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
CURRENT/OASIS7 was a randomized, quadruple-masked, phase 3 factorial trial evaluating two clopidogrel treatment regimens and two ASA dose levels in patients with acute coronary disease or unstable angina. The registry reports an enrollment of 25086.0 participants across four treatment combinations.
| Feature | CURRENT/OASIS7 |
|---|---|
| Phase | Phase 3 |
| Population | Patients with acute coronary disease or unstable angina |
| Design | Randomized, factorial, quadruple-masked |
| Primary purpose | Treatment |
| Enrollment | 25086 |
| Interventions | Clopidogrel and acetylsalicyclic acid (ASA) |
| Registered primary endpoints | 4 |
| ClinicalTrials.gov | NCT00335452 |
| Lead sponsor | Sanofi |
2. Clinical Question
The trial addressed two treatment questions within a factorial framework: whether the clopidogrel treatment regimen influenced the first occurrence of cardiovascular death, myocardial infarction, or stroke during the first 30 days, and whether ASA dose level influenced the same composite endpoint.
Population
Patients enrolled with acute coronary disease or unstable angina.
Intervention
Clopidogrel treatment regimens combined with either low-dose or high-dose ASA.
Comparator
The alternative clopidogrel treatment regimen and the alternative ASA dose level, depending on the factorial comparison.
Primary question
How do clopidogrel treatment regimen and ASA dose level affect the 30-day first occurrence of cardiovascular death, MI, or stroke?
3. Trial Design
The registry identifies the design model as factorial. Statistically, that structure is important because the trial evaluates two treatment dimensions within the same randomized framework: clopidogrel treatment regimen and ASA dose level. The registry also reports a specific interaction analysis between the two treatment factors.
Standard clopidogrel + low-dose ASA
- Clopidogrel 300/75/75 mg
- ASA low dose
Standard clopidogrel + high-dose ASA
- Clopidogrel 300/75/75 mg
- ASA high dose
High-dose clopidogrel + low-dose ASA
- Clopidogrel 600/150/75 mg
- ASA low dose
High-dose clopidogrel + high-dose ASA
- Clopidogrel 600/150/75 mg
- ASA high dose
4. Endpoints
| Registered endpoint | Time frame | Definition / analysis role |
|---|---|---|
| First Occurrence of CV Death / MI / Stroke — Clopidogrel Treatment Regimen Comparison | 30 days | The first occurrence of cardiovascular death, myocardial infarction, or stroke between randomization and Day 30, inclusive. |
| First Occurrence of CV Death / MI / Stroke — ASA Dose Comparison | 30 days | First occurrence of the same composite endpoint comparing ASA dose levels. |
| First Occurrence of CV Death / MI / Stroke — Interaction Clopidogrel Treatment Regimen and ASA Dose Level | 30 days | Assessment of the relationship between the two factorial treatment dimensions. |
| First Occurrence of CV Death / MI / Stroke — Clopidogrel Treatment Regimen Comparison in PCI Subgroup | 30 days | Clopidogrel regimen comparison among randomized patients who underwent PCI during the study. |
Primary composite endpoint definition
The registry defines the primary composite endpoint as the first occurrence of any of three events: cardiovascular death, myocardial infarction, or stroke. Cardiovascular death includes any death with a clear cardiovascular or unknown cause. Myocardial infarction is a new MI, nonfatal or fatal. Stroke is a new focal neurologic deficit thought to be vascular in origin, with signs or symptoms lasting more than 24 hours, nonfatal or fatal.
The registry states that these events were reported between randomization and Day 30, inclusive, and validated by the blinded Event Adjudication Committee.
5. Statistical Methodology
Intention-to-treat analysis
The principal clopidogrel treatment-regimen analyses used the intent-to-treat (ITT) population. The registry defines this population as all randomized patients irrespective of whether they received study medication, underwent PCI, or otherwise complied with the study protocol.
Using the randomized population preserves the treatment assignment created by randomization. This is particularly important when treatment exposure, treatment discontinuation, PCI status, or protocol adherence could differ after randomization.
Stratified log-rank testing
The primary time-to-event comparisons used two-sided stratified log-rank tests. For the overall clopidogrel comparison, stratification was by ASA dose level and qualifying condition, defined as UA/NSTEMI or STEMI. For the ASA comparison, stratification was by clopidogrel treatment regimen. The interaction analyses were stratified by qualifying condition.
The log-rank test evaluates whether the observed event-time experience differs between randomized groups over the specified follow-up period. Stratification allows the comparison to account for prespecified or clinically important factors without treating them as though they were the randomized treatment itself.
Stratified Cox proportional-hazards model
The registry states that relative risk reduction for the clopidogrel and ASA comparisons was estimated using a stratified Cox proportional-hazards model. For the overall clopidogrel comparison, the model controlled for ASA dose level and qualifying condition. For the ASA comparison, it controlled for clopidogrel treatment regimen.
The registry reports the effect measure as a relative risk reduction (%). The reported estimates should therefore be interpreted as percentage reductions relative to the comparison regimen, rather than as absolute percentage-point differences in event probability.
Chi-squared interaction test
The interaction between clopidogrel treatment regimen and ASA dose level was also assessed using an interaction chi-squared test of the Cox proportional-hazards model. This addresses a different question from the main treatment-effect comparisons: whether the association for one treatment factor differs according to the level of the other factor.
Logistic regression for major bleeding
Major bleeding was analyzed as a binary endpoint using logistic regression. The clopidogrel dose-regimen model included terms for ASA dose level and qualifying condition. The ASA dose-level model included a term for clopidogrel treatment regimen.
Time-to-event outcomes
Log-rank testing and Cox modeling use information about when the first event occurred and can accommodate censored observations.
Binary outcomes
Logistic regression treats major bleeding as an event/no-event outcome and can adjust for the factorial treatment structure and qualifying condition specified in the registry.
6. Primary Results: Clopidogrel Treatment Regimen
The primary clopidogrel comparison evaluated the first occurrence of cardiovascular death, MI, or stroke during 30 days. The ITT population was compared using a two-sided stratified log-rank test, with the relative risk reduction estimated from a stratified Cox proportional-hazards model.
30-day CV death / MI / stroke
Relative risk reduction: high-dose treatment regimen (600/150/75 mg) versus standard treatment regimen (300/75/75 mg)
95% CI: −5.8% to 16.6% · P = 0.3037
The point estimate of 6.1% represents the estimated relative risk reduction for the high-dose clopidogrel treatment regimen compared with the standard treatment regimen over the 30-day analysis period. In relative terms, it corresponds to an estimated 6.1% lower risk measure for the high-dose regimen in the fitted analysis.
The estimate does not mean that 6.1% of patients avoided cardiovascular death, MI, or stroke, and it does not provide an absolute risk difference. The 95% confidence interval extends from −5.8% to 16.6%, so the statistical uncertainty includes both a possible relative increase and a possible relative reduction in the endpoint.
The two-sided P = 0.3037 addresses the strength of evidence against the null hypothesis under the specified test; it does not measure the size or clinical importance of the treatment effect. The analysis is also conditional on the Cox-model framework and the stratified log-rank comparison.
7. Primary Results: ASA Dose Comparison
The second major factorial comparison evaluated low-dose versus high-dose ASA among the ASA-treated population. The registry defines this population as patients randomized and receiving at least one dose of ASA, analyzed according to the treatment group to which they were allocated by the AReS.
30-day CV death / MI / stroke
Relative risk reduction: ASA high dose versus ASA low dose
95% CI: −9.2% to 14.0% · P = 0.6047
The point estimate of 3.1% is the estimated relative risk reduction for high-dose ASA compared with low-dose ASA over 30 days. As with the clopidogrel estimate, this is a relative measure and should not be read as a 3.1-percentage-point absolute reduction in event probability.
The 95% CI of −9.2% to 14.0% spans zero on the relative-risk-reduction scale. The interval therefore includes the possibility of a relative increase as well as a relative reduction. Its width also shows that the point estimate should not be interpreted as a precise estimate of the underlying treatment effect.
The P = 0.6047 is a test statistic outcome, not a measure of treatment-effect magnitude. The analysis used a two-sided stratified log-rank test and a stratified Cox model controlling for clopidogrel treatment regimen.
8. Interaction Between Clopidogrel and ASA Dose
A central statistical feature of the factorial design is the explicit assessment of whether the clopidogrel treatment effect differed according to ASA dose level.
| ASA dose level | Comparison | Relative risk reduction | 95% CI | P-value |
|---|---|---|---|---|
| Low dose | 600/150/75 mg vs 300/75/75 mg clopidogrel | −6.5% | −26.0% to 9.9% | 0.4579 |
| High dose | 600/150/75 mg vs 300/75/75 mg clopidogrel | 17.6% | 2.2% to 30.5% | 0.0262 |
Low-dose ASA stratum
Clopidogrel comparison with low-dose ASA
95% CI: −26.0% to 9.9% · P = 0.4579
Within the low-dose ASA stratum, the reported relative risk reduction was −6.5%. Because the confidence interval ranges from −26.0% to 9.9%, the estimate is compatible with a range extending from a relative increase to a relative reduction in the composite endpoint.
This is a stratum-specific comparison. It should not be interpreted by itself as proof that clopidogrel has no effect overall, nor should the P-value be used as a direct test of whether the low-dose and high-dose ASA strata differ from each other. The formal interaction question is addressed separately.
High-dose ASA stratum
Clopidogrel comparison with high-dose ASA
95% CI: 2.2% to 30.5% · P = 0.0262
Within the high-dose ASA stratum, the estimated relative risk reduction was 17.6%, with a 95% CI from 2.2% to 30.5%. On the reported relative-risk-reduction scale, the entire interval is above zero.
The P-value of 0.0262 is evidence against the null hypothesis for this specified stratum-specific comparison under the two-sided stratified log-rank test. It does not quantify the magnitude of the effect and does not, by itself, establish that the high-dose ASA result is statistically different from the low-dose ASA result. That latter question requires the interaction analysis.
Formal interaction analysis
Clopidogrel × ASA interaction
95% CI reported: 95%
Interaction chi-squared test of the Cox proportional-hazards model
The registry reports an interaction chi-squared test of the Cox proportional-hazards model with P = 0.0355. This test is designed to address whether the clopidogrel treatment effect varies with ASA dose level. It is conceptually different from comparing the individual P-values of the low-dose and high-dose strata.
9. Primary Results: PCI Subgroup
The registry also reports a prespecified primary analysis comparing clopidogrel treatment regimens among randomized patients who underwent PCI during the study. This analysis remained an ITT analysis within the PCI subgroup.
30-day CV death / MI / stroke in PCI subgroup
Relative risk reduction: 600/150/75 mg versus 300/75/75 mg clopidogrel
95% CI: 1.2% to 26.3% · P = 0.0332
The PCI-subgroup estimate corresponds to a 14.7% relative risk reduction for the high-dose clopidogrel regimen compared with the standard regimen over 30 days. The 95% CI of 1.2% to 26.3% describes uncertainty around that relative estimate under the specified analysis.
The P = 0.0332 is the result of the two-sided stratified log-rank test. It does not measure the magnitude of the effect. The registry states that the analysis was stratified by ASA dose level and qualifying condition and that no adjustment was made.
This analysis should also be distinguished from the overall randomized comparison. A subgroup defined by undergoing PCI is not identical to the full randomized population, so its result answers a narrower question.
10. Secondary Results: Major Bleeding
Major bleeding was analyzed as a binary endpoint using logistic regression. The clopidogrel comparison used the ITT population; the ASA comparison used the treated patient population defined in the registry.
| Endpoint | Comparison | Method | 95% CI | P-value |
|---|---|---|---|---|
| Occurrence of Major Bleeding | Clopidogrel 300/75/75 mg + ASA vs 600/150/75 mg + ASA | Logistic regression | Two-sided; 95% | 0.012 |
| Occurrence of Major Bleeding | Clopidogrel + ASA low dose vs Clopidogrel + ASA high dose | Logistic regression | Two-sided; 95% | 0.945 |
Clopidogrel dose-regimen comparison
Major bleeding
Logistic regression with terms for ASA dose level and qualifying condition
The registry reports a P-value of 0.012 for the comparison of major bleeding between the two clopidogrel treatment regimens. Because the registry does not provide an effect estimate or confidence interval for this analysis in the posted statistical-analysis record, the P-value alone cannot quantify the size of the difference.
ASA dose comparison
Major bleeding
Logistic regression with a term for clopidogrel treatment regimen
The registry reports a P-value of 0.945 for the ASA dose-level comparison. Again, the P-value does not provide an effect size, and the posted analysis does not provide a corresponding estimate and confidence interval for this endpoint.
11. Safety: Serious Adverse Events by Arm
The registry reports serious adverse events by treatment arm as affected participants divided by participants at risk.
| Treatment arm | Affected / at risk |
|---|---|
| Clopidogrel 300/75/75 mg + ASA Low Dose | 322 / 6312 |
| Clopidogrel 300/75/75 mg + ASA High Dose | 329 / 6254 |
| Clopidogrel 600/150/75 mg + ASA Low Dose | 372 / 6267 |
| Clopidogrel 600/150/75 mg + ASA High Dose | 332 / 6253 |
These arm-level counts describe the serious-adverse-event burden reported by the registry. They are not interchangeable with the formal major-bleeding analyses because major bleeding is a specific binary outcome analyzed with logistic regression, whereas serious adverse events are presented here as affected and at-risk counts by treatment combination.
12. Post-Hoc Result: Stent Thrombosis
The registry identifies the stent-thrombosis analysis as post hoc. It compared clopidogrel treatment regimens using a two-sided stratified log-rank test and estimated relative risk reduction using a stratified Cox proportional-hazards model.
30-day stent thrombosis
Relative risk reduction: 600/150/75 mg versus 300/75/75 mg clopidogrel
95% CI: 16.2% to 45.8% · P = 0.0004
The reported estimate of 32.6% is a relative risk reduction for stent thrombosis associated with the high-dose clopidogrel treatment regimen compared with the standard regimen over 30 days. The 95% CI of 16.2% to 45.8% gives the statistical uncertainty around that relative estimate under the fitted analysis.
The P = 0.0004 indicates strong evidence against the null hypothesis for this specified post-hoc comparison, but the P-value is not itself an estimate of clinical magnitude. Because the registry classifies this analysis as post hoc, its inferential role differs from that of the registered primary endpoints.
13. Statistical Methods Explained
Why was a factorial design used?
A factorial design allows two treatment dimensions to be evaluated within the same randomized experiment. Here, the two dimensions were clopidogrel treatment regimen and ASA dose level. The four treatment combinations create the four cells of the factorial structure.
The design also makes it possible to ask an interaction question: does the effect associated with the clopidogrel regimen differ depending on whether ASA is given at a low or high dose? The registry reports such an interaction analysis explicitly.
Why was a log-rank test used?
The primary composite endpoint was a time-to-event outcome: the first occurrence of cardiovascular death, MI, or stroke during 30 days. A log-rank test compares the event-time experience between groups while accounting for the timing of events and censoring.
The registry used stratified versions of the log-rank test. Stratification permits the treatment comparison to be evaluated within levels of specified factors and then combined into an overall comparison.
What does a 6.1% relative risk reduction mean?
The 6.1% estimate is a relative measure comparing the high-dose and standard clopidogrel treatment regimens. It does not mean that the absolute probability of cardiovascular death, MI, or stroke fell by 6.1 percentage points. An absolute risk difference would require the corresponding event risks for both groups.
Why is the confidence interval more informative than the point estimate alone?
The point estimate is only one estimate of the treatment effect. The 95% confidence interval describes the uncertainty around that estimate under the specified statistical framework. For the overall clopidogrel comparison, the interval extends from −5.8% to 16.6%, illustrating that the observed estimate is compatible with a range of possible relative effects.
Why does the P-value not measure effect size?
A P-value summarizes the evidence against a null hypothesis under a specified statistical model and testing procedure. It does not tell the reader whether the treatment effect is large, small, clinically important, or precisely estimated. Effect estimates and confidence intervals provide that additional information.
Why is the interaction analysis important?
In a factorial trial, the two treatment factors can potentially influence one another. The question is not simply whether one subgroup has a small P-value and another subgroup has a large P-value. The appropriate statistical question is whether the estimated treatment effects differ across levels of the other factor. The registry's interaction chi-squared test addresses that question directly.
14. Understanding the Primary Analysis Populations
| Analysis | Population | Why it matters |
|---|---|---|
| Clopidogrel primary comparison | ITT: all randomized patients | Preserves the randomized treatment comparison irrespective of treatment receipt, PCI, or protocol compliance. |
| ASA primary comparison | ASA-treated population | Includes randomized patients who received at least one dose of ASA and retains their allocated treatment group. |
| PCI subgroup | Randomized patients who underwent PCI | Addresses the clopidogrel comparison within a narrower population defined by PCI during the study. |
| Major bleeding: clopidogrel comparison | ITT | Maintains the randomized comparison for the binary safety endpoint. |
| Major bleeding: ASA comparison | Treated patient population | Includes randomized patients receiving at least one dose of ASA. |
The population definition is part of the estimand. Two analyses using the same treatment arms can answer different questions when the analysis populations differ. The PCI subgroup, for example, is not simply another name for the overall ITT population.
15. Reading the Relative Risk Reduction Estimates
| Analysis | Estimate | 95% CI | P-value |
|---|---|---|---|
| Overall clopidogrel regimen | 6.1% | −5.8% to 16.6% | 0.3037 |
| ASA dose | 3.1% | −9.2% to 14.0% | 0.6047 |
| Clopidogrel regimen + low-dose ASA | −6.5% | −26.0% to 9.9% | 0.4579 |
| Clopidogrel regimen + high-dose ASA | 17.6% | 2.2% to 30.5% | 0.0262 |
| Clopidogrel regimen in PCI subgroup | 14.7% | 1.2% to 26.3% | 0.0332 |
| Stent thrombosis, post hoc | 32.6% | 16.2% to 45.8% | 0.0004 |
The estimates should be read together with their confidence intervals, analysis populations, and endpoint roles. In particular, the post-hoc stent-thrombosis result and the registered primary-endpoint analyses do not have the same confirmatory status.
For each result, first identify what population was analyzed, then what endpoint was measured, then what effect measure was used, followed by the confidence interval and finally the P-value. This prevents a statistically significant P-value from being interpreted without understanding the population or magnitude of the estimated effect.
16. Stratification and Why It Matters
The registry repeatedly specifies stratification by clinically relevant factors. For the overall clopidogrel comparison, the log-rank test was stratified by ASA dose level and qualifying condition. The Cox model likewise controlled for these factors. The ASA comparison was stratified by clopidogrel treatment regimen and controlled for that regimen in the Cox model.
ASA dose level
Used as a stratification or adjustment factor when estimating the overall clopidogrel treatment-regimen effect.
Qualifying condition
UA/NSTEMI or STEMI was used as a stratification factor in the relevant clopidogrel time-to-event analyses.
Clopidogrel regimen
Used as the stratification or adjustment factor for the primary ASA dose comparison.
Interpretive consequence
The reported effect estimates are not simple unadjusted comparisons; they arise from stratified time-to-event analyses.
Stratification does not turn the stratification factors into new treatment effects. Instead, it changes how the treatment comparison is constructed statistically. This is especially important in a factorial design, where one treatment factor is part of the comparison and the other is part of the design structure.
17. Limitations and Interpretation Issues
- Composite endpoint: cardiovascular death, MI, and stroke are combined into a single first-event endpoint. A composite estimate does not by itself describe the contribution of each component.
- Relative effect reporting: the primary registry analyses report relative risk reduction estimates, but the posted analyses do not provide corresponding absolute event risks in the statistical-analysis records presented here.
- Factorial interaction: the presence of different stratum-specific estimates makes the formal interaction analysis important. Individual stratum P-values should not substitute for an interaction test.
- Analysis-population differences: the principal clopidogrel comparison uses ITT, the ASA comparison uses an ASA-treated population, and the PCI analysis uses randomized patients who underwent PCI.
- PCI subgroup: the PCI analysis addresses a narrower subgroup than the overall randomized comparison and should not be treated as though it were the same estimand.
- Post-hoc analysis: the stent-thrombosis analysis is explicitly identified as post hoc, so it should be distinguished from the registered primary endpoints.
- Proportional-hazards framework: the registry uses a stratified Cox proportional-hazards model for the relative risk-reduction estimates. Interpretation therefore depends on the model's time-to-event framework and its proportional-hazards assumption.
- Multiplicity: the registry reports multiple primary analyses, an interaction analysis, secondary analyses, and a post-hoc analysis. The P-values should therefore be interpreted in the context of the distinct endpoint roles rather than as a single undifferentiated collection of hypothesis tests.
- Effect estimates versus P-values: statistical significance and treatment-effect magnitude are different quantities and should be interpreted separately.
18. Why This Trial Matters Statistically
CURRENT/OASIS7 is a particularly useful teaching example because the trial's statistical structure goes beyond a simple two-arm comparison. It combines randomization, quadruple masking, factorial allocation, time-to-event analysis, stratification, Cox modeling, interaction testing, binary safety analysis, and a post-hoc endpoint.
| Concept | How it appears in CURRENT/OASIS7 |
|---|---|
| Randomization | Randomized phase 3 allocation across four treatment combinations. |
| Factorial design | Two treatment dimensions: clopidogrel treatment regimen and ASA dose level. |
| Blinding | Quadruple masking. |
| ITT analysis | Used for the principal clopidogrel time-to-event comparison and other specified analyses. |
| Time-to-event endpoint | First occurrence of CV death, MI, or stroke over 30 days. |
| Log-rank test | Primary time-to-event comparisons. |
| Cox model | Estimation of relative risk reduction with stratification/control variables. |
| Risk reduction | Relative risk reduction reported as the principal effect measure. |
| Interaction | Formal clopidogrel-treatment-regimen × ASA-dose interaction analysis. |
| Logistic regression | Major bleeding analyses. |
| Subgroup analysis | Clopidogrel treatment-regimen comparison in the PCI subgroup. |
| Post-hoc analysis | Stent thrombosis comparison. |
The factorial structure is especially instructive. If the trial had only compared two clopidogrel regimens, the analysis would answer one primary treatment question. By simultaneously varying ASA dose, the trial also creates an opportunity to determine whether the clopidogrel effect changes according to ASA dose. That makes interaction testing an integral part of the statistical interpretation rather than an optional afterthought.
19. A Practical Framework for Interpreting the Results
Step 1: Identify the estimand
Determine whether the result concerns the overall randomized population, the ASA-treated population, a PCI subgroup, or another analysis population.
Step 2: Identify the endpoint
Distinguish the 30-day composite time-to-event endpoint from major bleeding and post-hoc stent thrombosis.
Step 3: Read the effect estimate
Relative risk reduction describes the treatment comparison on a relative scale; it is not an absolute risk difference.
Step 4: Read the confidence interval
The interval indicates the precision of the estimate and shows whether the plausible relative-effect range includes zero on the RRR scale.
Step 5: Read the P-value
The P-value describes evidence against the relevant null hypothesis; it is not an effect-size metric.
Step 6: Check interaction and endpoint role
In a factorial trial, examine the formal interaction analysis and distinguish primary, secondary, subgroup, and post-hoc results.
This sequence prevents several common statistical errors. For example, a reader should not conclude that a high-dose clopidogrel effect is different between ASA strata merely because one stratum has P = 0.0262 and the other has P = 0.4579. The formal interaction analysis is the appropriate test of whether the treatment effects differ.
20. Registry Reporting Summary
| Registry feature | Reported value |
|---|---|
| Results posted | Yes |
| Outcome measures posted | 7 |
| Statistical analyses posted | 9 |
| Primary-endpoint analyses | 6 |
| Primary analyses with estimate + CI | 5 |
| Normalized methods | Chi-squared test; Log-rank test; Logistic regression |
| Effect-measure family | Risk ratio / relative risk |
| Hypothesis type | Superiority |
The registry therefore contains formal statistical analyses for the principal composite endpoint, including estimates and confidence intervals for the main time-to-event comparisons. It also provides logistic-regression analyses for major bleeding and a post-hoc time-to-event analysis of stent thrombosis.
21. Related Tutorials
Learn more about the methods used in this trial:
22. Related Calculators
23. Sources
- ClinicalTrials.gov: NCT00335452 — CURRENT/OASIS7.
- PubMed: PMID 20818903.
- PubMed: PMID 20818910.
- PubMed: PMID 20817281.
- PubMed: PMID 20817280.
- PubMed: PMID 42508308.
Continue through the Clinical Biostats knowledge graph
Connect this trial's design, endpoints, and statistical methods to deeper tutorials and practical statistical tools.