A Study of RoActemra/Actemra (Tocilizumab) in Patients With Rheumatoid Arthritis Who Have an Inadequate Response to DMARDs or Anti-TNF: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of RoActemra/Actemra (Tocilizumab) in Patients With Rheumatoid Arthritis Who Have an Inadequate Response to DMARDs or Anti-TNF (NCT01326962), a na phase 3 study. The primary outcome is Disease Activity as Measured by Disease Activity Score 28 (DAS28).

2. Clinical Question

The clinical question is descriptive rather than comparative: what DAS28 disease-activity measurements and safety observations were recorded during tocilizumab treatment in this enrolled group? The design does not directly estimate how those observations would differ from outcomes under another treatment or no study treatment. This distinction matters because changes in a single group can reflect multiple influences, including the natural course of illness, ongoing background DMARD therapy, and expectations in an open-label study. The registry record does not establish how those influences were addressed analytically.

3. Trial Design

This was a Phase 3, open-label, single-group study sponsored by Hoffmann-La Roche. It enrolled 28 participants with rheumatoid arthritis who had an inadequate response to DMARDs or anti-TNF. The registry lists the study as completed, with a start date of 2011-11-30 and a primary completion date of 2013-05-12. These dates describe the study record and do not by themselves explain the timing of each participant's assessments.

All participants were assigned to receive tocilizumab; there was no randomized allocation and no control arm. Treatment was specified as 8 mg/kg intravenously, capped at 800 mg, every 4 weeks for a total of 6 infusions. Non-biologic DMARD therapy could continue throughout the study. Anticipated time receiving study treatment was 24 weeks. The primary outcome timeframe, however, is stated as up to 1 year. The registry record does not explain the schedule of outcome assessments across that timeframe or how follow-up after treatment was organized.

The lack of masking is consistent with the open-label description: participants and study personnel were not blinded to the assigned study treatment. This does not automatically invalidate a measured outcome, but it can affect subjective assessments, reporting, and decisions about care or follow-up. DAS28 combines joint counts, an inflammatory marker, and a participant global assessment, so its components do not all have the same susceptibility to knowledge of treatment. The record does not describe steps taken to reduce potential assessment bias.

The registry identifies the primary outcome as DAS28 and lists secondary outcomes concerning adverse events, adverse-event or serious-adverse-event-related discontinuation, ACR response, C-reactive protein abnormality, and erythrocyte sedimentation rate abnormality. It lists 13 outcome measures overall. The record does not report a protocol-level statistical plan, sample-size justification, planned interim analyses, or detailed analysis populations beyond the intent-to-treat definition attached to the primary outcome. Without a control group, a randomized treatment contrast cannot be estimated from this design. The study can document outcomes among participants who received treatment, but its structure does not isolate the effect of treatment from other explanations for observed outcomes.

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: The intent-to-treat population (ITT) consisted of all consented participants enrolled in the study, who had received any part of an infusion of study medication. Where, 'n' is equal to (=) number of participants analyzed at particular point of time.

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 Disease Activity as Measured by Disease Activity Score 28 (DAS28), measured over Up to 1 year.

The primary outcome is disease activity measured by the Disease Activity Score 28 (DAS28). The registry describes DAS28 as a combined index for rheumatoid arthritis that includes swollen and tender joint counts, the erythrocyte sedimentation rate (ESR), and a participant global assessment of disease activity using a visual analog scale. The joint counts use 28 joints; the record gives a range of 0-28 for each count. ESR is expressed in millimeters per hour, and the global assessment has a range of 1-100 mm. The resulting DAS28 scale ranges from 0 to 10, with higher scores representing higher disease activity.

A composite measure is used because rheumatoid arthritis activity is not captured by a single sign or laboratory value. Joint counts describe clinical findings, ESR represents an acute-phase response, and the global assessment captures the participant's overall view of disease activity. Combining these domains provides a broader measure than any one component alone. The components can nevertheless move differently, and a summary score may not show which component accounts for a particular value.

The registry specifies a median and inter-quartile range as the summary statistics. In general, the median describes the central position of observed values and is less influenced by extreme observations than the mean. The inter-quartile range describes the spread across the central portion of the observations. These statistics can be useful when a distribution is skewed or when outlying values are present. The record does not give the observed median or inter-quartile range, nor does it specify a baseline-to-follow-up contrast. A numerical interpretation of the primary outcome is therefore not possible from the posted registry information.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry describes a single-group, open-label study, with DAS28 summarized by a median and inter-quartile range. It does not report inferential analyses for the outcomes. The following concepts explain how such summaries and designs are generally read; they are not claims about analyses that the investigators performed.

What does a median describe?

In general, the median is the central observation after values are ordered. It is a measure of location and is relatively resistant to extreme values. This can make it useful when outcomes are skewed or contain unusually high or low measurements. A median does not describe the full distribution, however. Two groups can have the same median while differing in their spread or in patterns among participants away from the center.

The inter-quartile range (IQR) complements the median by describing dispersion across the central half of the observed values. A narrow IQR indicates that the central observations are relatively clustered; a wider IQR indicates more variation among them. The IQR is not a confidence interval. It describes the spread of participant observations, whereas a confidence interval, in general, expresses uncertainty about an estimated quantity. The registry record does not report an IQR value or a confidence interval for the primary outcome.

What does DAS28 combine?

The DAS28 combines swollen and tender joint counts, ESR, and participant global assessment. Its scale runs from 0 to 10, and higher scores indicate higher disease activity. The component measures represent different aspects of disease activity, so a combined score can provide a compact summary while also concealing component-specific patterns. A change in the composite could arise from movement in one component, several components, or a mix of changes. Component-level results would be needed to distinguish those possibilities, and the primary outcome description does not provide those results.

Because the outcome is described as a median, interpretation concerns the central DAS28 value in the analyzed participants at a given point or under a defined contrast. The registry does not specify whether its primary summary refers to a particular visit, a change from baseline, or another summary across the stated timeframe. It also says that the number analyzed may differ by time point. Without those details and the actual values, readers cannot tell whether an apparent distributional summary represents the same set of participants over time.

Why does the single-group design limit inference?

A single-group study can describe measurements observed among participants receiving an intervention, but it lacks a concurrent counterfactual group. A counterfactual is the outcome that the same participants would have experienced under a different treatment condition; it cannot be directly observed for each person. Randomized control groups help estimate this contrast by making treatment groups comparable in expectation. This study had no comparator, so a pre-post difference, if one were reported, could also reflect natural fluctuation, continued background therapy, changes in behavior or care, or other time-related influences.

Open-label status means treatment assignment was known. General methodological concerns include possible effects on participant-reported outcomes, clinician assessments, follow-up, and reporting of adverse events. DAS28 includes a participant global assessment, which is self-reported, alongside joint counts and ESR. The registry record does not describe masking of assessors, procedures for standardized measurement, or other steps intended to mitigate potential bias. These omissions do not show that bias occurred; they limit assessment of how it was managed.

How should statistical significance and precision be understood?

In general, a p-value evaluates how compatible data are with a specified statistical null model, conditional on the model and its assumptions. It does not quantify the probability that a treatment works, and it does not measure clinical importance. A confidence interval expresses a range of values reasonably compatible with an estimate under the analysis assumptions; its width conveys precision. Neither a p-value nor a confidence interval is reported for the primary outcome in the registry record. Since the numerical outcome estimate itself is also absent, statistical significance and precision cannot be evaluated.

The record lists 13 outcome measures and no statistical analyses. It does not describe a prespecified testing hierarchy, a multiplicity adjustment, or the handling of missing outcome data. When multiple outcomes are examined, general statistical practice distinguishes the primary question from secondary and exploratory questions and considers the risk of chance findings across repeated testing. The absence of a reported approach means that the intended inferential interpretation of secondary outcomes cannot be assessed from the registry record.

What do the planning assumptions mean?

The stated sample-size illustration uses a two-sided significance level of 0.05, equal allocation between two arms, a normal approximation, and an effect defined as a fraction of the observed effect. These are assumptions for hypothetical planning, not the design of this study, which enrolled a single group. Such illustrations can show how required group sizes depend on a chosen effect and analysis assumptions, but they are not observed power and do not establish that the actual study was adequately sized. The registry record does not report the trial's own sample-size justification. Any planning illustration should therefore remain separate from interpretation of the enrolled participants and their outcomes.

For count outcomes, a general consideration is whether the variability is greater than a simple model expects; this is called over-dispersion. If it is present, model-based standard errors and uncertainty intervals can be misleading unless the analysis accommodates it. The registry record does not report count-outcome models or analyses, so it is not possible to say whether over-dispersion was relevant or addressed in this study. This is a general methodological point, not a description of an analysis in the record.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists 13 outcome measures, including the primary DAS28 outcome and secondary measures concerning adverse events, discontinuation related to adverse events or serious adverse events, ACR response, C-reactive protein abnormality, and erythrocyte sedimentation rate abnormality. It does not report a statistical analysis plan, a testing hierarchy, or a method for addressing multiplicity. When several outcomes are tested, the chance of at least one apparently notable result can increase; in general, prespecified priorities or an adjustment strategy help clarify how multiple results should be interpreted. No such approach is reported in the registry record.

Missing-data methods are also not reported. The primary outcome description defines the intent-to-treat population as consented participants enrolled in the study who received any part of an infusion of study medication. It further notes that the number analyzed may vary at a particular time point. The record does not state how many participants contributed data at each assessment, why any observations were missing, or whether missing values were imputed, excluded, or handled with another method. Different approaches can produce different estimates when missingness is related to participant condition or outcome.

The registry record does not specify whether analyses used all available observations, a fixed subset, or any additional analysis population. Nor does it report sensitivity analyses to assess the influence of missing data or alternate analysis decisions. These omissions limit evaluation of how robust outcome summaries may be. They do not establish that data were missing or that a particular method was inappropriate; they mean that the handling of these issues is not reported in the registry record.

11. Safety Results

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

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 2% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: hypotension (3.6%), skin rash (3.6%), neutropenia (3.6%), thrombocytpenia (3.6%).

The safety record identifies 28 participants at risk for both serious and other adverse events. It records 2 participants with a serious adverse event and 4 participants with an adverse event that was not serious. These are participant counts within the study group; they are not comparisons with an untreated or alternative-treatment group. The record does not report exposure-adjusted rates, event timing, or a detailed account of each participant's events.

The serious adverse-event terms listed are hypotension, skin rash, neutropenia, and thrombocytpenia, each with a reported frequency of 3.6%. The record does not establish that these terms were caused by tocilizumab. It also does not clarify whether the terms refer to separate participants or whether one participant could contribute more than one term. The listed terms should not be treated as an exhaustive event narrative.

Death information is not reported in the registry record. An outcome concerning adverse-event or serious-adverse-event-related discontinuation is listed, but its numerical result is not reported in the registry information. The available counts therefore characterize recorded participants with events, while leaving important questions about event course, attribution, and discontinuation unanswered.

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

The study describes tocilizumab treatment in a defined group of participants with rheumatoid arthritis and an inadequate response to DMARDs or anti-TNF. It establishes that the registry intended to assess disease activity using DAS28 and to collect safety outcomes. It also records 28 participants at risk and reports participant counts for serious and other adverse events. These details are useful for understanding the scope and structure of the study.

The registry record does not report a numerical DAS28 result. It therefore cannot show the magnitude or direction of primary outcome measurements, how those measurements varied across participants, or how precisely an estimate was made. A stated outcome and timeframe are not substitutes for posted outcome values. The absence of a comparator also means that any within-group outcome pattern, even if available elsewhere, would not by itself establish that treatment caused it.

The safety counts describe events among participants in this study, not a comparative risk difference. They cannot show whether event occurrence was more or less frequent than it would have been under another treatment or without study treatment. The listed serious-event terms and frequencies do not provide event timing, severity details beyond the serious classification, individual-level overlap, or causal attribution. The record also does not report the number of deaths.

Thus, the registry information supports a description of the study design and selected safety findings, but it does not support a numerical conclusion about the primary disease-activity outcome or a causal treatment-effect estimate. Statistical significance, clinical importance, and precision cannot be judged without outcome estimates and appropriate uncertainty measures. This distinction is especially important for a small, single-group, open-label study: the design can document observations, but it cannot provide a like-for-like treatment comparison.

Planning calculations are separate from observed evidence. The stated sample-size assumptions describe hypothetical planning with two arms, even though this trial had one group. Such illustrations do not mean that the study enrolled two arms, nor do they represent observed power. The trial's own sample-size justification is not reported in the registry record.

13. Why This Trial Matters Statistically

This study is statistically instructive because its design and its available outcome information need to be kept distinct. The record specifies a clear clinical measure, DAS28, and defines the scale's components and direction. Yet a well-defined endpoint alone does not provide evidence about the observed outcome: the primary numerical result is not reported. This illustrates why readers need both an outcome definition and the associated estimate, analysis population, and uncertainty information.

The single-group design also demonstrates the difference between documenting change and estimating a treatment contrast. A set of measurements collected during treatment can describe what happened among enrolled participants. A control group is needed to provide a concurrent comparison for assessing whether a difference is attributable to treatment rather than time, background care, or other influences. Since participants could continue non-biologic DMARD therapy, that context is also relevant when interpreting any within-group observation.

Open-label assessment highlights how endpoint composition matters. DAS28 combines joint counts, ESR, and participant global assessment. These elements capture complementary aspects of disease activity, but some are more subjective than others. Knowing which components contribute to a composite helps readers understand both its usefulness and potential sensitivity to measurement context. The registry record does not describe assessor masking or procedures to address those concerns.

Finally, the record offers participant counts for serious and other adverse events, while leaving other safety details unreported. Reading these as descriptive counts rather than comparative risks prevents overinterpretation. The contrast between listed outcome categories and absent numerical primary results also underscores the importance of complete registry reporting. Without posted estimates, missing-data methods, and analysis details, readers cannot evaluate precision, multiplicity, or the robustness of conclusions.

14. 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 limitations specific to this study constrain interpretation. The trial enrolled 28 participants and had a single treatment group, with no randomized comparator. Any outcome pattern among participants therefore cannot distinguish treatment-associated changes from the course of disease, background therapy, or other influences. The registry also identifies the study as open-label, so participants and study personnel knew treatment was being given. The record does not describe whether assessors were independent or whether specific safeguards were used to limit possible assessment or reporting bias.

The primary outcome is identified as DAS28 with a timeframe of up to 1 year, but the registry record does not provide its numerical result, an assessment schedule, or a clearly specified contrast such as change from baseline. It notes that the number analyzed may vary by time point, yet does not give the corresponding counts or describe missing-data handling. The protocol's sample-size justification and detailed statistical analysis plan are not reported, nor are interim analyses or a multiplicity approach.

The safety information is also limited. Participant counts are reported for serious and other adverse events, but event timing, duration, exposure-adjusted rates, and treatment attribution are not provided. Four serious-event terms are listed with the same frequency, but the record does not describe whether they occurred in distinct participants or whether participants could have multiple terms. Death information is not reported in the registry record. The listed outcome for adverse-event-related discontinuation has no numerical result in the information.

Finally, the treatment regimen and the primary outcome timeframe are not identical in length: treatment was anticipated for 24 weeks, while the primary outcome timeframe is stated as up to 1 year. The record does not explain the follow-up schedule connecting these periods. This leaves uncertainty about which observations contribute to the primary summary and how post-treatment measurements, if any, were incorporated.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: A Study of RoActemra/Actemra (Tocilizumab) in Patients With Rheumatoid Arthritis Who Have an Inadequate Response to DMARDs or Anti-TNF — NCT01326962.
• Registry record last updated 2017-08-16; status last verified 2017-07.
• Results first posted 2016-03-28.
• 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.

18. Record Summary

The registry describes a completed Phase 3, open-label, single-group study of tocilizumab in 28 participants with rheumatoid arthritis who had an inadequate response to DMARDs or anti-TNF. Treatment was specified as 8 mg/kg intravenously, to a maximum of 800 mg, every 4 weeks for 6 infusions; non-biologic DMARD therapy could continue. Anticipated treatment time was 24 weeks.

The primary outcome was DAS28 disease activity, summarized as a median and inter-quartile range over a timeframe of up to 1 year. The record does not report a numerical primary result or statistical analyses. It records 2 participants with a serious adverse event and 4 with an adverse event that was not serious, among 28 at risk. The study has no control group, and the registry record does not report the trial's sample-size justification, missing-data methods, or multiplicity approach.