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
PROTECTED TAVR was a randomized, parallel-design trial evaluating the Sentinel® Cerebral Protection System during transcatheter aortic valve replacement. The registered primary endpoint was the rate of stroke through 72 hours post TAVR procedure or discharge, whichever came first.
| Feature | PROTECTED TAVR |
|---|---|
| Trial name | PROTECTED TAVR: Stroke PROTECTion With SEntinel During Transcatheter Aortic Valve Replacement |
| ClinicalTrials.gov identifier | NCT04149535 |
| Therapeutic area | Cardiology |
| Condition | Stroke |
| Status | Completed |
| Phase | NA |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | None |
| Primary purpose | Treatment |
| Enrollment | 3000 |
| Lead sponsor | Boston Scientific Corporation |
| Sponsor type | Industry |
| Start date | 2020-02-07 |
| Primary completion | 2022-02-01 |
2. Clinical Question
The central statistical question was whether use of the Sentinel® Cerebral Protection System during TAVR was associated with a different rate of stroke through 72 hours post TAVR procedure or discharge, whichever came first, compared with TAVR without the Sentinel system.
Population
Patients enrolled in the PROTECTED TAVR randomized trial of TAVR in the setting of the registered condition of stroke.
Intervention
Sentinel® Cerebral Protection System during TAVR.
Comparator
TAVR without the Sentinel® Cerebral Protection System.
Primary question
What is the rate of stroke through 72 hours post TAVR procedure or discharge, whichever comes first, in the randomized comparison?
3. Trial Design
Sentinel cerebral protection
- Sentinel® Cerebral Protection System during TAVR
- Randomized parallel-trial assignment
- At risk for the reported serious-adverse-event summary: 1501
Control treatment
- TAVR without the Sentinel® Cerebral Protection System
- Randomized parallel-trial assignment
- At risk for the reported serious-adverse-event summary: 1499
The registry identifies the allocation as randomized and the design model as parallel. Because masking is listed as none, the trial was not described in the registry as a masked comparison.
4. Endpoints
| Endpoint | Registry definition | Time frame |
|---|---|---|
| Primary endpoint | The Rate of Stroke Through 72 Hours Post TAVR Procedure or Discharge (Whichever Comes First) | <=72 hours |
Primary endpoint definition
The registered definition includes all stroke: hemorrhagic, ischemic, or undetermined status; disabling or nondisabling. Events were to be adjudicated by an independent Clinical Events Committee (CEC) using Neurologic Academic Research Consortium (NeuroARC) definitions.
This is a binary clinical endpoint assessed within a specified early post-procedure window. The time frame is therefore part of the endpoint itself: a stroke occurring outside the registered window would not be the same endpoint as a stroke occurring through 72 hours or discharge, whichever came first.
5. Statistical Methodology
ClinicalTrials.gov identifies the trial as randomized with a parallel design, but the ClinicalTrials.gov record does not provide a populated the posted statistical analyses record. Consequently, this page does not attribute a formal statistical test, effect estimate, confidence interval, or p-value to the PROTECTED TAVR registry record.
For a randomized comparison of two groups, the resulting event rates can be compared using an appropriate binary-outcome method. The choice between a risk difference, risk ratio, odds ratio, or another model depends on the prespecified statistical analysis plan and the assumptions of the chosen method.
Why the endpoint is statistically different from a time-to-event endpoint
The registered primary endpoint is defined as a rate of stroke through a fixed early window, not as a survival-type time-to-event endpoint extending indefinitely. A patient is classified according to whether a qualifying stroke occurred within the specified period. That makes the primary analysis fundamentally different from a Kaplan-Meier analysis of long-term survival.
Independent event adjudication
The endpoint definition specifies adjudication by an independent Clinical Events Committee using NeuroARC definitions. From a statistical perspective, adjudication provides a standardized mechanism for determining whether reported clinical events meet the prespecified endpoint definition before they enter the primary endpoint analysis.
Randomized comparison
Because the trial is randomized, the principal inferential comparison is between the randomized intervention strategies. A simple event-rate comparison describes the observed data; a formal inferential analysis adds an estimate of the treatment contrast and its uncertainty.
Risk difference
The absolute difference between the two stroke probabilities. It answers how much the observed event probability differs between randomized groups.
Risk ratio
The ratio of the stroke probability in one randomized group to that in the other. It expresses a relative rather than absolute difference.
The ClinicalTrials.gov record does not specify which of these measures, or another measure, was prespecified as the formal treatment-effect estimate. They therefore should not be presented as if one particular model or effect measure had been reported by ClinicalTrials.gov.
6. Planned Analysis
The registry identifies the primary outcome as the rate of all stroke through 72 hours post TAVR procedure or discharge, whichever comes first. The endpoint includes hemorrhagic, ischemic, and undetermined strokes, as well as disabling and nondisabling strokes, with adjudication by an independent CEC using NeuroARC definitions.
How an endpoint of this type is typically analyzed
For a binary endpoint measured over a common fixed follow-up window, the analysis generally begins with the number of patients experiencing the event and the number of patients contributing to the relevant analysis population in each randomized group. The event probability can then be summarized for each arm and contrasted using an absolute or relative effect measure.
| Analysis component | Typical role | What is available here |
|---|---|---|
| Event count | Number of patients with a qualifying stroke | Not reported in the ClinicalTrials.gov record |
| Event rate | Proportion with qualifying stroke during the registered window | Primary endpoint is registered; formal result not reported in the ClinicalTrials.gov record |
| Treatment contrast | Risk difference, risk ratio, odds ratio, or another prespecified measure | Not specified in the ClinicalTrials.gov record |
| Confidence interval | Quantifies uncertainty around the treatment contrast | Not reported |
| P-value | Tests a prespecified null hypothesis under a specified statistical framework | Not reported |
Accordingly, the appropriate interpretation of the available registry information is that the trial has a registered and defined primary clinical endpoint, and results are posted, but the ClinicalTrials.gov record does not provide the formal comparison needed to reconstruct the statistical result. ClinicalTrials.gov does not report a formal statistical comparison in the posted results.
7. Statistical Methods Explained
Why is randomization important for this endpoint?
Randomization establishes the treatment assignment before the outcome is observed. In a randomized trial, differences in the observed stroke rate can therefore be interpreted within the framework created by randomized treatment assignment, subject to adherence, missing data, outcome ascertainment, and other features of the trial.
Why does the 72-hour window matter?
The time window determines which events belong to the primary endpoint. A fixed endpoint window makes the analysis a comparison of event occurrence during a defined period rather than a comparison of indefinite follow-up. The phrase "or discharge (Whichever Comes First)" also means that the endpoint definition incorporates the earlier of those two stopping points.
Why use an adjudication committee?
Clinical events can require interpretation before they can be classified consistently. An independent Clinical Events Committee provides a standardized review process. Here, the registry specifically identifies the CEC and the NeuroARC definitions as the basis for determining qualifying stroke events.
What would a risk difference tell us?
A risk difference subtracts the event probability in one group from that in the other. It is an absolute measure. For example, a difference of 0.01 would correspond to a one-percentage-point difference in event probability. No such numerical treatment contrast is reported in the ClinicalTrials.gov record for this trial, so no numerical risk difference should be inferred.
What would a risk ratio tell us?
A risk ratio divides the event probability in one group by the event probability in the other. A value below 1 would indicate a lower observed event probability in the numerator group relative to the denominator group. A risk ratio is not interchangeable with a risk difference because the same relative effect can correspond to different absolute changes depending on the baseline event probability.
Why is a confidence interval important?
A confidence interval communicates statistical uncertainty around an estimated treatment contrast. A narrow interval generally indicates greater statistical precision than a wide interval, while the interval itself does not describe the range of outcomes an individual patient might experience. Because the ClinicalTrials.gov record contains no treatment-effect estimate, there is no confidence interval to interpret for the primary endpoint here.
Why should a p-value not be treated as an effect size?
A p-value addresses evidence against a specified null hypothesis under a particular statistical model. It does not directly describe the magnitude of a clinical difference. A complete interpretation of a binary endpoint therefore considers the event rates, an appropriate effect measure, its confidence interval, and the prespecified hypothesis-testing framework together.
8. Safety
The ClinicalTrials.gov record reports serious adverse events by randomized arm as affected patients over the corresponding number at risk.
| Safety measure | TAVR With Sentinel | TAVR Without Sentinel |
|---|---|---|
| Serious adverse events | 285/1501 | 315/1499 |
The serious-adverse-event figures are descriptive counts over the registry-reported at-risk populations. The ClinicalTrials.gov record does not include a formal statistical comparison of serious adverse events, event definitions beyond the registry-reported category, follow-up duration for the safety summary, or confidence intervals. Therefore, the counts should not be converted into a comparative safety conclusion without those additional analysis details.
9. Analysis Population and Missing Data Considerations
The ClinicalTrials.gov record identifies an enrollment of 3000 and reports serious adverse events using 1501 and 1499 patients as the corresponding numbers at risk. It does not provide the detailed analysis-population definitions, exclusions, protocol deviations, missing endpoint assessments, or imputation rules needed to reconstruct the primary efficacy analysis.
| Question | What the ClinicalTrials.gov record establishes |
|---|---|
| Randomized enrollment | 3000 |
| Number of trial arms | 2 |
| Sentinel safety denominator | 1501 |
| Control safety denominator | 1499 |
| Primary endpoint analysis population | Not specified in the ClinicalTrials.gov record |
| Missing-data method | Not specified in the ClinicalTrials.gov record |
| Imputation method | Not specified in the ClinicalTrials.gov record |
For a short, fixed endpoint window, missing outcome assessments can matter because the denominator and classification of patients without observed events affect the estimated event rate. An analysis plan would normally specify how incomplete assessments, withdrawals, protocol deviations, and other forms of missingness are handled.
10. Interim Analysis, Multiplicity, and Other Design Topics
The ClinicalTrials.gov record does not identify an interim-analysis plan, alpha-spending strategy, multiplicity adjustment, stratification factors, crossover rules, Bayesian methods, or a non-inferiority margin.
| Design topic | Available information |
|---|---|
| Interim analysis | Not specified in the ClinicalTrials.gov record |
| Multiplicity | Not specified in the ClinicalTrials.gov record |
| Stratification | Not specified in the ClinicalTrials.gov record |
| Crossover | Not specified in the ClinicalTrials.gov record |
| Bayesian methods | Not specified in the ClinicalTrials.gov record |
| Non-inferiority margin | Not specified in the ClinicalTrials.gov record |
| Missing-data / imputation method | Not specified in the ClinicalTrials.gov record |
This distinction is important. A randomized design alone does not establish the details of the inferential framework. Those details normally come from the statistical analysis plan, protocol, or prespecified registry analysis record. Where the ClinicalTrials.gov record is silent, this page does not infer them from the trial's name, intervention, or general clinical-trial conventions.
11. Statistical Interpretation of the Primary Endpoint
The primary endpoint is a clinically defined binary event measure over a short, prespecified post-procedure window. Its definition is unusually explicit about the stroke categories included and about independent adjudication. These features help establish a consistent outcome definition before comparing randomized groups.
The ClinicalTrials.gov record does not provide the primary endpoint event counts, event rates, treatment-effect estimate, confidence interval, or p-value. Therefore, the magnitude and statistical uncertainty of the randomized treatment contrast cannot be reconstructed from the ClinicalTrials.gov record without adding information from outside the ClinicalTrials.gov record.
A trial can have results posted while a particular exported analysis field remains empty. "Results posted" therefore does not by itself mean that every statistical quantity needed for an independent numerical analysis is present in the ClinicalTrials.gov recordset. A rigorous statistical review should preserve that distinction rather than filling gaps with remembered publication results.
12. Limitations
- No formal statistical analyses reported: no formal statistical analyses were posted, so the primary endpoint's treatment effect, confidence interval, and p-value cannot be reported from the ClinicalTrials.gov record.
- Limited endpoint reporting: one primary endpoint is reported, but no numerical primary outcome result is included in the statistical-analysis record.
- Analysis population not fully defined: the ClinicalTrials.gov record does not specify the formal efficacy analysis population or exclusions.
- Missing-data rules are unavailable: no imputation or missing-endpoint strategy is provided in the ClinicalTrials.gov record.
- Multiplicity information is unavailable: the registry extract does not specify whether any multiplicity adjustment was used for the primary or other endpoints.
- Interim monitoring information is unavailable: no interim-analysis or alpha-spending plan is included in the ClinicalTrials.gov record.
- Safety comparison is descriptive here: serious adverse-event counts are reported by arm, but the ClinicalTrials.gov record does not include a formal comparative safety analysis.
- External-result restriction: this page intentionally does not import numerical results from the linked publications because the task requires all trial numbers to come exclusively from the ClinicalTrials.gov record.
13. Why This Trial Matters Statistically
PROTECTED TAVR provides a useful teaching example because it illustrates a randomized comparison centered on a tightly defined clinical event rather than a long-term time-to-event endpoint. The statistical challenge is not simply calculating a percentage; it is defining exactly which events qualify, who belongs in the analysis population, when follow-up stops, and how the randomized treatment contrast is quantified.
| Concept | How it appears in PROTECTED TAVR |
|---|---|
| Randomization | Patients were assigned in a randomized two-arm parallel design. |
| Binary endpoint | The primary endpoint is the rate of stroke during a specified early follow-up window. |
| Endpoint adjudication | An independent CEC adjudicated qualifying strokes using NeuroARC definitions. |
| Time-window definition | The primary endpoint covers through 72 hours post TAVR procedure or discharge, whichever comes first. |
| Absolute effect | A risk difference would describe the absolute separation between event probabilities if that effect measure were prespecified. |
| Relative effect | A risk ratio or odds ratio could describe the relative contrast if specified by the analysis plan. |
| Confidence interval | An interval around the treatment contrast would quantify statistical precision. |
| Safety analysis | Serious adverse events are reported separately by randomized arm. |
| Analysis transparency | The absence of a populated statistical-analysis record demonstrates why registry results and formal statistical estimates should not be treated as interchangeable. |
14. Randomization as the Core Statistical Design Feature
The trial's statistical concepts in the ClinicalTrials.gov record identifies Randomization as the tutorial concept. Randomization is especially important when interpreting a two-arm clinical trial because it defines the comparison before outcomes are observed.
The formula is conceptual rather than a reported PROTECTED TAVR result. The ClinicalTrials.gov record does not provide the primary endpoint event frequencies needed to calculate the difference.
Randomization does not eliminate every source of uncertainty. It does not guarantee identical groups in every measured characteristic, and it does not automatically resolve missing outcome data, deviations from assigned treatment, or limitations in outcome ascertainment. What it does provide is a prespecified allocation mechanism that supports causal comparison under the trial's intended analysis framework.
15. Related Statistical Tutorials
Learn more about the methods used in this trial:
16. Related Statistical Calculators
17. Sources
- ClinicalTrials.gov: NCT04149535 — PROTECTED TAVR.
- Linked publication: PubMed record for PMID 39471309.
- Linked publication: PubMed record for PMID 36121045.
Continue through Clinical Biostats
Build from the trial's randomized design to tutorials on clinical-trial methods and calculators for common effect measures.
18. Record Summary
PROTECTED TAVR is a completed randomized, parallel-design trial enrolling 3000 patients and comparing TAVR with the Sentinel® Cerebral Protection System against TAVR without the device. Its registered primary endpoint is the rate of all stroke through 72 hours post TAVR procedure or discharge, whichever comes first, with hemorrhagic, ischemic, undetermined, disabling, and nondisabling strokes included and adjudicated by an independent Clinical Events Committee using NeuroARC definitions.
The statistical lesson is that a clearly defined randomized endpoint is only the beginning of the analysis. To quantify the treatment effect, the analyst needs the relevant event counts and denominators, the prespecified effect measure, its confidence interval, and the hypothesis-testing framework. Those formal statistical-analysis quantities are not contained in the posted results for this record, so this page deliberately stops short of assigning an unreported numerical treatment effect.