This page separates information reported in the ClinicalTrials.gov record from statistical interpretation. The registry reports that results are posted, but the ClinicalTrials.gov record does not contain formal statistical-analysis estimates, confidence intervals, or p-values. It explains what the registered endpoints measure and which statistical methods would normally be appropriate for them.
1. Trial at a Glance
ALTITUDE was a randomized, double-blind, parallel phase 3 trial of aliskiren versus placebo in participants with type 2 diabetes mellitus and cardiovascular disease. The trial enrolled 8606 participants and registered 18 primary endpoints across a core active-treatment phase and an extension phase.
| Feature | ALTITUDE |
|---|---|
| Trial name | Aliskiren Trial in Type 2 Diabetes Using Cardiovascular and Renal Disease Endpoints (Core and Extension Phases) |
| Phase | Phase 3 |
| Therapeutic area | Nephrology |
| Conditions | Type 2 Diabetes Mellitus; Cardiovascular Disease |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | Double |
| Primary purpose | Treatment |
| Interventions | Aliskiren; Placebo |
| Enrollment | 8606 |
| Study status | Terminated |
| Lead sponsor | Novartis Pharmaceuticals |
| Sponsor type | Industry |
| Start | 2007-10 |
| Primary completion | 2013-02 |
| ClinicalTrials.gov | NCT00549757 |
2. Clinical Question
The trial addresses whether treatment with aliskiren, compared with placebo, affects the occurrence of cardiovascular and renal disease outcomes in participants with type 2 diabetes mellitus and cardiovascular disease.
Population
Participants with type 2 diabetes mellitus and cardiovascular disease enrolled in the phase 3 ALTITUDE trial.
Intervention
Aliskiren.
Comparator
Placebo.
Primary question
How does aliskiren compare with placebo with respect to the registered cardiovascular and renal disease endpoints?
3. Trial Design
Aliskiren
- Study intervention: aliskiren
- Randomized treatment arm
- Evaluated during the core active-treatment phase
- Participants also contributed to the extension-phase assessment where applicable
Placebo
- Study intervention: placebo
- Randomized comparator arm
- Evaluated during the core active-treatment phase
- Participants also contributed to the extension-phase assessment where applicable
The randomized parallel-group structure is important statistically because treatment assignment is established before outcome observation. The double-blind design is likewise relevant because knowledge of treatment assignment can influence participant behavior, clinical assessment, reporting, and other aspects of trial conduct.
4. Randomization and Blinding
ClinicalTrials.gov describes ALTITUDE as randomized, parallel, and double masked. These design features address different sources of bias and should not be treated as interchangeable concepts.
5. Registered Primary Endpoints
ALTITUDE registered 18 primary endpoints. The ClinicalTrials.gov record identifies all of them as binary outcomes, but their time frames describe the time from randomization or a defined extension-phase cut-off to the first event. That combination makes the statistical structure closer to time-to-first-event analysis than to a simple end-of-study proportion comparison.
| Endpoint | Time frame | Type |
|---|---|---|
| Percentage of Participants With Occurrence of Primary Composite Endpoint (Core : Active Treatment Phase) | Time from randomization to the first event (Maximum 50 months) | Binary |
| Percentage of Participants With Cardiovascular (CV) Death (Core: Active Treatment Phase) | Time from randomization to the first event (Maximum 50 months) | Binary |
| Percentage of Participants With Resuscitated Sudden Death (Core: Active Treatment Phase) | Time from randomization to the first event (Maximum 50 Months) | Binary |
| Percentage of Participants With Fatal/Non-fatal Myocardial Infarction (MI) (Core: Active Treatment Phase) | Time from randomization to the first event (Maximum 50 Months) | Binary |
| Percentage of Participants With Fatal/Non-fatal Stroke (Core: Active Treatment Phase) | Time from randomization to the first event (Maximum 50 Months) | Binary |
| Percentage of Participants With Onset of End-stage Renal Disease (ESRD) (Core: Active Treatment Phase) | Time from randomization to the first event (Maximum 50 Months) | Binary |
| Percentage of Participants With Doubling of Baseline Serum Creatinine Concentration, Sustained for at Least One Month (Core: Active Treatment Phase) | Time from randomization to the first event (Maximum 50 Months) | Binary |
| Percentage of Participants With Unplanned Hospitalization for Heart Failure (Core: Active Treatment Phase) | Time from randomization to the first event (Maximum 50 Months) | Binary |
| Percentage of Participants With All Cause Mortality (Core: Active Treatment Phase) | Time from randomization to the first event (Maximum 50 months) | Binary |
| Percentage of Participants With Occurrence of Primary Composite Endpoint (Extension Phase) | From cut-off date (20Dec2011/End of Treatment (EOT)) to the first event after cut-off date (9 months in average) | Binary |
| Percentage of Participants With Cardiovascular (CV) Death (Extension Phase) | from cut-off date (20Dec2011/End of Treatment (EOT)) to the first event after cut-off date (9 months in average) | Binary |
| Percentage of Participants With Resuscitated Sudden Death (Extension Phase) | From cut-off date (20Dec2011/End of Treatment (EOT)) to the first event after cut-off date (9 months in average) | Binary |
| Percentage of Participants Fatal/Non-fatal Myocardial Infarction (MI) (Extension Phase) | From cut-off date (20Dec2011/End of Treatment (EOT)) to the first event after cut-off date (9 month in average) | Binary |
| Percentage of Participants With Fatal/Non-fatal Stroke (Extension Phase) | From cut-off date (20Dec2011/End of Treatment (EOT)) to the first event after cut-off date (9 months in average) | Binary |
| Percentage of Participants With Onset of End-stage Renal Disease (ESRD) (Extension Phase) | From cut-off date (20Dec2011/End of Treatment (EOT)) to the first event after cut-off date (9 months in average) | Binary |
| Percentage of Participants Doubling of Baseline Serum Creatinine Concentration, Sustained for at Least One Month (Extension Phase) | From cut-off date (20Dec2011/End of Treatment (EOT)) to the first event after cut-off date (9 months in average) | Binary |
| Percentage of Participants With Unplanned Hospitalization for Heart Failure (Extension Phase) | From cut-off date (20Dec2011/End of Treatment (EOT)) to the first event after cut-off date (9 months in average) | Binary |
| Percentage of Participants With All Cause Mortality (Extension Phase) | from cut-off date (20Dec2011/End of Treatment (EOT)) to the first event after cut-off date (9 months in average) | Binary |
Core-phase composite endpoint
The registered core-phase composite endpoint was defined as the first event of the following:
- Cardiovascular (CV) death
- Resuscitated sudden death
- Non-fatal myocardial infarction (MI)
- Non-fatal stroke
- Unplanned hospitalization for heart failure (HF)
- Onset of end-stage renal disease (ESRD) or death due to renal failure
The registry states that onset of ESRD was defined as initiation of dialysis or renal transplantation, with the registry-reported definition also specifying the renal laboratory criterion and renal death as described below.
Extension-phase composite endpoint
The extension-phase composite used the same broad cardiovascular and renal event structure and measured the first event after the 20Dec2011/End of Treatment (EOT) cut-off. The registry time frame describes this extension assessment as 9 months in average.
6. Endpoint Definitions
Resuscitated sudden death
The registry defines resuscitated sudden death as occurring when a subject experiences sudden death or cardiac arrest and is successfully resuscitated by cardioversion, defibrillation, or cardiopulmonary resuscitation with a meaningful recovery of consciousness. The definition excludes known transient losses of consciousness such as seizure or vasovagal episodes that do not reflect significant cardiac dysfunction.
End-stage renal disease
ESRD is defined as initiation of dialysis, renal transplantation, a serum creatinine concentration above 6.0 mg/dL (530 µmol per liter), or renal death.
Doubling of baseline serum creatinine
The registered definition requires doubling of the baseline serum creatinine concentration sustained for at least one month. To fulfill the endpoint, the serum creatinine concentration also had to be above the upper limit of normal according to the central laboratory. The upper limit of normal specified in the registry is 1.20 mg/dL for men and 0.91 mg/dL for women.
7. Statistical Methodology
The ClinicalTrials.gov record does not provide formal statistical-analysis results for the registered endpoints. They report that results have been posted, but no formal statistical analyses were posted. Therefore, no hazard ratio, confidence interval, p-value, event-rate comparison, or treatment-effect estimate can be reported here as a trial result.
The endpoint definitions nevertheless reveal the principal statistical structure. The core endpoints are described as the time from randomization to the first event, with a maximum of 50 months. The extension endpoints are described as the time from the 20Dec2011/End of Treatment (EOT) cut-off to the first event after that cut-off. For this type of endpoint, time-to-event methods are generally more informative than treating every participant simply as an event/no-event observation at a common administrative time point.
Here, T represents time from the relevant origin to the first occurrence of the endpoint. Participants who have not experienced the event by the end of their available follow-up may be censored rather than treated as if they experienced no risk after follow-up ended.
Kaplan-Meier estimation
For a time-to-first-event endpoint, the Kaplan-Meier method can estimate the probability of remaining event-free over time while accounting for censoring. It also permits graphical comparison of the event-free experience of the randomized groups.
At each observed event time, di is the number of events and ni is the number at risk immediately beforehand.
Log-rank comparison
A log-rank test is commonly used to compare time-to-event distributions between randomized groups. It evaluates whether the observed pattern of events over follow-up differs between groups under the assumptions of the test. The ClinicalTrials.gov record does not state that a log-rank test was used for ALTITUDE, so this is presented as the typical analysis rather than as a reported ALTITUDE analysis.
Cox proportional-hazards modeling
A Cox proportional-hazards model is another standard approach for estimating a relative treatment effect for a time-to-event endpoint. It can produce a hazard ratio and confidence interval while accommodating censored observations. Again, the ClinicalTrials.gov record does not report that a Cox model was used, so no Cox estimate is attributed to ALTITUDE here.
A hazard ratio is a relative time-to-event measure. It is not a risk ratio, an absolute risk difference, or the percentage of participants who benefit.
Binary endpoint reporting
The registry labels the primary endpoints as binary. A simple binary analysis could compare the proportion of participants experiencing an event. However, the registered time frames explicitly refer to the time from randomization or the extension cut-off to the first event. If follow-up differs between participants, a pure binary comparison at a single common time point can discard information about when events occurred and how long participants were observed.
8. Planned Analysis
Core active-treatment phase
The core-phase endpoints measure the time from randomization to the first event, with a maximum of 50 months. For endpoints of this type, a standard analysis would typically estimate event-free survival with Kaplan-Meier methods and compare treatment groups using a time-to-event test such as the log-rank test, with a Cox model commonly used to estimate a hazard ratio and its confidence interval.
The primary composite endpoint would normally be analyzed according to its prespecified composite definition: the first occurrence of any component event. Component endpoints such as cardiovascular death, myocardial infarction, stroke, ESRD, hospitalization for heart failure, and all-cause mortality can be analyzed separately, but those component analyses answer narrower questions than the composite.
Extension phase
The extension endpoints use a different time origin: the 20Dec2011/End of Treatment (EOT) cut-off. The registered time frame then follows participants to the first event after that cut-off, with the registry description stating 9 months in average.
This distinction is statistically important. An extension-phase analysis beginning at a post-treatment cut-off is not automatically interchangeable with the original randomized comparison beginning at randomization. The estimand, risk set, follow-up period, and interpretation of the resulting comparison depend on exactly who remains under observation and how treatment exposure is defined after the cut-off.
Composite endpoint
For a composite endpoint, the analysis normally treats the first qualifying component event as the endpoint event. A participant experiencing myocardial infarction before a later hospitalization for heart failure would contribute the time to the first qualifying event rather than two independent first-event observations. This preserves the registered "first event" definition but means the composite is a mixture of clinically distinct outcomes.
Component endpoints
Separate analyses of cardiovascular death, resuscitated sudden death, myocardial infarction, stroke, ESRD, doubling of serum creatinine, hospitalization for heart failure, and all-cause mortality provide more specific information about individual outcomes. However, multiple endpoint comparisons create a multiplicity problem: without an appropriate prespecified testing strategy, a collection of nominal significance tests can produce an inflated probability of at least one apparently positive result by chance.
9. Results Availability and Reporting Boundary
ClinicalTrials.gov reports Results Posted: Yes and Outcome Measures Posted: 26 for ALTITUDE. The ClinicalTrials.gov record, however, do not include statistical-analysis records. This creates an important distinction between the availability of posted outcome information and the availability of formal comparative statistical estimates in the ClinicalTrials.gov record.
| Information category | Supplied data | Interpretation |
|---|---|---|
| Results posted | Yes | The registry record indicates that results are posted. |
| Outcome measures posted | 26 | Outcome information is reported in the registry data. |
| Statistical analyses | None reported | No formal treatment-effect estimate can be reproduced from the registry-reported statistical-analysis field. |
| Confidence intervals | None reported | |
| P-values | None reported | No formal p-value should be attributed to the trial from the registry record. |
This reporting boundary is deliberate. A statistical analysis page should distinguish between a result that is actually reported and an analysis that would ordinarily be performed for an endpoint of that type.
10. Safety Results
The ClinicalTrials.gov record provides serious adverse-event counts by arm for both the core and extension phases. These figures are reported as affected participants over participants at risk.
| Phase | Arm | Serious adverse events | At risk |
|---|---|---|---|
| Core phase | Aliskiren | 1988 | 4272 |
| Core phase | Placebo | 1893 | 4285 |
| Extension phase | Aliskiren | 704 | 3773 |
| Extension phase | Placebo | 682 | 3817 |
The ClinicalTrials.gov record does not provide a formal statistical comparison of these serious adverse-event counts. Accordingly, the table should be read descriptively rather than as evidence of a statistically tested difference between treatment groups.
The numerator/denominator format is useful because the number affected is presented alongside the number at risk. However, these figures alone do not describe the timing, duration, severity distribution beyond the serious-adverse-event classification, recurrent events, exposure time, or statistical uncertainty around the between-group difference. Those additional quantities would be required for a more complete safety analysis.
11. Core Phase vs Extension Phase
The trial's statistical structure is unusual enough that the two phases should be kept conceptually separate.
| Feature | Core active-treatment phase | Extension phase |
|---|---|---|
| Time origin | Randomization | 20Dec2011/End of Treatment (EOT) cut-off |
| Event definition | First event | First event after the cut-off |
| Registered maximum / average time frame | Maximum 50 months | 9 months in average |
| Composite endpoint | CV death, resuscitated sudden death, non-fatal MI, non-fatal stroke, unplanned hospitalization for HF, ESRD or renal death | Same broad composite structure |
| Statistical-analysis record reported | No | No |
The different time origins matter because a hazard ratio or event probability calculated from randomization answers a different question from one calculated after a later treatment or follow-up cut-off. Combining the phases without identifying their respective risk sets could lead to an incorrect interpretation.
12. Statistical Methods Explained
Why is randomization important for this trial?
Randomization establishes treatment assignment independently of participants' subsequent outcomes. In a randomized comparison, systematic baseline differences are less likely to be caused by treatment assignment than they would be in an observational comparison. The validity of that protection depends on the integrity of the randomization and subsequent analysis.
Why is a time-to-event method appropriate for the core endpoints?
The registry defines the core endpoints using the time from randomization to the first event. Participants may therefore have different amounts of observable follow-up, and some may reach the end of observation without experiencing the event. Kaplan-Meier and related survival methods preserve information about when events occur and accommodate right censoring.
What does a hazard ratio mean?
A hazard ratio compares the estimated instantaneous event rates between two groups within a time-to-event model. An HR below 1 would indicate a lower estimated instantaneous event rate in the aliskiren group relative to placebo under that model. It would not mean that the same percentage of participants avoided the event.
Why does the composite endpoint need careful interpretation?
The composite combines cardiovascular and renal events that are not necessarily equally frequent or equally consequential. A treatment effect on the composite can therefore be driven disproportionately by one or more component events. The component-specific endpoints help reveal what contributes to the composite, but multiple component analyses also raise multiplicity considerations.
Why should the extension analysis not simply be combined with the core analysis?
The extension phase changes the time origin to the 20Dec2011/End of Treatment (EOT) cut-off and evaluates events after that point. A post-cut-off analysis has a different risk set from an analysis beginning at randomization. Treating the two periods as if they had the same starting point could mix different estimands.
Why are the posted results not enough to report a hazard ratio here?
The ClinicalTrials.gov recordset says that results are posted but provides no entries in the statistical-analysis field. A hazard ratio requires event and censoring information plus a defined statistical model or an explicitly reported estimate.
Why does double masking matter statistically?
Masking is primarily a design feature rather than an estimator. Its purpose is to reduce the possibility that knowledge of treatment assignment changes participant behavior, treatment administration, outcome assessment, or other aspects of trial conduct. It complements randomization rather than replacing it.
13. Multiplicity and the 18 Primary Endpoints
ALTITUDE registered 18 primary endpoints, spanning core and extension phases. This is statistically important because a large family of primary endpoints creates a need for a clearly prespecified multiplicity strategy if formal confirmatory claims are to be made.
Why multiplicity matters
If many hypotheses are tested independently at a conventional significance level, the probability of observing at least one apparently positive result by chance can increase.
Composite vs components
The composite endpoint and its individual components are related but distinct questions. The statistical interpretation should identify which hypothesis is being tested rather than treating every endpoint as interchangeable evidence.
Core vs extension
Endpoints measured from randomization and endpoints measured after the 20Dec2011/End of Treatment cut-off have different time origins and should not be silently pooled.
Registry limitation
The ClinicalTrials.gov record does not state a multiplicity adjustment, hierarchical testing procedure, alpha allocation, or formal statistical-analysis results.
Because the registry-reported statistical-analysis field is empty, this page does not assign a statistical hierarchy to the 18 registered primary endpoints or imply that any particular endpoint achieved statistical significance.
14. Missing Data, Censoring, and Analysis Populations
The ClinicalTrials.gov record does not specify the trial's missing-data or imputation procedures, analysis populations, censoring rules, or treatment-discontinuation handling. These omissions matter particularly for time-to-event endpoints.
Censoring is not the same as no event
If a participant reaches the end of available follow-up without experiencing the event, a time-to-event analysis typically treats that participant as censored at the last relevant observation rather than assuming the participant would never experience the event.
Why analysis-population definitions matter
A randomized efficacy analysis ordinarily preserves the original treatment assignment, while safety analyses may depend on actual exposure. The appropriate population depends on the estimand and the prespecified statistical analysis plan. Because the registry-reported ALTITUDE data do not specify these populations, this page does not assign one.
Missing-event information
For a composite cardiovascular and renal endpoint, incomplete follow-up can affect the event history. The validity of the analysis therefore depends not only on how the event is defined but also on how censoring and incomplete observations are handled.
15. Interim Analysis and Bayesian Methods
The ClinicalTrials.gov record does not report an interim-analysis strategy, alpha-spending procedure, stopping boundary, Bayesian model, posterior probability criterion, or other formal sequential-testing method.
| Design topic | What the ClinicalTrials.gov record supports |
|---|---|
| Interim analysis | Not specified in the ClinicalTrials.gov record |
| Alpha spending | Not specified in the ClinicalTrials.gov record |
| Stopping boundary | Not specified in the ClinicalTrials.gov record |
| Bayesian methods | Not specified in the ClinicalTrials.gov record |
| Missing-data imputation | Not specified in the ClinicalTrials.gov record |
| Stratification factors | Not specified in the ClinicalTrials.gov record |
| Non-inferiority margin | Not applicable to the registry-reported description of this randomized comparison; no non-inferiority framework is reported |
| Crossover | Not specified in the ClinicalTrials.gov record |
These topics should not be filled in from general knowledge of the trial. The absence of a the ClinicalTrials.gov record means that the page can explain why these methods matter without claiming that a particular procedure was used.
16. Clinical Biostats Interpretation of the Registered Endpoint Structure
The primary composite endpoint asks whether a participant experiences the first occurrence of one of several prespecified cardiovascular or renal events. The time frame begins at randomization in the core phase, making the endpoint inherently longitudinal rather than simply a snapshot at the end of the trial.
The composite endpoint does not by itself distinguish which component is responsible for a treatment-group difference. A composite result must therefore be interpreted alongside its component endpoints.
If a formal treatment-effect estimate were available, its confidence interval would describe statistical uncertainty around the estimated effect under the specified model and sampling framework. Because no statistical-analysis estimates are reported here, no confidence interval can be reported for ALTITUDE on this page.
A p-value addresses evidence against a specified null hypothesis under a statistical model; it does not quantify the magnitude or clinical importance of an effect. Effect estimates and their confidence intervals are required to describe the size and precision of a treatment difference.
17. Serious Adverse Events: Statistical Perspective
The serious-adverse-event data provide four descriptive numerator/denominator pairs:
| Phase | Aliskiren | Placebo |
|---|---|---|
| Core phase | 1988/4272 | 1893/4285 |
| Extension phase | 704/3773 | 682/3817 |
These data allow the reader to see both the number affected and the number at risk. They do not, by themselves, establish whether the observed difference is statistically significant or whether the groups differ after accounting for exposure duration or other design features.
For safety endpoints, analysts may consider risk differences, risk ratios, odds ratios, time-to-first-serious-adverse-event analyses, or recurrent-event methods depending on the exact estimand and data structure. The appropriate choice depends on the prespecified safety analysis plan and the nature of the adverse-event data. No such formal analysis is reported here.
18. Trial Timeline
Trial start
The ALTITUDE trial began in October 2007 according to the ClinicalTrials.gov record.
Randomized active-treatment assessment
The core primary endpoints measure time from randomization to the first event, with a maximum follow-up of 50 months.
Extension-phase cut-off
The extension endpoints use the 20Dec2011/End of Treatment (EOT) cut-off as their time origin.
Primary completion
The ClinicalTrials.gov record identifies February 2013 as the primary completion date.
19. What a Formal Primary Analysis Would Need to Report
A complete statistical result for the core composite endpoint would ideally identify the analysis population, number of events in each randomized group, censoring rules, estimated event-free curves or event probabilities, the treatment-effect measure, its confidence interval, the hypothesis test and p-value, and the prespecified multiplicity framework.
| Reporting element | Why it matters |
|---|---|
| Event counts | Shows how much information contributed to the treatment comparison. |
| Time-to-event estimate | Describes the event experience over follow-up rather than only at a single administrative time point. |
| Hazard ratio or other effect measure | Quantifies the relative treatment effect under the selected analysis. |
| Confidence interval | Describes statistical precision around the estimated treatment effect. |
| P-value | Provides a hypothesis-testing measure when a formal test is prespecified. |
| Analysis population | Defines which participants contribute to the analysis. |
| Censoring rules | Determine how incomplete follow-up contributes to the time-to-event analysis. |
| Multiplicity strategy | Determines how multiple endpoint tests affect confirmatory error control. |
The registry-reported ALTITUDE data do not provide these elements for the primary endpoints. This page therefore stops at the boundary of the available statistical evidence rather than filling the gaps with reconstructed or remembered results.
20. Important Limitations and Interpretation Issues
- No statistical-analysis records reported: the ClinicalTrials.gov record contains no formal treatment-effect estimates, confidence intervals, or p-values.
- Posted results are not equivalent to a registry-reported comparative analysis: the record says results are posted, but the statistical-analysis field provided here is empty.
- Composite endpoint interpretation: the primary composite contains multiple cardiovascular and renal outcomes that can contribute differently to the overall endpoint.
- Multiple primary endpoints: 18 registered primary endpoints create an important multiplicity issue that cannot be resolved from the ClinicalTrials.gov record.
- Different time origins: core endpoints begin at randomization, whereas extension endpoints begin at the 20Dec2011/End of Treatment cut-off.
- Time-to-event structure: the registered first-event time frames imply that simple end-of-study proportions may discard information about event timing and censoring.
- Safety interpretation: serious-adverse-event counts are reported descriptively, without a formal comparative analysis.
- Missing statistical-plan details: the ClinicalTrials.gov record does not specify interim analyses, alpha spending, stratification, imputation, crossover handling, or Bayesian methods.
21. Why This Trial Matters Statistically
ALTITUDE is a useful statistical teaching case because it illustrates how a large randomized clinical trial can contain a complex family of cardiovascular and renal time-to-event endpoints. It also demonstrates why the endpoint label alone is not enough to determine the analysis: the registry calls the outcomes binary, but the registered definitions specify time from randomization or a later cut-off to the first event.
| Concept | How it appears in ALTITUDE |
|---|---|
| Randomization | Randomized phase 3 parallel-group design |
| Blinding | Double masking |
| Time-to-event analysis | Core endpoints use time from randomization to the first event |
| Composite endpoint | CV death, resuscitated sudden death, MI, stroke, HF hospitalization, ESRD or renal death |
| Renal endpoints | ESRD and sustained doubling of baseline serum creatinine |
| Multiple endpoints | 18 registered primary endpoints |
| Extension analysis | Post-cut-off first-event endpoints beginning at 20Dec2011/EOT |
| Safety analysis | Serious adverse events reported by treatment arm and phase |
| Statistical reporting boundary | Results posted, but no statistical-analysis records reported |
22. Composite Endpoints: A Deeper Statistical View
A composite endpoint is attractive because it can increase the number of observed events and capture several clinically relevant outcomes within one primary outcome. But statistical efficiency does not automatically imply simple clinical interpretation.
In ALTITUDE, the composite includes cardiovascular death, resuscitated sudden death, myocardial infarction, stroke, hospitalization for heart failure, and renal outcomes. These events differ in clinical meaning and may occur at very different frequencies.
Advantage
Combining prespecified events into a first-event endpoint can provide a broader measure of cardiovascular and renal disease burden and can increase the number of endpoint events available for analysis.
Interpretive caution
A composite treatment effect can be driven primarily by one component. The overall result should therefore be examined together with component-specific outcomes.
First-event principle
The registered definition focuses on the first qualifying event, so later events do not become additional primary endpoint events for that participant.
Multiplicity
Analyzing the composite and multiple components produces a family of related questions. Formal confirmatory interpretation depends on the prespecified testing strategy.
23. Extension-Phase Interpretation
The extension phase deserves particular care because its registered time frame begins after the 20Dec2011/End of Treatment (EOT) cut-off rather than at randomization.
Changing the time origin changes the risk set and the interpretation of the resulting time-to-event quantity. Core and extension analyses should therefore be described separately unless a formal statistical framework explicitly connects them.
The registry-reported extension time frame is 9 months in average. That phrase should be retained as a registry description rather than silently converted into a different follow-up statistic.
24. Related Tutorials
Learn more about the methods used in this trial:
25. Related Statistical Calculators
26. Sources
- ClinicalTrials.gov: NCT00549757 — ALTITUDE.
- Linked publication: PubMed PMID 40526444.
- Linked publication: PubMed PMID 37872654.
- Linked publication: PubMed PMID 32964800.
- Linked publication: PubMed PMID 26774608.
- Linked publication: PubMed PMID 23121378.
Continue through the Clinical Biostats statistical pathway
Use the trial's design concepts as a starting point for deeper study of randomization, blinding, and clinical-trial methodology.
27. Record Summary
ALTITUDE was a randomized, double-blind, parallel phase 3 trial enrolling 8606 participants with type 2 diabetes mellitus and cardiovascular disease and comparing aliskiren with placebo. Its registered primary outcomes comprise 18 endpoints covering cardiovascular and renal events during core and extension phases. The core endpoints are defined around time from randomization to the first event, while the extension endpoints use the 20Dec2011/End of Treatment (EOT) cut-off as their time origin.
The central statistical lesson is that endpoint definitions determine the appropriate analysis. Although the registry classifies the primary endpoints as binary, the time-to-first-event wording points toward survival-analysis methods such as Kaplan-Meier estimation and, typically, time-to-event comparison and Cox modeling. The composite structure creates additional interpretation and multiplicity considerations, while the extension phase changes the time origin and therefore the estimand.
The ClinicalTrials.gov record reports that results are posted, but they contain no statistical-analysis records. The serious-adverse-event counts are presented exactly as reported in the registry for the core and extension phases, with no unsupported formal comparison.