The ClinicalTrials.gov record identifies CORONA as a completed phase 3 randomized trial with an enrollment of 206 patients. The registry reports the primary endpoint definition and time frame, but no formal statistical analyses are posted.
1. Trial at a Glance
CORONA was a randomized, double-masked, parallel-group phase 3 treatment trial evaluating levilimab against placebo in patients with severe COVID-19. The registry records an enrollment of 206 patients and a primary endpoint assessing sustained clinical recovery at day 14.
| Feature | CORONA |
|---|---|
| Acronym | CORONA |
| ClinicalTrials.gov identifier | NCT04397562 |
| Official trial title | A Clinical Trial of the Efficacy and Safety of Levilimab (BCD-089) in Patients With Severe COVID-19 |
| Phase | Phase 3 |
| Status | COMPLETED |
| Condition | COVID-19 |
| Allocation | RANDOMIZED |
| Design model | PARALLEL |
| Masking | DOUBLE |
| Primary purpose | TREATMENT |
| Enrollment | 206 |
| Number of arms | 2 |
| Interventions | Levilimab; Placebo |
| Lead sponsor | Biocad |
| Sponsor type | INDUSTRY |
2. Clinical Question
The clinical question is whether treatment with levilimab produces a higher proportion of patients with sustained clinical recovery by day 14 than placebo in patients with severe COVID-19.
Population
Patients with severe COVID-19 enrolled in the phase 3 CORONA trial.
Intervention
Levilimab (BCD-089).
Comparator
Placebo.
Primary question
What proportion of patients achieve sustained clinical recovery at day 14, and how does that proportion compare between the randomized treatment groups?
3. Trial Design
Levilimab
- Drug intervention
- Randomized parallel-group assignment
- Double-masked trial design
Placebo
- Placebo intervention
- Randomized parallel-group assignment
- Double-masked trial design
The registry describes the study as randomized, double-masked, and parallel. These design features are important statistically because the treatment comparison is anchored to randomized assignment, while masking is intended to reduce the influence of treatment knowledge on patient care, assessment, and reporting.
4. Trial Timeline
Study start
The registry lists April 29, 2020 as the study start date.
Primary completion
The registry lists July 3, 2020 as the primary completion date.
Registry status
The ClinicalTrials.gov record lists the study status as completed.
5. Primary Endpoint
| Endpoint | Time frame | Registry definition |
|---|---|---|
| Proportion of patients with sustained clinical recovery | Day 14 | Sustained clinical recovery is defined as either an improvement of at least 2 categories relative to baseline on a 7-Category Ordinal Scale of Clinical Status or reaching categories "Discharged" / "Hospitalized, not requiring supplemental oxygen - no longer requires ongoing medical care" at day 14. |
Understanding the endpoint
The endpoint is a binary, day-14 clinical outcome: each patient can ultimately be classified according to whether the registry's definition of sustained clinical recovery is met at day 14.
The definition has two routes to recovery. A patient can qualify through an improvement of at least 2 categories relative to baseline on the 7-Category Ordinal Scale of Clinical Status, or through reaching either of the specified clinical-status categories at day 14.
This structure converts a longitudinal clinical-status assessment into a prespecified binary endpoint for the primary comparison.
6. Statistical Methodology
The ClinicalTrials.gov record identifies the primary endpoint but does not post a formal statistical analysis for it. For a binary endpoint defined as a proportion at a fixed time point, the central statistical task is to compare the proportion meeting the sustained-recovery definition between the levilimab and placebo groups.
Comparing proportions
A standard analysis of this type of endpoint begins with the number of patients meeting the recovery definition and the number evaluated in each randomized group. The treatment effect can then be expressed using an absolute difference in proportions, a relative risk, an odds ratio, or another prespecified effect measure, with an appropriate confidence interval.
The risk difference describes the absolute difference in the proportions achieving sustained clinical recovery at day 14.
Confidence intervals
A confidence interval is important because a treatment-group difference calculated from a finite randomized sample is an estimate rather than a perfectly known population quantity. The interval communicates statistical precision around the estimated treatment effect.
Hypothesis testing
A conventional hypothesis test for a binary primary endpoint would assess whether the observed treatment-group difference is compatible with a prespecified null hypothesis. The p-value from such a test would describe evidence against that null hypothesis under the specified statistical model; it would not describe the magnitude or clinical importance of the treatment effect.
Analysis population
Randomized trials generally distinguish the population used for efficacy inference from populations defined by actual treatment exposure or evaluability. The registry information available here identifies randomization and enrollment but does not provide additional analysis-population definitions for the primary endpoint.
7. Planned Analysis
The primary endpoint is the proportion of patients with sustained clinical recovery at day 14. Because this is a binary proportion endpoint assessed at a fixed time point, a typical primary analysis would compare the proportions achieving recovery in the two randomized groups and report an effect measure with its confidence interval.
| Statistical element | Educational interpretation |
|---|---|
| Outcome | Binary sustained-clinical-recovery status at day 14 |
| Comparison | Levilimab versus placebo |
| Primary summary | Proportion achieving sustained clinical recovery in each group |
| Effect measure | A difference in proportions is a direct absolute measure; relative risk or odds ratio can provide a relative comparison |
| Uncertainty | A confidence interval should accompany the treatment-effect estimate |
| Hypothesis test | A prespecified test can assess compatibility with a null treatment effect |
Why the day-14 time point matters
A fixed assessment time creates a clinically interpretable snapshot: the analysis asks how many patients have reached the registry-defined recovery state by a specific point in follow-up. It is therefore different from a time-to-event endpoint, where the precise time at which recovery occurs contributes directly to the analysis.
Why the ordinal scale definition matters
The endpoint is not simply a subjective statement that a patient "improved." It uses a structured 7-Category Ordinal Scale of Clinical Status and specifies a threshold of at least 2 categories of improvement relative to baseline. The alternative route to endpoint success also uses explicitly named clinical-status categories. This prespecification reduces ambiguity in classifying the binary outcome.
8. Statistical Methods Explained
What does "proportion of patients" mean statistically?
It is the number of patients who satisfy the endpoint definition divided by the relevant number of patients evaluated. The resulting quantity is a probability estimate for achieving the specified outcome within the study population under the defined assessment conditions.
Why use a risk difference?
A risk difference directly answers an absolute question: how many more or fewer patients per unit of the analysis population achieved sustained clinical recovery in one group than the other? It is particularly intuitive when the outcome is a fixed-time binary endpoint.
What would a relative risk add?
A relative risk compares the probability of recovery between groups as a ratio. It answers a different question from the risk difference. For example, a relative risk of 1 would indicate equal estimated probabilities, whereas a value above 1 would indicate a higher estimated probability in the numerator group.
Why is the p-value not enough?
A p-value addresses evidence against a null hypothesis under the chosen testing framework. It does not tell the reader how large the treatment effect is. Two studies can have similar effect estimates but different p-values because of differences in sample size or variability. Conversely, a statistically detectable difference can be small in absolute terms.
Why does the confidence interval matter?
The confidence interval places the estimated treatment effect in a precision framework. A narrow interval indicates greater statistical precision than a wide interval, all else equal. The interval should be interpreted alongside the absolute proportions and the treatment-effect measure rather than as a substitute for them.
Why does randomization matter for this analysis?
Randomization creates the treatment-group comparison before outcomes are observed. When the randomized comparison is appropriately analyzed, differences in the primary endpoint can be interpreted in relation to the assigned interventions without relying solely on observational adjustment for measured baseline characteristics.
9. Interpreting the Primary Endpoint
What the endpoint measures
Whether a patient satisfies the registry-defined sustained-clinical-recovery criteria at day 14.
What it does not measure
It does not directly measure a continuous degree of improvement or the exact time at which recovery occurred.
Why absolute proportions matter
The percentage recovered in each randomized group provides the direct clinical scale of the observed outcome.
Why relative effects matter
A relative measure can describe the treatment-group relationship, but it should be read together with the underlying proportions.
The primary endpoint is fundamentally a comparison of two probabilities at a prespecified time point. A statistically complete presentation would therefore show the day-14 sustained-recovery proportion in each randomized group, the absolute and/or relative treatment effect, its confidence interval, and the corresponding hypothesis-test result when appropriate.
The registry does not post those statistical results. Consequently, the primary endpoint can be explained statistically from its definition, but the registry record does not establish a numerical treatment effect from the information available here.
10. Endpoint Definition and Statistical Consequences
The CORONA endpoint combines two clinically meaningful pathways into one binary outcome. The first is a change from baseline of at least 2 categories on the 7-Category Ordinal Scale of Clinical Status. The second is attainment of either of two specified day-14 clinical-status categories.
| Component | Statistical role |
|---|---|
| Baseline ordinal scale | Provides the reference state for determining whether improvement of at least 2 categories occurred. |
| At least 2-category improvement | Defines one route to sustained clinical recovery. |
| "Discharged" | Defines one qualifying day-14 clinical status. |
| "Hospitalized, not requiring supplemental oxygen - no longer requires ongoing medical care" | Defines another qualifying day-14 clinical status. |
| Day 14 | Fixes the primary assessment time. |
This type of composite classification has an important analytical advantage: the endpoint is defined before the statistical comparison is made. The same criteria are applied to the randomized groups, so the treatment comparison is based on a common outcome definition rather than on treatment-specific interpretations of improvement.
11. Missing Data and Endpoint Classification
A day-14 binary endpoint creates a practical statistical issue whenever a patient's outcome cannot be classified at the prespecified assessment. Depending on the prespecified analysis plan, missing endpoint observations can require rules for censoring, exclusion, imputation, or sensitivity analysis.
The ClinicalTrials.gov record does not report a missing-data or imputation strategy. That distinction matters because different assumptions about unobserved day-14 outcomes can change the estimated proportion achieving recovery.
12. Randomization and Causal Interpretation
The study is explicitly described as randomized. Randomization is the core design mechanism supporting a causal comparison between levilimab and placebo, provided that the randomized analysis and follow-up are handled appropriately.
The two-arm parallel design means that each participant is assigned to one of the two intervention groups rather than moving through both treatments as part of a crossover design. This makes the primary comparison a between-group comparison at day 14.
13. Why a Fixed-Time Proportion Differs From Survival Analysis
The CORONA primary endpoint is assessed at day 14 as a proportion. This is statistically different from an endpoint such as time to recovery, where the analysis would use the timing of each patient's event and account explicitly for censoring.
| Feature | Day-14 proportion endpoint | Time-to-event endpoint |
|---|---|---|
| Primary quantity | Whether the endpoint is achieved by the specified assessment | Time until the event occurs |
| Assessment | Day 14 | Uses follow-up time across observations |
| Typical effect measures | Risk difference, relative risk, odds ratio | Hazard ratio, survival probability, median time |
| Censoring | Depends on endpoint-assessment and missing-data rules | Central feature of standard survival analysis |
Neither framework is inherently more informative in every setting. The appropriate choice depends on the scientific question. CORONA's registered primary endpoint asks specifically about the proportion meeting the sustained-recovery definition at day 14.
14. Statistical Interpretation of a Potential Treatment Effect
If the observed sustained-recovery proportion were higher under levilimab, the risk difference would quantify the absolute increase in the probability of meeting the registry-defined endpoint at day 14. It would not describe the degree of improvement among patients who failed to meet the binary endpoint.
A relative risk greater than 1 would indicate a higher estimated probability of sustained recovery under levilimab than under placebo. It would not mean that every patient experiences the same proportional increase in recovery probability.
The confidence interval would describe statistical uncertainty around the estimated treatment effect under the selected analysis framework. It would not represent the range of outcomes expected for individual patients.
A p-value would quantify the compatibility of the observed data with a prespecified null hypothesis under the statistical testing framework. It would not measure the size, clinical importance, or probability that the treatment works.
15. Limitations
- No posted statistical results: the registry identifies the primary endpoint and its definition but does not post a formal statistical analysis or numerical primary-endpoint result.
- Fixed-time assessment: the primary endpoint summarizes recovery status at day 14 and therefore does not itself describe the exact timing of recovery before that assessment.
- Binary classification: converting clinical status into a sustained-recovery indicator simplifies a potentially more detailed clinical trajectory.
- Endpoint construction: the primary definition contains more than one route to endpoint success, so interpretation depends on applying the registry criteria consistently.
- Missing-data methodology: the registry information available here does not report a formal imputation or missing-endpoint strategy.
- Analysis population: enrollment is reported as 206, but the registry information available here does not provide detailed primary-analysis population counts.
- Generalizability: the trial specifically concerns patients with severe COVID-19; statistical conclusions from this population do not automatically apply to other clinical populations.
16. Why This Trial Matters Statistically
CORONA is a useful statistical teaching case because its primary endpoint illustrates an important alternative to time-to-event analysis: a prespecified clinical-status outcome evaluated as a proportion at a fixed follow-up time.
| Concept | How it appears in CORONA |
|---|---|
| Randomization | The study is randomized. |
| Parallel-group design | The study uses two parallel intervention arms. |
| Masking | The registry describes the trial as double-masked. |
| Binary endpoint | The primary endpoint is the proportion of patients meeting the sustained clinical recovery definition. |
| Fixed-time assessment | The primary endpoint is evaluated at day 14. |
| Ordinal clinical scale | One route to endpoint success requires an improvement of at least 2 categories on a 7-Category Ordinal Scale of Clinical Status. |
| Composite endpoint logic | Endpoint success can occur through either of two registry-defined clinical pathways. |
| Effect estimation | Binary outcomes can be described using absolute and relative treatment-effect measures with confidence intervals. |
| Missing-data considerations | A fixed day-14 endpoint requires prespecified rules for observations that cannot be classified. |
17. Results Status
Primary endpoint status on ClinicalTrials.gov
The registry reports the primary endpoint as the proportion of patients with sustained clinical recovery at day 14.
No estimate, confidence interval, or p-value is posted in the statistical-analysis information for this endpoint.
This is an important distinction between having a registered endpoint and having a posted statistical result. A registered endpoint specifies what the study intends to measure. A statistical result requires the observed endpoint data and an analysis framework sufficient to quantify the treatment comparison.
18. Sources
- ClinicalTrials.gov: NCT04397562 — CORONA.
Continue through Clinical Biostats
Explore additional clinical-trial analyses, statistical tutorials, and quantitative tools across the Clinical Biostats site.
19. Record Summary
CORONA is a completed phase 3 randomized, double-masked, parallel-group trial evaluating levilimab versus placebo in patients with severe COVID-19. The study enrolled 206 patients and registered a primary endpoint of the proportion of patients with sustained clinical recovery at day 14. Sustained clinical recovery is defined by either an improvement of at least 2 categories relative to baseline on a 7-Category Ordinal Scale of Clinical Status or reaching one of the specified discharge or no-longer-requiring-ongoing-medical-care categories at day 14.
From a statistical perspective, the endpoint is a fixed-time binary clinical outcome. Its natural analysis focuses on comparing the proportion of patients meeting the endpoint between randomized groups and expressing the treatment difference with an appropriate effect measure and confidence interval. Interpretation should distinguish the absolute probability of recovery from relative measures and from hypothesis-test results.