This page separates the registered trial design from statistical interpretation. The ClinicalTrials.gov record identifies one primary functional outcome measured at 90 days; it does not report posted statistical analyses for that endpoint.
1. Trial at a Glance
NOR-TEST 2, formally titled the Norwegian Tenecteplase Stroke Trial 2, was a randomized, double-masked, parallel-group phase 3 treatment trial comparing tenecteplase and alteplase in the setting of acute stroke and ischemic stroke.
| Feature | NOR-TEST 2 |
|---|---|
| Official title | The Norwegian Tenecteplase Stroke Trial 2 |
| Phase | Phase 3 |
| Status | Terminated |
| Enrollment | 201 |
| Conditions | Stroke, Acute; Cerebrovascular Disorders; Ischemic Stroke |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | Double |
| Primary purpose | Treatment |
| Number of arms | 2 |
| Interventions | Tenecteplase; Alteplase |
| Lead sponsor | Haukeland University Hospital |
| Sponsor type | Other |
| Start date | 2019-10-28 |
| Primary completion date | 2021-09-30 |
| ClinicalTrials.gov | NCT03854500 |
2. Clinical Question
The central statistical question is whether the proportion of patients achieving a favorable functional outcome at 90 days differs between the tenecteplase and alteplase treatment groups.
Population
The registry identifies the conditions as acute stroke, cerebrovascular disorders, and ischemic stroke.
Intervention
Tenecteplase.
Comparator
Alteplase.
Primary question
What is the difference in the proportion of patients with favorable functional outcome at 90 days between the randomized treatment groups?
3. Trial Design
Tenecteplase
- Drug intervention
- Evaluated in the randomized comparison
- Primary outcome assessed at 90 days
Alteplase
- Drug comparator
- Evaluated in the randomized comparison
- Primary outcome assessed at 90 days
The parallel-group structure is statistically important because each participant contributes to one randomized treatment comparison rather than moving between treatment conditions. Randomization is intended to establish comparable treatment groups in expectation, allowing differences in the prespecified outcome to be interpreted within a randomized comparative framework.
4. Endpoints
| Endpoint | Registry definition | Time frame |
|---|---|---|
| Proportion of patients with favorable functional outcome | Modified Rankin Scale 0-1 (favorable= 0-1, unfavorable 2-6) | 90 days |
The primary endpoint is therefore a binary functional outcome. Each patient is classified into one of two categories at the specified assessment time: favorable functional outcome, defined as Modified Rankin Scale 0-1, or unfavorable functional outcome, defined as Modified Rankin Scale 2-6.
The analysis target is the proportion of participants meeting the favorable-outcome definition at 90 days in each randomized treatment group.
This endpoint differs fundamentally from a time-to-event endpoint. The analysis is based on whether the specified outcome is present at a fixed assessment time rather than on the time until an event such as death or disease progression occurs.
5. Statistical Methodology
Analysis of a binary proportion
A binary endpoint such as favorable versus unfavorable functional outcome can be summarized as the proportion favorable within each randomized treatment group. A treatment comparison can then be expressed using an absolute risk difference, risk ratio, or odds ratio, each of which describes a different aspect of the same two-group comparison.
The registry does not report a posted statistical analysis specifying which effect measure or inferential model was used for NOR-TEST 2.
Confidence intervals
For a randomized binary endpoint, a confidence interval provides an uncertainty range around the estimated treatment effect. The appropriate interval depends on the chosen effect measure and statistical method. For example, a confidence interval for a risk difference is on the probability scale, whereas a confidence interval for an odds ratio is on the odds-ratio scale.
Hypothesis testing
A formal comparison can test a null hypothesis of no difference between the randomized treatment groups. For a risk difference, that null value is zero. For a risk ratio or odds ratio, the corresponding null value is one.
A statistical test evaluates compatibility with the null hypothesis. The test result should be interpreted alongside the estimated effect and its confidence interval rather than treated as a measure of clinical importance.
Regression-based analysis
For a binary outcome, logistic regression is a standard modeling approach. It can estimate the association between treatment assignment and the probability of a favorable outcome while allowing additional prespecified covariates to be included when appropriate.
However, the ClinicalTrials.gov record does not report a posted statistical analysis specifying that logistic regression, a particular covariate adjustment strategy, or another binary-outcome model was used for the primary endpoint.
Randomization and the treatment comparison
Because allocation was randomized, the primary treatment comparison is naturally framed around the randomized groups. This is different from an observational comparison in which treatment assignment may depend on prognosis, clinical characteristics, or treatment selection.
6. Planned Analysis
The registry specifies that the primary endpoint is the proportion of patients with favorable functional outcome at 90 days, with favorable defined as Modified Rankin Scale 0-1 and unfavorable defined as 2-6.
For an endpoint of this type, a typical primary analysis would compare the two treatment-group proportions using a prespecified two-group method and report an effect estimate with a confidence interval. Depending on the prespecified statistical plan, this could involve a comparison of proportions, a regression model for the binary outcome, or another appropriate method.
What is measured
Whether each participant has a Modified Rankin Scale value of 0-1 at the 90-day assessment.
What is compared
The proportion with favorable functional outcome between the randomized tenecteplase and alteplase groups.
What the effect represents
A difference in the probability of meeting the registry-defined favorable functional outcome criterion.
What remains unspecified
The registry does not report a posted statistical analysis defining the exact inferential model, effect measure, confidence interval method, or p-value for the primary endpoint.
7. Statistical Methods Explained
What makes the primary endpoint a binary endpoint?
The Modified Rankin Scale is converted by the registry definition into two outcome categories for the primary endpoint. Values of 0-1 are classified as favorable, while values of 2-6 are classified as unfavorable. The statistical target is therefore a proportion rather than a continuous mean or a time-to-event quantity.
Why compare proportions between the two randomized groups?
The primary question concerns the proportion of participants achieving the prespecified functional outcome under each treatment assignment. Comparing these proportions directly addresses the clinical question at the 90-day assessment.
What does a risk difference tell us?
A risk difference expresses the absolute separation between two probabilities. If the favorable-outcome probability were 0.60 in one group and 0.50 in another, the risk difference would be 0.10, meaning a ten-percentage-point difference. Such an example illustrates the scale of the measure and is not a result from NOR-TEST 2.
What does an odds ratio tell us?
An odds ratio compares the odds of favorable outcome between treatment groups rather than directly comparing probabilities. An odds ratio of one represents equal odds. Values above or below one indicate higher or lower odds in the numerator group, respectively. Odds ratios can differ materially from risk ratios when the outcome is common.
Why does the confidence interval matter?
An estimated treatment difference is subject to sampling uncertainty. A confidence interval communicates the precision of the estimate under the statistical model and sampling framework. A narrow interval indicates greater statistical precision than a wide interval, although precision and clinical importance are separate concepts.
Why should the p-value not be treated as the treatment effect?
A p-value addresses evidence against a specified null hypothesis under a statistical model. It does not describe the magnitude of the treatment difference. The estimated proportion in each group and an appropriate effect measure are needed to understand the size and direction of a treatment effect.
8. Interpreting a Fixed-Time Functional Outcome
The choice of a 90-day binary functional endpoint creates a particularly clear interpretation: the analysis asks whether a patient has reached the prespecified favorable functional category at one defined follow-up time.
A risk difference directly describes the change in the proportion achieving favorable functional outcome. It is expressed on the probability scale and is often straightforward to interpret clinically.
A risk ratio describes how the probability of favorable outcome in one group compares with the probability in the other group. It is relative rather than absolute, so it should be considered alongside the underlying proportions.
An odds ratio compares odds rather than probabilities. It is a valid measure for a binary endpoint, but it should not be casually described as a risk ratio, particularly when the outcome is frequent.
The endpoint is specifically defined at 90 days. It does not, by itself, describe how quickly a favorable outcome occurred or what happened after the 90-day assessment.
9. Analysis Population and Randomization
The registry identifies NOR-TEST 2 as randomized, with 201 participants enrolled across two arms. Randomization provides the foundation for comparing outcome proportions between treatment assignments.
| Design feature | Statistical implication |
|---|---|
| Randomized allocation | Supports a comparative analysis based on treatment assignment rather than treatment selection. |
| Parallel design | Participants remain within their randomized treatment comparison rather than switching between trial arms as part of the design. |
| Double masking | Can reduce the influence of treatment knowledge on participants, investigators, and outcome assessment. |
| Two arms | Creates a direct randomized comparison between tenecteplase and alteplase. |
| 201 enrolled | Determines the amount of information available for estimating the treatment-group proportions and their uncertainty. |
Randomization does not guarantee identical baseline characteristics in finite samples. Its statistical purpose is to make treatment assignment independent of measured and unmeasured prognostic factors in expectation, creating a principled basis for causal comparison.
10. Missing Data and the 90-Day Assessment
Because the primary endpoint is defined at 90 days, the treatment comparison depends on determining each participant's functional-outcome category at that assessment. Participants without an evaluable 90-day outcome create a missing-data issue that must be handled according to the prespecified analysis plan.
Possible approaches in clinical-trial statistics include defining an analysis population with explicit outcome availability rules, using prespecified assumptions about missing outcomes, or applying a model-based method. The appropriate approach depends on the trial's statistical analysis plan and the reasons for missingness.
This matters because the treatment-group proportion is a descriptive quantity only when the denominator and handling of unavailable assessments are clearly defined. A formal confirmatory analysis should therefore distinguish the observed-data summary from the prespecified inferential procedure.
11. Multiplicity and Other Design Considerations
The registry identifies one registered primary endpoint: the proportion of patients with favorable functional outcome at 90 days. It does not report a posted statistical analysis describing an endpoint hierarchy, multiplicity adjustment, interim-analysis procedure, alpha-spending strategy, or Bayesian framework.
| Design topic | Registry information |
|---|---|
| Primary endpoint | One registered primary endpoint is identified. |
| Multiple primary endpoints | Not reported in the registered primary-endpoint information. |
| Multiplicity adjustment | No posted statistical analysis is reported. |
| Interim analysis | No posted statistical analysis is reported. |
| Bayesian methods | No posted statistical analysis is reported. |
| Stratification | No stratification factors are reported in the ClinicalTrials.gov record. |
| Non-inferiority margin | No non-inferiority margin is reported in the ClinicalTrials.gov record. |
These distinctions are important because the same binary endpoint can support different statistical designs. Superiority, non-inferiority, equivalence, and adaptive frameworks require different hypotheses and different interpretations of confidence intervals and decision thresholds.
12. Limitations
- No posted statistical analysis: the registry identifies the primary endpoint but does not provide a posted formal statistical analysis with an effect estimate, confidence interval, or p-value.
- Limited endpoint information: the registered information specifies the primary outcome and its definition but does not provide additional endpoint-level analytical detail.
- Fixed-time outcome: the 90-day binary endpoint summarizes functional status at a specified time and does not itself characterize the trajectory of recovery.
- Missing-data considerations: the registry does not report the statistical procedure that would be used for participants without an evaluable 90-day outcome.
- Unreported design details: the registered information does not specify statistical stratification factors, multiplicity procedures, interim-analysis methods, or a non-inferiority margin.
- Terminated status: the ClinicalTrials.gov record lists the trial status as terminated. The registry information presented here does not establish the reason for termination or the implications for the primary endpoint.
- No outcome inference: the absence of posted numerical analyses means the registered endpoint should not be interpreted as evidence of a particular treatment effect.
13. Why This Trial Matters Statistically
NOR-TEST 2 is a useful teaching example because its primary endpoint illustrates a different statistical problem from the time-to-event analyses commonly encountered in oncology and other clinical-trial settings. The endpoint is a clinically defined binary outcome measured at a fixed follow-up time.
| Concept | How it appears in NOR-TEST 2 |
|---|---|
| Randomization | Participants were randomized to two treatment arms. |
| Parallel-group comparison | The registry specifies a parallel design. |
| Binary endpoint | Functional outcome is classified as favorable or unfavorable using the Modified Rankin Scale definition. |
| Fixed-time analysis | The primary endpoint is assessed at 90 days. |
| Risk difference | A natural absolute measure for comparing favorable-outcome proportions. |
| Risk ratio | A relative measure comparing favorable-outcome probabilities. |
| Odds ratio | An alternative model-based measure for the binary endpoint. |
| Confidence interval | Quantifies uncertainty around an estimated treatment effect. |
| Hypothesis testing | Can assess compatibility of the observed comparison with a no-difference hypothesis. |
| Missing data | 90-day outcome availability is relevant to the analysis population and inferential method. |
| Masking | Double masking can reduce bias associated with knowledge of treatment assignment. |
14. Clinical Interpretation vs Statistical Interpretation
Statistical interpretation
The registered primary outcome is a two-category functional endpoint evaluated at 90 days. The appropriate treatment comparison concerns the probability of meeting the favorable-outcome definition in the randomized groups.
Clinical interpretation
The endpoint is intended to distinguish patients with Modified Rankin Scale 0-1 from those with values 2-6 at the specified 90-day assessment. Its clinical meaning therefore depends directly on the registry-defined functional threshold.
These two levels of interpretation should remain distinct. Statistical analysis determines the estimated difference between treatment groups and the uncertainty around that estimate. Clinical interpretation considers what the endpoint represents for functional status.
15. What the Primary Endpoint Does — and Does Not — Mean
The primary endpoint measures the proportion of patients with favorable functional outcome, where favorable is defined as Modified Rankin Scale 0-1, at 90 days.
The endpoint does not directly measure the time required to reach a favorable outcome, the entire distribution of Modified Rankin Scale values, or outcomes beyond the specified 90-day assessment.
Converting the Modified Rankin Scale into favorable and unfavorable categories makes the endpoint easy to summarize as a proportion, but it also discards the distinctions among individual values within each category.
A hypothesis test alone would not communicate how large the treatment-group difference is. A complete statistical presentation should pair the treatment-group proportions with an interpretable effect measure and its confidence interval.
16. A Statistical Framework for Reading the Result
If numerical results become available, the most informative sequence for evaluating this primary endpoint would be to examine the treatment-group proportions first, followed by an absolute and/or relative effect estimate, its confidence interval, and the corresponding hypothesis test.
This sequence prevents a common statistical error: allowing a p-value to become the headline while leaving the magnitude of the treatment difference and its uncertainty unclear.
The numerical values would need to come from the actual analysis. The framework itself is general for a randomized binary endpoint and is not an outcome estimate for NOR-TEST 2.
17. Trial Timeline
Trial start
The registry records the start date of NOR-TEST 2 as 2019-10-28.
Primary completion
The registry records the primary completion date as 2021-09-30.
Terminated
The ClinicalTrials.gov record lists the overall study status as terminated.
18. Sources
- ClinicalTrials.gov: NOR-TEST 2, NCT03854500.
- PubMed: PubMed record.
- PubMed: PubMed record.
Continue through the Clinical Biostats knowledge graph
Explore additional Clinical Biostats material on clinical-trial design, binary endpoints, treatment effects, confidence intervals, and statistical interpretation.
19. Record Summary
NOR-TEST 2 is a randomized, double-masked, parallel-group phase 3 treatment trial with 201 enrolled participants and two drug interventions, tenecteplase and alteplase. Its registered primary endpoint is the proportion of patients with favorable functional outcome at 90 days, with favorable defined as Modified Rankin Scale 0-1 and unfavorable defined as 2-6.
Statistically, this makes the trial a clear example of a randomized binary-outcome comparison. The central quantities are the favorable-outcome proportions in the two treatment groups, the corresponding treatment effect on an absolute or relative scale, and the confidence interval describing uncertainty. Because the ClinicalTrials.gov record does not post a formal statistical analysis for the endpoint, those numerical quantities are not reported here as trial outcomes.