This page separates the registered trial design and endpoint from statistical interpretation. The ClinicalTrials.gov record identifies the trial as a randomized, parallel, double-masked phase 3 treatment study with 15603 participants and a primary composite cardiovascular endpoint.
1. Trial at a Glance
CHARISMA was a randomized phase 3 clinical trial studying clopidogrel (SR25990) in the setting of arteriosclerosis. The registry identifies a treatment purpose, parallel assignment, and double masking, with an enrollment of 15603 participants.
| Feature | CHARISMA |
|---|---|
| Trial acronym | CHARISMA |
| Full title | Clopidogrel for High Atherothrombotic Risk and Ischemic Stabilization, Management and Avoidance (CHARISMA) |
| ClinicalTrials.gov identifier | NCT00050817 |
| Phase | Phase 3 |
| Status | COMPLETED |
| Condition | Arteriosclerosis |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | Double |
| Primary purpose | Treatment |
| Enrollment | 15603 |
| Intervention | Clopidogrel (SR25990) |
| Lead sponsor | Sanofi |
| Sponsor type | Industry |
| Start date | 2002-10 |
2. Clinical Question
The registry identifies arteriosclerosis as the condition under study and clopidogrel (SR25990) as the intervention. The primary endpoint is the occurrence of myocardial infarction, stroke, or cardiovascular death.
Population
Participants enrolled in the phase 3 CHARISMA trial with the registry-listed condition of arteriosclerosis.
Intervention
Clopidogrel (SR25990).
Comparator
ClinicalTrials.gov does not report the comparator in the trial information summarized here.
Primary question
Whether treatment with clopidogrel is associated with the occurrence of myocardial infarction, stroke, or cardiovascular death as defined by the registered primary endpoint.
3. Trial Design
A randomized parallel design separates participants into treatment assignments at the start of the comparison and then follows those groups prospectively. Double masking is intended to reduce the influence of knowledge of treatment assignment on trial conduct and assessment.
The registry does not report the allocation ratio, treatment-group sample sizes, dosing schedule, treatment duration, follow-up duration, or comparator in the information summarized here.
4. Registered Primary Endpoint
| Endpoint | Registry definition |
|---|---|
| Primary endpoint | Occurrence of myocardial infarction,stroke or cardiovascular death. |
This is a composite clinical endpoint. Its components are myocardial infarction, stroke, and cardiovascular death. The registered wording combines these events into one primary outcome rather than specifying a single component as the primary endpoint.
Why a composite endpoint changes the statistical question
The primary event is the occurrence of any component of the composite endpoint. Statistically, the first qualifying component event can determine when a participant reaches the composite endpoint.
A composite endpoint can increase the number of observed primary events compared with a single component endpoint, which can provide more information for a time-to-event comparison. Interpretation also requires attention to the individual components because the components can differ substantially in clinical importance and frequency.
5. Statistical Methodology
Time-to-event analysis
The registered endpoint is an occurrence of clinical events, making it naturally suited to a time-to-event framework when the timing of events and censoring are available. Rather than analyzing only whether an event occurred by a fixed calendar date, survival analysis uses the amount of follow-up available for each participant.
For a composite endpoint such as myocardial infarction, stroke, or cardiovascular death, a conventional time-to-first-event analysis would treat the first qualifying component as the event of interest. Participants who have not experienced the composite endpoint by their last observation can contribute censored follow-up.
Kaplan-Meier estimation
A Kaplan-Meier estimator is commonly used to describe the probability of remaining free of the event over time. It can accommodate participants whose follow-up ends before the event occurs.
Here, di represents the number of events at event time ti, while ni represents the number of participants at risk immediately before that time.
Log-rank comparison
When two randomized treatment groups are being compared on a time-to-event endpoint, the log-rank test is a standard method for comparing their event-time distributions. It uses the ordering of observed event times and the numbers at risk rather than reducing the entire follow-up period to a single binary outcome.
Cox proportional-hazards model
A Cox proportional-hazards model is another standard approach for estimating a relative treatment effect for a time-to-event endpoint. Its principal effect measure is the hazard ratio.
A hazard ratio is a relative, model-based measure of event occurrence over time. It is not an absolute risk difference, a probability of benefit, or the percentage of participants who avoid the event.
The registry information summarized here does not specify whether Kaplan-Meier estimation, log-rank testing, a Cox model, or another statistical procedure was used for the primary endpoint. These methods describe the standard statistical framework for an endpoint of this type rather than a reported CHARISMA analysis.
6. Statistical Methods Explained
What makes the primary endpoint a time-to-event endpoint?
The endpoint is defined by the occurrence of myocardial infarction, stroke, or cardiovascular death. If the timing of each participant's first qualifying event is recorded, the analysis can use both whether an event occurred and how long participants remained event-free.
Why might a Kaplan-Meier curve be useful?
A Kaplan-Meier curve can display the estimated proportion of participants who remain free of the composite endpoint as follow-up progresses. It also makes censoring visible through the changing number of participants at risk over time.
What would a hazard ratio describe?
A hazard ratio would compare the estimated instantaneous rates of the composite event between randomized groups under a Cox proportional-hazards model. A value below 1 would correspond to a lower estimated hazard in the numerator treatment group, while a value above 1 would correspond to a higher estimated hazard.
Why is a composite endpoint different from analyzing myocardial infarction alone?
The composite endpoint counts any qualifying component as an event. A participant experiencing myocardial infarction reaches the composite endpoint even if the participant does not subsequently experience stroke or cardiovascular death. An analysis of myocardial infarction alone would answer a narrower question.
Why should the individual components also be examined?
A composite result combines its component outcomes into a single endpoint. The clinical interpretation can depend on how events are distributed among those components. Examining myocardial infarction, stroke, and cardiovascular death separately can therefore provide additional context, although separate analyses introduce their own statistical considerations.
Why does randomization matter for statistical inference?
Randomization creates the foundation for comparing treatment groups without relying on observed baseline characteristics to determine treatment assignment. Under appropriate conduct and analysis, differences between randomized groups can be interpreted in relation to treatment assignment rather than simply as associations with treatment received.
7. Composite Endpoint Interpretation
The primary endpoint combines three clinically distinct outcomes: myocardial infarction, stroke, and cardiovascular death. The statistical event is therefore broader than any single component.
Myocardial infarction
One of the three qualifying components of the registered primary endpoint.
Stroke
One of the three qualifying components of the registered primary endpoint.
Cardiovascular death
One of the three qualifying components of the registered primary endpoint.
Composite outcome
The registered endpoint is the occurrence of myocardial infarction, stroke, or cardiovascular death.
For a composite endpoint, the overall treatment comparison and the component-specific comparisons answer related but different questions. The composite endpoint can be statistically efficient when events are sufficiently frequent and clinically coherent, but its interpretation should remain tied to the exact components specified in the protocol or registry.
8. Analysis Population and Missing Follow-Up
The registry information summarized here identifies the total enrollment as 15603 but does not report the treatment-group sample sizes or define a specific efficacy analysis population.
For randomized clinical trials, an intention-to-treat analysis commonly retains participants according to their randomized assignment. This preserves the comparison created by randomization and avoids changing the treatment groups based on post-randomization events.
For a time-to-event endpoint, incomplete follow-up is commonly handled through censoring. Censoring allows a participant's available event-free follow-up to contribute information without treating the participant as though an event occurred at the end of follow-up.
ClinicalTrials.gov does not report a missing-data or imputation method in the information summarized here. For a time-to-event endpoint, ordinary right-censoring is conceptually different from imputing an unobserved event time.
9. Proportional-Hazards Considerations
If a Cox proportional-hazards model were used, one important assumption would be that the relative hazard between treatment groups is reasonably stable over time. When that assumption is appropriate, a single hazard ratio provides a concise summary of the relative event rate.
A hazard ratio is not equivalent to a ratio of cumulative event probabilities. It summarizes relative event rates under a time-to-event model. If hazards vary substantially in their relative separation over time, a single hazard ratio can become less descriptive of the full survival experience.
Alternative approaches can include reporting time-specific event-free probabilities, restricted mean survival time, or models that explicitly accommodate non-proportional hazards. The registry information summarized here does not specify whether any such alternative methods were used in CHARISMA.
10. Randomization and Masking
| Design feature | Registry description | Statistical significance |
|---|---|---|
| Allocation | Randomized | Supports a prospective comparison based on treatment assignment. |
| Design model | Parallel | Participants remain in their randomized treatment pathways rather than serving as their own controls. |
| Masking | Double | Reduces the potential influence of treatment knowledge on trial conduct and assessment. |
| Primary purpose | Treatment | Indicates that the trial's stated purpose was treatment evaluation. |
Randomization and masking address different sources of potential bias. Randomization concerns treatment assignment, whereas masking concerns knowledge of that assignment after participants enter the trial.
11. Trial Scale and Statistical Information
CHARISMA enrolled 15603 participants, making the registered study substantial in size for a phase 3 randomized trial. Large enrollment can provide more opportunities to observe relatively uncommon clinical events, although statistical precision ultimately depends on the number and timing of qualifying events as well as follow-up.
Enrollment alone does not determine the precision of a treatment-effect estimate. For time-to-event outcomes, the number of observed events is particularly important. The registry information summarized here does not report the number of primary endpoint events, treatment-group event counts, effect estimates, confidence intervals, or p-values.
12. Results Status
The registry identifies the trial as completed and specifies the primary endpoint, but no posted statistical analysis provides a treatment-effect estimate, confidence interval, or p-value for that endpoint.
Consequently, the available registry information supports interpretation of the design and endpoint structure, but it does not provide a numerical estimate of the comparative effect of clopidogrel on the registered composite endpoint.
13. Planned Analysis
The registered primary endpoint is the occurrence of myocardial infarction, stroke, or cardiovascular death. For an endpoint of this type, the typical primary analysis would be based on time to the first occurrence of any qualifying component, with participants who remain free of the composite event contributing follow-up until censoring.
A standard analysis could use Kaplan-Meier methods to estimate event-free survival, a log-rank test to compare randomized groups, and a Cox proportional-hazards model to estimate a hazard ratio with a confidence interval. The exact method used for CHARISMA is not specified in the registry information summarized here.
The statistical objective is to preserve the timing information rather than reduce all follow-up to a single yes/no outcome.
Because the primary endpoint is composite, a complete statistical assessment would ordinarily consider both the overall composite endpoint and the behavior of its individual components. The registry information summarized here does not provide those component-level analyses.
14. What a Hazard Ratio Would Mean in This Trial
If a hazard ratio were reported for the CHARISMA primary endpoint, it would represent the estimated relative rate of experiencing the composite endpoint over time between the specified treatment groups. A hazard ratio below 1 would indicate a lower estimated event hazard for the treatment group used in the numerator of the comparison.
A hazard ratio would not mean that the same percentage of participants experienced a corresponding reduction in their individual probability of myocardial infarction, stroke, or cardiovascular death. It would also not establish that each component of the composite changed by the same relative amount.
A confidence interval would quantify uncertainty around the estimated relative treatment effect. A narrow interval would indicate greater statistical precision than a wide interval, all else being equal. The interval would not describe the range of individual patient outcomes.
If a p-value were reported, it would address evidence against a specified null hypothesis under the chosen statistical model and testing framework. It would not measure the magnitude or clinical importance of the treatment effect. Effect size and uncertainty require the estimate and its confidence interval.
15. What Is Not Reported in the Registry Record
The ClinicalTrials.gov information summarized here does not report several details that would ordinarily be important for a complete numerical statistical reconstruction.
| Statistical detail | Registry information summarized here |
|---|---|
| Treatment-group sample sizes | Not reported |
| Comparator treatment | Not reported |
| Primary endpoint event count | Not reported |
| Treatment-group event counts | Not reported |
| Hazard ratio | Not reported |
| Confidence interval | Not reported |
| p-value | Not reported |
| Follow-up duration | Not reported |
| Secondary endpoint results | Not reported |
| Serious adverse events by arm | Not reported |
| Stratification factors | Not reported |
| Interim analysis plan | Not reported |
| Missing-data or imputation method | Not reported |
| Bayesian methods | Not reported |
These omissions matter because a numerical treatment-effect estimate cannot be interpreted independently of its analysis population, event definition, follow-up, censoring rules, and statistical model.
16. Limitations
- Limited numerical results in the registry record: no treatment-effect estimate, confidence interval, or p-value is posted in the statistical analyses information summarized here.
- Comparator not specified: the registry information summarized here identifies clopidogrel as the intervention but does not identify the comparator.
- Event counts unavailable: the number of participants experiencing myocardial infarction, stroke, cardiovascular death, or the composite endpoint is not reported in the summarized registry information.
- Component interpretation: the primary endpoint combines three clinically distinct outcomes, and their individual contributions cannot be assessed from the registered endpoint alone.
- Analysis method not specified: standard time-to-event methods can be described for this endpoint type, but the exact CHARISMA analysis method is not identified in the registry information summarized here.
- Follow-up not specified: without a reported follow-up period, the temporal context of the endpoint cannot be quantified from this record alone.
- Analysis populations not specified: the registry information summarized here does not identify a specific intention-to-treat, per-protocol, or safety analysis population.
17. Why This Trial Matters Statistically
CHARISMA provides a useful teaching example for understanding how a large randomized phase 3 trial can be structured around a composite cardiovascular time-to-event endpoint.
| Concept | How it appears in CHARISMA |
|---|---|
| Randomization | The trial uses randomized allocation. |
| Parallel design | The registry identifies a parallel design model. |
| Double masking | The registry identifies double masking. |
| Composite endpoint | Myocardial infarction, stroke, or cardiovascular death form the registered primary endpoint. |
| Time-to-event analysis | The occurrence of clinical events is naturally suited to analysis incorporating event timing and censoring. |
| Kaplan-Meier estimation | A standard method for describing event-free survival over time. |
| Log-rank testing | A standard method for comparing time-to-event distributions between randomized groups. |
| Cox regression | A standard model for estimating a hazard ratio for a time-to-event endpoint. |
| Component outcomes | The composite contains myocardial infarction, stroke, and cardiovascular death. |
| Statistical precision | Depends on observed events, follow-up, censoring, and the treatment comparison rather than enrollment alone. |
18. Sources
- ClinicalTrials.gov: CHARISMA — NCT00050817.
Continue through the Clinical Biostats clinical-trial library
Clinical trial pages connect study design, endpoints, and statistical methodology to a broader collection of clinical-trial analyses and educational resources.
19. Record Summary
CHARISMA is a completed randomized phase 3 treatment trial in arteriosclerosis with 15603 enrolled participants. The registry identifies a parallel, double-masked design and clopidogrel (SR25990) as the intervention. Its registered primary endpoint is the occurrence of myocardial infarction, stroke, or cardiovascular death.
From a statistical perspective, the endpoint is naturally framed as a time-to-event composite outcome. Kaplan-Meier estimation, log-rank comparison, and Cox proportional-hazards modeling are standard approaches for this class of endpoint, although the exact CHARISMA statistical methods are not specified in the registry information summarized here.
The ClinicalTrials.gov record identifies the trial as completed but does not post formal statistical analyses in the information summarized here. Accordingly, the central statistical lesson from the registry record is the relationship between randomized trial design, a composite cardiovascular endpoint, and the time-to-event methods ordinarily used to analyze such outcomes.