This page separates reported trial information from statistical interpretation. The ClinicalTrials.gov record identifies the registered primary endpoint and posted outcomes, but does not provide a formal statistical-analysis result table.
View the HPS2-THRIVE ClinicalTrials.gov record.
1. Trial at a Glance
HPS2-THRIVE was a randomized, parallel-group, quadruple-masked, phase 3 prevention trial conducted in participants with cardiovascular disease and related conditions. The ClinicalTrials.gov record reports 25,673 participants and two treatment arms, with a primary endpoint of major vascular event during the scheduled treatment period.
| Feature | HPS2-THRIVE |
|---|---|
| Trial name | HPS2-THRIVE |
| Brief title | Treatment of HDL to Reduce the Incidence of Vascular Events HPS2-THRIVE |
| Phase | Phase 3 |
| Status | Completed |
| Therapeutic area | Cardiology |
| Conditions | Cardiovascular Disease; Peripheral Arterial Disease; Diabetes Mellitus; Coronary Heart Disease |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | Quadruple |
| Primary purpose | Prevention |
| Enrollment | 25,673 |
| Lead sponsor | University of Oxford |
| Trial start | 2007-01 |
| Primary completion | 2012-10 |
2. Clinical Question
The registered trial question concerns whether treatment involving ER niacin/laropiprant, in the context of the study's lipid-lowering treatment strategy, affects the incidence of major vascular events during the scheduled treatment period.
Population
The registered conditions were cardiovascular disease, peripheral arterial disease, diabetes mellitus, and coronary heart disease.
Intervention
ER niacin/laropiprant was one of the registered interventions. Simvastatin and ezetimibe/simvastatin were also listed as study interventions.
Comparator
The ClinicalTrials.gov record identifies a placebo arm corresponding to the comparison with ER niacin/laropiprant.
Primary question
Does the randomized treatment comparison affect the occurrence of a major vascular event during the scheduled treatment period?
3. Trial Design
ER Niacin/Laropiprant
- ER niacin/laropiprant
- Randomized parallel-group assignment
- Quadruple-masked trial design
Placebo
- Placebo comparison
- Randomized parallel-group assignment
- Quadruple-masked trial design
The registry-reported intervention field also lists simvastatin and ezetimibe/simvastatin. Because the provided registry extract does not specify their allocation structure, dose, timing, or relationship to the two-arm randomized comparison in enough detail to reconstruct a fuller treatment schema, this page does not infer an additional factorial or multi-arm design.
4. Trial Population and Clinical Setting
The registered conditions identify a population with established or clinically important vascular risk involving cardiovascular disease, peripheral arterial disease, diabetes mellitus, and coronary heart disease. The primary purpose is explicitly listed as prevention.
The ClinicalTrials.gov record does not provide a baseline characteristics table, demographic distributions, disease-duration information, or subgroup-specific enrollment counts. Those characteristics are therefore not used to construct additional statistical comparisons.
5. Primary Endpoint
| Endpoint | Registered definition | Time frame | Results posted |
|---|---|---|---|
| Major Vascular Event | Non-fatal myocardial infarction or coronary death, non-fatal or fatal stroke, or revascularisation | During scheduled treatment period (median duration 3.9 years) | Yes |
The endpoint is a composite of several clinically important vascular outcomes. A participant can therefore contribute to the composite by experiencing any qualifying component event. The composite is useful for capturing a broad vascular outcome, but its interpretation depends on understanding which components contribute to the observed difference between treatment groups.
The registry definition combines myocardial infarction, coronary death, stroke and revascularisation into a single primary outcome.
6. Results Availability
The ClinicalTrials.gov record indicates that results have been posted and that five outcome measures are posted. However, the registry-reported statistical-analysis field contains no statistical analyses.
| Results component | Information available in the ClinicalTrials.gov record |
|---|---|
| Results posted | Yes |
| Outcome measures posted | 5 |
| Formal statistical analyses reported | None |
| Primary endpoint results | Endpoint identified; formal effect estimate not reported |
| Confidence intervals | Not posted on ClinicalTrials.gov for the primary endpoint |
| P-values | Not posted on ClinicalTrials.gov for the primary endpoint |
| Primary endpoint event counts by randomized arm | Not reported |
7. Planned Analysis
Because the ClinicalTrials.gov record does not contain a formal statistical-analysis result for the primary endpoint, the appropriate educational approach is to explain how a randomized major-vascular-event endpoint of this type would ordinarily be analyzed without presenting an unreported treatment effect.
Time-to-event analysis
For a composite vascular endpoint observed over a scheduled treatment period, a standard analysis would treat time from randomization to the first qualifying major vascular event as a time-to-event outcome. Participants who have not experienced the endpoint by the end of their available follow-up can contribute censored follow-up.
Kaplan-Meier estimation
Kaplan-Meier methods can estimate the event-free survival function over time. This is useful when participants have different amounts of follow-up and when the analysis needs to account for censoring rather than simply dividing the number of events by the number randomized.
Log-rank comparison
A log-rank test is commonly used to compare time-to-event distributions between randomized treatment groups. If the trial's prespecified analysis used such a test, it would evaluate evidence against equality of the underlying event-time distributions.
Cox proportional-hazards model
A Cox proportional-hazards model can estimate a hazard ratio comparing the instantaneous event rates between randomized groups. A hazard ratio below 1 would indicate a lower estimated instantaneous event rate in the intervention group under the fitted model; it would not mean that the same percentage of participants experienced a benefit.
Composite endpoint interpretation
A formal analysis should also examine the components of the composite. A statistically different composite endpoint can be driven primarily by one component rather than producing comparable effects across myocardial infarction, coronary death, stroke and revascularisation.
8. How a Major Vascular Event Analysis Works
Start at randomization
The randomized comparison begins at treatment assignment. This preserves the treatment groups created by randomization for the primary efficacy comparison.
Identify the first qualifying event
The composite endpoint is triggered by a qualifying non-fatal myocardial infarction, coronary death, stroke, or revascularisation according to the registered definition.
Account for censoring
Participants without an observed event by the end of their available observation contribute information until censoring rather than being treated as if they had experienced an event.
Compare randomized groups
A time-to-event model or test can then quantify the difference between the intervention and comparator groups.
The important statistical distinction is between event incidence and time to event. If all participants had exactly the same follow-up and no censoring, a simple risk comparison might be sufficient. Clinical trials typically require methods that account for the timing of events and differing observation periods.
9. Safety Results
The ClinicalTrials.gov record reports serious adverse events by randomized arm. The registry caveat states that participants may have serious adverse events reported in more than one category; consequently, the total number of participants affected across reported categories is less than the total obtained by summing category-specific counts.
| Safety measure | ER Niacin/Laropiprant | Placebo |
|---|---|---|
| Participants affected by serious adverse events | 7,137 / 12,838 | 6,762 / 12,835 |
The denominators posted on ClinicalTrials.gov for the serious-adverse-event data are 12,838 and 12,835. The counts should be interpreted as participants affected, not as the number of individual serious-adverse-event episodes.
10. Statistical Methods Explained
Why is randomization important?
Randomization creates treatment groups by assignment rather than by clinical choice. In principle, this balances measured and unmeasured prognostic factors in expectation and supports a causal comparison of the randomized strategies. The HPS2-THRIVE registry identifies the allocation as randomized and the design as parallel.
What does quadruple masking add?
Quadruple masking indicates that the trial used a high level of blinding. Masking can reduce the opportunity for knowledge of treatment assignment to influence behavior, assessment, reporting, or other aspects of trial conduct. The ClinicalTrials.gov record does not specify which four participant roles were masked, so no more specific attribution is made here.
Why is a composite endpoint used?
A composite endpoint can capture several clinically relevant event types in a single outcome. For HPS2-THRIVE, the registered major vascular event combines myocardial infarction or coronary death, stroke, and revascularisation. The statistical advantage is that more participants may experience at least one component, but the clinical interpretation requires attention to the individual components.
What would a hazard ratio mean here?
If a Cox model were used, the hazard ratio would compare the estimated instantaneous rate of the first major vascular event between randomized groups. A value below 1 would correspond to a lower estimated hazard in the intervention group. It would not be equivalent to the percentage of participants who avoid an event, and it would not by itself describe the absolute difference in event probability.
Why does the median treatment duration matter?
The registered primary endpoint is assessed during the scheduled treatment period, with a median duration of 3.9 years. This provides important context for the time horizon over which the primary endpoint was defined. It should not be converted into a mean follow-up time or treated as though every participant was observed for exactly 3.9 years.
Why are confidence intervals important?
A confidence interval around a treatment-effect estimate describes statistical uncertainty under the specified model and sampling framework. It is different from a range containing the effects experienced by individual participants. Because the registry-reported statistical-analysis field contains no formal estimates, no confidence interval is presented for the primary endpoint here.
Why should the composite components be examined separately?
Major vascular event is not a single biological event. It combines myocardial infarction, coronary death, stroke and revascularisation. A treatment effect on the composite can arise from different patterns across these components, so a complete statistical interpretation should distinguish the overall composite from its component outcomes when those results are available.
11. Randomization and Masking
Randomization and masking are central design features of HPS2-THRIVE. The registry describes the allocation as RANDOMIZED, the design model as PARALLEL, and the masking as QUADRUPLE.
| Design feature | Registered information | Statistical importance |
|---|---|---|
| Allocation | Randomized | Supports comparison of treatment strategies while reducing systematic allocation differences. |
| Design model | Parallel | Participants remain in parallel randomized treatment groups rather than serving as their own controls. |
| Masking | Quadruple | Reduces the potential influence of treatment knowledge on trial conduct and assessment. |
| Primary purpose | Prevention | Frames the major vascular event endpoint as a preventive outcome. |
These design characteristics are important when interpreting a treatment comparison because the statistical analysis should preserve the randomized structure rather than attempting to recreate treatment groups from observed clinical characteristics.
12. What the Primary Endpoint Can and Cannot Establish
The registered major vascular event endpoint measures whether a participant experiences at least one qualifying event during the scheduled treatment period: non-fatal myocardial infarction or coronary death, non-fatal or fatal stroke, or revascularisation.
A formal randomized comparison could estimate the difference in the time-to-event distribution, relative event rate, or absolute event probability between ER niacin/laropiprant and placebo, depending on the prespecified statistical method.
The registry-reported statistical-analysis field does not provide a formal treatment-effect estimate, confidence interval, p-value, or treatment-specific primary-endpoint event count. Those values cannot be inferred from the enrollment numbers or the serious-adverse-event data.
13. Analysis Population Considerations
For a randomized prevention trial, the primary efficacy analysis would ordinarily preserve randomized treatment assignment through an intention-to-treat framework unless the prespecified protocol specifies another primary population. The ClinicalTrials.gov record identifies randomization but does not provide the exact analysis-population definitions.
Randomized comparison
The fundamental efficacy comparison should retain the treatment assignment created by randomization.
Safety comparison
Safety summaries can be based on treatment exposure, but the precise safety-analysis population is not specified in the ClinicalTrials.gov record.
Censoring
A time-to-event analysis must define when participants are censored if they have not experienced a major vascular event.
Component events
The composite endpoint should be accompanied by component-specific analyses when available so that the nature of the observed effect can be understood.
14. Multiplicity and Multiple Endpoints
The ClinicalTrials.gov record reports five outcome measures, but they do not specify a multiplicity-adjustment procedure, endpoint hierarchy, alpha allocation, or formal testing sequence.
| Question | What the ClinicalTrials.gov record establishes |
|---|---|
| Primary endpoint | One registered primary endpoint: Major Vascular Event. |
| Posted outcome measures | Five outcome measures are reported as posted. |
| Multiplicity adjustment | Not specified in the ClinicalTrials.gov record. |
| Endpoint hierarchy | Not specified beyond the registered primary endpoint. |
| Alpha allocation | Not specified. |
This distinction matters because statistical significance across several outcomes cannot automatically be interpreted as though each test had an independent full type I error allowance. A complete confirmatory analysis would need the prespecified multiplicity strategy.
15. Interim Analysis, Missing Data, and Bayesian Methods
Interim analysis
The ClinicalTrials.gov record does not identify an interim-analysis plan or alpha-spending procedure.
Missing data / imputation
The ClinicalTrials.gov record does not specify a missing-data or imputation method for the primary endpoint.
Bayesian methods
No Bayesian statistical method is identified in the ClinicalTrials.gov record.
Stratification
No stratification variables are identified in the ClinicalTrials.gov record.
These omissions are not assumptions about what the investigators did. They are boundaries on what can be concluded from the registry extract. A complete statistical analysis plan would be required to describe these features definitively.
16. Non-Inferiority and Crossover
Non-inferiority methodology and treatment crossover are not identified in the ClinicalTrials.gov record. Consequently, this page does not apply a non-inferiority margin, discuss a non-inferiority hypothesis, or interpret the trial through a crossover-adjusted analysis.
17. Serious Adverse Events: Statistical Interpretation
The serious-adverse-event counts provide a safety comparison with denominators for each randomized arm. They can be expressed as affected participants out of participants at risk, but the ClinicalTrials.gov record does not provide a formal comparative statistical analysis.
| Arm | Affected | At risk | Interpretation |
|---|---|---|---|
| ER Niacin/Laropiprant | 7,137 | 12,838 | Participants affected by serious adverse events among those at risk. |
| Placebo | 6,762 | 12,835 | Participants affected by serious adverse events among those at risk. |
These are safety frequencies rather than a formal estimate of comparative treatment effect. A formal analysis could compare proportions, rates, or time to first serious adverse event depending on the prespecified safety methodology. The ClinicalTrials.gov record does not specify which approach was used.
18. Statistical Interpretation of the Trial Design
HPS2-THRIVE is particularly useful statistically because the registered design combines a large randomized prevention population with a composite time-to-event endpoint and extensive masking. The resulting analysis requires attention to several distinct layers of evidence rather than a single summary number.
| Statistical concept | How it appears in HPS2-THRIVE |
|---|---|
| Randomization | Participants were randomized to parallel treatment groups. |
| Masking | The trial was quadruple masked. |
| Composite endpoint | Major vascular event combines myocardial infarction/coronary death, stroke and revascularisation. |
| Time horizon | Primary endpoint assessed during the scheduled treatment period, with median duration of 3.9 years. |
| Time-to-event methods | Kaplan-Meier, log-rank and Cox methods would ordinarily be candidates for this type of endpoint, subject to the prespecified analysis plan. |
| Safety analysis | Serious adverse events are reported by randomized arm with affected/at-risk counts. |
| Multiplicity | Five outcome measures are posted, but the ClinicalTrials.gov record does not specify the multiplicity strategy. |
| Missing data | No missing-data or imputation strategy is reported. |
| Bayesian analysis | No Bayesian method is identified in the ClinicalTrials.gov record. |
19. Limitations
- No formal statistical analyses reported: the registry-reported statistical-analysis field is empty, so no treatment-effect estimate, confidence interval or p-value is presented for the primary endpoint.
- Limited endpoint results: although results are posted, the ClinicalTrials.gov record identifies the primary endpoint without providing its treatment-specific event counts or comparative statistics.
- Composite endpoint: major vascular event combines several distinct clinical outcomes, so interpretation of the composite should not be assumed to apply identically to every component.
- Missing methodological detail: the ClinicalTrials.gov record does not identify the exact time-to-event model, stratification variables, censoring rules, missing-data strategy or multiplicity procedure.
- Safety overlap: participants may have serious adverse events reported in more than one category, so affected participants cannot be summed across categories without accounting for overlap.
- No subgroup data: subgroup-specific estimates and baseline subgroup distributions are not reported.
- No non-inferiority information: no non-inferiority margin or non-inferiority testing framework is identified.
- No crossover information: the ClinicalTrials.gov record does not identify a crossover design or crossover-adjusted analysis.
- No Bayesian analysis: Bayesian methods are not identified in the ClinicalTrials.gov record.
20. Why This Trial Matters Statistically
HPS2-THRIVE provides a useful teaching example of how a large prevention trial should be translated from a registry description into a statistical analysis framework without confusing reported facts with assumptions about the analysis.
| Concept | How it appears in HPS2-THRIVE |
|---|---|
| Randomization | Randomized allocation in a parallel-group phase 3 design. |
| Blinding | Quadruple masking. |
| Prevention endpoint | Major vascular event during the scheduled treatment period. |
| Composite outcome | Myocardial infarction/coronary death, stroke or revascularisation. |
| Time-to-event analysis | Appropriate framework for an endpoint defined over follow-up time. |
| Kaplan-Meier estimation | Potential method for estimating event-free survival over time. |
| Hazard ratio | Potential relative measure from a Cox model if prespecified. |
| Confidence interval | Would quantify uncertainty around an estimated treatment effect. |
| Composite interpretation | Requires examination of individual endpoint components. |
| Safety analysis | Serious adverse events are reported by arm with affected/at-risk counts. |
| Multiplicity | Five outcome measures are posted, but the adjustment strategy is not reported. |
21. Clinical Interpretation vs Statistical Interpretation
Statistical interpretation
The registry establishes a randomized, parallel, quadruple-masked phase 3 comparison with a composite primary endpoint measured during a median scheduled treatment duration of 3.9 years. The registry-reported extract does not provide the formal comparative estimate.
Clinical interpretation
The primary endpoint represents clinically meaningful vascular outcomes: myocardial infarction or coronary death, stroke, and revascularisation. The clinical meaning of any treatment effect would depend on the magnitude, precision and component-specific pattern of the formal analysis.
This separation is important. A trial can have a sophisticated design even when a particular registry extract does not expose every statistical result. The correct response is to distinguish what is documented from what would ordinarily be analyzed.
22. A Deeper Look at the Composite Endpoint
Myocardial infarction and coronary death
These components represent coronary outcomes within the major vascular event definition. Because the registered endpoint treats them as components of a single composite, an analysis of the composite does not automatically establish an identical treatment effect for each component.
Stroke
The registered definition includes both non-fatal and fatal stroke. This broadens the stroke component to include fatal events rather than restricting the endpoint to survivors with non-fatal stroke.
Revascularisation
Revascularisation is also included in the composite. Unlike death, it is an intervention-dependent clinical event and may have different frequency and clinical interpretation from myocardial infarction or stroke.
Why component balance matters
If the components differ substantially in frequency, severity, or treatment responsiveness, the overall composite can be disproportionately influenced by one component. A complete statistical report therefore benefits from showing the composite alongside its component events whenever those data are available.
23. Understanding the 3.9-Year Time Frame
The primary endpoint time frame is explicitly stated as during the scheduled treatment period (median duration 3.9 years).
Trial start
The ClinicalTrials.gov record identifies January 2007 as the trial start.
Primary endpoint observation
Major vascular events were defined during the scheduled treatment period, with a median duration of 3.9 years.
Primary completion
The ClinicalTrials.gov record identifies October 2012 as the primary completion date.
The distinction between a median duration and an identical duration for every participant is important. A median of 3.9 years indicates the middle of the distribution of treatment durations; it does not imply that all participants contributed exactly 3.9 years of observation.
24. What a Complete Statistical Result Would Normally Contain
For an endpoint such as major vascular event, a complete statistical report would ordinarily allow the reader to reconstruct both the absolute and relative treatment comparison.
| Result element | Why it matters | Available in the ClinicalTrials.gov record? |
|---|---|---|
| Events by treatment group | Shows the observed burden of the composite endpoint. | No |
| At-risk population by treatment group | Provides the denominator for absolute event frequency. | No for the primary endpoint |
| Time-to-event estimate | Accounts for timing and censoring. | No |
| Hazard ratio | Provides a relative time-to-event treatment effect if Cox modeling was prespecified. | No |
| 95% confidence interval | Quantifies uncertainty around the effect estimate. | No |
| P-value | Addresses the specified null hypothesis under the statistical testing framework. | No |
| Component outcomes | Shows what drives the composite. | No |
Not having these values in the registry-reported statistical-analysis field does not justify reconstructing them from the enrollment or safety numbers. The two sets of numbers answer different questions.
25. Sources
- ClinicalTrials.gov: HPS2-THRIVE, NCT00461630.
- Linked publication: PubMed PMID 31805983.
- Linked publication: PubMed PMID 31447131.
- Linked publication: PubMed PMID 30821829.
- Linked publication: PubMed PMID 29449329.
- Linked publication: PubMed PMID 27407053.
26. Related Tutorials
Learn more about the methods used in this trial:
27. Related Calculators
Continue through Clinical Biostats
Connect the design and statistical concepts in HPS2-THRIVE with deeper tutorials and statistical calculators.
28. Record Summary
HPS2-THRIVE is a randomized, parallel, quadruple-masked phase 3 prevention trial with 25,673 enrolled participants and a registered primary endpoint of major vascular event during the scheduled treatment period, with a median duration of 3.9 years. The composite endpoint consists of non-fatal myocardial infarction or coronary death, non-fatal or fatal stroke, or revascularisation. The ClinicalTrials.gov record also report serious adverse events by treatment arm. Because the registry-reported statistical-analysis field contains no formal analyses, this page intentionally does not supply an unreported hazard ratio, confidence interval, p-value, or primary-endpoint event count.