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
AcT was a randomized, parallel, phase 3 treatment trial evaluating tenecteplase compared with alteplase in patients with acute ischemic stroke. The registered primary endpoint was freedom from disability, defined as a modified Rankin Scale score of 0-1, assessed by telephone follow-up between 90-120 days.
| Feature | AcT |
|---|---|
| Trial acronym | AcT |
| ClinicalTrials.gov identifier | NCT03889249 |
| Title | Alteplase Compared to Tenecteplase in Patients With Acute Ischemic Stroke |
| Phase | Phase 3 |
| Status | Completed |
| Start date | 2019-12-10 |
| Primary completion date | 2022-04-26 |
| Lead sponsor | University of Calgary |
| Sponsor type | Other |
| Conditions | Stroke, Acute; Thromboses, Intracranial |
| Enrollment | 1,600 |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | Single |
| Primary purpose | Treatment |
2. Clinical Question
The central statistical question is whether patients randomized to tenecteplase differ from patients randomized to alteplase in the probability of achieving freedom from disability, defined by a modified Rankin Scale score of 0-1, by telephone follow-up between 90-120 days.
Population
Patients with acute ischemic stroke, with thromboses, intracranial also listed among the trial conditions.
Intervention
Tenecteplase.
Comparator
Alteplase.
Primary question
Does the randomized treatment assignment affect the proportion of patients with an mRS score of 0-1 by the specified 90-120-day follow-up window?
3. Trial Design
Tenecteplase
- Drug intervention
- One of the two randomized parallel treatment groups
Alteplase
- Drug intervention
- One of the two randomized parallel treatment groups
The registry identifies the study as randomized, parallel, single-masked, and phase 3. These design features establish the basic statistical framework: treatment groups are formed through randomization, the groups are followed in parallel, and the registered purpose is treatment evaluation.
4. Primary Endpoint
| Endpoint | Registry definition | Time frame |
|---|---|---|
| Modified Rankin Scale (mRS) 0-1 (freedom from disability) | The modified Rankin Scale (mRS) is a commonly used scale for measuring the degree of disability or dependence in the daily activities of people who have suffered a stroke or other causes of neurological disability. The mRS is a range from 0-6. 0=No symptoms, 1=No significant disability. Able to carry out all usual activities, despite some symptoms 2=Slight disability. Able to look after own affairs without assistance, but unable to carry out all previous activities.3=Moderate disability. Requires some help, but able to walk unassisted4=Moderately severe disability. Unable to attend to own bodily needs without assistance, and unable to walk unassisted.5=Severe disability. Requires constant nursing care and attention, bedridden, incontinent.6=Dead | By telephone Follow-up between 90-120 days |
What the endpoint measures
The primary endpoint converts the six-point mRS scale into a clinically interpretable binary outcome: whether the participant has an mRS score of 0-1. In the registry's definition, an mRS score of 0 represents no symptoms and a score of 1 represents no significant disability, with the patient able to carry out all usual activities despite some symptoms.
This endpoint is therefore different from analysing the complete six-level mRS distribution. A binary analysis focuses specifically on the proportion of participants who reach the prespecified threshold of 0-1 by the assessment window.
5. Statistical Methodology
Binary endpoint analysis
The registered primary endpoint is binary: each participant is classified according to whether the mRS score is 0-1 or is above 1 at the 90-120-day telephone follow-up. A treatment comparison for a binary endpoint commonly begins with the observed proportion achieving the endpoint in each randomized group.
The absolute risk difference can be expressed as pT − pA. Other measures, such as a risk ratio or odds ratio, describe the same binary outcome on a relative scale.
The registry record identifies the endpoint and assessment window but does not report a posted statistical analysis for the endpoint. Consequently, the record establishes what was measured but does not specify a numerical treatment effect, confidence interval, or p-value.
Why the analysis population matters
Because this is a randomized trial, the principal treatment comparison is naturally framed around randomized treatment assignment. The advantage of this approach is that randomization is the mechanism intended to balance measured and unmeasured prognostic factors between groups at baseline, subject to chance variation.
For a binary functional outcome, analysis choices can also depend on how missing outcome assessments are handled and whether the statistical model adjusts for prespecified baseline variables. Those details can affect the numerical estimate and its uncertainty and therefore need to be specified before interpreting a formal result.
Confidence intervals
A confidence interval around a treatment-effect estimate quantifies statistical uncertainty under the selected model and sampling framework. For example, a confidence interval around a risk difference describes plausible values for the underlying difference under that framework; it does not describe the range of outcomes an individual patient might experience.
P-values
A p-value addresses compatibility between the observed data and a specified null hypothesis under the statistical model. It does not measure the magnitude of the treatment effect. A small p-value can accompany a small absolute difference in a sufficiently large study, while a clinically meaningful difference can be estimated imprecisely in a smaller study.
6. Planned Analysis
The primary endpoint is the proportion of patients with an mRS score of 0-1 by telephone follow-up between 90-120 days. For an endpoint of this type, a formal comparison would typically evaluate the binary outcome between the two randomized treatment groups and report an effect estimate together with a measure of statistical uncertainty.
| Statistical component | Educational interpretation |
|---|---|
| Outcome | Whether the participant has mRS 0-1 at the specified follow-up assessment. |
| Treatment comparison | Compare the probability of mRS 0-1 between the tenecteplase and alteplase groups. |
| Absolute effect | A risk difference directly describes the percentage-point separation between the two groups. |
| Relative effect | A risk ratio compares the probabilities; an odds ratio compares the corresponding odds. |
| Uncertainty | A confidence interval quantifies uncertainty around the selected effect estimate. |
| Hypothesis test | A formal test can assess the null hypothesis of no treatment difference, with the exact test depending on the prespecified statistical plan. |
For this endpoint, the choice between a simple unadjusted comparison and an adjusted model is a methodological decision that should be tied to the prespecified analysis plan. Likewise, missing 90-120-day assessments require an explicitly defined handling strategy rather than an assumption that every enrolled participant necessarily contributes an observed primary endpoint.
7. Statistical Methods Explained
Why is mRS 0-1 a binary endpoint?
The full mRS ranges from 0 to 6, but the registered primary endpoint focuses on the threshold of 0-1. This produces a yes/no outcome: the participant either reaches the defined category of freedom from disability or does not. That structure makes proportions and binary-effect measures natural ways to summarize the primary outcome.
What does an absolute risk difference mean?
If pT is the proportion with mRS 0-1 after tenecteplase and pA is the corresponding proportion after alteplase, then the risk difference is pT − pA. It is expressed on the probability scale and directly represents the difference in the proportion meeting the endpoint.
What is the difference between a risk ratio and an odds ratio?
A risk ratio compares two probabilities directly. An odds ratio instead compares the odds of the endpoint between groups. When the outcome is common, the odds ratio can look substantially different from the risk ratio even when both describe the same underlying treatment comparison.
Why does the 90-120-day window matter?
The endpoint is not defined simply as disability status at an unspecified follow-up. The registry specifies a telephone follow-up between 90-120 days. The assessment window is therefore part of the endpoint definition and determines when the primary outcome is evaluated.
Why does randomization matter statistically?
Randomization creates the treatment groups according to an allocation mechanism rather than allowing treatment selection to be determined by patient or clinician characteristics. This supports a causal comparison of treatment assignment, although chance imbalance, protocol deviations, missing outcomes, and other design or analysis issues can still affect the final estimate.
Why can a binary endpoint lose information?
Converting the mRS to mRS 0-1 versus greater than 1 discards distinctions among scores 2, 3, 4, 5, and 6. A participant with mRS 2 and a participant with mRS 6 are both classified as not achieving the primary endpoint. The binary endpoint can therefore be clinically clear while containing less information than an analysis of the complete ordinal mRS distribution.
8. Understanding the Modified Rankin Scale
| mRS score | Registry definition |
|---|---|
| 0 | No symptoms |
| 1 | No significant disability. Able to carry out all usual activities, despite some symptoms |
| 2 | Slight disability. Able to look after own affairs without assistance, but unable to carry out all previous activities |
| 3 | Moderate disability. Requires some help, but able to walk unassisted |
| 4 | Moderately severe disability. Unable to attend to own bodily needs without assistance, and unable to walk unassisted |
| 5 | Severe disability. Requires constant nursing care and attention, bedridden, incontinent |
| 6 | Dead |
The primary endpoint groups scores 0 and 1 together. This means that the primary statistical comparison is specifically about the probability of being in the least-disabled two categories of the scale at the specified follow-up.
mRS 0-1
No symptoms or no significant disability according to the registry definitions.
mRS 2-6
All remaining mRS categories, including increasing levels of disability and death.
This distinction is important when interpreting a treatment effect. A difference in the proportion achieving mRS 0-1 does not by itself describe how the remaining participants are distributed across mRS scores 2 through 6.
9. Analysis of the Primary Endpoint
The statistical analysis of this trial's primary endpoint is fundamentally a comparison of two probabilities at a prespecified follow-up window. The most direct descriptive quantities are the proportion with mRS 0-1 in each treatment group and their difference.
This expression describes an absolute treatment contrast on the probability scale. The corresponding relative measures can be used when specified by the statistical analysis plan.
The interpretation should remain tied to the endpoint actually registered. A higher proportion of mRS 0-1 would mean that more participants reached the defined freedom-from-disability category at the specified assessment. It would not establish that every patient experienced an improvement, nor would it describe the entire mRS distribution.
The primary endpoint is clinically interpretable because it identifies a specific functional state rather than an abstract laboratory measure. Statistically, however, the endpoint is binary and therefore compresses the six-level mRS into two categories. That simplification makes treatment proportions straightforward to compare but means that the primary endpoint alone does not describe the severity distribution among participants who do not achieve mRS 0-1.
10. What a Confidence Interval Would Tell Us
When a formal treatment-effect estimate is available, its confidence interval is essential to interpretation. For a risk difference, the interval describes uncertainty around the difference between the two treatment-group probabilities. For a ratio measure, it describes uncertainty around the relative effect on that scale.
Precision
Narrower intervals indicate greater statistical precision under the selected analysis, while wider intervals indicate greater uncertainty.
Not individual prognosis
A confidence interval around a treatment effect is not a range of outcomes that an individual patient should expect.
The enrollment of 1,600 participants provides the trial with a substantial randomized population, but sample size alone does not determine the precision of the final estimate. Precision also depends on the observed endpoint frequency, missing assessments, the effect measure, and the statistical model.
11. Why the P-Value Is Not the Effect Size
A formal hypothesis test and an effect estimate answer different questions. The effect estimate describes the magnitude and direction of the treatment contrast. The p-value describes the compatibility of the observed data with a specified null hypothesis under the chosen statistical framework.
| Quantity | Primary question it addresses |
|---|---|
| Proportion with mRS 0-1 | How frequently was the primary outcome observed in each group? |
| Risk difference | How many percentage points separate the two groups? |
| Risk ratio | How large is the probability in one group relative to the other? |
| Odds ratio | How do the odds of achieving mRS 0-1 compare? |
| Confidence interval | How precisely is the treatment effect estimated? |
| P-value | How compatible are the observed data with the specified null hypothesis? |
For a clinical trial, the effect estimate and its confidence interval generally provide more information about magnitude and precision than a p-value considered by itself.
12. Missing Data and Follow-Up
The primary endpoint requires a telephone follow-up between 90-120 days. This makes ascertainment of the follow-up outcome an important component of the analysis.
If an mRS assessment is unavailable for a participant, that participant's outcome cannot simply be treated as equivalent to either mRS 0-1 or mRS above 1 without an explicit analytical justification. Statistical handling of missing outcomes can include prespecified assumptions, sensitivity analyses, or model-based approaches, depending on the trial's statistical analysis plan.
The registry information presented here identifies the follow-up window but does not provide a posted statistical analysis specifying a missing-data or imputation method.
13. Randomization and Causal Interpretation
Randomization is central to the causal interpretation of a two-arm treatment comparison. Before outcomes are observed, treatment assignment is determined through the randomized design rather than by observed clinical characteristics.
Because AcT is identified as randomized and parallel, the primary statistical comparison can be framed as a between-group treatment comparison rather than a comparison of two observationally selected populations.
14. Masking and Outcome Assessment
The registry identifies the trial as single-masked. The record summarized here does not identify which study participants or study personnel were masked, so the masking designation should not be expanded beyond the registry wording.
The primary endpoint is assessed by telephone follow-up. Functional outcomes such as the mRS require classification of patient disability according to defined criteria, making consistency of assessment important to the reliability of the endpoint.
15. What the Registry Does and Does Not Establish
| Question | What the registry record establishes |
|---|---|
| Was the trial randomized? | Yes. The allocation is listed as randomized. |
| Was the design parallel? | Yes. The design model is listed as parallel. |
| Was the study masked? | Yes. The registry lists single masking. |
| How many participants were enrolled? | 1,600. |
| How many treatment arms were there? | 2. |
| What interventions were evaluated? | Tenecteplase and alteplase. |
| What was the registered primary endpoint? | mRS 0-1, defined as freedom from disability. |
| When was the primary endpoint assessed? | By telephone follow-up between 90-120 days. |
| Are numerical primary efficacy results posted? | No posted statistical analyses are present in the registry record. |
This distinction is important in statistical reporting. Trial registration describes the prespecified study framework, while a posted statistical analysis would provide the numerical treatment estimates and inferential results needed to quantify the observed treatment comparison.
16. Important Limitations and Interpretation Issues
- No posted statistical analyses: the ClinicalTrials.gov record identifies the primary endpoint but does not provide a posted numerical treatment effect, confidence interval, or p-value.
- Binary endpoint: mRS 0-1 versus other outcomes does not preserve the full six-level mRS distribution.
- Follow-up window: the primary endpoint is assessed between 90-120 days rather than at a single fixed day.
- Missing outcome assessments: incomplete follow-up can affect estimation if participants do not have an evaluable primary endpoint.
- Analysis specification: the appropriate formal model, covariate adjustment, hypothesis-testing procedure, and missing-data strategy should be established from the statistical analysis plan rather than inferred from the endpoint name alone.
- Masking detail: the registry identifies single masking but the record summarized here does not specify which parties were masked.
- Ordinal information: the primary endpoint collapses the mRS scale into a binary threshold and therefore does not distinguish among mRS scores 2 through 6.
- Generalizability: the applicability of results to other stroke populations depends on the trial's eligibility criteria and clinical setting, details not contained in the registry information summarized here.
17. Why This Trial Matters Statistically
AcT is a useful teaching case because its primary endpoint illustrates a central issue in clinical-trial statistics: the same clinical outcome can be represented at different levels of detail, and the choice of representation determines which treatment effects can be directly estimated.
| Concept | How it appears in AcT |
|---|---|
| Randomization | The trial uses randomized allocation to compare two drug interventions. |
| Parallel design | The two treatment groups are evaluated as parallel arms. |
| Binary endpoint | The registered primary outcome is mRS 0-1 versus not meeting that threshold. |
| Functional outcome | The primary endpoint measures disability and dependence after stroke. |
| Time frame | The primary outcome is assessed by telephone follow-up between 90-120 days. |
| Absolute treatment effect | A risk difference can describe the percentage-point separation in mRS 0-1 rates. |
| Relative treatment effect | A risk ratio or odds ratio can describe the treatment contrast on a relative scale. |
| Confidence intervals | They would quantify uncertainty around the selected treatment-effect estimate. |
| P-values | They would address a specified null hypothesis rather than quantify clinical magnitude. |
| Missing data | The telephone follow-up requirement makes outcome ascertainment relevant to the analysis. |
| Ordinal versus binary analysis | Binary mRS 0-1 analysis simplifies interpretation but discards distinctions among higher mRS categories. |
18. A Deeper Statistical View of the Primary Endpoint
The primary endpoint can be represented as an indicator variable for each participant:
The treatment comparison then concerns the probability that Y = 1 in each randomized group.
This representation clarifies exactly what the primary analysis is trying to estimate. It does not estimate the mean mRS score. It does not estimate the probability of every individual mRS category separately. It estimates the probability of crossing the prespecified threshold of 0-1.
Absolute versus relative effects
Suppose a treatment comparison produced proportions pT and pA. Three common summaries would answer different questions:
| Measure | Formula | Interpretation |
|---|---|---|
| Risk difference | pT − pA | Percentage-point difference in the probability of mRS 0-1 |
| Risk ratio | pT / pA | Relative probability of mRS 0-1 |
| Odds ratio | [pT/(1−pT)] / [pA/(1−pA)] | Relative odds of mRS 0-1 |
These measures are mathematically related but are not interchangeable. In particular, an odds ratio should not automatically be described as though it were a risk ratio or a percentage-point difference.
19. Why the Full mRS Scale Still Matters
The registry defines the mRS as a six-level scale from 0 through 6. The primary endpoint uses only the distinction between 0-1 and 2-6. That choice makes the primary question easy to communicate but creates an important statistical tradeoff.
Advantage of the binary threshold
The endpoint has a clear clinical interpretation: the proportion of participants reaching the predefined category of freedom from disability.
Information lost
The endpoint does not distinguish mild disability from severe disability or death among participants outside the 0-1 category.
An analysis of the full ordinal mRS can preserve more information because the six categories retain their ordering. But such an analysis answers a different statistical question from the registered binary primary endpoint. It should therefore not be substituted for the primary endpoint without an explicit statistical rationale.
20. Timeline
Trial start
The registered start date for AcT.
Randomized treatment comparison
The trial evaluates tenecteplase and alteplase in a randomized, parallel design with single masking.
Primary endpoint assessment
The modified Rankin Scale 0-1 endpoint is assessed by telephone follow-up during this window.
Primary completion
The registered primary completion date.
21. Interpreting the Trial Without a Numerical Result
The registry establishes a randomized phase 3 comparison between tenecteplase and alteplase with a prespecified functional primary endpoint. The primary endpoint is a binary indicator of whether a participant reaches mRS 0-1 by telephone follow-up between 90-120 days.
A treatment effect cannot be quantified without a posted numerical comparison. In particular, the registry record summarized here does not provide an observed difference in mRS 0-1 rates, a confidence interval, or a p-value.
Even without a posted result, the endpoint definition reveals the central estimand: the difference between two randomized treatment groups in the probability of reaching the mRS 0-1 category within the specified follow-up window. That makes the trial a useful example of binary functional-outcome analysis in randomized clinical research.
22. Sources
- ClinicalTrials.gov: AcT (NCT03889249), Alteplase Compared to Tenecteplase in Patients With Acute Ischemic Stroke.
Continue through Clinical Biostats
Explore additional clinical trial results, statistical tutorials, and analysis tools from Clinical Biostats.
23. Record Summary
AcT is a randomized, parallel, phase 3 treatment trial enrolling 1,600 participants and comparing tenecteplase with alteplase in patients with acute ischemic stroke. Its registered primary endpoint is modified Rankin Scale 0-1 (freedom from disability), assessed by telephone follow-up between 90-120 days.
Statistically, the trial provides a clear example of a binary functional endpoint embedded within an ordinal clinical scale. The primary comparison concerns the probability of reaching mRS 0-1, while interpretation of the treatment effect would require the observed treatment-group proportions together with an appropriate effect estimate and measure of uncertainty. The ClinicalTrials.gov record identifies the trial design and endpoint but contains no posted statistical analyses for the primary endpoint.