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
COMPARE-ACUTE was a randomized, quadruple-masked, phase 1 crossover study evaluating the acute effects of dihydroergotamine mesylate on pulmonary arterial pressure in healthy participants. The registry lists 24 enrolled participants and 6 arms.
| Feature | COMPARE-ACUTE |
|---|---|
| Study acronym | COMPARE-ACUTE |
| ClinicalTrials.gov identifier | NCT01089062 |
| Phase | Phase 1 |
| Status | Completed |
| Start date | March 2010 |
| Primary completion date | September 2010 |
| Condition | Healthy |
| Allocation | Randomized |
| Design model | Crossover |
| Masking | Quadruple |
| Primary purpose | Treatment |
| Enrollment | 24 |
| Number of arms | 6 |
| Lead sponsor | Allergan |
| Sponsor type | Industry |
2. Clinical Question
The central question was whether the acute administration of dihydroergotamine mesylate could be compared with placebo conditions with respect to the area under the pulmonary arterial systolic pressure curve during the first 2 hours after the first dose.
Population
Healthy participants enrolled in a phase 1 pharmacodynamic study.
Interventions
MAP0004, IV dihydroergotamine mesylate, IV placebo (saline), and placebo inhaler were among the interventions listed in the registry.
Design
Randomized, quadruple-masked crossover study with 6 arms.
Primary question
How does pulmonary arterial systolic pressure over the first 2 hours after the first dose compare under the randomized treatment conditions?
3. Trial Design
The crossover designation is the central statistical feature of this study. Unlike a parallel-group trial, a crossover study allows participants to contribute observations under more than one treatment condition. This can make within-participant comparisons statistically informative because each participant can provide a degree of their own control.
Why crossover design changes the analysis
In a conventional parallel-group comparison, the primary statistical contrast is between different groups of participants. In a crossover study, treatment comparisons can instead be based partly on changes within the same participants across treatment conditions. The statistical model therefore needs to respect the dependence among observations contributed by the same participant.
A crossover analysis also requires attention to the order in which treatments are received. Treatment sequence, period, and possible carryover effects can matter because the observation obtained under one treatment is not necessarily independent of observations obtained from the same participant during other periods.
4. Interventions and Study Conditions
MAP0004
- Listed in the registry as a drug intervention.
- Evaluated within the randomized crossover design.
IV Dihydroergotamine Mesylate
- Listed in the registry as a drug intervention.
- Evaluated for its acute pharmacodynamic effect.
IV Placebo (Saline)
- Listed in the registry as a drug intervention.
- Provides a placebo condition for comparison.
Placebo Inhaler
- Listed in the registry as a drug intervention.
- Included among the randomized study conditions.
The registry identifies 6 arms while listing 4 intervention names. That distinction is important: the number of arms describes the randomized study structure, whereas the intervention list identifies the drug interventions recorded for the study. The available registry information does not provide a complete arm-by-arm treatment-sequence table.
5. Primary Endpoint
| Endpoint | Registry definition | Time frame |
|---|---|---|
| AUC(0-2hrs) of Pulmonary Arterial Systolic Pressure (PASP) Over Time Post 1st Dose | AUC(0-2hrs), the area under the curve from time 0 to 2 hours post-first dose, in pulmonary arterial systolic pressure millimeters of mercury times minutes (mmHg*min). PASP is the highest pressure exerted on the walls of the pulmonary artery. | 2 hours from time of first dose |
The primary endpoint is an integrated pharmacodynamic measure rather than a single blood-pressure observation at one time point. AUC summarizes the accumulated PASP over the specified observation window. For this endpoint, a larger AUC corresponds to a larger accumulated pressure-time measurement over the 2-hour period, while a smaller AUC corresponds to a smaller accumulated measurement.
The registry expresses the endpoint in mmHg*min. The integral emphasizes why measurements across the complete time window contribute to the endpoint rather than a single measurement determining the result.
6. Results
The registry indicates that results have been posted for the primary endpoint, but it does not provide a posted statistical analysis for that endpoint in the available record. The registry therefore does not report an estimate, confidence interval, or p-value for the AUC(0-2hrs) of PASP endpoint.
| Primary endpoint | Result status in the registry |
|---|---|
| AUC(0-2hrs) of PASP Over Time Post 1st Dose | Results posted; no posted statistical analysis with an estimate, confidence interval, or p-value. |
For a continuous pharmacodynamic endpoint of this type, a formal analysis would normally compare the treatment-condition AUC measurements while accounting for the crossover structure. Depending on the prespecified statistical analysis plan, that could involve a crossover-specific linear model or another repeated-measures approach that accounts for participant, period, and treatment effects. The appropriate inferential target is the treatment contrast, not simply the difference between two unadjusted collections of observations.
The primary endpoint is a quantitative exposure-over-time measure of PASP, so the key statistical question is whether the treatment conditions differ in their mean or otherwise modeled AUC over the first 2 hours. A treatment contrast would describe the estimated difference between conditions on the mmHg*min scale.
That contrast would not by itself mean that every participant experienced the same change in pulmonary arterial pressure, nor would it establish that an observed difference at one isolated time point persisted throughout the full 2-hour period.
A confidence interval, if reported, would quantify statistical uncertainty around the estimated treatment contrast. A p-value would address evidence against a specified null hypothesis; it would not measure the magnitude or clinical importance of the effect. Because the study uses a crossover design, interpretation also depends on appropriate handling of within-participant dependence and treatment-period structure.
7. Statistical Methodology
Why AUC is appropriate for a pharmacodynamic time course
PASP is measured over time following the first dose. AUC converts that trajectory into a single summary quantity representing the accumulated PASP during the prespecified 2-hour window. This can be useful when the scientific question concerns the overall acute response rather than a single maximum or a single prespecified time point.
Continuous endpoint analysis
AUC is a continuous outcome. In a parallel-group study, a simple comparison of group means might be possible if its assumptions were appropriate. In a crossover study, however, the observations from the same participant are correlated. A valid analysis should account for that repeated structure rather than treating every observation as independent.
The exact estimand depends on the prespecified crossover analysis. The important principle is that the treatment comparison should respect the randomized crossover structure.
Within-participant comparison
One of the statistical advantages of a crossover design is that between-person variation can be reduced when treatment conditions are compared within the same participant. If two people have substantially different baseline PASP levels, comparing each person's response under different conditions can remove part of that stable individual variability from the treatment contrast.
Period effects
Because treatment conditions occur in different periods, the analysis should distinguish treatment effects from period effects. A systematic difference between the first and second treatment periods could otherwise be mistaken for a treatment difference.
Sequence effects and carryover
The order in which treatments are administered can also matter. If the effect of an earlier treatment persists into a later period, the subsequent observation may reflect both treatments. A crossover analysis therefore needs a design and analysis framework appropriate for the possibility of treatment-sequence and carryover effects.
Masking
The study is described as quadruple-masked. Masking can reduce the opportunity for knowledge of treatment assignment to influence treatment administration, participant behavior, outcome assessment, or other aspects of study conduct. The statistical value of masking is therefore primarily through reduction of potential bias rather than through a change in the mathematical form of the AUC calculation.
8. Statistical Methods Explained
Why use a crossover design?
A crossover design allows the same participant to receive different randomized treatment conditions. For a short-term pharmacodynamic question, this can be efficient because each participant can contribute information to more than one treatment comparison and can serve as their own reference for part of the analysis.
Why not analyze all observations as independent?
Repeated observations from the same participant are correlated. Treating them as independent can underestimate uncertainty because it assumes more independent information than the study actually contains. Crossover-specific methods account for the repeated contribution of participants.
What does AUC(0-2hrs) measure?
It summarizes PASP over the period from time 0 through 2 hours after the first dose. Because the endpoint integrates the entire trajectory, two treatments can have similar AUCs even if their pressure profiles differ at individual time points.
Why does the unit mmHg*min matter?
PASP is expressed as a pressure, while time is measured in minutes. Integrating pressure over time therefore produces pressure multiplied by time. The resulting mmHg*min value represents accumulated pressure over the specified interval rather than pressure at a single instant.
What would a treatment difference in AUC mean?
A difference would quantify how the accumulated PASP over the 2-hour window differs between treatment conditions under the statistical model. It would not automatically describe the maximum PASP, the pressure at every time point, or the response of every individual participant.
Why are period and sequence important?
In crossover studies, treatment is observed in a particular temporal order. A difference between periods or a residual effect from a previous treatment can distort a simple treatment comparison. The analysis therefore needs to separate treatment effects from other components of the crossover structure when those components are relevant to the prespecified model.
9. Interpreting an AUC Result
Because no numerical statistical estimate is reported in the registry's posted statistical-analysis record, the most useful way to interpret the endpoint is to understand what a future or separately reported treatment contrast would represent.
Difference in AUC
An absolute difference is expressed in mmHg*min and describes the modeled separation between treatment conditions over the 2-hour window.
Ratio of AUCs
If a ratio were reported, a value below 1 would indicate a lower modeled AUC under the numerator treatment relative to the comparator.
Confidence interval
A confidence interval would describe uncertainty around the estimated treatment contrast under the selected statistical model.
P-value
A p-value would quantify compatibility with a specified null hypothesis; it would not quantify the size or practical importance of an effect.
For a continuous AUC endpoint, the magnitude and direction of the treatment contrast are central. Statistical significance and effect magnitude are different concepts: a small p-value does not imply a large effect, while a large effect estimate with substantial uncertainty may still provide imprecise evidence.
The treatment contrast should be interpreted in the context of the crossover design. The same participants contribute information under multiple conditions, so within-participant dependence and treatment-period structure are part of the statistical evidence rather than technical details that can simply be ignored.
10. Safety Results
The registry reports serious adverse events by arm for three treatment groups. No affected participants are reported among the participants at risk in any of those three groups.
| Safety measure | Treatment A | Treatment B | Treatment C |
|---|---|---|---|
| Serious adverse events, affected / at risk | 0/20 | 0/20 | 0/20 |
These values describe serious adverse events in the three reported treatment groups. A value of 0/20 means that no affected participants were reported among 20 participants at risk in that arm. It does not establish that the probability of a serious adverse event is exactly zero; an event rate observed as zero in a finite sample still has statistical uncertainty.
With no observed events, the observed proportion is zero, but the underlying event probability may still be greater than zero. Confidence intervals for a binomial proportion are useful when quantifying the uncertainty associated with zero observed events.
The registry does not provide a complete arm-by-arm safety table beyond the three reported serious-adverse-event entries, so broader safety comparisons are not supported by the available record.
11. Analysis Populations and What the Registry Reports
The registry identifies an enrollment of 24 participants and describes the study as randomized and crossover. It does not provide, in the available statistical-analysis record, a detailed analysis-population definition such as a modified intention-to-treat population, per-protocol population, or complete crossover-completer population.
| Feature | Registry information |
|---|---|
| Enrollment | 24 |
| Allocation | Randomized |
| Design model | Crossover |
| Masking | Quadruple |
| Number of arms | 6 |
| Formal posted statistical analysis | None |
For a crossover study, the distinction between all randomized participants and participants with evaluable observations in each treatment period can be especially important. A formal analysis would normally specify how incomplete treatment sequences, missing period measurements, and protocol deviations are handled. Those details are not reported in the available statistical-analysis record.
12. Limitations
- No numerical endpoint analysis: the registry reports that results are posted for the primary endpoint, but no formal statistical analysis with an estimate, confidence interval, or p-value is posted in the available record.
- Crossover complexity: treatment, period, sequence, and participant effects can be intertwined, so a simple parallel-group comparison may not appropriately represent the design.
- Carryover: if a treatment effect persists into a subsequent period, the resulting observation may not represent the later treatment alone. The available registry information does not provide a detailed carryover assessment.
- Incomplete arm detail: the registry reports 6 arms but lists 4 intervention names, without providing a complete arm-by-arm sequence table in the available information.
- Limited safety reporting: serious adverse-event counts are available for three treatment groups, but a complete safety profile is not provided.
- Healthy population: the registry identifies the condition as healthy. Pharmacodynamic findings in healthy participants do not automatically describe effects in patients with pulmonary or other clinical conditions.
- Short observation window: the primary endpoint covers 2 hours from the time of the first dose, so it is specifically an acute pharmacodynamic measure.
- Sample size: enrollment was 24 participants. Estimates from a study of this size can be statistically uncertain, particularly for uncommon safety events or complex interaction effects.
13. Why This Trial Matters Statistically
COMPARE-ACUTE is a useful teaching example because its scientific question and design require the analyst to think about the structure of the data before choosing a statistical test. The primary endpoint is continuous and integrates a physiological response over time, while the randomized crossover design means that observations from the same participant are related.
| Concept | How it appears in COMPARE-ACUTE |
|---|---|
| Randomization | Participants were randomly allocated within a crossover study. |
| Crossover design | The registry identifies crossover as the design model. |
| Within-participant dependence | The same participant can contribute observations under different treatment conditions. |
| Continuous endpoint | The primary endpoint is an AUC measured in mmHg*min. |
| Time-course integration | PASP is summarized over the 0–2 hour period rather than at one isolated time point. |
| Period effects | Treatment comparisons in crossover studies need to distinguish treatment from temporal effects. |
| Sequence effects | The order of treatment exposure can be relevant to interpretation. |
| Carryover | A previous treatment could influence a later period if its effect persists. |
| Masking | The study is described as quadruple-masked. |
| Safety denominators | Serious adverse events are reported as affected / at risk for three arms. |
14. A Practical Statistical Framework for This Study
The statistical analysis should begin with the estimand: the treatment contrast in the 0–2 hour PASP AUC under the randomized crossover conditions. Once that target is defined, the analysis needs to respect the fact that multiple observations may come from the same participant.
Step 1 · Define the outcome
Use AUC(0-2hrs) of PASP as the primary continuous pharmacodynamic outcome.
Step 2 · Preserve randomization
Compare treatment conditions according to their randomized assignment and prespecified crossover structure.
Step 3 · Account for repeated measures
Use an analysis that recognizes that observations from the same participant are correlated.
Step 4 · Evaluate temporal structure
Consider treatment period and sequence, and address carryover when relevant to the prespecified design.
A linear mixed-effects model is one common framework for this type of problem because it can represent treatment effects while accounting for participant-level dependence and other design factors. An alternative crossover-specific model may also be appropriate. The exact model should be determined from the prespecified statistical analysis plan rather than selected solely after examining the observed results.
This is a conceptual representation rather than a claim about the model actually used in COMPARE-ACUTE. It illustrates why a crossover analysis can include treatment, period, sequence, and participant-level components.
15. Missing Data and Crossover Completion
Missing observations can have a different consequence in a crossover study than in a parallel-group study because a participant may have usable data for one treatment period but not another. The appropriate analysis depends on why observations are missing, how much information remains for each participant, and the prespecified handling of incomplete sequences.
For an AUC endpoint, missing intermediate measurements can also affect the calculated area under the curve itself. If the complete time course is unavailable, the method used to calculate or estimate AUC becomes part of the analysis. The available registry information does not describe a specific imputation method or missing-data strategy.
16. Multiplicity and Multiple Treatment Conditions
The registry reports 6 arms and several named interventions, creating the possibility of multiple treatment comparisons. When more than one treatment contrast is evaluated against a common comparator, or when several pairwise comparisons are tested, the probability of obtaining at least one statistically significant result by chance can increase.
Multiplicity control is therefore relevant if multiple confirmatory hypotheses are tested. Possible approaches include prespecified hierarchical testing, adjusted significance thresholds, or other multiplicity-control procedures. The available record does not report a formal multiplicity strategy for the primary endpoint.
17. What the Primary Endpoint Can and Cannot Establish
What AUC can describe
The accumulated PASP measurement from time 0 through 2 hours after the first dose, summarized on the mmHg*min scale.
What AUC does not describe alone
It does not identify the exact PASP at every time point or necessarily identify the maximum observed PASP.
What randomization contributes
Randomization supports a causal comparison of the randomized treatment conditions when the study is appropriately conducted and analyzed.
What phase 1 means statistically
The study is identified as phase 1. Its design and endpoint focus on an acute pharmacodynamic question rather than establishing a broad clinical-outcome treatment effect.
18. Study Timeline
Study start
The registry lists March 2010 as the study start date.
Primary completion
The registry lists September 2010 as the primary completion date.
Registry status
The study is listed as completed, with results posted in the registry.
19. Clinical Interpretation vs Statistical Interpretation
Statistical interpretation
The primary outcome is a continuous AUC measure in a randomized crossover design. The appropriate treatment comparison must account for repeated observations within participants and the temporal structure of crossover treatment.
Clinical interpretation
The endpoint concerns acute pulmonary arterial systolic pressure over 2 hours after the first dose. Any treatment contrast would therefore describe an acute pharmacodynamic effect over that specified window rather than a long-term clinical outcome.
Keeping these two levels separate is particularly important for a phase 1 pharmacodynamic study. A statistically detectable difference in an acute physiological endpoint and a clinically meaningful treatment benefit are related but distinct questions.
20. Important Statistical Interpretation Issues
- Within-person correlation: repeated treatment observations from the same participant are not independent.
- Period effects: changes associated with time or study period can be confounded with treatment unless the design and analysis account for them.
- Sequence effects: treatment order can affect interpretation when multiple treatment conditions are administered sequentially.
- Carryover: residual treatment effects can complicate later-period comparisons if they persist beyond the preceding treatment period.
- AUC interpretation: an integrated endpoint summarizes the complete 2-hour response but does not describe the exact shape of the underlying PASP curve.
- Multiplicity: six arms create the potential for multiple comparisons, making the prespecified hypothesis structure important.
- Finite safety sample: observing 0 serious adverse events among 20 participants in an arm does not imply that the underlying event probability is zero.
- Analysis population: the available record does not provide a detailed definition of the population used for the posted endpoint result.
21. Related Tutorials
Learn more about the methods used in this trial:
22. Related Calculators
23. Sources
- ClinicalTrials.gov: NCT01089062 — COMPARE-ACUTE.
- PubMed: PubMed record.
Continue through the Clinical Biostats knowledge graph
Explore clinical-trial methods, statistical tutorials, and analytical tools connected to randomized study design and quantitative endpoints.
24. Record Summary
COMPARE-ACUTE is a phase 1 randomized, quadruple-masked crossover study in 24 healthy participants with 6 arms. Its primary endpoint, AUC(0-2hrs) of pulmonary arterial systolic pressure over time after the first dose, is a continuous integrated pharmacodynamic measure expressed in mmHg*min. The statistical structure is therefore driven by two features: the continuous AUC outcome and the repeated, within-participant observations created by the crossover design.
The registry reports that results have been posted for the primary endpoint, but the available statistical-analysis record contains no formal endpoint-level estimate, confidence interval, or p-value. For interpretation, the central statistical issues are consequently the correct handling of crossover dependence, treatment period and sequence, potential carryover, and the distinction between an acute physiological endpoint and a broader clinical treatment outcome.