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
EUROMAX was a randomized, parallel-group, open-label phase 3 trial in acute coronary syndrome. The trial enrolled 2,198 participants and compared bivalirudin with standard-of-care heparins with optional GPI. The registered primary endpoint was binary and assessed within 30 days.
| Feature | EUROMAX |
|---|---|
| Trial name | EUROMAX |
| Brief title | European Ambulance Acute Coronary Syndrome (ACS) Angiography Trial |
| Phase | Phase 3 |
| Condition | Acute Coronary Syndrome |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | None |
| Primary purpose | Treatment |
| Enrollment | 2,198 |
| Interventions | Bivalirudin; Heparin |
| Primary endpoint type | Binary |
| Results posted | Yes |
| Statistical analyses posted | 1 |
| Primary analysis | Chi-squared test with relative risk |
| Hypothesis type | Superiority |
| ClinicalTrials.gov | NCT01087723 |
| Lead sponsor | The Medicines Company |
| Sponsor type | Industry |
2. Clinical Question
The registered trial question can be framed around whether bivalirudin differed from standard-of-care heparins with optional GPI for the incidence of the prespecified composite of death and non-CABG major bleeding within 30 days.
Population
Participants in a phase 3 randomized trial for acute coronary syndrome.
Intervention
Bivalirudin.
Comparator
Standard of Care: Heparins With Optional GPI.
Primary question
Is the 30-day incidence of the registered composite endpoint different between the randomized treatment groups under a superiority framework?
3. Trial Design
Bivalirudin
- Randomized intervention group.
- Primary endpoint assessed within 30 days.
Standard of Care: Heparins With Optional GPI
- Randomized comparator group.
- Primary endpoint assessed within 30 days.
Trial timeline
Trial start
The registry lists March 2010 as the study start.
Primary completion
The registry lists August 2013 as the primary completion date.
4. Endpoints
The registry identifies one primary endpoint. It is a binary endpoint with a 30-day time frame.
| Endpoint | Time frame | Type | Registered definition |
|---|---|---|---|
| The Composite Incidence of Death and Non-coronary Artery Bypass Graft (CABG) Major Bleeding | Within 30 days | Binary | A participant was defined to have had a composite event if the participant experienced at least 1 of the 2 components (death or non-CABG major bleeding) of the composite. Incidence = the number of participants to experience the event / total number of at risk participants × 100. Death was defined as death from any cause at any time. Non-CABG major bleeding was defined as any 1 of the following: intra-cranial, retroperitoneal, intraocular, access site hemorrhage requiring radiological or surgical intervention, reduction in hemoglobin (Hb) concentration of >4 grams/deciliter (g/dL) without an overt source of bleeding, reduction in hemoglobin concentration of >3 g/dL with an overt source of bleeding; re-intervention for bleeding, or use of any blood product transfusion. |
5. Analysis Population and Statistical Framework
The posted primary statistical analysis specifies an intention-to-treat population: participants who were randomized and signed an informed consent form (ICF). The analysis compared bivalirudin with Standard of Care: Heparins With Optional GPI.
| Element | Registered / posted specification |
|---|---|
| Analysis population | Participants who were randomized and signed an ICF; ITT population |
| Endpoint type | Binary |
| Statistical method | Chi-squared |
| Effect measure | Relative risk |
| Hypothesis type | Superiority |
| Confidence interval | 95%, two-sided |
6. Results
The registry posts one formal statistical analysis, corresponding to the primary endpoint. The analysis uses a chi-squared test and reports a relative risk of 0.60 with a two-sided 95% confidence interval of 0.43 to 0.82 and a p-value of 0.0014.
Primary Composite Endpoint
Relative risk for the composite endpoint
95% CI: 0.43–0.82 · P = 0.0014
Chi-squared test · ITT population · Superiority analysis
| Endpoint | Analysis population | Method | Effect measure | Estimate | 95% CI | P-value |
|---|---|---|---|---|---|---|
| Composite incidence of death and non-CABG major bleeding within 30 days | Randomized and ICF-signed participants; ITT | Chi-squared | Relative risk | 0.60 | 0.43–0.82 | 0.0014 |
The reported relative risk of 0.60 means that the estimated probability of experiencing the registered composite endpoint in the bivalirudin group was 0.60 times the corresponding probability in the Standard of Care: Heparins With Optional GPI group, under the reported analysis. Expressed as a simple relative comparison, this corresponds to an estimated 40% lower risk of the composite endpoint.
That statement is about the composite endpoint, not about death alone or major bleeding alone. Because a participant could meet the composite through either component, the relative risk does not tell us from the reported analysis how much of the observed difference was attributable to death versus non-CABG major bleeding.
The 95% confidence interval of 0.43–0.82 describes uncertainty around the estimated relative risk under the statistical analysis. It does not mean that 95% of individual participants experienced a risk between 0.43 and 0.82 times the comparator risk.
The p-value of 0.0014 addresses the statistical evidence against the null comparison specified by the test. It does not measure the magnitude, clinical importance, or certainty of the treatment effect. Effect size and precision are better communicated by the relative risk and its confidence interval.
The analysis is based on an ITT population and a binary 30-day outcome. The ClinicalTrials.gov record does not report the arm-specific primary-event counts or percentages, so the relative risk should not be converted into an absolute risk difference from the information on this page.
7. How to Read the Primary Result
Relative rather than absolute effect
A relative risk of 0.60 describes the ratio of the two event probabilities. It does not by itself show the absolute difference in event probability between the groups.
Confidence interval
The interval from 0.43 to 0.82 shows the precision of the estimated relative risk. Its interpretation is tied to the statistical model and sampling framework.
Statistical significance
The p-value of 0.0014 provides evidence against the null comparison under the reported chi-squared analysis. It is not an effect-size measure.
Composite outcome
The endpoint combines death and non-CABG major bleeding. A significant composite result does not establish that both components changed to the same degree.
8. Statistical Methodology
Chi-squared test for a binary endpoint
The posted primary analysis uses a chi-squared test. This is a categorical-data method for comparing observed outcome distributions between treatment groups. For a binary endpoint, the underlying data can be represented as a two-by-two table: treatment group by whether the participant experienced the composite event.
The event-status categories are whether the participant experienced the registered composite endpoint within 30 days. The treatment groups are bivalirudin and Standard of Care: Heparins With Optional GPI.
Risk ratio
The effect measure reported for the primary analysis is relative risk, also called a risk ratio. It compares the probability of the binary outcome in the intervention group with the probability in the comparator group.
An RR of 1 represents equal observed risk. An RR below 1 indicates lower observed risk in the numerator group, while an RR above 1 indicates higher observed risk in the numerator group.
Intention-to-treat analysis
The posted analysis uses participants who were randomized and signed an ICF, identified as the ITT population. An ITT framework maintains the treatment assignment established by randomization rather than redefining treatment groups according to later exposure or outcome.
Two-sided confidence interval
The reported confidence interval is 95% and two-sided. For the relative risk, the interval is 0.43–0.82. Because the null value for a risk ratio is 1, an interval entirely below 1 is consistent with a treatment-group difference in the direction of lower relative risk for the bivalirudin group.
Superiority hypothesis
The posted analysis is classified as a superiority analysis. This is different from a non-inferiority framework: the objective is not to establish that bivalirudin is no worse than a comparator within a prespecified margin. The reported result is evaluated as a superiority comparison using the stated statistical method and effect measure.
9. Statistical Methods Explained
What does a risk ratio of 0.60 mean?
A risk ratio of 0.60 means that the estimated probability of the primary composite endpoint in the bivalirudin group was 60% of the probability in the Standard of Care: Heparins With Optional GPI group. Equivalently, the ratio corresponds to an estimated 40% relative reduction in risk. It does not provide the absolute number of events prevented.
Why was a chi-squared test used?
The registered primary endpoint is binary: a participant either experienced the composite event within 30 days or did not. A chi-squared test is designed for comparing categorical outcome distributions between groups. The registry specifically identifies chi-squared as the posted method for this primary analysis.
Why does the confidence interval matter?
The point estimate of 0.60 is only one estimate of the treatment-group risk ratio. The 95% confidence interval of 0.43–0.82 communicates the statistical uncertainty around that estimate. A narrower interval would generally indicate greater precision than a wider interval, all else being equal.
Why is the p-value not the effect size?
The p-value of 0.0014 describes the statistical evidence against the null comparison under the reported testing framework. It does not say that the treatment effect is 0.14%, nor does it quantify the clinical magnitude of the result. The risk ratio describes relative effect magnitude, while the confidence interval describes uncertainty around that estimate.
Why does the ITT population matter?
Randomization is intended to establish comparable treatment groups at the beginning of the trial. An ITT analysis preserves that randomized comparison by analyzing participants according to the randomized framework specified for the primary analysis. This is especially important when interpreting a superiority trial because excluding participants after randomization can alter the comparability created by randomization.
What does the composite endpoint hide?
A composite endpoint combines multiple clinical outcomes into one binary outcome. This can increase the number of participants experiencing the endpoint, but the overall composite effect can be influenced differently by its components. Here, the registry defines the composite as death or non-CABG major bleeding; the reported statistical analysis does not provide separate effect estimates for those two components.
10. Primary Endpoint: Statistical Interpretation in Context
The relative risk of 0.60 is a ratio of event probabilities, not a hazard ratio. Therefore, it should be interpreted as a comparison of the probability of the binary 30-day composite outcome rather than as an instantaneous event-rate comparison over time.
The 95% CI of 0.43–0.82 provides the reported range of uncertainty around the relative-risk estimate. The interval remains below the null value of 1, which is consistent with the direction of the reported superiority result.
The two-sided P = 0.0014 is the formal statistical result reported for the chi-squared comparison. It should be read alongside the effect estimate and confidence interval rather than used as a substitute for either one.
The result applies to the registered composite of death and non-CABG major bleeding within 30 days. It should not be restated as though it were a result for mortality alone or bleeding alone.
11. Safety
The trial data provide serious adverse-event counts by randomized treatment arm. These data are reported as affected participants divided by participants at risk.
| Arm | Serious adverse events | Affected / at risk |
|---|---|---|
| Bivalirudin | Serious adverse events | 145/1099 |
| Standard of Care: Heparins With Optional GPI | Serious adverse events | 129/1094 |
12. Statistical Relationship Between the Primary Endpoint and Safety
The primary endpoint itself includes non-CABG major bleeding as one of its two components, while the safety information in the ClinicalTrials.gov record reports serious adverse events. These are not interchangeable statistical concepts.
| Measure | Role | Interpretation |
|---|---|---|
| Death + non-CABG major bleeding | Registered primary endpoint | Binary composite evaluated within 30 days and formally analyzed using chi-squared with relative risk. |
| Serious adverse events | Safety information | Reported as affected participants / participants at risk by arm; no formal comparative analysis is reported in the ClinicalTrials.gov record. |
This distinction prevents an important statistical error: a safety category should not be treated as though it were automatically identical to a component of the primary efficacy analysis. Endpoint definitions determine which participants are counted and therefore which statistical question is being answered.
13. What This Analysis Does Not Establish
- It does not isolate the components: the relative risk of 0.60 applies to the composite endpoint, not separately to death and non-CABG major bleeding.
- It does not provide an absolute risk difference: the registry-reported statistical analysis reports the relative risk and its confidence interval, but does not provide arm-specific primary-event percentages or counts.
- It does not imply individual-level benefit: a group-level relative risk does not mean every participant experienced the same proportional change in risk.
- It does not equate statistical significance with clinical importance: the p-value describes evidence under the test framework, while clinical interpretation also depends on the magnitude and nature of the outcome.
- It does not make a claim about unreported endpoints: the ClinicalTrials.gov record contains nine posted outcome measures but provide a formal statistical analysis for only the primary endpoint summarized here.
- It does not establish a mechanism: the statistical comparison describes the observed treatment-group difference; it does not by itself explain why the difference occurred.
14. Limitations
Limited formal results
The ClinicalTrials.gov record contains one posted statistical analysis. Consequently, this page cannot construct a complete comparative results table for all nine posted outcome measures.
No component-specific estimates
The formal analysis posted on ClinicalTrials.gov for the primary endpoint does not provide separate relative-risk estimates for death and non-CABG major bleeding.
Absolute effect unavailable
The registry-reported statistical analysis gives a relative risk, confidence interval, and p-value but not the arm-specific primary-event counts or percentages needed to report an absolute risk difference.
Safety analysis scope
Serious adverse-event counts are reported by arm, but no comparative statistical test or confidence interval is provided for that safety measure.
The trial is also described as open-label. The primary endpoint is nevertheless a binary clinical outcome defined in the registry, and the registry-reported analysis uses the randomized ITT population. The consequences of lack of masking cannot be quantified from the registry-reported statistical analysis alone.
15. Why This Trial Matters Statistically
EUROMAX is a useful teaching example because its primary analysis is relatively direct: randomization produces two comparison groups, the primary outcome is binary, and the registry reports both a categorical-data test and a clinically interpretable relative-risk measure.
| Concept | How it appears in EUROMAX |
|---|---|
| Randomization | The trial uses randomized allocation. |
| Parallel design | The two intervention groups are evaluated in a parallel-group design. |
| Binary endpoint | The primary endpoint records whether a participant experienced the composite within 30 days. |
| Composite endpoint | Death and non-CABG major bleeding are combined into one primary outcome. |
| Intention-to-treat | The primary analysis population consists of participants who were randomized and signed an ICF. |
| Chi-squared test | The registry reports chi-squared as the primary statistical method. |
| Risk ratio | The treatment effect is reported as relative risk. |
| Confidence interval | A two-sided 95% CI accompanies the relative-risk estimate. |
| Superiority testing | The posted analysis is classified under a superiority hypothesis. |
| Safety analysis | Serious adverse events are reported separately by randomized arm. |
16. A Worked Statistical Reading of the EUROMAX Result
The primary outcome is whether a participant experienced at least one component of the registered composite endpoint within 30 days.
The randomized comparison is bivalirudin versus Standard of Care: Heparins With Optional GPI.
The posted analysis uses participants who were randomized and signed an ICF, described as the ITT population.
The registry reports a chi-squared test for the binary endpoint.
The estimated risk of the composite endpoint in the bivalirudin group was 0.60 times the risk in the Standard of Care group.
The confidence interval describes precision around the relative-risk estimate, while the p-value describes the statistical evidence under the reported superiority test.
17. Primary Result: What the Numbers Say Together
The three central statistical quantities should be read together rather than separately:
| Quantity | Reported value | What it contributes |
|---|---|---|
| Relative risk | 0.60 | Magnitude and direction of the relative difference in the binary composite endpoint. |
| 95% CI | 0.43–0.82 | Uncertainty and precision around the estimated relative risk. |
| P-value | 0.0014 | Statistical evidence under the reported chi-squared superiority analysis. |
Considering the three together gives a more complete statistical description than the p-value alone. The effect estimate is below the null value of 1, the confidence interval remains below 1, and the reported p-value is 0.0014. None of these quantities, by themselves or together, supplies the absolute clinical risk without the underlying arm-specific event probabilities.
18. Important Interpretation Issues
Relative risk is not a hazard ratio
The EUROMAX primary analysis reports a relative risk, not a hazard ratio. Relative risk compares probabilities over the specified binary endpoint window. A hazard ratio instead compares instantaneous event rates over time in a time-to-event model. The two measures should not be used interchangeably.
The 30-day window matters
The registered primary endpoint is explicitly defined as occurring within 30 days. The reported relative risk therefore addresses the probability of the composite during that specified time frame.
The composite matters
Death and non-CABG major bleeding are clinically distinct events. A single composite estimate is useful for the prespecified primary statistical question, but it cannot reveal the separate treatment effect on each component when those component-specific analyses are not reported.
The ITT population matters
Because the analysis is specified in the ITT population, the primary comparison is tied to randomized assignment. This is a central feature of the causal interpretation of randomized superiority trials.
19. Record Summary
EUROMAX was a completed randomized phase 3 parallel trial in acute coronary syndrome with 2,198 participants and two treatment arms. Its registered primary endpoint was the composite incidence of death and non-CABG major bleeding within 30 days, defined as a binary participant-level outcome. The registry-reported formal analysis used the ITT population, a chi-squared test, and relative risk under a superiority framework.
Primary statistical result
95% CI: 0.43–0.82 · P = 0.0014
Primary composite endpoint: death and non-CABG major bleeding within 30 days
The most important statistical lesson is that the result should be interpreted at the level at which it was analyzed: a randomized comparison of a binary composite endpoint. The relative risk communicates the direction and magnitude of the group-level difference, the confidence interval communicates its precision, and the p-value communicates statistical evidence under the reported test. Serious adverse events are separately reported by arm and should not be conflated with the primary composite endpoint.
20. Related Tutorials
Learn more about the methods used in this trial:
21. Related Statistical Calculators
22. Sources
- ClinicalTrials.gov: NCT01087723 — EUROMAX.
- PubMed: PMID 29225903.
- PubMed: PMID 28273285.
- PubMed: PMID 27287251.
- PubMed: PMID 27165710.
- PubMed: PMID 26995053.
Continue through Clinical Biostats
Explore statistical tutorials, calculators, and additional clinical trial analyses to deepen your understanding of trial design and interpretation.