This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the results posted in the ClinicalTrials.gov record; calculations added by Clinical Biostats are labelled as such.
1. Trial at a Glance
This page covers the registry results for Endovascular Therapy Following Imaging Evaluation for Ischemic Stroke 3 (NCT02586415), a randomized na study. The primary outcome is The Distribution of Scores on the Modified Rankin Scale (mRS) at Day 90.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT02586415 |
| Phase | NA |
| Condition | Stroke, Acute, Cerebral Infarction |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | None (open label) |
| Enrollment | 182 |
| Arms in the primary outcome | 2 |
| Lead sponsor | Gregory W Albers |
| Results posted | Yes |
2. Clinical Question
Population
The study concerned patients with acute ischemic stroke who had an ICA or MCA M1 occlusion and a Target Mismatch Profile. The registry summary describes treatment initiation between 6-16 hours after last seen well for patients who could be randomized. The primary outcome population is identified as intention to treat.
Intervention
Endovascular thrombectomy therapy used one or more listed thrombectomy devices together with standard medical therapy. The listed devices include Trevo Retriever, Solitaire™ FR Revascularization Device, Penumbra thrombectomy system, and Covidien MindFrame Capture Revascularization Device.
Comparator
The comparator was standard medical therapy alone. The registry describes a parallel, randomized trial with no masking.
Primary question
The study tested whether the distribution of modified Rankin Scale scores at Day 90 differed between the treatment strategies. The registry describes the hypothesis as superiority of thrombectomy devices plus medical management for clinical outcomes in the selected population.
3. Trial Design
This was a randomized, parallel-group trial listed as terminated. The registry reports enrollment of 182 participants. It records the study start as 2016-04 and primary completion as 2017-08-23. The phase is listed as NA, and masking is listed as none. The treatment arms were endovascular thrombectomy therapy and medical management alone.
The intervention combined standard medical therapy with endovascular thrombectomy using one or more of the devices listed in the protocol. The comparator was standard medical therapy alone. The registry’s brief summary describes participants with a Target Mismatch Profile and an ICA or MCA M1 occlusion, with endovascular treatment to be initiated between 6-16 hours after last seen well. It does not describe further eligibility criteria in the record summarized for this trial.
The primary outcome population is identified as intention to treat. The registry does not report the randomization procedure, allocation concealment, sample-size justification, or further details about the analysis population. It also does not report why the trial was terminated. Those omissions limit what can be concluded about the conduct and planning of the trial from the registry record alone.
The posted primary outcome is the Day 90 distribution of modified Rankin Scale scores. The registry also posts a secondary count of participants with scores of 0-2 at Day 90, described as functional independence. The broad ordinal primary outcome and the binary secondary outcome express related but not identical aspects of post-stroke disability.
4. Participant Flow and Accounting
The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.
Before assignment: After obtaining written informed consent, participants were randomized if they met all clinical and imaging eligibility requirements and could undergo initiation of endovascular therapy between 6 and 16 hours after the time that they had last been known to be well.
Overall Study
| Stage | Endovascular Thrombectomy Therapy | Medical Management |
|---|---|---|
| STARTED | 92 | 90 |
| 24 Hour Assessment | 91 | 90 |
| Day 30 Assessment | 81 | 70 |
| Day 90 Assessment | 78 | 65 |
| COMPLETED | 78 | 65 |
| NOT COMPLETED | 14 | 25 |
Reasons for not completing this period:
| Reason | Endovascular Thrombectomy Therapy | Medical Management |
|---|---|---|
| Death | 13 | 23 |
| Lost to Follow-up | 1 | 2 |
Analysis population for the primary outcome, as defined in the registry: intention to treat
5. Baseline Characteristics
The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.
| Characteristic | Unit | Endovascular Thrombectomy Therapy | Medical Management | Total |
|---|---|---|---|---|
| Participants analysed | 92 | 90 | 182 | |
| Age, Continuous | years | 70 (59 to 78.5) | 71 (59 to 80) | 70.5 (59 to 80) |
| Sex: Female, Male: Female | Participants | 46 | 46 | 92 |
| Sex: Female, Male: Male | Participants | 46 | 44 | 90 |
| Ethnicity (NIH/OMB): Hispanic or Latino | Participants | 14 | 10 | 24 |
| Ethnicity (NIH/OMB): Not Hispanic or Latino | Participants | 77 | 80 | 157 |
| Ethnicity (NIH/OMB): Unknown or Not Reported | Participants | 1 | 0 | 1 |
| Race (NIH/OMB): American Indian or Alaska Native | Participants | 0 | 1 | 1 |
| Race (NIH/OMB): Asian | Participants | 3 | 3 | 6 |
| Race (NIH/OMB): Native Hawaiian or Other Pacific Islander | Participants | 0 | 0 | 0 |
| Race (NIH/OMB): Black or African American | Participants | 10 | 5 | 15 |
| Race (NIH/OMB): White | Participants | 78 | 80 | 158 |
| Race (NIH/OMB): More than one race | Participants | 0 | 0 | 0 |
| Race (NIH/OMB): Unknown or Not Reported | Participants | 1 | 1 | 2 |
| NIHSS, continuous | units on a scale | 16 (10 to 20) | 16 (12 to 21) | 16 (11 to 21) |
6. Primary Endpoint
The registry lists the primary outcome as The Distribution of Scores on the Modified Rankin Scale (mRS) at Day 90, measured over Day 90.
| Element | Registry entry |
|---|---|
| Result type | Median |
| Measure of spread | Inter-Quartile Range |
| Unit | score on a scale |
| Endpoint classification | not reduced to a single comparison |
The primary endpoint is the distribution of modified Rankin Scale (mRS) scores at Day 90. The mRS is an ordinal measure of disability and dependence in daily activities after stroke or another cause of neurological disability. Its scale runs from 0-6, from no symptoms to death. Intermediate scores describe increasing levels of disability and need for assistance.
An ordinal endpoint retains information across the scale rather than reducing results to a single cut-off. This can be useful when an intervention may affect the overall pattern of disability, including shifts among several levels of function. The mRS is commonly used to describe stroke outcomes, as the registry record notes. A median and inter-quartile range provide a compact description of the score distribution, but do not show every feature of that distribution.
The posted arm summaries report a median of 3 with an inter-quartile range of 1 to 4 for endovascular thrombectomy therapy, and a median of 4 with an inter-quartile range of 3 to 6 for medical management. These summaries describe the observed scores; they are not, by themselves, a complete comparison of the distributions. The reported inferential analysis is a Cochran-Mantel-Haenszel analysis with an odds ratio, confidence interval, and p-value. The registry does not report details of its score coding or the direction convention used for the odds ratio, so the numerical odds ratio should be read alongside the registry’s stated superiority analysis rather than interpreted as an intuitively oriented probability.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The posted primary summaries show a median mRS score of 3 (inter-quartile range 1 to 4) in the endovascular thrombectomy therapy arm and 4 (inter-quartile range 3 to 6) in the medical management arm at Day 90. These statistics describe the central score and the middle half of the observations in each arm. They do not describe every participant’s outcome or establish how much of the distribution differs at each score.
The registry reports a Cochran-Mantel-Haenszel odds ratio of 2.77, with a 95% confidence interval from 1.63 to 4.70 and p < 0.001, and labels the analysis as superiority. The confidence interval expresses the uncertainty around the estimated odds ratio under the reported analysis. Its bounds are above 1, and the reported p-value indicates that the result is not readily compatible with the null hypothesis used for that analysis. The registry does not state the odds-ratio coding direction or the detailed statistical model, so the estimate should not be translated into a specific probability of a better score without those details.
The descriptive medians and the reported inferential result should be considered together. The medians summarize each arm, while the reported analysis addresses the comparison across the ordinal outcome. Neither the p-value nor the confidence interval measures the clinical importance of the difference. The registry record does not report a prespecified threshold for clinical importance or a detailed account of how the ordinal scores entered the model.
The posted secondary endpoint counts participants with mRS scores of 0-2 as a measure of functional independence. Its unadjusted recalculation is consistent in broad direction with the primary result, but it is a separate binary comparison and should not be treated as a replication of the primary ordinal analysis. Because the secondary comparison is unadjusted and the registry does not report a multiplicity plan, it carries less confirmatory weight.
The analysis population is identified as intention to treat, but the registry does not explain how missing Day 90 scores were handled or whether all randomized participants contributed observed outcome data. The result therefore describes the posted analysis, with the reported uncertainty, and must be interpreted in light of the outcome and analysis details that are absent from the registry record.
9. Statistical Methods Explained
The primary endpoint is ordinal: mRS scores have an order, but the distance between adjacent scores should not automatically be treated as equal. A median and inter-quartile range describe the location and spread of scores without requiring that the outcome be continuous and normally distributed. They are useful summaries, but an ordinal distribution can differ in ways that are not visible from its median alone. The registry reports those summaries for each arm and separately reports an inferential comparison.
What does the reported odds ratio represent?
The registry names the primary method as Cochran-Mantel-Haenszel and reports an odds ratio. In general, Cochran-Mantel-Haenszel methods compare groups while accounting for strata, or can be used with ordered outcomes, depending on the specific implementation. The registry record does not describe the strata, score coding, or exact model specification used. It therefore is not possible to reconstruct the precise interpretation of the odds ratio from the record alone. In particular, its direction should not be converted into a statement about the odds of a better outcome without knowing the coding convention.
The reported 95% confidence interval gives a range of values compatible with the data and analysis under the method’s assumptions. It is not a range that guarantees the true value lies within its limits, and it does not describe the chance that a particular treatment effect is true. A p-value addresses compatibility with a specified null hypothesis; it is not the probability that the null hypothesis is true, nor a measure of the size or importance of an effect. The registry reports p < 0.001 for the primary analysis and labels it superiority.
Why is the secondary comparison different?
The posted secondary endpoint reduces mRS to a binary outcome, with scores of 0-2 counted as functional independence. The recalculation is a simple, unadjusted comparison based on the posted counts. It reports a risk difference and related measures, with a p-value. This changes the question from whether the overall ordinal distributions differ to whether the proportion on one side of a selected threshold differs. Dichotomizing an ordinal scale can make a result easier to state, but it discards distinctions among scores within each side of the threshold.
The secondary comparison’s planning illustration uses a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation. It presents participants per arm for 80% and 90% power under scenarios that use 100%, 75%, and 50% of the observed effect. These are planning illustrations based on the observed effect, not the trial’s own sample-size justification and not observed power. The registry record does not report the trial’s original sample-size assumptions or justification.
Why does over-dispersion matter in general?
For count outcomes, over-dispersion means that the observed variability exceeds what a simple model assumes. If present and ignored, standard errors can be too small and confidence intervals too narrow. The posted primary endpoint is an ordinal score and the secondary endpoint is binary, not a count-rate outcome; the registry does not report an over-dispersion assessment. This general issue should not be taken to imply that over-dispersion affected this trial’s reported analysis.
What can be inferred about the analysis plan?
The record identifies an intention-to-treat population for the primary outcome and gives one reported primary analysis. It does not report the detailed analysis plan, handling of missing observations, any interim analyses, or a multiplicity adjustment. A reader can interpret the statistics that are posted, but cannot establish from this record whether all analysis decisions were prespecified or how unreported outcomes were treated. The absence of those details is an uncertainty about documentation, not evidence that a particular method was or was not used.
10. Analyses Reported by the Sponsor
| Method | Estimate | 95% CI | P-value |
|---|---|---|---|
| Cochran-Mantel-Haenszel | Odds Ratio (OR) 2.77 | 1.63 to 4.70 | < 0.001 |
The registry reports the primary analysis as a Cochran-Mantel-Haenszel odds ratio of 2.77, with a 95% confidence interval from 1.63 to 4.70 and p < 0.001. That is the trial’s posted primary inferential result and is the basis for the registry’s superiority designation. The record does not provide enough model detail to independently reproduce that analysis from the arm medians and inter-quartile ranges.
The recalculated comparison applies to the posted secondary binary endpoint, not the primary ordinal endpoint. It is an unadjusted calculation from the participant counts. It can provide a transparent comparison of that secondary outcome, but it is not interchangeable with the registry’s primary analysis and does not verify the primary odds ratio. The registry record does not post a separate sponsor analysis for the secondary outcome.
No discrepancy is flagged between a sponsor result and a recalculation. Where a discrepancy exists, the sponsor’s own analysis takes precedence; that issue does not arise in the posted results for this trial. The secondary recalculation remains descriptive and exploratory in interpretation, particularly because a multiplicity adjustment is not reported.
11. Multiplicity, Missing Data and Analysis Populations
Multiplicity concerns arise when multiple outcomes, time points, comparisons, or analyses are tested. Conducting several tests without adjustment can increase the chance that at least one result appears statistically notable by chance. A prespecified testing hierarchy or an adjustment can control that risk, depending on the plan. The registry lists 8 outcome measures and one reported analysis, and it posts one secondary endpoint result. It does not report a multiplicity plan, hierarchy, or adjustment. The record therefore does not establish how the primary and secondary outcomes were treated as a family of tests.
The primary endpoint is identified as an intention-to-treat analysis. This usually refers to analyzing participants in the groups to which they were randomized, but the registry does not describe the specific implementation or provide a participant flow for the Day 90 outcome. It does not report how missing mRS scores were handled, whether imputation was used, or whether sensitivity analyses assessed the effects of missing data. The number analyzed for each outcome is posted, but those counts alone do not reveal why any outcome data may be absent or how missing observations were treated.
The trial status is listed as terminated. The reason for termination is not reported in the registry record. The record also does not describe interim analyses, stopping boundaries, or whether any interim review influenced termination. These omissions prevent a reader from assessing whether early stopping or repeated looks at accumulating results affected the reported inference. No such procedure should be assumed from the status alone.
The registry does not report the original sample-size justification or the detailed statistical analysis plan. The posted planning illustration for the secondary effect is based on the observed effect and is not evidence of the planned sample size or observed power. These reporting gaps do not invalidate the stated estimates, but they leave important questions about prespecification, missingness, and the full testing strategy unresolved.
12. Safety Results
The registry reports serious adverse events as affected participants divided by participants at risk.
| Arm | Participants with serious adverse events | Participants at risk | Percentage |
|---|---|---|---|
| Endovascular Thrombectomy Therapy | 40 | 92 | 43.5% |
| Medical Management | 48 | 90 | 53.3% |
Deaths
| Arm | Deaths (participants affected of at risk) |
|---|---|
| Endovascular Thrombectomy Therapy | 13 of 92 |
| Medical Management | 23 of 90 |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 0% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| Endovascular Thrombectomy Therapy | 57 of 92 |
| Medical Management | 59 of 90 |
Most frequent serious adverse event terms
| Event term | Organ system | Endovascular Thrombectomy Therapy | Medical Management |
|---|---|---|---|
| Stroke in evolution | Nervous system disorders | 3 of 92 | 9 of 90 |
| Pneumonia aspiration | Respiratory, thoracic and mediastinal disorders | 3 of 92 | 5 of 90 |
| Brain oedema | Nervous system disorders | 4 of 92 | 3 of 90 |
| Cerebrovascular accident | Nervous system disorders | 2 of 92 | 5 of 90 |
| Haemorrhagic transformation stroke | Nervous system disorders | 3 of 92 | 3 of 90 |
| Neurological decompensation | Nervous system disorders | 2 of 92 | 3 of 90 |
| Sepsis | Infections and infestations | 1 of 92 | 4 of 90 |
| Haemorrhage intracranial | Nervous system disorders | 2 of 92 | 2 of 90 |
| Pulmonary embolism | Respiratory, thoracic and mediastinal disorders | 1 of 92 | 3 of 90 |
| Acute respiratory failure | Respiratory, thoracic and mediastinal disorders | 1 of 92 | 2 of 90 |
Most frequent non-serious adverse event terms
| Event term | Organ system | Endovascular Thrombectomy Therapy | Medical Management |
|---|---|---|---|
| Haemorrhagic transformation stroke | Nervous system disorders | 15 of 92 | 10 of 90 |
| Urinary tract infection | Infections and infestations | 4 of 92 | 10 of 90 |
| Atrial fibrillation | Cardiac disorders | 6 of 92 | 5 of 90 |
| Pyrexia | General disorders | 7 of 92 | 4 of 90 |
| Hypotension | Vascular disorders | 5 of 92 | 5 of 90 |
| Headache | Nervous system disorders | 5 of 92 | 4 of 90 |
| Pneumonia aspiration | Respiratory, thoracic and mediastinal disorders | 4 of 92 | 5 of 90 |
| Cerebral haemorrhage | Nervous system disorders | 4 of 92 | 4 of 90 |
| Hypertension | Vascular disorders | 4 of 92 | 4 of 90 |
| Hypokalaemia | Metabolism and nutrition disorders | 3 of 92 | 5 of 90 |
Serious events with the highest arm-level rate: stroke in evolution (10.0%), cerebrovascular accident (5.6%), pneumonia aspiration (5.6%), sepsis (4.4%), brain oedema (4.3%).
The posted safety counts list serious adverse events in 40 of 92 participants in the endovascular thrombectomy therapy arm and 48 of 90 participants in the medical management arm. Other adverse events are listed for 57 of 92 and 59 of 90 participants, respectively. Deaths are listed for 13 and 23 participants, respectively. These are observed counts in the registry record; they do not by themselves establish that the treatment strategy caused or prevented any event.
The registry lists serious adverse event terms including Stroke in evolution (10.0%), Cerebrovascular accident (5.6%), Pneumonia aspiration (5.6%), Sepsis (4.4%), Brain oedema (4.3%), and Haemorrhagic transformation stroke (3.3%). The record does not provide arm-specific counts for these terms in the data, nor does it explain whether participants could contribute more than one event term. These percentages should not be interpreted as a comparative treatment effect.
Safety counts depend on definitions, observation periods, and how events are collected and attributed. Further details on those procedures are not reported in the registry record. The counts are useful descriptions of the posted safety experience, but should be read alongside the limitations in the reporting rather than as a complete accounting of comparative risk.
13. What the Result Does, and Does Not, Mean
The registry record reports a randomized comparison in a selected population with acute ischemic stroke and reports a Day 90 primary outcome on the mRS. The arm medians and the reported Cochran-Mantel-Haenszel analysis describe the observed outcome distributions and their statistical comparison. The confidence interval conveys uncertainty around the reported odds ratio, and the p-value reflects evidence against the null hypothesis used in that analysis. These statistics do not state the probability that a particular patient will have a particular outcome.
The registry labels the primary analysis as superiority. That designation describes the statistical conclusion recorded for the trial’s comparison; it does not, by itself, quantify clinical importance or establish how the result applies outside the selected study population. The odds-ratio coding direction and full model details are not reported, which limits a more specific interpretation of its magnitude. The endpoint summaries and the inferential analysis should be considered together, while recognizing that the summaries cannot capture every feature of the score distributions.
The secondary functional-independence outcome is a threshold-based summary of the same scale, not an independent replication. Its unadjusted result is compatible in broad direction with the primary analysis, but an unadjusted secondary finding carries less confirmatory weight when the testing strategy and multiplicity adjustment are not reported.
The results do not establish that every participant benefited, describe outcomes beyond the stated follow-up time, or resolve how missing observations were handled. Nor do the adverse-event counts alone establish causation or a comparative safety effect. The registry record does not report the reason for termination, the sample-size justification, or several other details needed to assess the full design and analysis plan. Interpretation should therefore stay close to the posted population, endpoints, and analyses.
14. Why This Trial Matters Statistically
Statistically, this trial illustrates the value of an ordinal endpoint when outcomes range across levels of disability. The mRS distribution can reflect changes across multiple levels rather than only a shift across one chosen threshold. Reporting medians and inter-quartile ranges gives readers a clear descriptive starting point, while an inferential analysis addresses the group comparison. Those summaries and the model-based estimate answer related but distinct questions.
The posted secondary endpoint illustrates the trade-off involved in dichotomizing an ordinal measure. A threshold such as mRS 0-2 produces a straightforward count of participants meeting a defined criterion, but it does not retain all the distinctions in the full scale. The secondary analysis can be read alongside the primary result, while preserving the distinction between the endpoint definitions and their inferential status.
The record also demonstrates why analysis details matter. An odds ratio for an ordinal outcome needs a clear coding direction and model description to support precise interpretation. The registry identifies a Cochran-Mantel-Haenszel method, but does not describe the specific implementation. The reported estimate can still be reported faithfully, while acknowledging the limits on reconstructing its meaning from the record alone.
Finally, the trial’s termination status, absent sample-size justification, and unreported details about missing data and multiplicity highlight the difference between a posted result and a fully documented statistical analysis plan. The secondary planning illustration is not the trial’s design calculation. Keeping observed estimates, planning assumptions, and unreported design decisions distinct is essential for a sound statistical reading.
15. Limitations
- The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
- Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.
Several trial-specific details needed for a fuller interpretation are not reported in the registry record. The reason the trial was terminated is absent, as are details of interim analyses and any stopping rules. The record does not provide the original sample-size justification or explain whether the planned enrollment changed. The planning illustration based on the observed secondary effect should not be treated as evidence of the original design assumptions.
The record identifies the primary population as intention to treat but does not explain the handling of missing Day 90 outcomes. It does not describe imputation, sensitivity analyses, or the number with observed outcome data at each time point beyond the posted arm summaries. Consequently, the effect of missing observations on the estimates cannot be assessed from the registry information alone.
The primary method is named, but the registry does not describe the exact model, any stratification variables, or the mRS coding convention for the odds ratio. That limits interpretation of the odds ratio’s direction and impedes independent reproduction. The multiplicity strategy is also not reported, so the confirmatory status of secondary tests cannot be established.
Generalizability is bounded by the described eligibility profile: patients with a Target Mismatch Profile and an ICA or MCA M1 occlusion, with treatment initiation between 6-16 hours after last seen well. The registry does not report additional detail about which groups were excluded or how closely the enrolled population reflected the target population. Safety reporting is similarly limited: event counts and listed terms are available, but event collection procedures, attribution, and arm-specific details for the listed terms are not described in the information summarized.
16. Related Tutorials
17. Related Statistical Calculators
18. Sources
- ClinicalTrials.gov: Endovascular Therapy Following Imaging Evaluation for Ischemic Stroke 3 — NCT02586415.
- Registry record last updated 2019-05-21; status last verified 2019-05.
- Results first posted 2018-09-19.
- Data retrieved from ClinicalTrials.gov on 2026-11-09.
- Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.
Continue through the Clinical Biostats statistical pathway
Use this trial's endpoint as a starting point for deeper study of effect measures, confidence intervals, and sample-size planning.
19. Record Summary
NCT02586415 is a randomized, parallel-group trial listed as terminated, with enrollment of 182 and no masking. It compared endovascular thrombectomy therapy plus standard medical therapy with standard medical therapy alone in selected patients with acute ischemic stroke. The registry lists the primary outcome as the Day 90 mRS distribution in an intention-to-treat population.
The reported primary analysis is a Cochran-Mantel-Haenszel odds ratio of 2.77, with a 95% confidence interval from 1.63 to 4.70 and p < 0.001, labeled as superiority. A secondary endpoint reports functional independence using mRS scores of 0-2. The record does not report the reason for termination, detailed missing-data methods, multiplicity handling, or the full primary model specification. Those gaps constrain interpretation of the posted results.