This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. The registry record reports that results have been posted, but the registry-reported statistical-analyses field contains no formal statistical analyses.
1. Trial at a Glance
GLOBAL LEADERS was a randomized, parallel phase 3 treatment trial in coronary artery disease. The trial enrolled 15991.0 participants and compared two treatment strategies involving ticagrelor, acetylsalicylic acid, and clopidogrel after stent implantation.
| Feature | GLOBAL LEADERS |
|---|---|
| Trial name | GLOBAL LEADERS |
| Phase | Phase 3 |
| Condition | Coronary Artery Disease (CAD) |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | None |
| Primary purpose | Treatment |
| Enrollment | 15991 |
| Arms | 2 |
| Status | Completed |
| Start | 2013-07-01 |
| Primary completion | 2015-11-09 |
| Lead sponsor | ECRI bv |
| Sponsor type | Industry |
2. Clinical Question
The registry describes GLOBAL LEADERS as a clinical study comparing two forms of anti-platelet therapy after stent implantation. The statistical question represented by the registered primary endpoint is whether the two randomized treatment strategies differ with respect to the occurrence of the composite of all-cause mortality or non-fatal new Q-wave myocardial infarction during the specified 2-year period.
Population
Participants with coronary artery disease enrolled in the phase 3 GLOBAL LEADERS study.
Interventions
The trial included ticagrelor, acetylsalicylic acid, and clopidogrel as study interventions.
Comparator
The trial compared an Experimental Treatment Strategy with a Reference Treatment Strategy.
Primary question
How do the two randomized treatment strategies compare for the registered 2-year composite endpoint?
3. Trial Design
Experimental strategy
The ClinicalTrials.gov record identifies this arm as the Experimental Treatment Strategy and report serious adverse events affecting 1459 of 7980 participants.
Reference strategy
The ClinicalTrials.gov record identifies this arm as the Reference Treatment Strategy and report serious adverse events affecting 1543 of 7988 participants.
4. Endpoints
| Endpoint | Registry definition | Time frame | Type |
|---|---|---|---|
| Primary endpoint | Number of Participants With a Composite of All-cause Mortality or Non-fatal New Q-wave Myocardial Infarction (MI) | 2 year | Binary |
The registered definition is: number of participants with a composite of all-cause mortality or non-fatal new Q-wave MI up to 2 years post randomisation. The endpoint is binary because each participant can be classified according to whether the composite event occurred within the specified time frame.
5. Statistical Methodology
Why the primary endpoint is a binary endpoint
The registry classifies the primary endpoint as binary. A binary endpoint reduces the longitudinal clinical experience during the 2-year period to an event indicator: whether the participant experienced the specified composite by the end of the defined assessment period.
This representation is a statistical explanation of the registered endpoint, not a reported analysis from the trial.
Typical comparison for a randomized binary endpoint
For a randomized two-arm trial with a binary primary endpoint, a formal analysis would normally compare the proportion of participants experiencing the endpoint between the randomized groups. Depending on the prespecified statistical analysis plan, this can be expressed using an absolute risk difference, a relative risk, an odds ratio, or an appropriate hypothesis test with a confidence interval.
For GLOBAL LEADERS, no formal statistical analyses were posted. Therefore, none of these effect measures can be attributed to the registry as a reported result from the ClinicalTrials.gov record.
Composite endpoints
The registered primary endpoint combines two clinically distinct outcomes: all-cause mortality and non-fatal new Q-wave myocardial infarction. A participant who experiences either component is counted as having the composite endpoint. The ClinicalTrials.gov record does not provide the component-specific event counts, so the relative contribution of each component cannot be evaluated from this record.
Randomization as the core statistical design feature
Randomization creates the principal framework for comparing the two treatment strategies. Because treatment assignment is randomized, a valid prespecified comparison can attribute differences in outcome to the treatment strategies to the extent that randomization, follow-up, endpoint ascertainment, and the statistical analysis are properly implemented.
What the registry data do not establish
The ClinicalTrials.gov record does not establish the numerical magnitude of any treatment effect for the primary endpoint. Results are posted, but no statistical analyses are reported in the ClinicalTrials.gov record. That distinction is important: the existence of a posted outcome measure is not equivalent to the availability of a formal between-group statistical estimate.
6. Results
The ClinicalTrials.gov record states that results have been posted and that the primary endpoint has results posted for it. However, the statistical analyses posted on ClinicalTrials.gov for the trial contain no analyses. Consequently, the available record supports reporting the endpoint and its design classification, but it does not support reporting a treatment-effect estimate, confidence interval, or p-value.
What can be concluded from the ClinicalTrials.gov record
| Result element | Available in the ClinicalTrials.gov record |
|---|---|
| Results posted | Yes |
| Primary endpoint results posted | Yes |
| Formal statistical analyses | None reported |
| Primary endpoint estimate | Not reported |
| Confidence interval | Not reported |
| P-value | Not reported |
| Component-specific event counts | Not reported |
The appropriate statistical conclusion from the ClinicalTrials.gov record is limited. The trial was randomized and had a prespecified binary composite primary endpoint assessed over 2 years, but the ClinicalTrials.gov record does not provide the numerical comparison needed to quantify a difference between the treatment strategies.
A confidence interval would ordinarily describe the precision of an estimated treatment effect; no such interval is reported here. Similarly, a p-value would address evidence against a prespecified null hypothesis under the chosen test, but it would not measure the size or clinical importance of an effect. Because neither an effect estimate nor a p-value is reported, neither can be interpreted without introducing information outside the specified trial data.
7. Planned Analysis
Because the ClinicalTrials.gov record contains a posted primary endpoint but no formal statistical analyses, the appropriate educational approach is to describe how this type of endpoint would ordinarily be analyzed without presenting an unreported result as if it came from GLOBAL LEADERS.
Step 1: Define the analysis population
A prespecified analysis would first establish which randomized participants are included in the primary efficacy analysis and how participants with incomplete follow-up are handled. The ClinicalTrials.gov record does not specify the analysis population or missing-data rules.
Step 2: Summarize the binary endpoint
The number and proportion of participants experiencing the composite endpoint would normally be reported separately for the Experimental Treatment Strategy and Reference Treatment Strategy.
Step 3: Estimate the between-group effect
Common measures for a binary endpoint include an absolute risk difference, relative risk, and odds ratio. The choice should follow the prespecified statistical analysis plan rather than being selected after viewing the observed data.
Step 4: Quantify uncertainty
A confidence interval would ordinarily accompany the effect estimate. Its interpretation depends on the effect measure: for example, a confidence interval for a risk difference is expressed on an absolute probability scale, while an interval for a relative risk is expressed on a ratio scale.
Step 5: Test the prespecified hypothesis
If the protocol specifies a null hypothesis and a type I error level, an appropriate statistical test can be used to evaluate the evidence against that null hypothesis. The resulting p-value should be interpreted together with the effect estimate and its confidence interval rather than as a standalone measure of treatment benefit.
8. Statistical Methods Explained
Why was randomization important?
Randomization provides the framework for comparing treatment strategies while reducing systematic differences in baseline characteristics that could otherwise confound the treatment comparison. The registry-reported trial identifies allocation as randomized and the design model as parallel.
Why use a parallel-group design?
In a parallel-group trial, participants remain associated with their assigned treatment strategy rather than serving sequentially as their own controls. This is particularly useful when treatment effects or clinical outcomes may persist and when the endpoint is defined over a fixed post-randomization period.
Why is the primary endpoint called binary?
The registry classifies the primary endpoint as binary because the analysis can classify each participant according to whether the specified composite occurred by 2 years. The underlying clinical events are time-dependent in reality, but the registered endpoint is expressed as the number of participants experiencing the composite within the defined period.
Why combine mortality and myocardial infarction into one endpoint?
A composite endpoint can capture more than one clinically important outcome while producing a single prespecified primary endpoint. Interpretation requires knowing which components contribute events, because two treatment strategies could have similar composite event rates while differing in the individual components. The ClinicalTrials.gov record does not provide component-specific counts.
Why are event counts needed before calculating an effect measure?
For a binary endpoint, an effect estimate requires the number of events and the number at risk in each treatment group. Without those quantities, a risk difference, relative risk, odds ratio, or corresponding confidence interval cannot be calculated from the ClinicalTrials.gov record.
Why is a p-value not enough?
A p-value addresses the statistical evidence against a specified null hypothesis under a particular analysis. It does not describe the magnitude of an effect or its clinical importance. A complete interpretation normally considers the estimated effect, confidence interval, event rates, and prespecified hypothesis together.
9. Safety Results
The ClinicalTrials.gov record provides serious adverse events by treatment strategy. These figures are reported as affected participants divided by participants at risk.
| Safety measure | Affected | At risk |
|---|---|---|
| Experimental Treatment Strategy | 1459 | 7980 |
| Reference Treatment Strategy | 1543 | 7988 |
The denominators differ slightly between the two reported safety groups. Because the ClinicalTrials.gov record does not provide a formal statistical analysis of serious adverse events, these counts should be treated as descriptive safety information rather than as evidence of a statistically significant difference.
The affected/at-risk figures allow the reader to understand the number of participants with serious adverse events and the corresponding safety population size. They do not, by themselves, establish causality, statistical significance, or a comparative safety effect.
A formal safety comparison would require a prespecified definition of the safety population and an appropriate comparison of event proportions or event rates. The ClinicalTrials.gov record does not report such an analysis.
10. Understanding the Primary Composite Endpoint
All-cause mortality
One component of the registered composite. The ClinicalTrials.gov record does not provide the number of deaths separately from myocardial infarctions.
Non-fatal new Q-wave MI
The second component of the registered composite. Component-specific event counts are not reported.
Composite event
A participant meeting either component contributes to the binary primary endpoint.
2-year assessment
The registry defines the endpoint as occurring up to 2 years post randomisation.
The composite structure creates an important interpretive distinction. A result for the composite does not automatically tell the reader whether its components behaved similarly. For a complete statistical interpretation, the component event counts and their prespecified analyses would be useful, but they are not included in the ClinicalTrials.gov record.
11. Missing Information That Matters Statistically
The absence of statistical analyses in the ClinicalTrials.gov record limits the level of quantitative interpretation that can responsibly be made. Several common elements of a full clinical-trial statistical report are not provided.
| Statistical element | Status in the ClinicalTrials.gov record | Why it matters |
|---|---|---|
| Primary event counts by arm | Not reported | Needed to quantify the binary treatment effect. |
| Effect estimate | Not reported | Defines the magnitude and direction of the between-group comparison. |
| Confidence interval | Not reported | Describes statistical precision around the effect estimate. |
| P-value | Not reported | Provides evidence against a specified null hypothesis under the selected test. |
| Analysis population | Not reported | Defines which randomized participants contribute to the primary analysis. |
| Missing-data rules | Not reported | Determines how incomplete follow-up affects the analysis. |
| Multiplicity strategy | Not reported | Determines how multiple hypotheses or endpoints affect type I error. |
| Interim analysis procedure | Not reported | Determines whether repeated looks at accumulating data require error control. |
| Stratification variables | Not reported | Could affect the prespecified treatment-effect model or test. |
| Bayesian methods | Not reported | No Bayesian analysis can be attributed to the ClinicalTrials.gov record. |
12. Non-Inferiority, Superiority, and the Direction of Testing
The ClinicalTrials.gov record does not state whether the primary statistical hypothesis was formulated as superiority, non-inferiority, equivalence, or another testing framework. Accordingly, no claim about the trial's formal success criterion is made here.
This distinction is important because the same binary endpoint can be analyzed under fundamentally different inferential frameworks. A non-inferiority analysis requires a prespecified margin and an interpretation based on whether the confidence interval remains within that margin. A superiority analysis instead evaluates evidence that one treatment differs from the other in the prespecified direction. Without the protocol-defined hypothesis and margin, it would be inappropriate to reconstruct that logic from the endpoint alone.
13. Crossover, Factorial Design, and Bayesian Methods
Crossover
The ClinicalTrials.gov record does not report a crossover design or a crossover analysis.
Factorial design
The registered design model is parallel, and the ClinicalTrials.gov record does not identify a factorial design.
Bayesian methods
No Bayesian statistical method is reported in the ClinicalTrials.gov record.
Interim analysis
The ClinicalTrials.gov record does not report an interim-analysis procedure or alpha-spending strategy.
14. Statistical Interpretation of the Available Safety Data
The Experimental Treatment Strategy has 1459 serious adverse events among 7980 participants at risk, while the Reference Treatment Strategy has 1543 among 7988 participants at risk.
These are descriptive counts. The ClinicalTrials.gov record does not provide a formal hypothesis test or confidence interval for the between-group safety comparison, so no statistical conclusion about a difference in serious-adverse-event risk should be drawn from these figures alone.
Comparing raw event counts without considering the number of participants at risk can be misleading. The ClinicalTrials.gov record appropriately pair each affected count with its corresponding at-risk count, allowing the reader to understand the scale of the safety populations.
The primary efficacy endpoint and serious adverse events address different questions. The efficacy endpoint concerns a specified composite clinical outcome over 2 years, whereas the safety figures summarize serious adverse events. Neither should be treated as a substitute for the other.
15. Limitations
- No formal primary analysis reported: no formal statistical analyses were posted even though results are posted.
- No primary event counts by arm: the ClinicalTrials.gov record does not provide the number of participants experiencing the composite endpoint in either treatment strategy.
- No treatment-effect estimate: a risk difference, relative risk, odds ratio, or other primary-effect measure cannot be reported without adding information outside the ClinicalTrials.gov record.
- No confidence interval or p-value: the ClinicalTrials.gov record does not provide the uncertainty or hypothesis-testing quantities needed for formal inferential interpretation.
- Composite endpoint limitation: component-specific counts for all-cause mortality and non-fatal new Q-wave MI are not reported.
- Analysis population not specified: the ClinicalTrials.gov record does not identify the formal primary efficacy analysis population or missing-data rules.
- Multiplicity not specified: no multiplicity procedure is included in the ClinicalTrials.gov record.
- Interim analysis not specified: no interim-monitoring or alpha-spending method is reported in the ClinicalTrials.gov record.
- Limited safety analysis: serious adverse-event counts are available, but no formal comparative safety analysis is reported.
- External results excluded by design: this page intentionally does not import numerical results from publications or other sources because the task restricts trial facts and numerical results to the ClinicalTrials.gov record.
16. Why This Trial Matters Statistically
GLOBAL LEADERS provides a useful teaching example for understanding how a large randomized clinical trial can be translated into a statistical analysis framework even when a registry-reported registry extract does not contain the formal comparative estimates. Its design links randomization, a parallel treatment comparison, a binary composite endpoint, a fixed follow-up period, and descriptive safety outcomes.
| Concept | How it appears in GLOBAL LEADERS |
|---|---|
| Randomization | The trial is registered as randomized. |
| Parallel-group design | The registered design model is parallel with 2 arms. |
| Binary endpoint | The registered primary endpoint is classified as binary. |
| Composite endpoint | The primary endpoint combines all-cause mortality and non-fatal new Q-wave MI. |
| Fixed time frame | The primary endpoint is defined up to 2 years post randomisation. |
| Effect estimation | A formal binary treatment-effect estimate would normally quantify the difference between strategies. |
| Confidence intervals | Would normally quantify uncertainty around a reported treatment-effect estimate. |
| Hypothesis testing | A prespecified statistical test would normally accompany the primary comparison. |
| Safety analysis | Serious adverse events are reported by treatment strategy with affected and at-risk counts. |
| Evidence reporting | The distinction between posted outcomes and posted statistical analyses is important for interpretation. |
17. Randomization as the Central Statistical Concept
Randomization is the explicit learning-pathway concept associated with this trial in the ClinicalTrials.gov record. Its importance extends beyond the act of assigning participants to groups. The validity of the eventual treatment comparison depends on maintaining the randomized comparison in the primary analysis and prespecifying how endpoint ascertainment, deviations, missing observations, and statistical testing are handled.
The arrow represents the statistical rationale for randomized trials. It does not imply that two particular randomized groups must be identical on every baseline characteristic.
For a binary endpoint, randomization makes a direct comparison of event proportions particularly interpretable. The key requirement is that the comparison remains anchored to the prespecified randomized groups and that the statistical analysis follows the protocol-defined framework.
18. Related Tutorials
Learn more about the methods used in this trial:
19. Related Statistical Calculators
20. Sources
- ClinicalTrials.gov: GLOBAL LEADERS, NCT01813435.
- PubMed: PMID 39875171.
- PubMed: PMID 35260381.
- PubMed: PMID 35229616.
- PubMed: PMID 35041147.
- PubMed: PMID 34533034.
Continue through the Clinical Biostats statistical pathway
Explore the broader methods used to understand randomized treatment comparisons, binary clinical endpoints, and statistical uncertainty.
21. Record Summary
GLOBAL LEADERS is a completed randomized phase 3 parallel-group trial in coronary artery disease with 15991.0 participants, 2 treatment arms, and a registered binary primary endpoint defined as the number of participants with a composite of all-cause mortality or non-fatal new Q-wave myocardial infarction up to 2 years post randomisation. The ClinicalTrials.gov record also report serious adverse events by treatment strategy. the posted statistical analyses, however, contains no formal analyses, so this record cannot support a numerical estimate of the primary treatment effect, a confidence interval, or a p-value without introducing information outside the specified trial data.