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
MITRAL evaluated transcatheter mitral valve replacement in patients with mitral valve disease. The registry describes the study as completed, non-randomized, single-group, unmasked, and treatment-focused, with 91 participants and three reported arms.
| Feature | MITRAL |
|---|---|
| Official title | Mitral Implantation of TRAnscatheter vaLves |
| Condition | Mitral Valve Disease |
| Intervention | Transcatheter Mitral Valve Replacement (device) |
| Phase | NA |
| Allocation | NON_RANDOMIZED |
| Design model | SINGLE_GROUP |
| Masking | NONE |
| Primary purpose | TREATMENT |
| Enrollment | 91 |
| Number of arms | 3 |
| Status | COMPLETED |
| Lead sponsor | Mayra Guerrero |
2. Clinical Question
The clinical question can be expressed as whether transcatheter mitral valve replacement can achieve technical success, reduce clinically important mitral regurgitation, and produce an acceptable mitral valve gradient after implantation in patients with mitral valve disease.
Population
Participants enrolled in the MITRAL study with the condition listed as mitral valve disease.
Intervention
Transcatheter Mitral Valve Replacement (device).
Comparator
The registry describes the allocation as non-randomized and the design model as single-group. It therefore does not define a randomized comparator group.
Primary questions
Can the procedure achieve the registry-defined technical-success criteria, maintain absence of significant mitral regurgitation, and produce an acceptable mitral valve gradient?
3. Trial Design
The registry simultaneously reports a single-group design and three arms. The arm-level safety data identify three groups: Native Mitral Valve With Severe MAC, Valve-in-Ring, and Valve-in-Valve. Because the study is explicitly described as non-randomized, those groups should not be interpreted as randomized treatment assignments.
Native Mitral Valve With Severe MAC
- Serious adverse events: 18/31 affected / at risk
Valve-in-Ring
- Serious adverse events: 13/30 affected / at risk
Valve-in-Valve
- Serious adverse events: 8/30 affected / at risk
Three reported groups
- Allocation: non-randomized
- Design model: single group
- Masking: none
4. Study Timeline
Study start
The registry lists February 25, 2015 as the study start date.
Primary completion
The registry lists December 2018 as the primary completion date.
Study status
The current registry status is COMPLETED.
5. Primary Endpoints
| Endpoint | Time frame | Registry definition |
|---|---|---|
| Technical Success at Exit From the Cath Lab | 30 days | Number of subject to achieve technical success at exit from cath lab, defined by successful vascular delivery and retrieval of transcatheter valve delivery system; deployment of single valve; correct position of transcatheter valve; adequate performance of prosthesis (MVA > 1.5 cm2) without residual MR grade ≥2 (+); no need for additional surgery or re-intervention; and patient leaves cath lab alive. |
| Absence of MR Grade 2 (+) or Greater | 30 days and 1 year | Number of subjects to have absence of MR grade 2 (+) or greater assessed with echocardiography. MR severity grading ranges from Grade 1=mild to Grade 4=severe. |
| Mitral Valve Gradient (MVG) | 30 days and 1 year | MVG assessed by echocardiography and measured in mmHg. |
These endpoints represent three different statistical data structures. Technical success and absence of MR grade 2 (+) or greater are naturally treated as binary outcomes, whereas mitral valve gradient is a continuous measurement. That distinction matters because the appropriate descriptive summaries and inferential models differ by endpoint type.
6. Planned Analysis
The registry marks all three primary endpoints as having results posted. However, no formal statistical analyses are posted in the record represented here. The available registry information therefore identifies the endpoints and their definitions but does not provide endpoint-specific estimates, confidence intervals, or p-values.
Technical Success
Technical success is a composite binary endpoint: a subject either meets the complete set of technical-success criteria or does not. A natural descriptive analysis would report the number and proportion achieving technical success, with a confidence interval for that proportion when precision is required.
Here, x represents subjects satisfying every component of the registry-defined technical-success definition and n represents the relevant analysis population.
Absence of MR Grade 2 (+) or Greater
This endpoint is also binary. The clinically relevant classification is whether echocardiographic assessment shows absence of MR grade 2 (+) or greater. Because measurements occur at 30 days and 1 year, the two time points should be kept distinct rather than treated as though they were a single observation.
Mitral Valve Gradient
Mitral valve gradient is measured in mmHg and is therefore continuous. A typical analysis would summarize the distribution at 30 days and 1 year using a measure of central tendency and variability appropriate to the observed data. A paired analysis could be considered when the same subjects contribute measurements at both time points and the assumptions of the selected method are appropriate.
7. Statistical Methodology
Binary endpoint analysis
Technical success and absence of MR grade 2 (+) or greater are binary endpoints. For a single-group study, the fundamental quantity is the observed proportion. Confidence intervals can then quantify uncertainty around that proportion.
The observed proportion p̂ describes the fraction of the analysis population meeting the endpoint. An exact or other appropriate binomial confidence interval can quantify sampling uncertainty.
A confidence interval around a proportion should not be confused with a prediction about an individual patient. It describes uncertainty in the estimated population proportion under the statistical model and sampling framework.
Continuous endpoint analysis
Mitral valve gradient is measured on a continuous scale in mmHg. For such an endpoint, the mean and standard deviation may be useful when the distribution is reasonably symmetric, while the median and interquartile range can be more informative for skewed measurements.
When both measurements are available for the same subject, the within-subject change can be analyzed directly rather than treating the two measurements as independent observations.
Repeated measurements
The registry specifies MVG and absence of MR grade 2 (+) or greater at both 30 days and 1 year. Repeated observations create a dependence structure: measurements from the same subject are generally not statistically independent. An analysis that compares time points should account for this within-subject relationship.
Composite technical success
The technical-success endpoint is especially important statistically because it combines several procedural conditions into one outcome. Failure of any required component prevents the subject from meeting the composite definition. Consequently, a high rate of success on individual components would not automatically imply an equally high composite technical-success rate.
8. Statistical Methods Explained
Why is technical success a binary endpoint?
The registry defines technical success through a complete list of required conditions. Each subject ultimately either satisfies all listed conditions or does not. That makes the endpoint binary even though the underlying procedure contains several distinct technical steps.
Why should the technical-success components not be analyzed as though they were independent endpoints?
The components are logically connected because the registry definition requires all of them for technical success. For example, successful valve deployment alone does not constitute technical success if another required condition is not met. The composite therefore summarizes the complete procedural outcome rather than simply averaging its individual components.
Why is absence of MR grade 2 (+) or greater a binary endpoint?
The endpoint converts the echocardiographic MR severity scale into a clinically defined threshold: whether MR grade 2 (+) or greater is absent. The underlying grading system has four levels, but the endpoint itself is a yes-or-no classification.
Why is MVG analyzed differently?
MVG is measured numerically in mmHg, so it retains quantitative information rather than reducing the observation to a yes-or-no category. Continuous measurements can therefore be summarized with quantities such as means, medians, standard deviations, or other measures of dispersion.
Why do the 30-day and 1-year assessments need to remain separate?
They represent different follow-up times. A 30-day result describes an earlier post-procedure state, whereas a 1-year result describes a later state. Combining them into one proportion or average could obscure clinically meaningful changes over time.
What does the non-randomized design mean for comparisons between the three groups?
Because the registry identifies the study as non-randomized, an observed difference between Native Mitral Valve With Severe MAC, Valve-in-Ring, and Valve-in-Valve groups cannot automatically be interpreted as a causal treatment effect. Group membership may be associated with underlying clinical or procedural characteristics.
9. Results
The registry marks each of the three primary endpoints as having results posted, but the record represented here contains no posted formal statistical analyses for those endpoints. The available information therefore supports identification of the endpoint definitions and their intended follow-up periods, but not numerical endpoint estimates, confidence intervals, or p-values.
Technical Success at Exit From the Cath Lab
Registry endpoint
Technical success was defined as successful delivery and retrieval, deployment of a single valve, correct valve position, adequate prosthesis performance, no residual MR grade ≥2 (+), no additional surgery or re-intervention, and leaving the cath lab alive.
The statistical quantity of primary interest for this endpoint is the proportion of subjects meeting the complete technical-success definition. Because no endpoint estimate is included here, the record does not establish a numerical technical-success rate.
What the endpoint means: technical success is a composite measure of whether the procedure met every requirement specified by the registry definition.
What it does not mean: it is not a measure of long-term survival, long-term valve performance, or comparative treatment efficacy.
Precision: a confidence interval around the success proportion would describe uncertainty in the estimated rate. No such interval is included in the available record.
Statistical significance: a p-value would not itself measure how large the technical-success rate was. The clinically interpretable quantity is the observed proportion, together with its uncertainty and the endpoint definition.
Absence of MR Grade 2 (+) or Greater
Registry endpoint
Echocardiographic assessment of absence of MR grade 2 (+) or greater, using an MR severity grading system ranging from Grade 1=mild to Grade 4=severe.
This endpoint is a binary echocardiographic outcome evaluated at two specified follow-up times. The corresponding statistical summaries would normally be proportions at 30 days and 1 year, with uncertainty quantified separately for each time point.
What the endpoint means: it asks whether a subject has avoided MR grade 2 (+) or greater at the specified assessment.
What it does not mean: absence of the threshold does not by itself summarize every aspect of valve function or establish a causal treatment effect.
Precision: confidence intervals would indicate uncertainty around the observed proportions. No numerical interval is included in the available record.
Time matters: the 30-day and 1-year assessments answer different questions and should not be interpreted as interchangeable measurements.
Mitral Valve Gradient
Registry endpoint
Mitral valve gradient assessed by echocardiography and measured in mmHg.
MVG is a continuous endpoint. A statistical report would ordinarily describe its distribution at each time point and, where appropriate, quantify within-subject change between 30 days and 1 year.
What the endpoint means: MVG provides a quantitative echocardiographic measurement of mitral valve gradient in mmHg.
What it does not mean: a single gradient value is not equivalent to a composite measure of procedural success or a patient-level clinical outcome.
Precision: confidence intervals around a mean, median, or change estimate would quantify uncertainty around the corresponding population quantity.
Statistical significance: a p-value for a change or comparison would answer a hypothesis-testing question; it would not describe the magnitude or clinical importance of the gradient itself.
10. Safety Results
The registry reports serious adverse events by the three named arms. These are affected / at-risk counts rather than randomized treatment comparisons.
| Reported arm | Serious adverse events | At risk | Affected / at risk |
|---|---|---|---|
| Native Mitral Valve With Severe MAC | 18 | 31 | 18/31 |
| Valve-in-Ring | 13 | 30 | 13/30 |
| Valve-in-Valve | 8 | 30 | 8/30 |
The affected / at-risk values describe the observed number of participants with serious adverse events relative to the number at risk in each reported group. Because allocation was non-randomized, these figures should be interpreted descriptively rather than as randomized comparative safety effects.
11. Understanding the Three Primary Endpoints Together
| Endpoint | Data type | Follow-up | Core statistical quantity |
|---|---|---|---|
| Technical Success at Exit From the Cath Lab | Binary composite | 30 days | Proportion achieving all technical-success criteria |
| Absence of MR Grade 2 (+) or Greater | Binary echocardiographic outcome | 30 days and 1 year | Proportion without MR grade 2 (+) or greater |
| Mitral Valve Gradient | Continuous echocardiographic measurement | 30 days and 1 year | Distribution and, where appropriate, change in MVG |
These endpoints complement one another. Technical success addresses whether the procedure met a detailed procedural definition. Absence of MR grade 2 (+) or greater addresses residual valve regurgitation. MVG provides a quantitative measure of valve gradient. A statistically complete interpretation should keep these constructs separate rather than treating them as interchangeable measures of one underlying outcome.
12. Why the Non-Randomized Design Matters
Randomization is the principal mechanism used in comparative trials to balance both measured and unmeasured prognostic factors, on average, between treatment groups. MITRAL is registered as NON_RANDOMIZED with a SINGLE_GROUP design. That changes the interpretation of any comparison between its three reported groups.
Descriptive comparison
The three groups can be described in terms of observed safety or outcome measures. Such comparisons are useful for characterizing the study population and observed outcomes.
Causal comparison
A causal statement that one group produced better outcomes than another requires stronger assumptions because group membership was not assigned randomly.
Confounding
Differences in baseline clinical characteristics or procedural circumstances could contribute to differences in observed outcomes between groups.
Statistical adjustment
Regression or other adjustment methods can account for measured covariates, but adjustment does not guarantee removal of unmeasured confounding.
13. Interpreting Confidence Intervals and P-Values
No endpoint-specific confidence intervals or p-values are included in the statistical analyses represented by this record. Nevertheless, understanding their roles is important when interpreting this type of study.
A confidence interval communicates statistical precision around an estimated quantity. Its width depends on factors such as sample size and variability; it does not describe the range of outcomes that every individual participant will experience.
For a binary endpoint such as technical success, the confidence interval would describe uncertainty around the estimated proportion. For MVG, it could describe uncertainty around a mean, median-related quantity, or mean change depending on the analysis.
A p-value quantifies how incompatible the observed data are with a specified null hypothesis under the assumptions of the test. It does not tell the reader how clinically important an observed effect is.
In a non-randomized study, a small p-value for a between-group comparison would also not by itself convert an observational association into a randomized causal effect.
14. Analysis Population and Missing Measurements
The registry information identifies an overall enrollment of 91 and arm-specific safety denominators of 31, 30, and 30. It does not provide a separate definition of an endpoint-specific analysis population or a formal missing-data strategy.
This distinction matters particularly for the 30-day and 1-year echocardiographic endpoints. If some participants lack an assessment at one of these time points, the denominator used for the analysis becomes an important part of the interpretation. The statistical meaning of an observed proportion depends on precisely which subjects are included.
15. Repeated Echocardiographic Assessment
The registry specifies echocardiographic assessment of MR at 30 days and 1 year and assessment of MVG at the same two time points. When the same participant contributes measurements at both times, the observations are paired within subject.
Cross-sectional question
What is the distribution or proportion at 30 days, considered independently of the 1-year measurement?
Longitudinal question
How does the measurement change from 30 days to 1 year within the same participants?
These questions require different statistical structures. Treating paired observations as independent can underestimate uncertainty because it ignores the correlation between repeated measurements from the same subject.
16. Statistical Interpretation of Safety Across the Three Groups
The reported serious-adverse-event counts are 18/31 for Native Mitral Valve With Severe MAC, 13/30 for Valve-in-Ring, and 8/30 for Valve-in-Valve. The denominators are close in size, but similar denominators do not remove the fundamental limitation created by non-randomized group assignment.
The most defensible descriptive statement is that the registry reports different observed serious-adverse-event frequencies across the three groups. A causal interpretation would require additional assumptions and information about differences between the groups.
The three affected / at-risk counts summarize observed safety experience within each reported group. They do not, by themselves, estimate a causal treatment effect, establish statistical significance, or quantify the probability that one procedural strategy is safer than another.
17. Limitations
- Non-randomized allocation: the registry explicitly identifies the study as non-randomized, so comparisons among the reported groups are susceptible to confounding.
- Single-group design: the registered design model is SINGLE_GROUP, which does not provide the same causal comparison framework as a randomized controlled trial.
- Three reported groups: the safety results are presented by three named groups, but group assignment should not be interpreted as randomized treatment assignment.
- Composite technical-success endpoint: technical success requires every listed criterion to be satisfied, so the composite can be driven by failure of any individual component.
- Repeated assessments: the 30-day and 1-year assessments represent different follow-up times and may involve different numbers of evaluable subjects.
- Endpoint-specific denominators: enrollment of 91 does not necessarily imply that every endpoint was assessed in all 91 participants at every time point.
- Safety interpretation: serious adverse-event counts describe observed events but do not by themselves establish comparative causality across the three non-randomized groups.
- Limited inferential detail: the available registry record does not provide formal statistical analyses for the primary endpoints.
- Clinical heterogeneity: the three named groups represent different procedural categories, making direct interpretation of between-group differences dependent on the clinical characteristics associated with each category.
18. Why This Trial Matters Statistically
MITRAL is a useful statistical teaching case because its endpoints illustrate several important principles of clinical-trial analysis without relying on a randomized treatment comparison. It demonstrates how procedural success, echocardiographic outcomes, repeated measurements, and safety counts require different statistical approaches.
| Concept | How it appears in MITRAL |
|---|---|
| Non-randomized design | The registry identifies allocation as NON_RANDOMIZED. |
| Single-group design | The registry identifies the design model as SINGLE_GROUP. |
| Binary endpoint | Technical success and absence of MR grade 2 (+) or greater are naturally binary outcomes. |
| Composite endpoint | Technical success requires multiple procedural criteria to be satisfied simultaneously. |
| Continuous endpoint | Mitral valve gradient is measured quantitatively in mmHg. |
| Repeated measures | MR and MVG are assessed at both 30 days and 1 year. |
| Confidence intervals | Would quantify uncertainty around binary proportions or continuous estimates. |
| Safety denominators | Serious adverse events are reported as affected / at-risk counts by group. |
| Confounding | Non-randomized group membership limits causal interpretation of between-group differences. |
| Endpoint hierarchy | Technical success, MR, and MVG measure distinct aspects of procedural and post-procedure performance. |
19. Statistical Concepts in This Trial
20. Related Statistical Calculators
21. Sources
- ClinicalTrials.gov: MITRAL — NCT02370511.
Continue through the Clinical Biostats knowledge graph
Explore related biostatistical methods, clinical-trial concepts, and statistical tools.
22. Record Summary
MITRAL provides a useful example of how statistical interpretation changes when a clinical study is non-randomized and registered as a single-group design. Its three primary endpoints span a composite binary procedural outcome, a binary echocardiographic outcome, and a continuous echocardiographic measurement. The 30-day and 1-year assessments additionally introduce a longitudinal dimension, while the serious-adverse-event counts illustrate the importance of keeping observed safety frequencies separate from causal comparisons.
The key statistical lesson is that the endpoint determines the analysis structure. Technical success is naturally summarized as a proportion; absence of MR grade 2 (+) or greater is another binary outcome requiring time-specific denominators; and MVG is a continuous measurement that can be evaluated cross-sectionally or through within-subject change. At the same time, the non-randomized design means that comparisons among the three reported groups should remain descriptive unless stronger causal assumptions and appropriate adjustment methods are justified.