56 0.32
Secondary Major Adverse Cardiovascular Event or Any Coronary Revascularization 0.95 0.81–1.12 0.57
Secondary Hospitalization for Congestive Heart Failure 0.89 0.60–1.31 0.54
Secondary Major Adverse Cardiovascular Event, Coronary Revascularization, Hospitalization for Congestive Heart Failure or All Cause Mortality 0.98 0.84–1.14 0.80
Other pre-specified Hospitalization for Unstable Angina That Led to Unplanned Coronary Revascularization 0.81 0.53–1.22 0.31
Other pre-specified Coronary Revascularization 0.92 0.75–1.12 0.38

All eight analyses used a log-rank test and a Cox proportional-hazard effect measure. The analyses were stratified according to the event and risk-status factors specified in the registry analysis notes. The confidence intervals were two-sided 95% intervals and were not adjusted for multiplicity.

22. Clinical Interpretation vs Statistical Interpretation

Statistical interpretation

The two primary Cox hazard-ratio estimates were 1.01 and 0.96, with two-sided 95% confidence intervals of 0.82–1.25 and 0.79–1.16. Both intervals include the null value of 1.00, and the corresponding P-values were 0.91 and 0.67.

Clinical interpretation

The ClinicalTrials.gov record does not provide absolute event rates, median event times, or participant-level outcome trajectories for the primary endpoints. Accordingly, this page does not infer an absolute magnitude of clinical benefit or harm from the hazard ratios alone.

The appropriate interpretation is therefore deliberately narrower than a simple statement based on a P-value. The trial provides estimated relative effects for the prespecified time-to-event endpoints, together with uncertainty intervals and a documented early-termination decision. Those pieces should be considered together.

23. Related Tutorials

Learn more about the methods used in this trial:

24. Related Statistical Calculators

25. Sources

Continue through the Clinical Biostats library

Explore tutorials and statistical calculators that explain the survival-analysis methods, confidence intervals, hypothesis testing, and clinical-trial concepts used in CIRT.

26. Record Summary

CIRT is a phase 3 randomized, parallel, triple-masked trial comparing methotrexate with placebo in cardiovascular disease. Its two primary endpoints were time-to-event composites, and both were analyzed using stratified log-rank tests with Cox proportional-hazard estimates. The primary MACE analysis produced an HR of 1.01 (95% CI 0.82–1.25; P = 0.91), while the primary analysis of MACE or hospitalization for unstable angina leading to urgent coronary revascularization produced an HR of 0.96 (95% CI 0.79–1.16; P = 0.67).

The posted secondary and additional pre-specified analyses likewise use survival-analysis methods, with hazard-ratio estimates ranging from 0.81 to 1.16 across the endpoints posted on ClinicalTrials.gov. Their confidence intervals were not adjusted for multiplicity. The registry also states that the trial was terminated early after crossing a pre-specified futility boundary for both the original and final primary endpoints and because of the lack of evidence of a reduction in hsCRP level with methotrexate treatment.

Clinical Biostats methodology: A trial-results page should distinguish the numerical estimate from its uncertainty, the hypothesis test from the effect size, and the statistical result from the design decision. For CIRT, the most important statistical features are the randomized comparison, time-to-event endpoint definitions, stratified log-rank testing, Cox hazard-ratio estimation, multiplicity caveat, safety data, and early futility termination.