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 SELECT2: A Randomized Controlled Trial to Optimize Patient's Selection for Endovascular Treatment in Acute Ischemic Stroke (NCT03876457), a randomized na study. The primary outcome is Degree of Disability/Dependence as Measured by the Modified Rankin Scale (mRS) Score.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT03876457 |
| Phase | NA |
| Condition | Acute Ischemic Stroke |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | Single |
| Enrollment | 352 |
| Arms in the primary outcome | 2 |
| Lead sponsor | University Hospitals Cleveland Medical Center |
| Results posted | Yes |
2. Clinical Question
Population
Adults with acute ischemic stroke due to a large-vessel occlusion in the distal internal carotid artery or middle cerebral artery M1, with a large core identified on CT, advanced perfusion imaging, or both. The registry summary describes selection by CT ASPECTS of 3 to 5, or by perfusion criteria of rCBF <30% on CT perfusion or ADC <620 on MRI with a volume of at least 50cc. Treatment was to occur within 0 to 24 hours from last known well.
Intervention
Endovascular thrombectomy plus medical management. The record identifies the thrombectomy intervention as a device.
Comparator
Medical management alone. The registry record does not provide further arm-specific treatment details.
Primary question
At 90 days, what degree of disability or dependence, measured by the modified Rankin Scale, was reported in each randomized group? The posted primary results give a median score and a 95% confidence interval for each group. They do not report a direct between-group estimate or test for this endpoint.
SELECT2 was a randomized comparison intended to evaluate the efficacy and safety of thrombectomy compared with medical management alone in the described stroke population. The clinical question is specifically about people with large-core stroke who met the registry’s imaging and treatment-window criteria; the results should not automatically be extended to people with other stroke presentations or selection criteria. The registry reports 352 participants enrolled. It does not report additional population characteristics in the primary outcome description, so the relevance of these results to particular subgroups cannot be assessed from that description alone.
3. Trial Design
SELECT2 is recorded as a completed, randomized, parallel-group trial with single masking. The registry lists University Hospitals Cleveland Medical Center as the sponsor and the phase as NA. Enrollment was 352. The registry dates are 2019-10-11 for study start and 2022-11-20 for primary completion.
The comparison was endovascular thrombectomy plus medical management versus medical management alone. The primary endpoint was the modified Rankin Scale score at 90 days. The primary results list 178 participants in the thrombectomy-plus-medical-management group and 174 in the medical-management group. These are the reported group sizes for the primary result; they should not be assumed to be the denominator for every secondary endpoint, where the posted denominators differ.
Random allocation is relevant because, in a well-conducted trial, it can make treatment groups comparable on average with respect to measured and unmeasured baseline factors. The record’s designation of single masking does not identify who was masked, and further masking details are not reported in the registry record. The registry record also does not provide arm-specific descriptions beyond the intervention names, so details of medical management and the procedural protocol cannot be described from this record.
The study is marked completed. The record includes primary and secondary outcome results, but the outcome data do not include reported comparative analyses for the primary endpoint. The record does not report the protocol’s sample-size justification, an interim-analysis plan, or enough information to establish which analysis population was intended for each endpoint. Those omissions limit assessment of how closely the posted outcome results correspond to the trial’s planned statistical analysis.
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.
Overall Study
| Stage | Endovascular Thrombectomy Plus Medical Management | Medical Management |
|---|---|---|
| STARTED | 178 | 174 |
| COMPLETED | 177 | 171 |
| NOT COMPLETED | 1 | 3 |
Reasons for not completing this period:
| Reason | Endovascular Thrombectomy Plus Medical Management | Medical Management |
|---|---|---|
| Lost to Follow-up | 1 | 3 |
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 Plus Medical Management | Medical Management | Total |
|---|---|---|---|---|
| Participants analysed | 178 | 174 | 352 | |
| Age, Continuous | years | 65.1 (SD 12.7) | 65.3 (SD 12.6) | 65.2 (SD 12.6) |
| Sex: Female, Male: Female | Participants | 71 | 74 | 145 |
| Sex: Female, Male: Male | Participants | 107 | 100 | 207 |
| Ethnicity (NIH/OMB): Hispanic or Latino | Participants | 29 | 31 | 60 |
| Ethnicity (NIH/OMB): Not Hispanic or Latino | Participants | 136 | 134 | 270 |
| Ethnicity (NIH/OMB): Unknown or Not Reported | Participants | 13 | 9 | 22 |
| Race (NIH/OMB): American Indian or Alaska Native | Participants | 0 | 1 | 1 |
| Race (NIH/OMB): Asian | Participants | 5 | 3 | 8 |
| Race (NIH/OMB): Native Hawaiian or Other Pacific Islander | Participants | 2 | 0 | 2 |
| Race (NIH/OMB): Black or African American | Participants | 26 | 24 | 50 |
| Race (NIH/OMB): White | Participants | 132 | 130 | 262 |
| Race (NIH/OMB): More than one race | Participants | 0 | 0 | 0 |
| Race (NIH/OMB): Unknown or Not Reported | Participants | 13 | 16 | 29 |
6. Primary Endpoint
The registry lists the primary outcome as Degree of Disability/Dependence as Measured by the Modified Rankin Scale (mRS) Score, measured over 90 days.
| Element | Registry entry |
|---|---|
| Result type | Median |
| Measure of spread | 95% Confidence Interval |
| Unit | scores on a scale |
| Endpoint classification | not reduced to a single comparison |
The primary endpoint is degree of disability or dependence on the modified Rankin Scale at 90 days. This scale runs from 0 to 6: 0 denotes no symptoms and 6 denotes death, with intermediate scores describing increasing limitations in daily activities and mobility. It is widely used after stroke because it captures a broad range of functional outcomes rather than focusing on a single symptom or clinical event.
The registry reports the endpoint as a median with a 95% confidence interval in each group. The thrombectomy-plus-medical-management group had a median score of 4, with a 95% confidence interval from 3 to 6. The medical-management group had a median score of 5, with a 95% confidence interval from 4 to 6. Lower scores correspond to less disability on this scale.
A median conveys the middle of the observed score distribution, but it does not show how participants were distributed across all seven possible scores. In particular, the same median can arise from different distributions of disability. The registry record does not report a direct treatment-effect estimate, confidence interval, or test for the primary comparison. Therefore, the group medians describe the posted results, but do not by themselves quantify the uncertainty in the difference between groups.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The posted primary result places the median 90-day mRS score at 4 in the thrombectomy-plus-medical-management group and 5 in the medical-management group. Because lower mRS scores indicate less disability, the group medians are descriptively different in the direction of a lower score in the thrombectomy group. The confidence intervals reported for the medians are 3 to 6 and 4 to 6, respectively. These are intervals for the group-level median estimates, not a confidence interval for the difference between treatment groups.
The registry record does not report a primary between-group estimate, its confidence interval, or a p-value. It is therefore not possible to use the posted primary statistics alone to assess the statistical uncertainty around a treatment contrast or to establish the result of a prespecified hypothesis test. Overlap or non-overlap of separate group confidence intervals is not a substitute for a direct comparison. The median also compresses an ordinal outcome into a single summary and does not reveal the full distribution of scores or how many participants experienced each level of disability.
Several secondary outcomes describe binary events or thresholds on the mRS. These are clinically interpretable summaries, but they do not replace the primary endpoint’s broader ordinal scale. Posted unadjusted comparisons include differences for functional independence and independent ambulation, while the symptomatic intracranial hemorrhage comparison is much less precise. The registry lists p < 0.001 for each of the two functional threshold comparisons and p = 0.549 for symptomatic intracranial hemorrhage. These results should be read as separate, unadjusted secondary comparisons, not as a single confirmatory analysis or as evidence that all outcomes have the same degree of certainty.
The primary and secondary results also use different denominators. The primary outcome lists 178 and 174 participants, while the functional threshold outcomes list 177 and 171. The reasons for these differences are not reported in the registry record. Without further information about missing outcome data and analysis populations, the effect of those differences on the estimates cannot be assessed.
Overall, the registry’s primary statistics describe a lower median mRS score in the thrombectomy group, but leave the direct inferential comparison unreported. The secondary estimates add descriptive information about selected outcomes, yet their unadjusted status, endpoint-specific denominators, and differing precision matter when weighing them. The record does not establish that every analysis was prespecified or that the posted comparisons account for multiplicity.
9. Statistical Methods Explained
Statistical methods are easiest to interpret by separating summaries within each group from estimates comparing groups. The primary endpoint is reported as a median and a 95% confidence interval for each group. The registry does not report a primary estimate of the between-group difference, a test statistic, or a p-value. Consequently, the posted primary summaries should not be mistaken for a complete inferential comparison.
What does a median describe?
The mRS is ordinal: its scores have an ordered meaning, but the gaps between adjacent scores are not necessarily equal in practical impact. A median identifies the midpoint of the observed ordered scores. It can communicate the central level of disability without assuming that the distance from one score to the next is constant. However, it omits information about the full distribution. Two groups can have the same median but different numbers of participants at the favorable and unfavorable ends of the scale.
The 95% confidence interval expresses uncertainty about the estimated group median under the method used to calculate it. It is not a range containing 95% of individual participant scores, and it is not the interval for a contrast between the groups. A direct comparison would require an estimate and uncertainty measure for the treatment contrast, ideally using the analysis specified for the ordinal endpoint. Such a primary comparison is not reported in the registry record.
How should binary comparisons be read?
For secondary outcomes reported as participant counts, a risk difference compares the event proportions between groups. Its sign depends on the order in which groups are subtracted. For a favorable outcome, fewer events in the medical-management group than in the thrombectomy group can therefore appear as a negative estimate when the comparison is expressed in that direction. The sign does not itself indicate benefit or harm without knowing the event definition and group order. A confidence interval describes the uncertainty around the risk difference; when it includes zero, the data are compatible with no difference on that scale under the stated method.
The registry’s secondary comparisons are marked unadjusted. Their p-values are not corrections for the number of outcomes examined, and they do not account for the possibility that selecting among multiple endpoints can produce chance findings. These results are useful for description and hypothesis generation, but the record does not establish them as confirmatory analyses.
Why does precision matter?
Precision depends on both the amount of information and the frequency of the event. Rare outcomes can yield unstable comparisons even when the group sizes are substantial, because only a few participants experience the event. The symptomatic intracranial hemorrhage estimate has a confidence interval spanning values on both sides of no difference, and the unadjusted p-value is 0.549. This reflects substantial uncertainty in that comparison; it does not establish equivalence between groups.
Confidence intervals also help distinguish an estimated effect from the range of effects that remain compatible with the data under the analysis. A narrow interval can be more informative than a broad one, but precision alone does not resolve bias, missing-data concerns, or the importance of the outcome. For the primary endpoint, the registry gives intervals around group medians, but does not give an interval for the difference in mRS between groups. The amount of uncertainty in the main treatment contrast therefore cannot be read directly from the posted primary result.
What do the planning illustrations mean?
The registry’s planning illustrations use the observed effect as a starting point and show hypothetical participant counts per arm for stated power targets under assumptions of a two-sided significance level of 0.05, equal allocation, and a normal approximation. Rows that plan for 75% or 50% of the observed effect explore smaller effects than the observed estimate. The resulting counts are planning illustrations, not evidence of the trial’s actual design requirements and not observed power. The registry’s own sample-size justification is not reported in the registry record.
These calculations depend on the assumed effect size and statistical model. For binary outcomes, event frequency affects the amount of information available; a rare event can require very large planned groups to detect a modest difference. A planning calculation based on an observed effect is especially uncertain because that estimate can be imprecise and may not represent the effect expected in a future study. The illustrations should therefore be interpreted as conditional scenarios, not as a retrospective test of whether SELECT2 was adequately powered.
The record lists no reported analyses for the posted secondary comparisons and does not provide the analysis plan, missing-data approach, or definitions of the analysis populations. Without those details, it is not possible to confirm whether the displayed estimates follow a prespecified method, whether covariates were included elsewhere, or how the primary ordinal outcome was analyzed for a formal treatment comparison.
10. Multiplicity, Missing Data and Analysis Populations
The registry record lists 21 outcome measures and 14 posted secondary outcomes, but it does not report a plan for controlling multiplicity. When several outcomes or analyses are examined, the chance of seeing at least one apparently strong result by chance can increase. Methods such as a prespecified hierarchy or an adjustment across tests can address that issue, but neither a procedure nor a testing order is reported in the registry record. Accordingly, the unadjusted secondary p-values should not be treated as if they were adjusted confirmatory evidence.
The record lists zero reported analyses for the primary and secondary results in the posted analysis fields, while providing some outcome summaries and simple unadjusted comparisons. It does not report whether additional analyses were conducted outside the registry, which model was planned for the primary mRS outcome, or whether the posted comparisons were prespecified. The protocol’s statistical analysis plan and interim-analysis plan are not reported in the registry record. Their absence means that readers cannot check whether the analyses were selected before the results were known.
Denominators vary by outcome. The primary endpoint includes 178 and 174 participants, whereas the two 90-day functional threshold outcomes include 177 and 171. Symptomatic intracranial hemorrhage uses 178 and 174. The reasons for differing denominators, the extent and handling of missing data, and any imputation or sensitivity analyses are not reported in the registry record. It is not possible to determine whether missingness was related to outcome or treatment assignment from the posted information.
The registry also does not state which analysis population was used for each result, such as whether participants were analyzed according to randomized assignment or another rule. Randomization supports the validity of a comparison when the analysis preserves the randomized groups, but the record does not give sufficient detail to assess that aspect of the posted results. These limitations do not establish that the analyses were inappropriate; they identify information needed for a fuller appraisal.
11. 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 Plus Medical Management | 129 | 178 | 72.5% |
| Medical Management | 123 | 174 | 70.7% |
Deaths
| Arm | Deaths (participants affected of at risk) |
|---|---|
| Endovascular Thrombectomy Plus Medical Management | 68 of 178 |
| Medical Management | 71 of 174 |
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 Plus Medical Management | 77 of 178 |
| Medical Management | 59 of 174 |
Most frequent serious adverse event terms
| Event term | Organ system | Endovascular Thrombectomy Plus Medical Management | Medical Management |
|---|---|---|---|
| Cerebral Edema with Midline Shift | Nervous system disorders | 64 of 178 | 57 of 174 |
| Neurological Deterioration | Nervous system disorders | 44 of 178 | 27 of 174 |
| Progression of Disease, Palliative Care Decision | Nervous system disorders | 25 of 178 | 31 of 174 |
| Respiratory failure | Reproductive system and breast disorders | 30 of 178 | 19 of 174 |
| Lower respiratory infection (pneumonia) | Respiratory, thoracic and mediastinal disorders | 18 of 178 | 21 of 174 |
| Urinary Tract Infection | Renal and urinary disorders | 12 of 178 | 13 of 174 |
| Sepsis, Systemic inflammatory response syndrome | Infections and infestations | 11 of 178 | 9 of 174 |
| Acute kidney injury | Renal and urinary disorders | 9 of 178 | 9 of 174 |
| Heart failure | Cardiac disorders | 7 of 178 | 8 of 174 |
| Delirium, Cognitive Dysfunction | Nervous system disorders | 9 of 178 | 5 of 174 |
Most frequent non-serious adverse event terms
| Event term | Organ system | Endovascular Thrombectomy Plus Medical Management | Medical Management |
|---|---|---|---|
| Cardiac Arrythmia | Cardiac disorders | 21 of 178 | 20 of 174 |
| Gastrointestinal Complications | Gastrointestinal disorders | 19 of 178 | 8 of 174 |
| Progression of Disease | Nervous system disorders | 16 of 178 | 11 of 174 |
| Other Renal Complications | Renal and urinary disorders | 14 of 178 | 7 of 174 |
| Evolution of infarction on the brain imaging | Nervous system disorders | 8 of 178 | 8 of 174 |
| Fever | General disorders | 7 of 178 | 9 of 174 |
| Musculoskeletal Complications | Musculoskeletal and connective tissue disorders | 9 of 178 | 6 of 174 |
| Anemia | Blood and lymphatic system disorders | 8 of 178 | 4 of 174 |
| Headache | General disorders | 9 of 178 | 2 of 174 |
| Skin Complications | Skin and subcutaneous tissue disorders | 10 of 178 | 1 of 174 |
Serious events with the highest arm-level rate: cerebral edema with midline shift (36.0%), neurological deterioration (24.7%), progression of disease, palliative care decision (17.8%), respiratory failure (16.9%), lower respiratory infection (pneumonia) (12.1%).
The registry reports serious adverse-event counts for both groups: 129 among 178 participants in the thrombectomy-plus-medical-management group and 123 among 174 participants in the medical-management group. It also reports other adverse-event counts of 77 among 178 and 59 among 174, respectively, and deaths of 68 and 71. These counts describe recorded events or participants as listed in the registry; the record does not provide a comparative statistical analysis for these safety summaries.
The most frequently listed serious adverse-event terms include cerebral edema with midline shift, neurological deterioration, progression of disease with a palliative-care decision, respiratory failure, lower respiratory infection (pneumonia), and urinary tract infection. The registry presents percentages for these terms, but it does not provide the group-specific breakdown for each term in the information. The terms may overlap, and the counts should not be assumed to represent mutually exclusive categories unless the registry specifies that structure.
Event counts alone do not establish a difference in safety between groups. Interpretation depends on definitions, time at risk, severity, attribution, completeness of reporting, and whether participants can contribute more than one event. Comparative confidence intervals and adjusted analyses for these safety data are not reported in the registry record.
12. What the Result Does, and Does Not, Mean
The primary result describes the 90-day mRS medians in the randomized groups and shows a lower median score in the thrombectomy-plus-medical-management group than in the medical-management group. This is a descriptive contrast on an ordered disability scale. It does not, by itself, provide a precise estimate of the treatment effect or the statistical uncertainty around the between-group difference.
The registry record does not report a direct primary comparison, a confidence interval for that comparison, or a primary p-value. As a result, the posted primary endpoint information cannot answer how compatible the observed group difference is with a range of possible treatment effects under a specified model. Separate confidence intervals around the group medians do not fill that gap. The record also does not show the full mRS score distribution, so the group-level medians cannot describe the proportions at every disability level.
The secondary results offer additional descriptions of selected functional and safety outcomes. Their interpretation is constrained by unadjusted testing, endpoint-specific denominators, and incomplete reporting of analysis methods and missing data. The functional threshold comparisons are consistent in broad direction with the lower median, but they are not substitutes for a reported primary ordinal comparison. The imprecise symptomatic intracranial hemorrhage comparison does not establish that the groups are equivalent on that outcome.
These registry results apply to the trial population described in the record: patients with acute ischemic stroke, the specified large-vessel occlusion and large-core imaging features, and treatment within the stated window. The record does not provide enough information to assess how the findings vary across patient subgroups or how fully the enrolled population represents people outside those eligibility features.
In short, the results describe what was posted for the randomized groups. They do not establish a universal treatment effect, resolve all questions about safety, or provide the planned primary inferential analysis. Further interpretation requires the complete statistical analysis and fuller outcome reporting, neither of which is reported in the registry record.
13. Why This Trial Matters Statistically
SELECT2 raises a statistically important question about how to evaluate interventions in acute ischemic stroke when imaging identifies a large core. The trial’s population is defined not only by the diagnosis and vessel location, but also by imaging features and a treatment window. Such selection criteria make it essential to describe the studied population precisely when interpreting results or considering how they relate to other settings.
The primary outcome uses the mRS, an ordered scale that captures a broad range of disability. Reporting medians is straightforward, but a median alone cannot show the full distribution or quantify a between-group effect. For ordinal outcomes, the choice of summary and model matters: analyses can use the ordering of all scores, whereas a single median or threshold discards some of that information. The registry record does not report the planned primary comparison, making this a useful example of why endpoint summaries and inferential analyses are not interchangeable.
The secondary endpoints also illustrate why statistical evidence needs to be read across the full outcome set. Threshold-based functional measures can complement an ordinal endpoint, while rare safety events may remain uncertain because few events contribute information. Unadjusted tests across multiple outcomes require caution, especially when the analysis plan and multiplicity approach are not reported.
Finally, planning calculations based on observed effects should not be confused with the trial’s actual design justification or with observed power. Their purpose is to show how assumptions about effect size influence hypothetical sample-size requirements. The registry does not report the study’s own sample-size justification, and its planning illustrations cannot resolve whether the trial was designed around the primary analysis that readers would want to interpret.
14. 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.
The main limitation for statistical interpretation is that the registry provides group medians and confidence intervals for the primary outcome but no direct estimate or test of the between-group difference. The mRS median does not show the full distribution of disability, and the reported confidence intervals do not describe uncertainty in the treatment contrast.
The record does not report the primary analysis method, statistical analysis plan, sample-size justification, interim analyses, or analysis populations. It also does not report the approach to missing data or explain why some secondary outcome denominators are smaller than the primary group sizes. These omissions prevent assessment of prespecification, robustness to missing observations, and whether the posted estimates preserve the randomized comparison.
The secondary comparisons are unadjusted, and a multiplicity strategy is not reported. Several outcomes are listed, but comparative results are not available for all of them. In particular, the available results should not be interpreted as a complete analysis of all secondary endpoints. The safety summaries provide counts and serious adverse-event terms but do not supply comparative estimates or detailed event-level definitions in the registry record.
Some outcome definitions depend on thresholds, which can simplify interpretation but discard distinctions across the full mRS scale. The rare symptomatic intracranial hemorrhage outcome has little precision in the posted comparison, so the absence of a small p-value cannot be taken as evidence of no difference. Conversely, the unadjusted functional comparisons do not resolve the missing primary inferential result.
Finally, the registry summary describes a particular population with large-vessel occlusion, large-core imaging features, and a specified treatment window. The posted information does not support conclusions about people outside those criteria or about differences within subgroups. These limitations concern what can be concluded from the registry record; they do not determine what additional analyses may appear in other trial documents.
15. Related Tutorials
16. Related Statistical Calculators
17. Sources
- ClinicalTrials.gov: SELECT2: A Randomized Controlled Trial to Optimize Patient's Selection for Endovascular Treatment in Acute Ischemic Stroke — NCT03876457.
- Registry record last updated 2024-11-21; status last verified 2024-11.
- Results first posted 2024-03-15.
- 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.
18. Record Summary
SELECT2 is a completed, randomized, parallel-group, single-masked trial sponsored by University Hospitals Cleveland Medical Center. The registry reports 352 participants enrolled and compares endovascular thrombectomy plus medical management with medical management alone in acute ischemic stroke involving a large-vessel occlusion and large core.
The primary endpoint is 90-day mRS. The posted medians are 4, with a 95% confidence interval from 3 to 6, in the thrombectomy group and 5, with a 95% confidence interval from 4 to 6, in the medical-management group. A direct primary comparison is not reported in the registry record. Secondary results include functional thresholds and symptomatic intracranial hemorrhage, but the comparisons are unadjusted and the registry does not report a multiplicity plan, missing-data approach, or analysis populations.