This page separates reported trial results from statistical interpretation. The numerical results and trial characteristics presented here are restricted to the ClinicalTrials.gov record. ClinicalTrials.gov provides the official trial registry record.
1. Trial at a Glance
N-MOmentum was a randomized, parallel-group, quadruple-masked phase 2/3 trial evaluating inebilizumab versus placebo in neuromyelitis optica and neuromyelitis optica spectrum disorders. The registry reports one primary time-to-event endpoint and five posted statistical analyses, including the primary Cox proportional-hazards analysis and four secondary analyses.
| Feature | N-MOmentum |
|---|---|
| Trial name | N-MOmentum |
| Brief title | A Clinical Research Study of Inebilizumab in Neuromyelitis Optica Spectrum Disorders |
| Phase | 2/3 |
| Status | COMPLETED |
| Therapeutic area | Neurology |
| Conditions | Neuromyelitis Optica and Neuromyelitis Optica Spectrum Disorders |
| Allocation | RANDOMIZED |
| Design model | PARALLEL |
| Masking | QUADRUPLE |
| Primary purpose | TREATMENT |
| Enrollment | 231 |
| Interventions | Inebilizumab; Placebo |
| Lead sponsor | MedImmune LLC |
| Sponsor type | INDUSTRY |
| ClinicalTrials.gov | NCT02200770 |
2. Clinical Question
The central statistical question was whether randomized assignment to inebilizumab, compared with placebo, was associated with a difference in the time to an Adjudication Committee (AC)-determined neuromyelitis optica spectrum disorder attack during the RCP.
Population
Participants in the N-MOmentum study with neuromyelitis optica or neuromyelitis optica spectrum disorders.
Intervention
Inebilizumab.
Comparator
Placebo.
Primary question
Does randomized treatment assignment affect the time to an AC-determined NMOSD attack during the RCP?
3. Trial Design
Inebilizumab
- Intervention: inebilizumab
- Compared with placebo
- Analyzed according to randomized treatment in the reported ITT analyses
Placebo
- Comparator: placebo
- Compared with inebilizumab
- Analyzed according to randomized treatment in the reported ITT analyses
The trial was randomized and parallel-group, with quadruple masking. The ClinicalTrials.gov record does not specify a factorial design, crossover design, non-inferiority margin, interim-analysis procedure, or Bayesian analysis, so those features are not attributed to N-MOmentum on this page.
4. Randomization and Analysis Population
The posted statistical analyses use an intention-to-treat population. The registry defines this population as participants who were randomized, received any study drug, and were grouped according to their randomized treatment.
| Analysis population | Registry definition / role |
|---|---|
| Intention-to-treat | All participants who were randomized, received any study drug, and were grouped according to randomized treatment. |
| Primary efficacy comparison | Placebo/Inebilizumab versus Inebilizumab/Inebilizumab as reported in the statistical-analysis grouping. |
5. Primary Endpoint
| Endpoint | Registry definition / time frame | Primary analysis |
|---|---|---|
| Time to Adjudication Committee (AC)-Determined Neuromyelitis Optica Spectrum Disorder (NMOSD) Attack During RCP | Day 1 (Baseline) through Day 197. The NMOSD attack is defined as the presence of new or worsening symptom(s) related to NMOSD that meet at least one of the 18 protocol-defined attack criteria. The criteria were developed with a panel of disease experts and with Food and Drug Administration input and were intended to be clinically meaningful, objective, quantifiable, and able to be used worldwide. | Cox proportional-hazards model |
The endpoint is a time-to-event outcome. That matters because the analysis uses not only whether an attack occurred, but also the time at which the event occurred and the information contributed by participants who did not experience the event during the relevant observation period.
6. Results: Primary Endpoint
The registry reports a formal superiority analysis of the primary time-to-event endpoint in the ITT population.
Time to AC-Determined NMOSD Attack During RCP
Hazard ratio
95% CI: 0.1496–0.4961 · P < 0.0001
Analysis: Cox proportional-hazards model · ITT population
| Primary endpoint | Effect estimate | 95% CI | P-value | Hypothesis |
|---|---|---|---|---|
| Time to AC-determined NMOSD attack during RCP | HR 0.272 | 0.1496–0.4961 | <0.0001 | Superiority |
An HR of 0.272 means that, under the fitted Cox model, the estimated instantaneous rate of the adjudicated NMOSD attack endpoint in the compared treatment group was 27.2% of the rate in the reference group. Expressed as a simple relative-hazard interpretation, this corresponds to an estimated 72.8% lower hazard because 1 − 0.272 = 0.728.
The HR does not mean that 72.8% of participants avoided an attack, that an individual participant's probability was reduced by exactly 72.8%, or that the absolute difference in event probability was 72.8 percentage points. Hazard is an instantaneous event-rate quantity, not an absolute risk measure.
The 95% CI of 0.1496 to 0.4961 describes uncertainty around the estimated hazard ratio under the analysis model and sampling framework. It does not describe the range of effects that individual participants would experience.
The P-value of <0.0001 addresses the statistical evidence against the null hypothesis specified by the superiority analysis. It is not a measure of effect size and does not tell us how clinically important the observed difference is.
Because the analysis is based on a Cox proportional-hazards model, interpretation of a single HR is most straightforward when the proportional-hazards assumption is reasonably appropriate. The ClinicalTrials.gov record does not report a formal assessment of that assumption, so this page does not infer one.
7. Secondary Endpoint Results
Four secondary statistical analyses are posted in the ClinicalTrials.gov record. They use the same ITT principle and compare Placebo/Inebilizumab with Inebilizumab/Inebilizumab.
EDSS Worsening
Odds ratio for worsening in EDSS score
95% CI: 0.1755–0.7059 · P = 0.0033
Day 1 (Baseline) through Day 197 · Logistic regression
The endpoint was the percentage of participants with worsening in Expanded Disability Severity Scale (EDSS) score from baseline to the last visit of RCP. The posted effect measure was an odds ratio.
An OR of 0.352 means that the modeled odds of the specified EDSS-worsening outcome were estimated at 35.2% of the odds in the reference group. In simple relative terms, that is an estimated 64.8% lower odds. Odds are not the same as probabilities or risks, so the OR should not be described as a 64.8% reduction in the percentage of participants with worsening.
The 95% CI of 0.1755–0.7059 quantifies uncertainty around the odds-ratio estimate. The P-value of 0.0033 provides evidence against the relevant null comparison but does not quantify the magnitude or clinical importance of the association.
Low-Contrast Visual Acuity
Mean difference in binocular score
95% CI: -2.0254–2.2941 · P = 0.9026
Day 1 (Baseline) through Day 197 · ANCOVA
The endpoint was change from baseline in Low-Contrast Visual Acuity Binocular Score to the last visit of RCP. The posted effect measure was the mean difference (net), analyzed using ANCOVA.
A mean difference of 0.134 indicates a very small estimated difference in the outcome scale between the compared groups under the reported ANCOVA analysis. It does not by itself establish that the two groups have identical individual responses.
The 95% CI of -2.0254 to 2.2941 spans zero, meaning the registry-reported estimate is compatible with both a negative and positive difference under the specified uncertainty framework. The P-value of 0.9026 does not provide evidence against the null comparison in this analysis. It also should not be interpreted as the probability that the null hypothesis is true.
The registry identifies ANCOVA as the method but the ClinicalTrials.gov record does not provide the model covariates or a full model specification. This page therefore does not infer additional adjustment variables.
Cumulative Number of Active MRI Lesions
Rate ratio
95% CI: 0.3866–0.8279 · P = 0.0034
From Screening (Day -28) to Day 197
The endpoint was the cumulative number of active Magnetic Resonance Imaging (MRI) lesions during RCP. The statistical-analysis record reports Negative Binomial Regression as the method and a rate ratio as the effect measure.
A rate ratio of 0.566 indicates that the modeled event rate was estimated at 56.6% of the rate in the reference group. In relative terms, that corresponds to an estimated 43.4% lower rate. A rate ratio is not a risk ratio and should not be interpreted as a 43.4% reduction in the probability that an individual participant develops a lesion.
The 95% CI of 0.3866–0.8279 expresses uncertainty around the rate-ratio estimate. The P-value of 0.0034 is evidence against the relevant null comparison but is not a measure of the size of the observed treatment effect.
Count outcomes can require methods that account for the distribution and exposure of observations. The registry explicitly reports Negative Binomial Regression for this endpoint; that reported method is retained here rather than substituting a different count-data model.
NMOSD-Related In-Patient Hospitalizations
Rate ratio
95% CI: 0.1257–0.7972 · P = 0.0146
Day 1 (Baseline) through Day 197
The endpoint was the number of NMOSD-related in-patient hospitalizations during RCP. The statistical-analysis record reports Negative Binomial Regression and a rate ratio.
A rate ratio of 0.317 means that the modeled hospitalization rate was estimated at 31.7% of the rate in the reference group. In relative terms, this corresponds to an estimated 68.3% lower rate. It does not mean that 68.3% fewer participants were hospitalized, because a hospitalization rate and a participant-level probability are different quantities.
The 95% CI of 0.1257–0.7972 describes uncertainty around the rate-ratio estimate. The P-value of 0.0146 addresses the statistical evidence for the comparison; it does not measure the magnitude or clinical importance of the rate difference.
8. Statistical Results Summary
| Endpoint | Type | Method | Effect | 95% CI | P-value |
|---|---|---|---|---|---|
| Time to AC-determined NMOSD attack during RCP | Time-to-event | Cox proportional-hazards model | HR 0.272 | 0.1496–0.4961 | <0.0001 |
| Worsening in EDSS score | Binary | Logistic regression | OR 0.352 | 0.1755–0.7059 | 0.0033 |
| Change from baseline in low-contrast visual acuity binocular score | Continuous | ANCOVA | Mean difference 0.134 | -2.0254–2.2941 | 0.9026 |
| Cumulative number of active MRI lesions | Continuous / count | Negative Binomial Regression | Rate ratio 0.566 | 0.3866–0.8279 | 0.0034 |
| NMOSD-related in-patient hospitalizations | Continuous / count | Negative Binomial Regression | Rate ratio 0.317 | 0.1257–0.7972 | 0.0146 |
These results span four different statistical estimands: a hazard ratio for time to attack, an odds ratio for a binary disability outcome, a mean difference for a continuous visual-acuity score, and rate ratios for count-type outcomes. Comparing the numerical magnitudes directly would therefore be misleading. Each estimate must be interpreted on the scale of its own endpoint.
9. Statistical Methodology
Cox proportional-hazards model
The primary endpoint is a time-to-event outcome, and the registry reports a Cox regression analysis. The Cox model estimates a relative hazard between treatment groups without requiring the baseline hazard function to take a particular parametric form.
For a binary treatment indicator, exp(β) represents the hazard ratio comparing the treatment groups under the fitted model.
Hazard ratio
The primary HR of 0.272 is a relative time-to-event measure. An HR below 1 indicates a lower estimated instantaneous event rate in the numerator treatment group relative to the reference group. It does not directly provide an absolute risk difference, median time to event, or probability of remaining event-free at a particular day.
Logistic regression
The EDSS worsening endpoint is binary, so the registry reports logistic regression. Logistic regression models the log odds of an outcome. Its exponentiated treatment coefficient is an odds ratio.
For a binary treatment indicator, exp(β1) is the modeled odds ratio.
ANCOVA
The low-contrast visual acuity endpoint is continuous and was analyzed using ANCOVA. In general, ANCOVA compares an outcome between groups while incorporating specified covariates into a linear model. This can improve precision when relevant baseline information explains variation in the outcome.
The ClinicalTrials.gov record identifies ANCOVA but does not provide the complete model specification or all covariates. The interpretation here therefore stays at the method level rather than assuming a particular covariate set.
Negative Binomial Regression
The MRI-lesion and hospitalization analyses report Negative Binomial Regression. This is a regression framework commonly used for count outcomes when a Poisson model's equal-mean-and-variance restriction is not appropriate. The ClinicalTrials.gov record identifies the reported method and rate-ratio effect measure; no additional distributional assumptions are imposed on the trial beyond that registry description.
Intention-to-treat analysis
The posted analyses use an ITT population consisting of randomized participants who received any study drug and were grouped according to randomized treatment. The principle preserves the treatment comparison established by randomization rather than redefining groups according to treatment actually received after randomization.
10. Statistical Methods Explained
Why was a Cox model used for the primary endpoint?
The primary endpoint measures time to an adjudicated NMOSD attack rather than simply whether an attack occurred. A Cox model is designed for this type of censored time-to-event data and produces a hazard ratio that summarizes the relative event rate between treatment groups under the model.
What does an HR of 0.272 mean?
It means the estimated instantaneous event rate under the fitted model was 27.2% of the reference rate. The arithmetic complement, 1 − 0.272, is 0.728, so the estimate can be described as a 72.8% lower estimated hazard. It does not mean a 72.8% absolute reduction in attack probability.
Why is the confidence interval important?
The point estimate is only one estimate of the treatment effect. The 95% CI of 0.1496–0.4961 shows the uncertainty around the primary HR. A narrower interval generally provides more precision than a wider one, while the location of the interval relative to 1 is relevant for interpreting a ratio measure.
Why is the EDSS endpoint analyzed with logistic regression?
The endpoint asks whether participants had worsening in EDSS score from baseline to the last visit of RCP. Because the outcome is expressed as a binary event, logistic regression is a natural model for the reported endpoint and yields an odds ratio.
Why should an odds ratio not be called a risk ratio?
An odds ratio compares odds, calculated as p/(1−p), whereas a risk ratio compares probabilities directly. The two measures can be numerically quite different, particularly when an outcome is common. Therefore, an OR of 0.352 should not automatically be described as a 64.8% reduction in risk.
Why use a rate ratio for MRI lesions and hospitalizations?
Both endpoints count events rather than simply recording whether one event occurred. A rate ratio compares modeled event rates between groups. For the MRI and hospitalization analyses, the registry reports Negative Binomial Regression and rate ratios, so the rate-ratio scale is the appropriate scale for interpreting the posted estimates.
Why does the ANCOVA result differ conceptually from the HR and OR?
ANCOVA produces a difference on the outcome's continuous scale, whereas the Cox model produces a hazard ratio and logistic regression produces an odds ratio. The mean difference of 0.134 therefore cannot be compared numerically with an HR of 0.272 or an OR of 0.352 as though they were the same type of effect measure.
11. P-values and Confidence Intervals
The five posted statistical analyses use two-sided 95% confidence intervals and report superiority hypotheses. The primary analysis has a P-value of <0.0001, while the secondary analyses report P-values of 0.0033, 0.9026, 0.0034, and 0.0146.
| Endpoint | 95% CI excludes the ratio null / difference null? | P-value | Statistical reading |
|---|---|---|---|
| Primary time to NMOSD attack | Yes; HR interval is below 1 | <0.0001 | Evidence against the null comparison |
| EDSS worsening | Yes; OR interval is below 1 | 0.0033 | Evidence against the null comparison |
| Low-contrast visual acuity | No; mean-difference interval includes 0 | 0.9026 | No evidence against the null comparison in this analysis |
| Active MRI lesions | Yes; rate-ratio interval is below 1 | 0.0034 | Evidence against the null comparison |
| In-patient hospitalizations | Yes; rate-ratio interval is below 1 | 0.0146 | Evidence against the null comparison |
12. Multiplicity and Multiple Endpoints
The ClinicalTrials.gov record identifies one primary endpoint and four posted secondary statistical analyses. All five analyses report superiority hypotheses and two-sided 95% confidence intervals.
| Analysis | Role | Effect measure | Reported P-value |
|---|---|---|---|
| Time to AC-determined NMOSD attack | Primary | Hazard ratio | <0.0001 |
| EDSS worsening | Secondary | Odds ratio | 0.0033 |
| Low-contrast visual acuity | Secondary | Mean difference | 0.9026 |
| Cumulative active MRI lesions | Secondary | Rate ratio | 0.0034 |
| NMOSD-related in-patient hospitalizations | Secondary | Rate ratio | 0.0146 |
The ClinicalTrials.gov record does not specify an alpha-allocation strategy, hierarchical testing procedure, gatekeeping strategy, or other formal multiplicity adjustment among the secondary endpoints. Accordingly, the secondary P-values are reported exactly as posted but are not characterized here as individually establishing a multiplicity-adjusted confirmatory claim.
13. Safety Results
The ClinicalTrials.gov record provides serious adverse event counts by arm. These are reported as affected participants divided by participants at risk.
| Arm | Serious adverse events, affected / at risk |
|---|---|
| Placebo/Inebilizumab | 24 / 56 |
| Inebilizumab/Inebilizumab | 28 / 174 |
The ClinicalTrials.gov record does not provide a formal statistical comparison of serious adverse events, a confidence interval, or a P-value. The affected/at-risk figures should therefore be treated as descriptive safety information rather than as a formal hypothesis test.
14. Registry Limitations and Safety Follow-Up
The registry contains an important caveat concerning the safety follow-up period (SFP). Only 1 participant from the Placebo/Inebilizumab arm rolled over to SFP, while no participant from the Inebilizumab/Inebilizumab arm rolled over to SFP.
The registry explains that, for EudraCT result posting, a study period with any one arm having zero participants started is not acceptable because of an EudraCT limitation. Consequently, the SFP was not included in the Participant Flow section.
15. Missing Data, Censoring, and Analysis Populations
The primary endpoint is a time-to-event endpoint, so participants who do not experience an adjudicated attack during the relevant observation period may contribute censored information. The ClinicalTrials.gov record does not specify the complete censoring rules, sensitivity analyses, or missing-data imputation strategy.
Censoring
Time-to-event analyses can incorporate participants whose event status is not observed through the entire endpoint window by using their available follow-up information up to censoring.
Binary outcomes
The EDSS endpoint is analyzed as a binary outcome using logistic regression rather than as a time-to-event endpoint.
Continuous outcomes
The low-contrast visual acuity endpoint is analyzed using ANCOVA on the change-from-baseline outcome.
Count outcomes
MRI lesions and hospitalizations are analyzed using Negative Binomial Regression with rate ratios reported as the effect measure.
No specific imputation method is provided in the ClinicalTrials.gov record. It would therefore be inappropriate to attribute a particular missing-data approach, such as multiple imputation or last-observation-carried-forward, to this trial without additional evidence.
16. Stratification, Interim Analysis, and Bayesian Methods
The ClinicalTrials.gov record identifies the methods used for the five posted analyses but do not report randomization stratification factors, an interim-analysis procedure, alpha spending, a non-inferiority margin, or Bayesian methodology.
| Design topic | What the ClinicalTrials.gov record supports |
|---|---|
| Randomization | Yes. Allocation is RANDOMIZED. |
| Parallel design | Yes. Design model is PARALLEL. |
| Masking | Yes. Masking is QUADRUPLE. |
| Superiority | Yes. The posted statistical analyses identify superiority as the hypothesis type. |
| Non-inferiority margin | Not reported in the ClinicalTrials.gov record. |
| Crossover | Not reported in the ClinicalTrials.gov record. |
| Factorial design | Not reported; the registered design model is parallel. |
| Interim analysis | Not reported in the ClinicalTrials.gov record. |
| Missing-data imputation | Not specified in the ClinicalTrials.gov record. |
| Bayesian methods | Not reported in the ClinicalTrials.gov record. |
| Randomization stratification | Not specified in the ClinicalTrials.gov record. |
This distinction is important because the absence of a detail from the registry extract is not evidence that the underlying protocol or statistical analysis plan lacked such a feature. It means only that the feature is not supported by the data used for this page.
17. Reading the Five Effect Measures Together
| Effect measure | Trial endpoint | Null value | How to read values below the null |
|---|---|---|---|
| Hazard ratio | Time to AC-determined NMOSD attack | 1 | Lower estimated instantaneous event rate |
| Odds ratio | EDSS worsening | 1 | Lower modeled odds of the binary outcome |
| Mean difference | Low-contrast visual acuity score | 0 | Direction depends on how the score is defined |
| Rate ratio | Active MRI lesions | 1 | Lower modeled event rate |
| Rate ratio | NMOSD-related hospitalizations | 1 | Lower modeled event rate |
The common feature is that each effect measure compares two randomized groups, but the mathematical quantity being compared differs. This is why an HR of 0.272, an OR of 0.352, and a rate ratio of 0.317 should not be placed on a single "effect-size" scale.
18. Primary Result: What the Hazard Ratio Does — and Does Not — Mean
The primary HR of 0.272 indicates a substantially lower estimated hazard of the adjudicated NMOSD attack endpoint in the compared group under the fitted Cox model. The arithmetic complement corresponds to a 72.8% lower estimated hazard relative to the reference group.
The HR does not state how many fewer participants experienced an attack, how many additional days participants remained attack-free, or the absolute probability of an attack by Day 197. Those quantities require absolute event or survival estimates, which are not provided in the ClinicalTrials.gov record.
The 95% CI of 0.1496–0.4961 provides a measure of statistical uncertainty around the estimated HR. It is important because the point estimate alone does not convey how precisely the treatment comparison was estimated.
The P-value of <0.0001 provides evidence against the relevant null hypothesis in the reported superiority analysis. It does not mean that there is a <0.01% probability that the treatment effect is due to chance, nor does it quantify clinical importance.
19. What the Secondary Results Add
The secondary analyses broaden the statistical picture beyond the primary attack endpoint.
Disability
The EDSS worsening analysis reports an OR of 0.352, with a 95% CI of 0.1755–0.7059 and P = 0.0033.
Visual function
The low-contrast visual acuity analysis reports a mean difference of 0.134, with a 95% CI of -2.0254–2.2941 and P = 0.9026.
MRI activity
The cumulative active MRI lesion analysis reports a rate ratio of 0.566, with a 95% CI of 0.3866–0.8279 and P = 0.0034.
Hospitalization
The NMOSD-related in-patient hospitalization analysis reports a rate ratio of 0.317, with a 95% CI of 0.1257–0.7972 and P = 0.0146.
These endpoints address different dimensions of disease activity and patient outcomes. Their statistical estimates are complementary rather than interchangeable. In particular, the continuous visual-acuity result should not be judged using the same effect-size interpretation used for the hazard ratio or rate ratios.
20. Clinical Interpretation vs Statistical Interpretation
Statistical interpretation
The primary Cox analysis produced an HR of 0.272 with a 95% CI of 0.1496–0.4961 and P < 0.0001. Three of the four reported secondary analyses also have effect intervals that exclude their null value, while the low-contrast visual acuity mean-difference interval includes zero.
Clinical interpretation
The ClinicalTrials.gov record address several clinically distinct outcomes: adjudicated NMOSD attacks, EDSS worsening, visual acuity, MRI lesions, and hospitalizations. Clinical meaning requires interpreting each outcome on its own scale rather than combining the estimates into a single numerical benefit measure.
21. Important Limitations and Interpretation Issues
- Registry-level detail: the ClinicalTrials.gov record provides the principal endpoint definitions and posted statistical-analysis results but not the full protocol or statistical analysis plan.
- Hazard-ratio assumptions: the primary analysis uses a Cox model. A single HR is most straightforward to interpret when the proportional-hazards assumption is reasonably appropriate; the ClinicalTrials.gov record does not report a formal assessment.
- Absolute effects: the primary statistical record supplies an HR and CI but does not provide absolute event probabilities or a median time to attack.
- Secondary multiplicity: four secondary analyses are reported, but the ClinicalTrials.gov record does not specify a multiplicity-adjustment strategy for those endpoints.
- Different estimands: HRs, ORs, mean differences, and rate ratios measure different quantities and should not be directly compared by numerical magnitude.
- Analysis populations: the posted efficacy analyses use the defined ITT population, while the ClinicalTrials.gov record is presented as affected participants divided by arm-specific numbers at risk.
- Safety follow-up reporting: only 1 participant from the Placebo/Inebilizumab arm and no participant from the Inebilizumab/Inebilizumab arm rolled over to SFP, and the registry excluded SFP from Participant Flow because of the stated EudraCT posting limitation.
- Missing-data methods: no specific imputation method is identified in the ClinicalTrials.gov record.
- Generalizability: the registry extract does not provide a baseline-characteristics table or detailed eligibility distribution, limiting additional assessment of population representativeness on this page.
22. Why This Trial Matters Statistically
N-MOmentum is a useful teaching case because the posted analyses demonstrate how the same randomized comparison can require different statistical models depending on the endpoint's structure.
| Concept | How it appears in N-MOmentum |
|---|---|
| Randomization | Participants were randomized in a parallel-group design. |
| Quadruple masking | The registered masking level is QUADRUPLE. |
| ITT analysis | Posted statistical analyses use the defined ITT population. |
| Time-to-event analysis | The primary endpoint measures time to an AC-determined NMOSD attack. |
| Cox model | The primary endpoint is analyzed using Cox proportional-hazards regression. |
| Hazard ratio | The primary treatment effect is reported as HR 0.272. |
| Logistic regression | EDSS worsening is analyzed as a binary endpoint using logistic regression. |
| Odds ratio | The EDSS analysis reports OR 0.352. |
| ANCOVA | Change from baseline in low-contrast visual acuity is analyzed using ANCOVA. |
| Mean difference | The visual-acuity analysis reports a mean difference of 0.134. |
| Negative Binomial Regression | Active MRI lesions and NMOSD-related hospitalizations are analyzed using the reported Negative Binomial Regression method. |
| Rate ratio | MRI lesions and hospitalizations are reported using rate ratios. |
| Confidence intervals | All five posted analyses report two-sided 95% confidence intervals. |
| P-values | Each posted analysis includes a P-value. |
| Superiority testing | The statistical analyses posted on ClinicalTrials.gov identify superiority as the hypothesis type. |
23. Related Tutorials
Learn more about the methods used in this trial:
24. Related Statistical Calculators
25. Sources
- ClinicalTrials.gov: N-MOmentum, NCT02200770.
- PubMed: PMID 39222408.
- PubMed: PMID 38760098.
- PubMed: PMID 37221052.
- PubMed: PMID 36370634.
- PubMed: PMID 35158461.
Continue through the Clinical Biostats statistical library
Use the trial's endpoint and model structure as a starting point for deeper study of survival analysis, regression, confidence intervals, and clinical-trial methodology.
26. Record Summary
N-MOmentum provides a compact example of how clinical-trial statistics adapt to different endpoint structures. The primary outcome is a time-to-event endpoint analyzed with a Cox proportional-hazards model, producing an HR of 0.272 with a 95% CI of 0.1496–0.4961 and P < 0.0001. The posted secondary analyses extend the statistical assessment through logistic regression for EDSS worsening, ANCOVA for change in low-contrast visual acuity, and Negative Binomial Regression for active MRI lesions and NMOSD-related hospitalizations.
The most useful statistical reading is therefore not simply that several P-values are small. It is that each result answers a different question and uses an effect measure matched to the endpoint: hazard ratio for time to attack, odds ratio for a binary disability outcome, mean difference for a continuous score, and rate ratio for count-type outcomes. Confidence intervals provide the corresponding uncertainty, while the ITT framework preserves the randomized treatment comparison used by the posted efficacy analyses.