Tocilizumab in the Treatment of Refractory Polymyositis and Dermatomyositis: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Tocilizumab in the Treatment of Refractory Polymyositis and Dermatomyositis (NCT02043548), a randomized phase 2 study. The primary outcome is Compare the Average Total Improvement Scores at Visits 2 Through 7 During the 6-month Treatment Period Between the Treatment and Placebo Arms.

2. Clinical Question

3. Trial Design

The registry describes a phase 2, randomized, triple-masked, parallel-group trial. It lists 36 participants as enrolled, with 18 assigned to the tocilizumab arm and 18 to the placebo arm in the posted outcome results. The trial status is completed. The listed start date is 2014-10-01, and the primary completion date is 2019-07-30.

The tocilizumab regimen was 8mg/kg by intravenous infusion every 4 weeks at six time points. The comparator was placebo, described as having no active drug. The record does not identify who was masked, so the specific masking arrangements are not reported in the registry record. The study is described as a multicenter pilot study in adults with refractory polymyositis and dermatomyositis.

The primary analysis population is described as all participants randomized, analyzed according to intention to treat. The primary outcome was an average total improvement score across visits at weeks 4, 8, 12, 16, 20, and 24. The posted results give an arm mean and standard deviation for each group, along with a registry-reported GEE p-value. The record does not report the details of the randomization procedure, allocation concealment, trial protocol, or the original sample-size justification.

These design features shape how the results should be read. Random assignment supports a comparative interpretation within the trial, while the pilot designation and the enrolled group size signal that the estimates may be imprecise. Masking may help limit some forms of bias, but the record does not provide enough detail to assess how masking was implemented or maintained. The trial’s findings therefore need to be considered alongside the reported uncertainty and the limits of the registry information.

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.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All participants randomized are included in intention to treat analysis

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.

6. Primary Endpoint

The registry lists the primary outcome as Compare the Average Total Improvement Scores at Visits 2 Through 7 During the 6-month Treatment Period Between the Treatment and Placebo Arms, measured over Week 4, 8, 12, 16, 20, and 24.

The primary endpoint is the average total improvement score at visits 2 through 7 during the six-month treatment period. The measure adds the improvement scores from six core set measures, using the 2016 American College of Rheumatology/European League Against Rheumatism myositis response criteria. It is reported as a mean on a scale from 0 to 100, and higher scores indicate a better outcome. The registry reports that the included participants’ average total improvement scores ranged from 5.0 to 67.1.

A composite score can bring information from multiple clinical measures into one summary, making it possible to compare the overall pattern of improvement between treatment groups. Repeated visits also allow the endpoint to reflect measurements across the treatment period rather than only one visit. The posted primary outcome is an average across the specified visits; it should not be mistaken for a single-time-point score or for a measure of each component considered separately.

The endpoint is relevant to the registry’s stated question about treatment response in refractory polymyositis and dermatomyositis. Its interpretation depends on the construction of the composite, the handling of measurements across visits, and the completeness of the data. Details about missing observations and the exact statistical model specification are not reported in the registry record. The results therefore describe the posted average score comparison, while those unreported details limit a fuller assessment of how the endpoint was analyzed.

7. Results Posted in the Registry

• Group A: Tocilizumab: mean 26.40 versus 29.30 on control; difference -2.90 (95% CI -13.88 to 8.08), standardized effect d = -0.17; z-test p = 0.605.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Comparison of Individual Average Core Set Measure in Subjects Over Time Between the 2 Arms (Repeated Measures Analysis)

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Group A: Tocilizumab: mean 130.70 versus 137.30 on control; difference -6.60 (95% CI -16.15 to 2.95), standardized effect d = -0.45; z-test p = 0.175.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Group A: Tocilizumab and the control. The interval gives the range of effects that remain compatible with the data.

The registry lists secondary questions covering time to first definition of improvement, adverse events, steroid-sparing effect, individual average core set measures over time, and effect size. The posted record includes a detailed repeated-measures result for an individual average core set measure. The numerical results for the other listed secondary questions are not reported in the registry record, so a complete account of the secondary findings cannot be drawn from the listed material.

The available secondary comparison and the primary result do not supply a consistent, complete set of evidence across all secondary domains. A secondary measure can add context by examining a related but distinct outcome, yet it does not replace the prespecified primary question. Results across outcomes may also differ because the measures capture different aspects of response and use different analytic approaches.

Secondary findings generally warrant more caution when many outcomes are tested. Without adjustment for multiple comparisons, the chance of obtaining at least one apparently unusual result can increase as more hypotheses are examined. The registry record does not report whether or how multiplicity was addressed. Accordingly, unadjusted secondary results should be viewed as supportive or exploratory rather than as stand-alone confirmation of a treatment difference.

The record’s listing of secondary outcomes is useful for understanding the broader questions the trial aimed to examine. However, because several results are not reported and the multiplicity strategy is not described, the secondary information cannot establish a broad pattern of benefit or lack of benefit across all outcomes. The primary comparison remains the central result for interpreting the trial’s stated main question.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Comparison of the Time to First Definition of Improvement (DOI) Between the 2 Arms
• Count of the Adverse Events Between the Treatment and Placebo Arms.
• Comparison of the Steroid-sparing Effect (Calculated Using Prednisone Dose Equivalents) Between the Treatment and Placebo Arms
• Comparison of Individual Average Core Set Measure in Subjects Over Time Between the 2 Arms (Repeated Measures Analysis)
• Magnitude of the Effect Size Between the Both Treatment Arms

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is continuous: it is a score on a scale, reported as a mean with a standard deviation. A mean describes the group’s average score, while a standard deviation describes how dispersed individual scores are around that average. These summaries provide different information. A group can have a mean near another group’s mean while still containing wide variation among participants.

What does the mean difference represent?

The recalculated mean difference compares the tocilizumab arm mean with the placebo arm mean, using placebo as the reference. The posted arm means are 26.4 and 29.3, and the recalculated difference is -2.9. Because higher scores indicate a better outcome, a negative difference means the tocilizumab arm’s average score was lower in this descriptive comparison. This is an unadjusted comparison based on the posted summary statistics; it does not account for possible covariates or reproduce the sponsor’s repeated-measures model.

The estimated difference is only one part of the result. Its uncertainty is described by a confidence interval, and the p-value evaluates compatibility with a specified null hypothesis under the analysis assumptions. These measures answer related but different questions. A p-value is not the size of an effect, and it does not tell readers whether any effect would be clinically important. The registry record does not state a threshold for a clinically important difference.

Why are the sponsor analysis and recalculation different?

The registry reports a GEE analysis for the primary outcome, with p = 0.86. Generalized estimating equations are a general approach for analyzing correlated observations, such as repeated measurements from the same participant. Because observations within a participant may be related, an analysis that accounts for this correlation can differ from a simple comparison of group summary values. The record does not report the GEE specification, its estimate, or its confidence interval. The unadjusted recalculation therefore serves as a separate summary-statistic comparison, not as a replacement for the sponsor’s model.

The posted outcome averages cover visits at weeks 4, 8, 12, 16, 20, and 24. Repeated-measure data can be analyzed in different ways, depending on the question and model assumptions. The registry identifies GEE but does not report enough detail to determine the working correlation structure, covariates, or handling of incomplete repeated measurements. Those aspects of the statistical plan are not reported in the registry record.

What do the standard deviations tell readers?

The standard deviation is a measure of variation in scores, not a measure of the uncertainty in the estimated group mean. For the primary outcome, each arm has a standard deviation of 16.8. The similarity of these arm-level standard deviations does not imply that every participant’s score was similar, nor does it show how scores changed for a given participant over time. Individual-level distributions and within-participant trajectories are not reported in the registry record.

Variation matters because it affects the precision of a group comparison. When scores vary widely, a given difference between group means may be estimated less precisely than it would be with less variation, all else being equal. The confidence interval expresses uncertainty around the recalculated difference; the standard deviations help describe the observed spread that contributes to that uncertainty.

How should the p-values be read?

The registry-reported GEE p-value for the primary outcome is 0.86. The separate, unadjusted recalculation has p = 0.605. Both are above conventional thresholds often used in hypothesis testing, but the registry does not report the threshold or full inferential plan for its own analysis. A p-value does not measure the probability that the null hypothesis is true. It summarizes how unusual data at least as inconsistent with the null would be under a specified model if the null hypothesis and model assumptions held.

The recalculated p-value is tied to an unadjusted comparison of the posted arm summaries. The sponsor’s p-value comes from the reported GEE analysis. Differences in analysis approach can yield different estimates of uncertainty and different p-values. Since the GEE estimate and confidence interval are not reported, the record does not allow a complete side-by-side comparison of the two methods.

Why does over-dispersion matter in general?

As a general statistical point, over-dispersion occurs when data vary more than a model assumes. If a model underestimates variation, its standard errors may be too small and its confidence intervals too narrow, which can make results appear more precise than they are. The primary endpoint is continuous, and the registry does not report an over-dispersion assessment or indicate that this was a concern. This general issue should not be taken as a claim that over-dispersion occurred in this trial.

More broadly, statistical models rely on assumptions about how outcomes vary and how observations are related. Readers can assess the reported result most confidently when the model, its assumptions, and its diagnostic checks are described. Those details are not reported in the registry record for the GEE analysis.

11. Confidence Intervals and Precision

The recalculated primary mean difference is -2.9 points, with a 95% confidence interval from -13.88 to 8.076. A confidence interval gives a range of values compatible with the observed data under the analysis method and its assumptions. It is not a range containing a stated proportion of individual participants’ scores, and it is not a probability statement about the true difference after the data have been observed.

The interval spans zero, so the simple unadjusted comparison does not clearly distinguish a lower average score in the tocilizumab arm from a higher one. It also includes differences in both directions that are larger in magnitude than the point estimate. This breadth signals limited precision: the point estimate alone should not be read as a definitive estimate of the size or direction of any underlying difference.

The arm means and standard deviations provide context for the interval. Both groups have a standard deviation of 16.8, while their posted means are relatively close. The standard deviations describe variation among scores; the confidence interval describes uncertainty around the estimated difference. These are not interchangeable quantities.

The interval is from a recalculation using posted summary statistics, not the sponsor’s GEE model. The registry reports a GEE p-value but no GEE estimate or confidence interval, so the precision of the sponsor analysis cannot be assessed from the reported values. The confidence interval should therefore be interpreted specifically as the uncertainty around the unadjusted recalculated difference. It does not capture every possible source of uncertainty, including limitations from missing-data handling or model specification that are not described in the registry record.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The sample-size figures are planning illustrations based on the observed effect, not the trial’s original design assumptions. The registry does not report the trial’s own sample-size justification or original design assumptions. The planning illustration uses a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. It considers effects equal to the observed effect and fractions of that effect. The figures are participants per arm, not total event counts.

The 80% power column describes the number of participants per arm estimated to give an 80% chance of detecting the specified effect under the illustration’s assumptions, if that effect were the true one. The 90% power column applies the same idea with a greater chance of detection, and therefore generally requires more participants per arm. These are planning quantities, not guarantees about what a future study would find.

The rows for smaller fractions of the observed effect require more participants per arm. Smaller differences are harder to distinguish from random variation under otherwise similar assumptions, so greater sample sizes are needed to detect them at the stated power level. The figures depend on the observed effect used as the starting point. Since the estimate from this trial is uncertain, projections based on it should be treated cautiously.

This is not observed power. It is a planning illustration using an effect estimate from the posted results; it does not measure the probability that the completed trial detected a true effect. The curve should be read as a visual display of the same planning relationship between assumed effect size and required participants per arm, rather than as an additional trial result.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = -2.900; pooled standard deviation σ = 16.800.
2. n = 2 × (1.96 + 0.842)² × 16.800² / -2.900² = 526.8, rounded up to 527 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
delta <- -2.9                      # observed difference in means
sigma <- 16.8                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(2043548)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 527 (80% power), 706 (90% power)
 75% of observed effect: n per arm = 937 (80% power), 1,254 (90% power)
 50% of observed effect: n per arm = 2,108 (80% power), 2,822 (90% power)
power.t.test: 527.8 
Simulated power at n = 527 per arm: 0.801 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Analyses Reported by the Sponsor

The registry reports the sponsor’s primary comparison as a GEE analysis, with p = 0.86. It does not provide the model estimate or confidence interval for that analysis. The separate recalculation from the posted group summaries gives a mean difference of -2.9, a 95% confidence interval from -13.88 to 8.076, and an unadjusted p-value of 0.605.

These are not duplicate versions of the same calculation. The recalculation is a simple unadjusted comparison of the posted arm summaries. The GEE result is the registry-reported analysis and may account for repeated observations in a way that the summary-statistic calculation does not. The registry does not report enough model detail to explain precisely why the results differ or to reproduce the GEE analysis.

The discrepancy flag is false, and the registry-reported result remains the sponsor’s analysis. The recalculation offers a transparent comparison based on the posted means and standard deviations, but it should not be presented as though it were the sponsor’s model-based estimate. Both reported p-values indicate that the respective analyses do not provide strong evidence against their null comparison, while neither establishes that the groups are equivalent or that a difference is absent.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists multiple outcome measures and analyses, including the primary comparison and several secondary questions. Testing multiple outcomes creates a multiplicity issue: the more hypotheses examined, the greater the chance that at least one result will appear statistically notable by chance alone. A prespecified testing order or adjustment procedure can help control that risk, but the statistical plan for multiplicity is not reported in the registry record.

As a result, secondary p-values should not be treated as though each were the only analysis conducted. They may be informative, but their interpretation depends on the full analysis plan and on whether the outcomes and comparisons were selected in advance. The record does not report whether there were interim analyses or how any interim testing would have been handled. Those details are not reported in the registry record.

The primary population is described as all randomized participants in an intention-to-treat analysis. This indicates that the stated analysis population was based on randomization rather than only on treatment completion. However, the record does not explain how missing outcome measurements were handled, whether all randomized participants contributed data at every visit, or whether any sensitivity analyses addressed missingness. These details are not reported in the registry record.

Missing data can affect repeated-measure analyses if the chance of a measurement being missing is related to the outcome or other participant characteristics. Different assumptions and methods for handling incomplete measurements can lead to different estimates. The registry identifies GEE for the sponsor’s primary analysis but does not provide the model specification or missing-data strategy, so it is not possible to assess those assumptions from the record.

The lack of reporting does not establish that multiplicity or missing data were mishandled. It means that readers cannot evaluate those aspects in detail. A fuller assessment would require the protocol and statistical analysis plan, including the planned outcome hierarchy, any adjustment strategy, analysis populations, and methods for incomplete data; these are not reported in the registry record.

16. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 1% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: bilateral sub-massive pulmonary embolism (5.6%).

The adverse-event summary reports serious adverse events in 1 participant in the tocilizumab arm and 0 participants in the placebo arm. The top serious adverse-event term listed is bilateral sub-massive pulmonary embolism, reported as 5.6%. The record does not specify the group attribution for that listed term, so it should not be assigned to an arm based on the summary alone.

Other adverse events were reported for 11 participants in the tocilizumab arm and 8 in the placebo arm. No deaths were reported in either arm. These counts are descriptive safety information; they do not by themselves establish that a treatment caused or prevented an event. The registry record does not provide further detail about event timing, severity beyond the serious-event classification, exposure-adjusted rates, or how events were adjudicated.

Safety counts should be read in the context of the trial’s size and duration. Small group counts can be sensitive to individual events, and a difference in counts is not necessarily evidence of a causal difference. The figures also do not summarize the full clinical context of each event. They describe what the registry reports, without supporting a broader safety conclusion.

17. What the Result Does, and Does Not, Mean

The primary result describes average total improvement scores during the trial’s six-month treatment period in the randomized tocilizumab and placebo groups. The posted means were 26.4 and 29.3, respectively, and the recalculated unadjusted difference was -2.9. The registry’s sponsor-reported GEE analysis gave p = 0.86. Together, these results do not provide clear evidence of a difference on the primary endpoint under the reported analyses.

That is not the same as demonstrating that the treatments are equivalent or that no difference exists. The recalculated confidence interval extends from -13.88 to 8.076, leaving uncertainty about both the direction and size of the difference. The sponsor’s GEE estimate and confidence interval are not reported, so its precision cannot be evaluated from the record. The trial therefore cannot be used to make a precise claim that the groups had identical outcomes.

The findings are specific to the population, treatment regimen, comparator, endpoint, and follow-up described in the registry. They do not by themselves answer questions about different doses, longer follow-up, other patient groups, or outcomes not measured in the trial. The registry does not report enough information about missing-data handling, model details, or multiplicity to resolve those methodological uncertainties.

The adverse-event summary reports serious and other adverse events in both arms, as well as no deaths. These counts are descriptive and do not establish causation or fully characterize safety. Similarly, the listed secondary outcomes do not provide a complete basis for an overall conclusion across every outcome domain, because results for several are not reported in the registry record.

In general, a trial result is best interpreted through its estimate, uncertainty, design, and reporting completeness rather than through a p-value alone. Here, the posted primary result is compatible with a range of possible differences, and the available record does not justify a definitive claim that tocilizumab improves or fails to improve the measured outcome.

18. Why This Trial Matters Statistically

This trial illustrates why a point estimate and its precision need to be read together. The primary arm means are close, but the recalculated confidence interval remains broad enough to include differences in either direction. A result that is not statistically significant does not establish equivalence; it may instead reflect an estimate that is too imprecise to distinguish among plausible effects.

It also illustrates the importance of separating a sponsor-reported model from a simple recalculation. The registry reports a GEE p-value for the repeated-measures analysis, while the unadjusted recalculation uses posted summary statistics. These methods address the comparison differently. Without the sponsor’s model estimate, confidence interval, and specification, the reader cannot fully assess how the repeated observations were handled.

The trial’s phase 2 pilot context makes estimation especially important. Pilot data can inform the scale and uncertainty of a treatment comparison, but planning figures derived from the observed effect should not be mistaken for evidence that a future study would achieve a particular result. The posted sample-size illustrations show how assumed effect size and desired power relate, while the original design justification is not reported in the registry record.

Finally, the record highlights the value of complete reporting. Details about missing data, multiplicity, model specification, and the results for several secondary outcomes are not reported. Their absence does not prove a flaw, but it limits the conclusions readers can draw. Clear reporting of estimates, confidence intervals, assumptions, and analysis decisions is central to evaluating clinical evidence.

19. 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.

Several trial-specific limitations affect interpretation. The primary comparison is based on a modest-sized pilot trial, with 18 participants in each arm in the posted results. The recalculated confidence interval is broad, so the estimate is not precise enough to rule out differences in either direction. The registry does not report a threshold for a clinically important change on the total improvement score, making it difficult to judge the clinical importance of any difference represented by the estimate.

The registry reports a GEE p-value for the primary outcome but does not give the GEE estimate or confidence interval, nor the details of the model specification. The unadjusted recalculation is therefore not a full replication of the sponsor analysis. Information about the correlation structure, covariate adjustment, and model checks is not reported in the registry record.

The record describes an intention-to-treat analysis but does not explain how missing outcome data were handled or whether all participants contributed measurements at each visit. It also does not report the multiplicity plan for the several outcomes and analyses. These omissions prevent a detailed assessment of the assumptions behind the repeated-measures analysis and the interpretation of secondary findings.

Results for several listed secondary outcomes are not reported in the registry record, limiting assessment of consistency across outcome domains. The safety summary provides counts but little context about timing, exposure, or event adjudication. The registry also does not identify the masked roles or provide detailed enrollment and eligibility characteristics. These reporting limits should be distinguished from evidence that the trial necessarily had problems; they define what can and cannot be evaluated from the record.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Tocilizumab in the Treatment of Refractory Polymyositis and Dermatomyositis — NCT02043548.
• Registry record last updated 2020-10-30; status last verified 2020-10.
• Results first posted 2020-10-30.
• 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.

23. Record Summary

NCT02043548 is a completed phase 2 randomized, triple-masked, parallel-group trial of tocilizumab versus placebo in adults with refractory polymyositis or dermatomyositis. The registry lists 36 enrolled participants and reports 18 participants in each arm. Tocilizumab was given by intravenous infusion at 8mg/kg every 4 weeks at six visits; the comparator was placebo.

The primary endpoint was average total improvement score across weeks 4, 8, 12, 16, 20, and 24, with higher scores indicating a better outcome. The posted means were 26.4 for tocilizumab and 29.3 for placebo. The sponsor-reported GEE analysis gave p = 0.86; an unadjusted recalculation gave a mean difference of -2.9, with a 95% confidence interval from -13.88 to 8.076 and p = 0.605.

The available results do not establish a clear primary-outcome difference, and the recalculated interval indicates substantial uncertainty. Important details about model specification, missing data, multiplicity, and several secondary results are not reported in the registry record.