This page separates information reported in the ClinicalTrials.gov record from statistical interpretation. The registry reports that results were posted and identifies five posted outcome measures, but the ClinicalTrials.gov record does not contain formal statistical analyses for the primary endpoint.
1. Trial at a Glance
ROADSTER 2 was a completed post-approval study of transcarotid artery revascularization in patients with significant carotid artery disease. The registry identifies one primary endpoint: the number of participants with procedural success at 30 days.
| Feature | ROADSTER 2 |
|---|---|
| Trial name | ROADSTER 2 |
| Brief title | POST-APPROVAL STUDY of TRANSCAROTID ARTERY REVASCULARIZATION in PATIENTS With SIGNIFICANT CAROTID ARTERY DISEASE |
| Condition | Carotid Artery Disease |
| Therapeutic area | Cardiovascular |
| Status | Completed |
| Enrollment | 692 |
| Start | 2015-10-23 |
| Primary completion | 2019-04-29 |
| Lead sponsor | Silk Road Medical |
| Sponsor type | Industry |
| Primary endpoint count | 1 |
| Primary endpoint type | Binary |
| Results posted | Yes |
| ClinicalTrials.gov | NCT02536378 |
2. Clinical Question
The registered study concerns transcarotid artery revascularization in patients with significant carotid artery disease. The primary statistical question represented in the ClinicalTrials.gov record is whether participants achieve the prespecified definition of procedural success within 30 days.
Population
Patients with significant carotid artery disease enrolled in the ROADSTER 2 post-approval study.
Intervention
Transcarotid artery revascularization using the procedural components specified in the registered definition of success.
Primary outcome
Number of participants with procedural success at 30 days.
Statistical question
How many enrolled participants satisfy the registry's composite definition of procedural success by 30 days?
3. Trial Design
The registry defines procedural success as a composite concept requiring three components: acute device success, technical success, and absence of a major adverse event through 30 days. The major adverse-event component is hierarchical and includes stroke, death, and myocardial infarction.
How the composite endpoint works
Procedural success is not simply a measurement of whether the device was inserted. The registry definition requires the simultaneous satisfaction of acute device success, technical success, and absence of a major adverse event through 30 days.
| Component | Registry definition |
|---|---|
| Acute device success | Successful insertion of the ENROUTE NPS and establishment of flow reversal |
| Technical success | Deployment of interventional tools |
| Safety component | Absence of a major adverse event through 30 days |
| Major adverse events | Hierarchical stroke/death/myocardial infarction |
| Overall procedural success | Acute device success + technical success + absence of the specified major adverse event through 30 days |
4. Endpoints
| Endpoint | Time frame | Type | Registry definition |
|---|---|---|---|
| Number of Participants With Procedural Success | 30 Days | Binary | Procedural success is defined as acute device success (successful insertion of the ENROUTE NPS and establishment of flow reversal), technical success (deployment of interventional tools) and the absence of a major adverse events (hierarchical stroke/death/myocardial infarction) through 30 days. |
The ClinicalTrials.gov record identifies one registered primary endpoint and five posted outcome measures. Only the primary endpoint definition is provided in the registry-reported primary-endpoint data, so no additional endpoint definitions are reproduced here.
5. Statistical Methodology
Binary primary endpoint
The registered primary endpoint is explicitly classified as binary. Each participant can therefore be classified according to whether the complete definition of procedural success was achieved by the 30-day assessment: success or not success.
Here, x represents the number of participants meeting the complete definition of procedural success and n represents the relevant analysis population.
For a single-arm binary endpoint, the primary descriptive quantity is the observed success proportion. Statistical inference, when prespecified, can then be expressed with a confidence interval for that proportion and potentially a hypothesis test against a prespecified benchmark. The ClinicalTrials.gov record does not identify such a benchmark, analysis population beyond the overall enrollment, confidence interval, or hypothesis-test result.
Composite endpoint structure
The statistical interpretation must respect the fact that procedural success is a composite endpoint. A participant must satisfy all required components to be classified as a procedural success. A failure of any required component prevents the participant from meeting the complete procedural-success definition.
Why this matters
The endpoint integrates procedural performance and short-term clinical safety rather than measuring a single technical event in isolation.
Interpretation
A procedural-success proportion represents the proportion meeting the full composite definition, not the proportion achieving only device or technical success.
Analysis population matters
The ClinicalTrials.gov record reports serious adverse events using both a per-protocol population and an intention-to-treat population: 27/632 for PP and 14/692 for ITT. These figures demonstrate why the denominator must always accompany a binary outcome. A proportion calculated among 632 participants answers a different descriptive question from one calculated among all 692 enrolled participants.
6. Planned Analysis
ClinicalTrials.gov reports that results were posted for the primary endpoint, but the ClinicalTrials.gov record contains no formal statistical analyses for that endpoint. Consequently, no primary endpoint estimate, confidence interval, p-value, or formal comparison is reported on this page.
What would ordinarily be measured
The registry says that the primary outcome is the number of participants with procedural success at 30 days. For a binary endpoint, the first descriptive analysis would ordinarily report the number of successes and the corresponding proportion in the prespecified analysis population.
What formal inference would ordinarily address
If the protocol specified a benchmark or performance goal, the observed procedural-success proportion could be compared with that prespecified value using an appropriate one-sample binary-outcome method. A confidence interval would quantify uncertainty around the estimated success proportion.
| Statistical element | What can be established from the ClinicalTrials.gov record |
|---|---|
| Endpoint | Number of participants with procedural success |
| Endpoint type | Binary |
| Time frame | 30 days |
| Formal statistical estimate | Not in the posted results |
| Confidence interval | Not in the posted results |
| P-value | Not in the posted results |
| Prespecified benchmark | Not reported in the ClinicalTrials.gov record |
7. Statistical Methods Explained
Why is the primary endpoint binary?
The registry defines procedural success as a yes-or-no outcome at 30 days. A participant either satisfies the complete definition—acute device success, technical success, and absence of the specified major adverse event—or does not.
Why is the composite definition important?
Because procedural success requires multiple conditions, the endpoint incorporates both technical and clinical elements. The overall success classification therefore cannot be interpreted as equivalent to any one component, such as successful device insertion.
What does a procedural-success proportion mean?
If the observed proportion were calculated as \(x/n\), it would describe the fraction of the specified analysis population meeting the complete procedural-success definition by 30 days. It would not describe the probability of success for every future patient or prove that each individual component has the same success probability.
Why does the denominator matter?
A binary outcome is inseparable from its analysis population. The ClinicalTrials.gov record illustrate this directly by reporting serious adverse events as 27/632 in the PP population and 14/692 in the ITT population. Even when the numerator and denominator are both reported, the interpretation depends on which population generated them.
Why would a confidence interval be useful?
A confidence interval around a binary success proportion communicates statistical precision. A point estimate alone does not indicate how much uncertainty surrounds the observed proportion. The ClinicalTrials.gov record does not contain a confidence interval for procedural success, so none is reported here.
A p-value requires a defined null hypothesis and a statistical test. The ClinicalTrials.gov record does not specify a benchmark for the procedural-success endpoint or provide a formal statistical analysis. Producing a p-value from incomplete information would therefore imply a hypothesis and analysis that are not documented in the ClinicalTrials.gov record.
Why are composite endpoints sometimes difficult to interpret?
A composite endpoint can combine events with different clinical meanings. ROADSTER 2's registered procedural-success definition combines acute device success, technical success, and absence of a hierarchical major adverse event. The overall result therefore needs to be understood as success across the entire predefined bundle rather than as a single isolated procedural measurement.
8. Safety Information Reported in the Supplied Data
the ClinicalTrials.gov record reports serious adverse events using two analysis populations. No treatment arms are provided in the ClinicalTrials.gov record, so these figures are presented by analysis population rather than assigned to intervention groups.
| Analysis population | Serious adverse events | At risk |
|---|---|---|
| PP (Per Protocol) | 27 affected | 632 |
| ITT (Intention to Treat) | 14 affected | 692 |
The two reported fractions use different analysis populations. The PP figure uses 632 participants at risk, whereas the ITT figure uses 692. They should therefore not be treated as two estimates from the same denominator or as a comparison between randomized treatment groups.
The ClinicalTrials.gov record does not provide additional serious-adverse-event definitions, treatment-arm assignments, confidence intervals, p-values, or comparative safety analyses. Those quantities are therefore not reconstructed on this page.
9. Analysis Populations and Denominators
| Population label | Participants represented in the ClinicalTrials.gov record | Reported role |
|---|---|---|
| ITT | 692 | Serious adverse events reported as 14/692 |
| PP | 632 | Serious adverse events reported as 27/632 |
The difference between 692 enrolled participants and 632 participants in the reported PP denominator is itself statistically important. Per-protocol analyses can involve a subset of the enrolled population, whereas an intention-to-treat population is generally anchored to the full enrolled or randomized population according to the protocol's definition. For ROADSTER 2, the ClinicalTrials.gov record establishes the denominators but do not provide the detailed eligibility or exclusion rules that generated the PP population.
10. What Can and Cannot Be Inferred From a Binary Endpoint
What it can describe
The proportion of the relevant analysis population satisfying the complete 30-day procedural-success definition.
What it cannot establish alone
Whether the procedure is superior, inferior, or equivalent to another strategy without a prespecified comparator or benchmark and formal analysis.
What the composite adds
It requires technical and clinical conditions to be satisfied simultaneously.
What the ClinicalTrials.gov record does not provide
A formal effect estimate, confidence interval, p-value, or prespecified statistical benchmark for the primary endpoint.
11. Important Statistical Interpretation Issues
- Composite endpoint: procedural success requires multiple components, so the overall endpoint should not be interpreted as a pure measure of technical device performance.
- Binary outcome: the primary endpoint is a success/failure classification evaluated at 30 days.
- Analysis population: the ClinicalTrials.gov record uses both PP and ITT denominators, demonstrating the importance of identifying the population underlying every proportion.
- No formal analysis reported:
- No benchmark reported: the ClinicalTrials.gov record do not identify a null success rate or performance goal against which the primary endpoint was formally tested.
- No treatment comparison reported: the ClinicalTrials.gov record does not identify treatment arms or a comparator for the primary endpoint analysis.
- 30-day time frame: procedural success is defined through a specific short-term follow-up window, so it should not be interpreted as a long-term clinical outcome.
12. Why the 30-Day Window Matters Statistically
The primary endpoint has a fixed 30-day time frame. That makes the outcome a time-bounded binary endpoint rather than a conventional time-to-event endpoint extending indefinitely through follow-up.
The classification depends on whether the registry's required procedural and clinical conditions are satisfied within the specified 30-day window.
This distinction matters because a 30-day procedural-success endpoint answers a narrower question than long-term survival or event-free survival. A participant classified as a procedural success has satisfied the specified conditions through the registered assessment window; that classification does not, by itself, describe outcomes beyond that window.
13. Why This Trial Matters Statistically
ROADSTER 2 provides a useful teaching example of how a clinical-trial analysis can be statistically meaningful even when the primary endpoint is not a time-to-event measure. The trial centers on a binary composite endpoint with a clearly defined 30-day assessment window.
| Concept | How it appears in ROADSTER 2 |
|---|---|
| Binary endpoint | Procedural success is classified as a yes/no outcome. |
| Composite endpoint | Success requires acute device success, technical success, and absence of a major adverse event through 30 days. |
| Fixed assessment window | The primary endpoint is evaluated at 30 days. |
| Analysis population | The ClinicalTrials.gov record distinguishes PP and ITT populations. |
| Denominator discipline | Safety counts are reported with their corresponding populations: 27/632 PP and 14/692 ITT. |
| Statistical inference | The registry-reported statistical-analysis field contains no formal primary endpoint analysis. |
| Registry interpretation | Posted results should be distinguished from formal comparative or hypothesis-testing results. |
14. Statistical Concepts in This Trial
Learn more about the methods used in this trial:
15. Related Statistical Calculators
16. Limitations
- No formal statistical analyses in the ClinicalTrials.gov record: no formal statistical analyses were posted, so this page cannot report a formal primary endpoint estimate, confidence interval, or p-value without introducing information outside the ClinicalTrials.gov recordset.
- Primary endpoint result details are incomplete: although results are identified as posted, the ClinicalTrials.gov record does not provide the number of procedural successes or the corresponding denominator for the primary endpoint.
- No comparator is reported: the provided ClinicalTrials.gov record does not identify treatment arms or a comparative intervention for the primary endpoint.
- No benchmark is reported: the data do not identify a prespecified performance goal or null hypothesis for procedural success.
- Safety data are population-specific: the reported serious-adverse-event counts use PP and ITT populations and should not be interpreted as an arm-to-arm comparison.
- Limited endpoint detail: the ClinicalTrials.gov record provides the primary endpoint definition but do not provide definitions for the remaining four posted outcome measures.
- No subgroup analyses reported: subgroup estimates cannot be reconstructed without additional data.
- No longitudinal efficacy estimates reported: the registry-reported primary endpoint is a 30-day binary outcome, and no longer-term statistical estimates are provided in the ClinicalTrials.gov record.
17. Record Timeline
Study start
ROADSTER 2 began on 2015-10-23 according to the ClinicalTrials.gov record.
Primary completion
The ClinicalTrials.gov record gives 2019-04-29 as the primary completion date.
Registry status
the ClinicalTrials.gov record lists the study status as completed.
Outcome reporting
The registry record is identified as having results posted, with five outcome measures posted in the ClinicalTrials.gov record.
18. A Practical Framework for Reading the Primary Endpoint
A useful statistical reading of ROADSTER 2 begins with the endpoint definition rather than with an assumed analysis method.
Step 1: Identify the unit
The unit is the participant, classified according to whether procedural success was achieved.
Step 2: Identify the window
The primary endpoint is evaluated through 30 days.
Step 3: Identify the composite
Acute device success, technical success, and absence of the specified major adverse event are all required.
Step 4: Identify the population
The denominator must correspond to the prespecified primary endpoint analysis population.
Only after these elements are established should the observed success proportion, confidence interval, or formal hypothesis test be interpreted. This sequence prevents a common statistical error: treating a composite clinical endpoint as though it were a simple technical measurement.
19. What the Supplied Data Support
| Question | Answer supported by the ClinicalTrials.gov record |
|---|---|
| How many participants were enrolled? | 692 |
| What was the primary endpoint? | Number of participants with procedural success |
| What was the time frame? | 30 days |
| What type of endpoint was it? | Binary |
| How was procedural success defined? | Acute device success + technical success + absence of a major adverse event through 30 days |
| What major adverse events were included? | Hierarchical stroke/death/myocardial infarction |
| Were results posted? | Yes |
| How many outcome measures were posted? | 5 |
| Was a formal statistical analysis posted on ClinicalTrials.gov for the primary endpoint? | No; no formal statistical analyses were posted |
| Were serious adverse events reported? | Yes: 27/632 PP and 14/692 ITT |
20. Why a Missing Statistical Analysis Should Not Be Filled In
Clinical trial registries can contain posted outcome data without providing every element needed to reproduce a formal statistical analysis. A statistically responsible analysis distinguishes between what is explicitly reported and what would normally be calculated.
For ROADSTER 2, the ClinicalTrials.gov record establishes the primary endpoint, its 30-day time frame, its binary structure, and its composite definition. They do not establish the observed number of procedural successes, a formal confidence interval, a p-value, or a prespecified performance benchmark. Reporting any of those quantities as though they came from the ClinicalTrials.gov record would create false precision.
This distinction is particularly important for educational statistical work. A method can be appropriate in principle without being an appropriate substitute for an actual reported result. The correct statistical response to missing analysis information is to identify the appropriate framework and preserve the boundary between documented evidence and methodological explanation.
21. Related Clinical Trial Results
22. Sources
- ClinicalTrials.gov: ROADSTER 2, NCT02536378.
- Linked publication: PubMed record — PMID 28335706.
- Linked publication: PubMed record — PMID 36563712.
- Linked publication: PubMed record — PMID 35381327.
- Linked publication: PubMed record — PMID 32811386.
Continue through Clinical Biostats
Explore statistical tutorials and calculators that explain the methods used to design, analyze, and interpret clinical trials.
23. Record Summary
ROADSTER 2 is a useful statistical case study because its registered primary endpoint is a 30-day binary composite outcome. Procedural success requires acute device success, technical success, and absence of a major adverse event defined hierarchically as stroke, death, or myocardial infarction. The ClinicalTrials.gov record reports an enrollment of 692 participants and identifies results as posted, but the ClinicalTrials.gov record contains no formal primary endpoint analysis. The appropriate interpretation therefore emphasizes endpoint construction, analysis-population discipline, binary-outcome estimation, and the distinction between a methodological framework and a documented inferential result.