This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the results posted in the ClinicalTrials.gov record; calculations added by Clinical Biostats are labelled as such.
1. Trial at a Glance
This page covers the registry results for A Study of Ocrelizumab in Comparison With Interferon Beta-1a (Rebif) in Participants With Relapsing Multiple Sclerosis (NCT01247324), a randomized phase 3 study. The primary outcome is Annualized Relapse Rate (ARR) in Participants With Relapsing Multiple Sclerosis (MS) at 96 Weeks.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT01247324 |
| Phase | Phase 3 |
| Condition | Relapsing Multiple Sclerosis |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | Triple |
| Enrollment | 821 |
| Arms in the primary outcome | 2 |
| Lead sponsor | Hoffmann-La Roche |
| Results posted | Yes |
2. Clinical Question
Population
Participants with relapsing multiple sclerosis were eligible for the study. The primary analysis population was the intent-to-treat population, defined in the registry record as all randomized participants. The record does not report further eligibility details in the trial summary.
Intervention
Participants assigned to ocrelizumab received 600 mg intravenously: 300 mg infusions on Days 1 and 15 for the first dose, followed by a single 600 mg infusion every 24 weeks. They also received placebo injections matching the interferon beta-1a schedule.
Comparator
The comparator was interferon beta-1a 44 mcg by subcutaneous injection three times per week, with placebo infusions matching the ocrelizumab infusion schedule. The double-dummy approach used matched placebo treatments to maintain masking across the different routes and schedules.
Primary question
The primary question was whether the annualized relapse rate differed between the randomized groups at 96 weeks. The registry defines this rate as total relapses in a treatment group divided by total participant-years of exposure to that treatment. The primary analysis included all randomized participants.
3. Trial Design
This was a Phase 3 randomized, parallel-group study sponsored by Hoffmann-La Roche. The registry record lists triple masking and describes the study as double-blind and double-dummy. The treatment design used matching placebo infusions or injections so participants in both groups followed an infusion schedule and an injection schedule, despite receiving different active treatments.
The planned double-blind treatment period was 96 weeks. Participants could enter a single-group, active-treatment, open-label extension after completing that period, provided they met the extension eligibility criteria. The registry reports an enrollment of 821 participants. The primary efficacy population was the intent-to-treat population, which included all randomized participants; the posted primary arm counts were 411 for interferon beta-1a and 410 for ocrelizumab.
Ocrelizumab was given intravenously, beginning with two 300 mg infusions on Days 1 and 15 and continuing with a single 600 mg infusion every 24 weeks. Interferon beta-1a was given by subcutaneous injection three times per week at 44 mcg. The double-blind period included matching placebo treatment for the route not used for active treatment in each group.
The primary endpoint was assessed at Week 96. The registry also lists secondary outcomes involving confirmed disability progression, confirmed disability improvement, MRI lesion counts, functional composite score and brain volume. Some disability-progression outcomes have a Week 108 timeframe. The reason for this timeframe, details of follow-up procedures, and the protocol’s sample-size justification are not reported in the registry record.
4. Participant Flow and Accounting
The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.
Before assignment: In error, the study completion status for 5 participants in WA21092 was registered as 'Sponsor Termination' in the study eCRF. However, all 5 participants were successfully enrolled into MN43964 and thus have completed the WA21092 study.
Overall Study
| Stage | Interferon Beta-1a 44 mcg SC | Ocrelizumab |
|---|---|---|
| STARTED | 411 | 410 |
| Entered Open-Label Extension | 326 | 352 |
| COMPLETED | 226 | 239 |
| NOT COMPLETED | 185 | 171 |
Reasons for not completing this period:
| Reason | Interferon Beta-1a 44 mcg SC | Ocrelizumab |
|---|---|---|
| Death | 6 | 10 |
| Physician Decision | 2 | 3 |
| Non-compliance with study drug | 3 | 0 |
| Lack of Efficacy | 24 | 14 |
| Non-compliance | 3 | 2 |
| Pregnancy | 5 | 10 |
| Protocol Violation | 1 | 1 |
| Reason not specified | 42 | 32 |
| Consent withdrawn by participant | 41 | 56 |
| Adverse Event | 49 | 35 |
| Lost to Follow-up | 3 | 6 |
| Study Terminated by Sponsor | 4 | 1 |
Analysis population for the primary outcome, as defined in the registry: Intent-to-treat (ITT) population included all randomized participants in the study.
5. Baseline Characteristics
The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.
Population: Intent-to-treat (ITT) population included all randomized participants in the study.
| Characteristic | Unit | Interferon Beta-1a 44 mcg SC | Ocrelizumab | Total |
|---|---|---|---|---|
| Participants analysed | 411 | 410 | 821 | |
| Age, Continuous | years | 36.9 (SD 9.3) | 37.1 (SD 9.3) | 37.0 (SD 9.3) |
| Sex: Female, Male: Female | Participants | 272 | 270 | 542 |
| Sex: Female, Male: Male | Participants | 139 | 140 | 279 |
6. Primary Endpoint
The registry lists the primary outcome as Annualized Relapse Rate (ARR) in Participants With Relapsing Multiple Sclerosis (MS) at 96 Weeks, measured over Week 96.
| Element | Registry entry |
|---|---|
| Result type | Number |
| Measure of spread | 95% Confidence Interval |
| Unit | relapses/participant year of treatment |
| Endpoint classification | not reduced to a single comparison |
The primary endpoint was annualized relapse rate (ARR) at 96 weeks, expressed as relapses per participant-year of treatment. It is a rate rather than a simple proportion: the numerator is the total number of relapses recorded in a treatment group, and the denominator is the total participant-years of exposure in that group. This accounts for the amount of time participants contributed to observation, rather than treating every participant as if they had exactly the same follow-up.
Relapse rate is used to summarize the frequency of relapses over a period of treatment. The annualized form makes rates interpretable on a per-year exposure scale, while the analysis can use each participant’s available exposure time. The registry identifies the intent-to-treat population as the analysis population, including all randomized participants.
The posted arm-specific estimates were 0.292 relapses per participant year for interferon beta-1a and 0.156 for ocrelizumab. Their 95% confidence intervals were 0.235 to 0.361 and 0.122 to 0.2, respectively. The reported comparison used a negative binomial model and gave a rate ratio with a 95% confidence interval. The rate estimates describe each group; the modeled comparison estimates the relative rate between groups while accounting for the model’s structure.
ARR is not a direct measure of disability, MRI activity, or quality of life. Those domains appear as separate outcomes in the registry. The endpoint therefore answers a focused question about relapse frequency over the treatment period, not every aspect of multiple sclerosis or every possible treatment effect.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The primary outcome estimates were 0.292 relapses per participant year in the interferon beta-1a group and 0.156 in the ocrelizumab group. The respective 95% confidence intervals were 0.235 to 0.361 and 0.122 to 0.2. These intervals describe uncertainty around the group-specific rate estimates; they are not ranges containing a fixed proportion of individual participants’ relapse rates.
The registry’s reported comparison used a negative binomial model. It gave a rate ratio of 0.536, with a 95% confidence interval from 0.4 to 0.719 and p < 0.001. The reported interval lies below 1, and the p-value indicates that the model-based comparison was inconsistent with a null rate ratio of 1 under the test’s assumptions. This is evidence about the statistical comparison, not a measure of the probability that either treatment is effective or ineffective.
The rate ratio is a relative comparison, whereas the arm estimates are rates on the outcome’s original scale. Both are useful, but they answer different questions. The rate ratio describes the relative rate in the comparison; the arm estimates show the observed scale of relapse rates in each randomized group. Neither by itself describes how individual participants experienced relapses. The registry does not report participant-level relapse distributions in the posted results.
Confidence intervals communicate precision as well as direction. A narrower interval would indicate a more precisely estimated quantity, while a wider interval would leave more uncertainty about its value. The interval around the rate ratio excludes 1, but it still spans a range of possible relative rate values. The p-value does not replace this range and does not establish the clinical importance of the difference.
The result should be read in the context of the endpoint definition and analysis population. ARR is based on total relapses divided by total participant-years, and the stated primary population includes all randomized participants. The registry does not report details of how missing relapse data, incomplete follow-up, or deviations from assigned treatment were handled. Those details can affect interpretation of an exposure-based rate and are relevant when judging how well the analysis reflects the randomized comparison.
The registry also reports disability-related and other secondary outcomes, but they measure different domains and use different statistical approaches. Their results do not change the definition of the primary endpoint. The strength of inference from secondary analyses depends in part on how many outcomes were tested and whether testing was adjusted for multiplicity; such a plan is not reported in the registry record. Accordingly, the primary model-based result is the clearest single statistical comparison reported for the trial, while the broader pattern requires more cautious interpretation.
9. Statistical Methods Explained
The primary endpoint is a count-derived rate, so a model for event counts is a natural analytic approach. The registry reports a negative binomial model for the primary comparison. In general, count outcomes can vary more than a simple Poisson model assumes; this extra variation is called over-dispersion. A negative binomial model allows greater variation in the counts and can be useful when participants’ event counts are heterogeneous. The registry record does not report its diagnostic checks, model specification details, or covariates.
What does the rate ratio describe?
A rate ratio compares event rates between groups. A value of 1 corresponds to equal rates under the comparison’s definition; a value below 1 indicates a lower rate in the numerator group relative to the denominator group, while a value above 1 indicates a higher rate. The registry reports a rate ratio of 0.536, with a 95% confidence interval from 0.4 to 0.719. The treatment-group ordering and modeling details are part of interpreting which rate is in the numerator. The arm-specific rates give useful context: 0.292 for interferon beta-1a and 0.156 for ocrelizumab.
How should the confidence interval and p-value be read?
A confidence interval gives a range of values compatible with the observed data and the method’s assumptions at the stated confidence level. It expresses estimation uncertainty, not the range of individual outcomes and not the probability that the true value lies in that particular interval. The primary comparison’s 95% confidence interval excludes 1. The reported p-value, p < 0.001, quantifies how unusual data at least as incompatible with the null hypothesis would be if that null and the analysis assumptions held. It is not the probability that the null hypothesis is true.
General statistical interpretation should consider the estimate, its interval, the endpoint, and the analysis method together. A small p-value alone does not indicate the size or clinical importance of an effect. The registry does not report a threshold for a clinically meaningful ARR difference, so statistical evidence should not be equated with clinical importance.
Why are the other outcomes analyzed differently?
The registry includes outcomes with different data types. Time to confirmed disability progression is a time-to-event outcome, for which the reported method was Log Rank and the comparison was expressed as a hazard ratio. A hazard ratio compares the event rates over follow-up under the method’s assumptions; it is not the same as a ratio of medians. For the Week 108 CDP outcomes, the reported hazard ratio confidence intervals and p-values come from the sponsor’s analysis. Median event times are not estimable in the posted arm summaries, so the full ranges do not supply median estimates.
Confirmed disability improvement is reported as a percentage of participants. The sponsor’s comparison used a stratified Cochran-Mantel-Haenszel chi-squared test and reported a stratified relative risk. A simple unadjusted recalculation can use the posted counts, but it does not reproduce the stratified analysis. The two estimates may differ because they answer the comparison using different methods, including different handling of stratification.
How should the planning illustrations be understood?
Planning illustrations based on an observed effect are hypothetical design calculations, not observed trial results and not retrospective measures of the trial’s power. The stated assumptions are a two-sided significance level of 0.05, equal allocation, and a normal approximation. The scenarios reduce the observed effect to stated fractions. For time-to-event outcomes, the illustrations refer to total events; for other outcomes, they refer to participants per arm. The required information is different because event-driven analyses depend on the number of events, whereas binary-outcome planning depends on the number of participants and the assumed event proportions.
These illustrations are conditional on the observed effect and the stated assumptions. They do not show that the trial was designed using those assumptions, because the protocol’s sample-size justification is not reported in the registry record. They should not be interpreted as observed power or as proof that a future study of a particular size will reproduce the posted effect.
What methodological details remain unclear?
The registry record does not report the model’s full specification, missing-data strategy, interim-analysis plan, or whether the primary hypothesis was adjusted for any other testing. The primary analysis population is specified as intent-to-treat, but equivalent details for each secondary analysis are not consistently stated in the record. These omissions limit independent assessment of analytic choices and of the potential effect of incomplete outcome data.
10. Analyses Reported by the Sponsor
| Method | Estimate | 95% CI | P-value |
|---|---|---|---|
| Negative Binomial Model | Rate Ratio 0.536 | 0.4 to 0.719 | < 0.001 |
The registry reports the primary comparison from a negative binomial model: rate ratio 0.536, 95% confidence interval 0.4 to 0.719, and p < 0.001. This is the sponsor’s reported model-based analysis and should be the main inferential result for the primary endpoint. The arm-specific ARR estimates offer descriptive context, but they do not replace the model-based comparison.
For confirmed disability improvement, the reported analysis used a stratified Cochran-Mantel-Haenszel chi-squared test and gave a relative risk of 1.61, with a 95% confidence interval from 1.11 to 2.33 and p = 0.011. An unadjusted recalculation from the posted summary counts gives a risk difference of 0.076, with a 95% confidence interval from 0.018 to 0.134 and p = 0.011. These are not identical estimands or methods: the recalculation is unadjusted, while the sponsor’s analysis is stratified and reports a relative risk. The small numerical difference in p-values is therefore not a direct conflict.
The registry does not flag a discrepancy between sponsor results and recalculated results. Where a sponsor model is reported, its analysis remains the primary reference because a simple recalculation cannot reproduce model adjustments or the full analysis specification. The time-to-event secondary comparisons were also reported by the sponsor using Log Rank methods.
11. Multiplicity, Missing Data and Analysis Populations
The registry defines the primary analysis population as intent-to-treat, including all randomized participants. That specification supports interpretation of the primary comparison as based on randomized assignment, although the record does not provide the details needed to establish how every participant’s outcome data contributed to the fitted model.
Missing-data methods are not reported in the registry record. The record does not explain how participants with incomplete follow-up, missing relapse information, or unavailable secondary outcomes were handled. It also does not report whether imputation, censoring rules, sensitivity analyses, or other procedures were used. For a rate based on participant-years of exposure, understanding exposure time and follow-up is important; the posted endpoint definition gives the rate calculation but not the full missing-data strategy.
The record identifies multiple outcomes, including a primary endpoint and a range of secondary endpoints, but a multiplicity plan is not reported. Multiplicity refers to the increased chance of a false-positive finding when several statistical hypotheses are tested. General approaches include defining a hierarchy of tests or adjusting significance thresholds, but the registry does not state that any particular approach was used. The reported secondary p-values therefore should not automatically be read as if they were the sole planned test.
Interim analyses and any stopping rules are not reported in the registry record. The sample-size justification and the assumptions used for the original design are also not reported. The planning illustrations based on observed effects are not evidence of the trial’s original design or observed power.
The primary endpoint is explicitly associated with the intent-to-treat population. The record does not give comparable population definitions for every secondary outcome, nor does it provide detailed analysis-set membership for the open-label extension. These gaps make it difficult to assess whether missingness or different analysis populations could affect how results across endpoints should be compared.
12. Safety Results
The registry reports serious adverse events as affected participants divided by participants at risk.
| Arm | Participants with serious adverse events | Participants at risk | Percentage |
|---|---|---|---|
| Interferon Beta -1a + Placebo (Double Blind Period) | 32 | 409 | 7.8% |
| Ocrelizumab + Placebo (Double Blind Period) | 28 | 408 | 6.9% |
| Interferon Beta-1a + Placebo (Open Label Extension) | 109 | 326 | 33.4% |
| Ocrelizumab + Placebo (Open Label Extension) | 116 | 352 | 33.0% |
Deaths
| Arm | Deaths (participants affected of at risk) |
|---|---|
| Interferon Beta -1a + Placebo (Double Blind Period) | 1 of 409 |
| Ocrelizumab + Placebo (Double Blind Period) | 0 of 408 |
| Interferon Beta-1a + Placebo (Open Label Extension) | 6 of 326 |
| Ocrelizumab + Placebo (Open Label Extension) | 11 of 352 |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 5% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| Interferon Beta -1a + Placebo (Double Blind Period) | 257 of 409 |
| Ocrelizumab + Placebo (Double Blind Period) | 241 of 408 |
| Interferon Beta-1a + Placebo (Open Label Extension) | 285 of 326 |
| Ocrelizumab + Placebo (Open Label Extension) | 286 of 352 |
Most frequent serious adverse event terms
| Event term | Organ system | Interferon Beta -1a + Placebo (Double Blind Period) | Ocrelizumab + Placebo (Double Blind Period) | Interferon Beta-1a + Placebo (Open Label Extension) | Ocrelizumab + Placebo (Open Label Extension) |
|---|---|---|---|---|---|
| COVID-19 PNEUMONIA | Infections and infestations | 0 of 409 | 0 of 408 | 17 of 326 | 20 of 352 |
| PNEUMONIA | Infections and infestations | 0 of 409 | 0 of 408 | 11 of 326 | 6 of 352 |
| COVID-19 | Infections and infestations | 0 of 409 | 0 of 408 | 6 of 326 | 8 of 352 |
| MULTIPLE SCLEROSIS RELAPSE | Nervous system disorders | 3 of 409 | 0 of 408 | 3 of 326 | 5 of 352 |
| PYREXIA | General disorders | 0 of 409 | 0 of 408 | 1 of 326 | 5 of 352 |
| APPENDICITIS | Infections and infestations | 2 of 409 | 0 of 408 | 1 of 326 | 3 of 352 |
| URINARY TRACT INFECTION | Infections and infestations | 1 of 409 | 0 of 408 | 4 of 326 | 1 of 352 |
| UTERINE LEIOMYOMA | Neoplasms benign, malignant and unspecified (incl cysts and polyps) | 2 of 409 | 0 of 408 | 3 of 326 | 1 of 352 |
| ANKLE FRACTURE | Injury, poisoning and procedural complications | 1 of 409 | 0 of 408 | 2 of 326 | 2 of 352 |
| INVASIVE DUCTAL BREAST CARCINOMA | Neoplasms benign, malignant and unspecified (incl cysts and polyps) | 0 of 409 | 2 of 408 | 2 of 326 | 1 of 352 |
Most frequent non-serious adverse event terms
| Event term | Organ system | Interferon Beta -1a + Placebo (Double Blind Period) | Ocrelizumab + Placebo (Double Blind Period) | Interferon Beta-1a + Placebo (Open Label Extension) | Ocrelizumab + Placebo (Open Label Extension) |
|---|---|---|---|---|---|
| INFUSION RELATED REACTION | Injury, poisoning and procedural complications | 31 of 409 | 126 of 408 | 83 of 326 | 71 of 352 |
| URINARY TRACT INFECTION | Infections and infestations | 56 of 409 | 52 of 408 | 80 of 326 | 101 of 352 |
| UPPER RESPIRATORY TRACT INFECTION | Infections and infestations | 35 of 409 | 59 of 408 | 89 of 326 | 90 of 352 |
| NASOPHARYNGITIS | Infections and infestations | 40 of 409 | 42 of 408 | 68 of 326 | 73 of 352 |
| COVID-19 | Infections and infestations | 0 of 409 | 0 of 408 | 96 of 326 | 96 of 352 |
| HEADACHE | Nervous system disorders | 54 of 409 | 33 of 408 | 46 of 326 | 40 of 352 |
| MULTIPLE SCLEROSIS RELAPSE | Nervous system disorders | 0 of 409 | 0 of 408 | 82 of 326 | 84 of 352 |
| SINUSITIS | Infections and infestations | 25 of 409 | 18 of 408 | 41 of 326 | 49 of 352 |
| ARTHRALGIA | Musculoskeletal and connective tissue disorders | 30 of 409 | 27 of 408 | 46 of 326 | 29 of 352 |
| DEPRESSION | Psychiatric disorders | 26 of 409 | 27 of 408 | 24 of 326 | 45 of 352 |
Serious events with the highest arm-level rate: covid-19 pneumonia (5.7%), pneumonia (3.4%), covid-19 (2.3%), pyrexia (1.4%), multiple sclerosis relapse (1.4%).
The registry reports serious adverse event counts separately for the double-blind period and the open-label extension. During the double-blind period, serious adverse events affected 32 of 409 participants in the interferon beta-1a plus placebo group and 28 of 408 in the ocrelizumab plus placebo group. The record also reports other adverse events for 257 of 409 and 241 of 408 participants, respectively. These are counts of participants with events, not event rates adjusted for time at risk.
In the open-label extension, serious adverse events affected 109 of 326 participants in the interferon beta-1a plus placebo group and 116 of 352 in the ocrelizumab plus placebo group. Other adverse events affected 285 of 326 and 286 of 352, respectively. Death counts were 1 and 0 during the double-blind period, and 6 and 11 in the extension groups. These counts are descriptive and do not establish that an event was caused by a treatment.
The extension groups represent a distinct period, and entry was optional for participants who completed double-blind treatment and met eligibility criteria. The counts should therefore not be treated as a direct randomized comparison across periods. The registry does not provide exposure-adjusted safety rates or sufficient detail in these figures to account for duration at risk. Serious event terms and listed percentages also require attention to their stated grouping and period when interpreting them; they should not be combined into a single overall comparison.
13. What the Result Does, and Does Not, Mean
The primary result is a model-based comparison of annualized relapse rates over the study period. The registry reports lower ARR in the ocrelizumab arm than in the interferon beta-1a arm, with a rate ratio of 0.536 and a 95% confidence interval from 0.4 to 0.719. The confidence interval excludes 1, and the reported p-value is p < 0.001. These statistics describe the posted trial data under the stated model; they do not, on their own, establish the size of a clinically meaningful benefit.
The result applies to the study population and treatment schedules in the registry record. The trial enrolled participants with relapsing multiple sclerosis and used a double-blind, double-dummy comparison. It does not directly answer questions about other populations, treatment schedules, outcomes, or follow-up periods not represented in the record.
ARR focuses on relapse frequency per participant-year. It does not summarize disability progression, MRI findings, daily functioning, or every possible safety outcome. The registry includes secondary outcomes in some of those domains, but they have their own uncertainty and multiplicity concerns. It would be inappropriate to treat one endpoint as a complete account of all outcomes.
The results also do not imply that every participant experienced the group-average rate, nor do the group estimates predict an individual participant’s future course. Confidence intervals describe uncertainty in estimates, not individual variation. The registry record does not report a clinical-importance threshold for ARR, full missing-data methods, or a complete multiplicity plan, all of which constrain interpretation.
The serious adverse event counts are descriptive and separated by treatment period. They should not be used to infer causality or to make an unadjusted comparison across the double-blind period and open-label extension. The open-label extension is not described as a new randomized comparison. Overall, the posted data support a statistical comparison of the specified outcome in the trial, while leaving other clinical and methodological questions unresolved.
14. Why This Trial Matters Statistically
This trial illustrates why the choice of statistical scale matters. ARR is an exposure-based rate, not simply the fraction of participants who experienced at least one relapse. The denominator of participant-years accounts for differing observation time, and the negative binomial model provides a rate-ratio comparison for count data. Reading the model estimate alongside the arm-specific rates helps distinguish a relative comparison from the observed rates on the endpoint’s own scale.
It also shows why endpoints with different data structures need different analyses. Relapse frequency was analyzed with a count model, disability progression with time-to-event methods, and disability improvement as a participant percentage. Each analysis targets a different quantity. A hazard ratio is not a median, and a relative risk is not a risk difference. Statistical results are clearer when the estimand and method are stated together.
The double-dummy design is another useful design feature to understand statistically. The groups received treatments by different routes and schedules, and matched placebo treatments were used to support masking. Random allocation and masking address different sources of bias: randomization supports comparability at assignment, while masking can reduce effects of knowledge of assigned treatment on behavior or assessment. The registry does not describe the practical success of masking.
Finally, the record demonstrates the limits of interpreting a collection of endpoints without a reported multiplicity plan or missing-data strategy. A convincing primary comparison can coexist with uncertainty about the evidential status of secondary findings and the influence of incomplete data. The distinction between sponsor analyses and simple recalculations also matters: a summary-statistic recalculation may help readers understand an effect, but it cannot reproduce a stratified or model-based analysis in full.
15. Limitations
- The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
- Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.
The registry record does not report the full statistical analysis plan, including the negative binomial model specification, covariates, diagnostics, or sensitivity analyses. Although the primary analysis population is identified as intent-to-treat, details about how missing relapse data and incomplete exposure were handled are not reported. These omissions limit assessment of robustness.
The record does not report the original sample-size justification, interim-analysis plan, stopping rules, or multiplicity adjustment. The number of secondary outcomes and differing outcome types make those omissions relevant when interpreting secondary p-values. The observed-effect planning illustrations are not a substitute for the original design rationale.
The planned double-blind treatment period was 96 weeks, while disability-progression outcomes are reported with a Week 108 timeframe. The registry does not explain the relationship between these timeframes in enough detail to assess follow-up and censoring procedures. The posted summaries also do not provide participant-level data, so subgroup patterns and distributional features cannot be examined from these results.
The double-blind comparison and open-label extension should be distinguished. The extension was optional for eligible completers, and its treatment groups are not presented as a new randomized comparison. Safety counts from the extension therefore cannot be directly compared with double-blind counts as if they arose from the same randomized population and follow-up conditions. Exposure-adjusted safety rates are not reported in these figures.
Finally, the trial concerns participants with relapsing multiple sclerosis and the specific schedules described in the registry. The record does not establish how well the results generalize to populations or treatment settings outside that study description. The absence of a reported clinical-importance threshold also means statistical evidence cannot, from this record alone, be translated into a judgment about clinical importance.
16. Related Tutorials
17. Related Statistical Calculators
18. Sources
- ClinicalTrials.gov: A Study of Ocrelizumab in Comparison With Interferon Beta-1a (Rebif) in Participants With Relapsing Multiple Sclerosis — NCT01247324.
- Registry record last updated 2024-03-04; status last verified 2024-02.
- Results first posted 2017-07-18.
- Data retrieved from ClinicalTrials.gov on 2026-11-09.
- Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.
Continue through the Clinical Biostats statistical pathway
Use this trial's endpoint as a starting point for deeper study of effect measures, confidence intervals, and sample-size planning.
19. Record Summary
The registry record describes a Phase 3 randomized, parallel-group comparison of ocrelizumab and interferon beta-1a in participants with relapsing multiple sclerosis. The study enrolled 821 participants and planned 96 weeks of double-blind treatment, with an optional open-label extension for eligible completers.
The primary endpoint was annualized relapse rate at Week 96 in the intent-to-treat population. The reported negative binomial model gave a rate ratio of 0.536, with a 95% confidence interval from 0.4 to 0.719 and p < 0.001. Secondary outcomes covered disability progression and improvement, MRI measures, functional score, and brain volume, although not all listed outcomes have detailed results in the record.
Important details, including the missing-data approach, multiplicity plan, interim analyses, and original sample-size justification, are not reported in the registry record. Safety counts are reported separately for the double-blind period and open-label extension and should be interpreted as descriptive figures rather than causal comparisons.