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
BASICS was a randomized, parallel-group clinical study evaluating intra-arterial treatment in the setting of basilar artery thrombosis, embolism, occlusion, and related stroke conditions. The registry identifies favourable outcome at day 90 as the primary endpoint, defined using the modified Rankin Score.
| Feature | BASICS |
|---|---|
| Official title | Basilar Artery International Cooperation Study |
| Study acronym | BASICS |
| ClinicalTrials.gov identifier | NCT01717755 |
| Phase | NA |
| Status | UNKNOWN |
| Enrollment | 282 |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | None |
| Primary purpose | Treatment |
| Number of arms | 2 |
| Intervention | Intra-arterial treatment (other) |
| Lead sponsor | Erik van der Hoeven |
| Sponsor type | Other |
2. Clinical Question
The central clinical-trial question is whether a randomized comparison involving intra-arterial treatment can be associated with a difference in functional outcome at day 90 among participants with basilar artery thrombosis, basilar artery embolism, stroke of the basilar artery, stroke, cerebrovascular disorders, or basilar artery occlusion.
Population
Participants enrolled under conditions including basilar artery thrombosis, basilar artery embolism, stroke of the basilar artery, stroke, cerebrovascular disorders, and basilar artery occlusion.
Intervention
Intra-arterial treatment (other).
Comparator
The registry identifies 2 randomized arms, but the registry information available here does not specify the treatment description for the comparator arm.
Primary question
Does the randomized treatment comparison differ with respect to favourable functional outcome at day 90?
3. Trial Design
Randomization is important statistically because it creates the basis for comparing the treatment groups as assigned rather than attempting to construct a treatment comparison from observational differences. The parallel design means that the randomized groups are followed as distinct treatment pathways rather than being assigned sequentially or in a crossover structure.
4. Trial Timeline
Study start
The registry lists October 2011 as the study start date.
Primary completion
The registry lists January 2019 as the primary completion date.
The registry lists the study status as UNKNOWN. The available registry record therefore supplies the study timing and design information above but does not establish a current recruitment or completion status beyond that recorded status.
5. Primary Endpoint
| Endpoint | Registry definition | Time frame |
|---|---|---|
| Favourable outcome | Favourable outcome at day 90 defined as a modified Rankin Score (mRS - functional scale) of 0-3. | Day 90 |
The primary endpoint is a binary functional outcome derived from the modified Rankin Score: participants with an mRS of 0-3 at day 90 are classified as having a favourable outcome, while participants outside that range are not classified as favourable for this endpoint.
This endpoint converts an ordinal functional scale into a clinically interpretable binary outcome. Statistically, that choice determines the natural primary comparison: the proportion of participants achieving the favourable-outcome definition in each randomized arm.
6. Planned Analysis
No statistical analyses are posted in the ClinicalTrials.gov record available for this analysis. The registry identifies favourable outcome at day 90 as the primary endpoint, but does not provide an observed treatment effect, confidence interval, or p-value for that endpoint.
Primary endpoint: favourable outcome
Favourable outcome is defined as modified Rankin Score (mRS) 0-3.
No treatment-effect estimate, confidence interval, or p-value is reported in the registry.
How this endpoint would typically be analysed
Because the primary endpoint is binary, a standard analysis would compare the proportion of participants with a favourable outcome between the randomized arms. Depending on the prespecified statistical analysis plan, this could be expressed using an absolute risk difference, a risk ratio, an odds ratio, or a regression model estimating one of these measures.
A two-group comparison can be represented conceptually as:
The treatment comparison can then be accompanied by a confidence interval and a hypothesis test, with the exact estimand and model determined by the prespecified analysis plan.
The randomized design provides the principal basis for causal interpretation. However, the appropriate analysis population, treatment of missing day-90 assessments, adjustment variables, and inferential framework should come from the prespecified statistical analysis plan. Those details are not reported in the registry information available here.
7. Statistical Methodology
Binary functional outcome
The primary endpoint is defined as favourable versus not favourable according to the day-90 mRS threshold. This is a binary endpoint even though the underlying modified Rankin Score is a multi-level functional scale.
For a randomized two-arm trial, the simplest descriptive analysis is to report the favourable-outcome proportion separately by randomized arm. The principal comparative measure can then be chosen according to the trial's estimand and analysis plan.
These measures describe the same binary outcome in different ways. They are not interchangeable and can have materially different numerical interpretations, particularly when the outcome is common.
Confidence intervals
A confidence interval around a treatment-effect estimate quantifies statistical uncertainty under the specified sampling and modeling framework. It does not describe the range of individual patient outcomes and does not establish that every patient would experience an effect of the estimated magnitude.
Hypothesis testing
A hypothesis test would address whether the observed difference is compatible with a prespecified null hypothesis under the chosen statistical model. A p-value is therefore evidence about compatibility with the null hypothesis; it is not a measure of clinical importance or effect size.
Randomization and causal interpretation
Randomization is the key design feature supporting a causal treatment comparison. If the randomized groups are analysed according to their assigned treatment, the comparison is protected from many forms of confounding that affect non-randomized studies.
Modified Rankin Score thresholding
The registry defines favourable outcome specifically as mRS 0-3 at day 90. The analysis therefore targets the probability of crossing that prespecified functional threshold rather than estimating a mean mRS or treating every adjacent change in the ordinal scale as numerically equivalent.
8. Statistical Methods Explained
Why is the primary endpoint binary?
The modified Rankin Score is a functional scale, but the registry defines the primary outcome as a favourable score of 0-3 at day 90. Once that threshold is applied, each participant contributes one of two endpoint states: favourable or not favourable.
What would a risk difference mean?
If the favourable-outcome probabilities were pT and pC, the risk difference would be pT − pC. A positive value would indicate a higher observed probability of favourable outcome in the intervention group; a negative value would indicate a lower observed probability. The registry does not report the values needed to calculate this quantity for BASICS.
What would a risk ratio mean?
A risk ratio compares the probability of favourable outcome between the randomized groups. A value of 1 would represent equal probabilities, while values above or below 1 would indicate relative differences in the probability of favourable outcome. The risk ratio should not be confused with an odds ratio.
Why might an odds ratio be used?
Logistic regression naturally models a binary endpoint through the odds of the event and can incorporate prespecified covariates if appropriate. Its odds ratio is a relative measure of odds, not a direct measure of relative probability. When favourable outcomes are common, the numerical difference between odds ratios and risk ratios can be substantial.
Why does the day-90 time frame matter?
The time frame defines when the functional status is assessed for the primary endpoint. Changing the assessment time could change the number of participants classified as favourable, so the day-90 definition is part of the estimand rather than merely an administrative detail.
Why is the mRS threshold important?
The threshold determines which functional states count as favourable. An analysis of the full ordinal mRS distribution would answer a different statistical question from the registered binary endpoint of mRS 0-3 at day 90.
9. What the Primary Endpoint Does — and Does Not — Mean
A favourable outcome means that the participant's modified Rankin Score at day 90 falls within the registry-defined range of 0-3. It is a functional endpoint rather than a direct measure of survival, infarct volume, vessel recanalization, or another clinical characteristic.
The endpoint does not distinguish among the individual mRS values within the favourable range. For this endpoint, an mRS of 0 and an mRS of 3 are both classified as favourable, while the statistical analysis of the registered endpoint treats them as the same binary outcome category.
If a comparative analysis were reported, the effect estimate would describe the magnitude and direction of the difference between randomized groups. The confidence interval would describe uncertainty around that estimate, while the p-value would address the prespecified hypothesis test. These quantities answer different questions.
The ClinicalTrials.gov record does not provide the favourable-outcome counts or rates by randomized arm, nor does it post a confidence interval or p-value for the primary endpoint.
10. Analysis Population and Missing Day-90 Assessments
Randomization establishes the treatment groups, but interpretation of a day-90 endpoint also depends on how participants with missing or unavailable outcome assessments are handled. The registry information available here does not specify the analysis population, the number of participants with missing day-90 mRS assessments, or the imputation method, if any.
These details can matter for a binary functional endpoint. For example, excluding participants with missing outcomes can change the denominator and potentially change the interpretation of the comparison. An intention-to-treat analysis generally preserves the randomized comparison, but the precise missing-data strategy still needs to be specified.
| Statistical issue | Why it matters | Registry information |
|---|---|---|
| Randomized population | Defines the treatment groups created by randomization. | 282 participants enrolled; allocation randomized. |
| Day-90 assessment | Determines whether the participant meets the registered endpoint definition. | Primary endpoint assessed at day 90. |
| Missing mRS | Can affect denominators and the estimated favourable-outcome proportion. | Not specified. |
| Imputation | Different assumptions about missing outcomes can produce different estimates. | Not specified. |
| Analysis population | Determines which participants enter the formal endpoint analysis. | Not specified. |
11. Stratification and Covariate Adjustment
Randomized trials sometimes stratify allocation or prespecify covariate adjustment to improve precision or account for important baseline characteristics. The registry information available here does not report any stratification factors or prespecified covariate-adjustment strategy for BASICS.
That distinction is important: randomization itself supplies the primary design protection against confounding, whereas regression adjustment is an additional analytical choice. A covariate-adjusted estimate should not be presented as though it were part of the registered analysis unless the statistical plan establishes that role.
12. Interim Analysis and Multiplicity
The available registry information does not describe an interim efficacy analysis, alpha-spending strategy, multiplicity adjustment, or endpoint hierarchy beyond identifying favourable outcome at day 90 as the primary endpoint.
Interim analysis
No interim-analysis method is reported in the registry information available here.
Multiplicity
No multiplicity-adjustment procedure is reported in the registry information available here.
Primary endpoint
Favourable outcome at day 90 is the registered primary endpoint.
Additional endpoints
No secondary endpoint definitions are included in the registry information available here.
This matters because the interpretation of a p-value depends partly on the testing framework. If multiple confirmatory hypotheses or repeated interim looks are part of a trial's design, the statistical procedure generally needs to account for that structure to preserve the intended type I error rate.
13. Non-Inferiority, Factorial, Crossover, and Bayesian Methods
The registry information identifies BASICS as a randomized parallel-group treatment study. It does not report a non-inferiority margin, factorial design, crossover design, or Bayesian analysis.
| Design or method | Registry information |
|---|---|
| Non-inferiority margin | Not reported. |
| Crossover | Not reported. |
| Factorial design | Not reported; the design model is identified as parallel. |
| Bayesian methods | Not reported. |
| Interim efficacy analysis | Not reported. |
| Multiplicity procedure | Not reported. |
These distinctions are statistically consequential. A non-inferiority trial requires an explicit margin and a different inferential framework from a conventional superiority comparison. A crossover design requires within-participant treatment-period considerations. A factorial design changes the interpretation of treatment effects and interactions. Bayesian analyses introduce prior distributions and posterior inference. None of those structures is identified in the registry information available here.
14. Safety and Adverse Events
The registry information available here does not provide serious adverse-event counts, adverse-event rates, treatment-emergent adverse events, or other safety estimates by randomized arm.
Safety and efficacy are analytically distinct. A functional outcome at day 90 addresses the registered efficacy endpoint, while adverse events would characterize harms associated with treatment exposure. A complete safety analysis would require arm-specific event counts, denominators, event definitions, severity categories, timing, and the relevant analysis population.
15. Limitations
- No posted statistical analysis: The ClinicalTrials.gov record does not post a formal statistical analysis for the registered primary endpoint.
- No reported treatment effect: The registry information available here does not provide favourable-outcome counts, rates, effect estimates, confidence intervals, or p-values by randomized arm.
- Limited design detail: The registry identifies randomization, parallel design, no masking, 2 arms, and 282 participants, but does not provide a detailed statistical analysis plan.
- Unknown analysis population: The registry information does not specify whether the primary endpoint analysis used an intention-to-treat population, a modified population, or another defined analysis set.
- Missing-data strategy not reported: The treatment of missing day-90 mRS assessments is not specified.
- No stratification information: The registry information does not identify randomization strata or covariates used in the primary analysis.
- No multiplicity framework reported: The available record does not specify whether any additional hypotheses were formally controlled for type I error.
- No detailed comparator description: The available intervention information identifies intra-arterial treatment (other) but does not provide the comparator arm's intervention description.
- Functional thresholding: The primary endpoint reduces the underlying mRS scale to a favourable versus non-favourable classification, which does not preserve all information in the ordinal scale.
16. Why This Trial Matters Statistically
BASICS is a useful statistical case study because its primary endpoint illustrates a common clinical-trial problem: translating an ordinal functional outcome into a prespecified binary endpoint measured at a clinically defined time point.
| Concept | How it appears in BASICS |
|---|---|
| Randomization | Participants were randomly allocated to treatment groups. |
| Parallel-group design | The registry identifies a parallel design with 2 arms. |
| Binary endpoint | Favourable outcome is defined as mRS 0-3. |
| Functional outcome | The primary endpoint uses the modified Rankin Score. |
| Fixed assessment time | The primary endpoint is assessed at day 90. |
| Effect measures | Binary outcomes can be expressed using risk differences, risk ratios, or odds ratios. |
| Confidence intervals | Would quantify uncertainty around a reported treatment-effect estimate. |
| Hypothesis testing | Would evaluate a prespecified null hypothesis for the treatment comparison. |
| Missing data | Day-90 outcome availability can affect the analysis and requires a prespecified strategy. |
| Analysis population | The interpretation depends on the population defined for the primary endpoint analysis. |
17. The Statistical Interpretation of a Binary Functional Endpoint
The most important statistical distinction in BASICS is between the underlying ordinal functional scale and the registered binary endpoint. The modified Rankin Score contains ordered categories, but the registered endpoint asks a narrower question: whether the participant's score at day 90 is within the 0-3 range.
This has two consequences. First, a binary analysis is straightforward to interpret because it estimates the probability of crossing a prespecified functional threshold. Second, information about differences within the favourable and unfavourable categories is not retained in that primary endpoint.
What the binary endpoint captures
Whether a participant has an mRS of 0-3 at day 90.
What it does not capture directly
The full distribution of individual mRS values across all functional categories.
Why absolute effects matter
An absolute difference in favourable-outcome probability is directly interpretable as a difference in the proportion reaching the prespecified functional threshold.
Why relative effects matter
A risk ratio or odds ratio can describe the relative contrast between randomized groups, but each uses a different mathematical scale.
18. Statistical Interpretation vs Clinical Interpretation
Statistical interpretation
The registered primary analysis concerns a binary functional outcome at a fixed time point. The randomized design provides the framework for comparing the probability of favourable outcome between treatment groups.
Clinical interpretation
A favourable outcome is defined specifically by the registry as an mRS of 0-3 at day 90. Its clinical meaning therefore depends on the functional state represented by that prespecified threshold and on the context of the population studied.
A statistical treatment effect and a clinical judgment are related but not identical. The statistical analysis quantifies the difference between randomized groups under the prespecified endpoint definition. Clinical interpretation considers what that difference means for functional outcomes and for the broader care context.
19. A Practical Analysis Framework for BASICS
If arm-specific day-90 outcome data and a prespecified statistical analysis plan were available, a transparent analysis would proceed in a sequence such as the following:
The order is important. Descriptive event rates establish the observed data before a relative effect is interpreted. The confidence interval should accompany the point estimate, and the p-value should be interpreted within the prespecified hypothesis-testing framework rather than treated as a standalone measure of clinical importance.
20. ClinicalTrials.gov Record and Publicly Reported Analysis
The ClinicalTrials.gov record identifies BASICS as a randomized, parallel-group study with 282 participants and a primary endpoint of favourable outcome at day 90, defined as mRS 0-3. It does not post a formal statistical analysis for that endpoint.
The registry therefore provides the core design and endpoint framework needed to understand the intended statistical question, but the numerical treatment effect for the primary endpoint is not contained in the record available here.
21. Sources
- ClinicalTrials.gov: NCT01717755 — BASICS.
- PubMed record: PubMed 34125952.
- PubMed record: PubMed 34010530.
- PubMed record: PubMed 23835026.
Continue through the Clinical Biostats trial library
Clinical trial analysis is most useful when the endpoint definition, randomized design, effect measure, uncertainty, and analysis assumptions are considered together.
22. Record Summary
BASICS is a randomized, parallel-group study with 282 participants and 2 arms, evaluating intra-arterial treatment in the setting of basilar artery thrombosis, basilar artery embolism, stroke of the basilar artery, stroke, cerebrovascular disorders, and basilar artery occlusion. The registered primary endpoint is favourable outcome at day 90, defined as a modified Rankin Score of 0-3.
Statistically, this is a binary functional endpoint derived from an ordinal scale. The randomized design provides the framework for comparing the probability of favourable outcome between treatment groups. A complete inferential analysis would normally pair arm-specific outcome rates with an appropriate effect measure, confidence interval, and prespecified hypothesis test, while also specifying the analysis population and missing-data strategy.