An Observational Study of the Impact of RoActemra/Actemra on Fatigue in Patients With Rheumatoid Arthritis (PEPS): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for An Observational Study of the Impact of RoActemra/Actemra on Fatigue in Patients With Rheumatoid Arthritis (PEPS) (NCT01185522), a registered phase not stated study. The primary outcome is Percentage of Participants With a Clinically Significant Improvement in Fatigue After 4 Months of Tocilizumab Treatment.

2. Clinical Question

The question focuses on fatigue, a patient-reported experience that may affect daily activities and function. The FACIT-Fatigue measure gives a structured way to record this experience and compare a participant’s score at inclusion with a later score. The outcome is expressed as the percentage meeting the stated improvement threshold, rather than as the average change in score.

This distinction matters for interpretation. A percentage above the threshold describes how often the defined individual-level change occurred in the analyzed population; it does not describe the size of change for every participant. Nor does a before-and-after observation by itself establish that treatment caused a change. In an observational study, disease course, other care, participant characteristics, and the circumstances of measurement may also be relevant. The registry record does not report a comparison group that could help separate these influences.

3. Trial Design

The registry describes PEPS as a prospective, observational study of tocilizumab and fatigue in rheumatoid arthritis. Its stated enrollment is 719. The record lists Hoffmann-La Roche as sponsor and gives the study status as completed. The recorded start is 2010-01 and the primary completion is 2011-05.

Participants with moderate to severe rheumatoid arthritis and an inadequate response to disease-modifying antirheumatic drugs or anti-tumor necrosis factor drugs were eligible under the study description. The treatment was administered according to standard care, and eligible participants receiving tocilizumab were observed in a real-life setting for 4 months. The registry describes a tocilizumab group; it does not describe a concurrent untreated or active comparator group. Allocation is not stated.

Because treatment followed routine care rather than a described random allocation, the design is suited to observing outcomes among people who received treatment in that setting. It does not have the balancing feature of randomization, which is intended to distribute known and unknown prognostic factors across comparison groups. The registry record does not report a randomized comparison, a control arm, or details of how treatment decisions were made.

The primary endpoint concerns the percentage of participants with a clinically significant improvement in fatigue at Month 4. The measure is the FACIT-Fatigue score, with a stated threshold of a change of >= 4 points from baseline. The record also lists baseline disease characteristics, fatigue scores, correlations between changes in fatigue measures, and time to onset of an improvement among secondary outcomes.

The registry lists 21 outcome measures and 25 analyses. Those counts do not establish how outcomes were prioritized in the analysis, whether analyses were prespecified, or whether any were exploratory. A detailed statistical analysis plan, sample-size justification, interim-analysis plan, and approach to multiplicity 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: Of 719 participants, 25 did not respect inclusion criteria (age less than 18 years or missing age, no or missing attestation of information about the study) and one did not receive any tocilizumab (RoActemra®) infusion.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Patient analysis population consisted of all participants included in the study, who respected inclusion criteria, received at least one tocilizumab infusion, had an evaluable FACIT-Fatigue score at inclusion, and at least one evaluable FACIT fatigue score during treatment.

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: Safety population consisted of all participants included in the study who respected the inclusion and non-inclusion criteria and received at least one tocilizumab infusion.

6. Primary Endpoint

The registry lists the primary outcome as Percentage of Participants With a Clinically Significant Improvement in Fatigue After 4 Months of Tocilizumab Treatment, measured over At Month 4.

The primary endpoint is the percentage of participants with a clinically significant improvement in fatigue after 4 months of tocilizumab treatment. Fatigue is assessed with the Functional Assessment of Chronic Illness Therapy (FACIT)-Fatigue questionnaire, a self-reported measure of fatigue and its impact on daily activities and function.

The questionnaire has 13 items, each using a 5-point scale from 0 (not at all) to 4 (very much). The registry explains that responses to the negatively stated items are reversed before the responses are summed. The total score ranges from 0 (worse score) to 52 (better score), so a higher score reflects improvement in the participant’s health status. A clinically relevant improvement is defined as a change of >= 4 points from baseline.

A threshold-based endpoint is useful because it describes the proportion of participants whose score change reaches a defined level of relevance, rather than relying only on an average score change. It can make the outcome easier to relate to the frequency of a meaningful individual change. However, it also reduces the detail in the underlying score: participants just below and just above the threshold are placed in different categories even if their changes are close.

The record says the analysis used last observation carried forward and refers to participants who completed a FACIT-Fatigue score at Month 4. The primary endpoint estimate and its 95% confidence interval are not reported in the registry record, so the observed proportion and its precision cannot be assessed from the registry information.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does the primary endpoint count?

The primary endpoint is a percentage: the share of the analyzed participants whose FACIT-Fatigue score improved by the prespecified threshold from baseline at Month 4. The registry defines clinically relevant improvement as a change of >= 4 points. A binary endpoint turns each participant’s change into a category: meeting or not meeting that threshold. This is different from analyzing the full score as a continuous measure, where the amount of change can retain more detail.

In general, a percentage is estimated by dividing the number of participants meeting the endpoint definition by the number included in that endpoint analysis. The exact analysis denominator and numerator are not reported in the registry record. Although a 95% confidence interval is named for the endpoint, its limits are also not reported. Without those values, the observed percentage and its statistical precision cannot be described.

Why use a confidence interval?

In general, a confidence interval communicates sampling uncertainty around an estimate. It is more informative than a point estimate alone because it indicates a range of values compatible with the data and the interval method. It does not capture every source of uncertainty. In particular, it does not automatically account for selection effects, measurement limitations, confounding, or errors in the endpoint definition.

For a percentage, precision depends in part on how many participants contribute endpoint information and on how the outcomes are distributed. But the relevant endpoint counts are not reported in the registry record. No confidence interval can be reconstructed from the enrollment figure alone, because enrollment is not necessarily the endpoint analysis population and the registry gives no outcome count.

What is the role of last observation carried forward?

The registry says the primary endpoint was performed using last observation carried forward (LOCF), and also refers to participants who completed a FACIT-Fatigue score at Month 4. LOCF is a missing-data procedure that uses a participant’s most recent recorded value as a substitute for an unavailable later value. In general, this procedure can retain participants with incomplete follow-up, but it imposes an assumption: the carried-forward value is treated as a reasonable stand-in for the missing later measurement.

That assumption may not match a participant’s actual course. Fatigue can change after the last recorded measurement, so carrying a score forward can either understate or overstate later fatigue depending on what would otherwise have happened. LOCF also does not remove the possibility that missingness is related to health status or treatment experience. The registry does not report the amount or pattern of missing endpoint data, or explain how the stated use of LOCF relates to the mention of Month 4 completers. Those details are not reported in the registry record.

What does the observational design imply?

In general, randomization helps make treatment groups comparable with respect to measured and unmeasured characteristics. This study is described as observational, and allocation is not stated. The record describes participants receiving tocilizumab according to standard care, rather than a randomized treatment assignment. Consequently, the endpoint can describe outcomes among the analyzed treated participants, but a change from baseline is not, on its own, an estimate of a treatment effect relative to an alternative.

Factors such as underlying disease severity, prior treatment response, other care, and changes over time can influence fatigue and may also relate to treatment receipt. This is general reasoning about observational research, not a claim that any particular factor biased this study. The registry record does not report an adjusted causal analysis or the variables used in any adjustment.

How should planning illustrations be read?

General sample-size calculations ask how much information a future study might need to detect a specified effect under stated statistical assumptions. The planning assumptions described for this record include a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; the target effect is a stated fraction of an observed effect. For time-to-event illustrations, the stated quantity is the total number of events; for other illustrations, it is participants per arm.

These are planning illustrations computed from an observed effect, not the trial’s own design. They should not be interpreted as observed power, as evidence that the study had a particular probability of detecting an effect, or as a substitute for the unreported endpoint estimate. The registry record does not report its own sample-size justification. No planning sample-size values are stated in the registry record information, so none can be described.

Why do multiple outcomes matter?

The registry lists 21 outcome measures and 25 analyses, including baseline disease characteristics and secondary fatigue measures. In general, testing or estimating many outcomes creates interpretive challenges. If multiple significance tests are conducted, the chance of at least one nominally unusual result can increase unless the analysis plan accounts for multiplicity. A multiplicity strategy may control error rates, prioritize outcomes, or treat some results as exploratory. The registry record does not report which approach, if any, was used.

Not every listed outcome necessarily represents a hypothesis test; some describe baseline characteristics or associations. The number of listed outcomes alone does not show how many inferential tests were performed, which analyses were prespecified, or whether findings were adjusted. Those details are not reported in the registry record. Interpreting any result would therefore require its estimate, uncertainty, analysis population, and context within the full set of analyses.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("Tocilizumab")
n      <- c(610)
events <- c(386)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  Tocilizumab  610  386  63.3  59.3  67.1  59.4  67.0  59.4  67.0

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 21 outcome measures and 25 analyses. It includes baseline disease characteristics, measures of fatigue and rheumatoid arthritis status, correlations between changes in fatigue measures, and time to onset of an improvement in FACIT-Fatigue score. These listings indicate a broad outcome set, but do not establish which outcomes were confirmatory, how the analyses were prioritized, or whether a multiplicity adjustment was used. Those details are not reported in the registry record.

In general, multiplicity matters when many hypotheses are tested because repeated testing can make apparently notable findings more likely by chance. A prespecified strategy can distinguish the primary question from secondary or exploratory questions, or otherwise control the interpretation of multiple tests. The record identifies a primary endpoint, but does not report a complete statistical analysis plan or describe an adjustment strategy. The count of analyses alone cannot establish that all were significance tests.

The primary endpoint description states that LOCF was used and refers to participants who completed a FACIT-Fatigue score at Month 4. The patient analysis population required an evaluable score at inclusion and at least one evaluable score during treatment, in addition to meeting inclusion criteria and receiving at least one infusion. The endpoint’s actual analysis denominator, missing-data count, and precise handling of incomplete Month 4 scores are not reported in the registry record.

In general, missing-data methods rely on assumptions about why observations are unavailable and what values would have been observed. LOCF substitutes a previously recorded measurement for a later missing one. Its use does not demonstrate that missingness was harmless or that the substituted score matches the unobserved score. The record does not report sensitivity analyses that would show how conclusions vary under alternative assumptions. It also does not report an interim-analysis plan or a detailed account of the analysis populations for all outcomes.

12. 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: arthritis bacterial (0.3%), diverticulitis (0.3%), neutropenia (0.3%), pulmonary embolism (0.3%), appendicitis (0.1%).

For the tocilizumab group, the registry lists 693 participants at risk for serious adverse events and 27 affected. For other adverse events, it lists 693 at risk and 117 affected. These are separate descriptive counts. The registry record does not state that the affected participants in one category are distinct from those in the other, so the counts should not be combined or treated as a total number of people with any event.

The record lists serious adverse-event terms with percentages: Arthritis bacterial, Diverticulitis, Neutropenia, and Pulmonary embolism at 0.3% each; Appendicitis and Cellulitis at 0.1% each. The listed term percentages describe what the record reports, but the registry does not provide further detail about event timing, recurrence, severity definitions, or whether individual participants experienced more than one term.

In general, an at-risk count identifies the group used as a denominator for a safety summary, while an affected count indicates the number recorded with the relevant category. These counts do not establish causation. In this observational study, no comparison group is described, so the figures cannot be used as a randomized comparison of treatment-related risk. Deaths are not reported in the registry record. The safety figures should be read as descriptive reports rather than estimates of causal effects.

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

The registry record describes a prospective, observational study of fatigue among people with moderate to severe rheumatoid arthritis who received tocilizumab according to standard care. It defines a primary endpoint based on a change in FACIT-Fatigue score at Month 4 and provides a threshold for classifying an improvement as clinically relevant. It also identifies a patient analysis population and describes use of LOCF.

The primary endpoint result itself is not reported in the registry record. The percentage meeting the threshold and its 95% confidence interval are not available in the registry information. Therefore, the record does not show how often the defined improvement occurred or how precisely that percentage was estimated. The listed safety counts and event terms do not fill that gap: they concern adverse events, not the fatigue endpoint.

Even if a before-and-after fatigue estimate were available, an observational study without a described comparison group would not by itself show that treatment caused the observed change. Participants may differ in health status or other factors, and fatigue may change over time for reasons not attributable to the treatment. This is a general limitation of interpreting uncontrolled observational outcomes, not a conclusion that a particular source of bias occurred in this study.

The defined threshold is useful for describing the proportion who reached a specified degree of score improvement, but it does not describe the full distribution of score changes. It also does not make the endpoint result equivalent to a change in disease activity, a safety outcome, or an overall treatment effect. Those questions require their own outcomes and analyses.

The record does not report a sample-size justification, detailed statistical analysis plan, multiplicity strategy, or sensitivity analysis for missing data. Planning illustrations based on an observed effect are hypothetical design calculations, not observed power and not results from this study. The appropriate conclusion is therefore limited: the registry specifies what was intended to be measured and provides some safety descriptions, but the primary fatigue finding and important details needed for a fuller statistical assessment are not reported in the registry record.

14. Why This Trial Matters Statistically

This study is statistically notable because it frames fatigue as a patient-reported outcome in a real-life observational setting. The FACIT-Fatigue score records both self-reported fatigue and its impact on daily activities and function. The endpoint translates score change into a percentage meeting a stated threshold, which is a different perspective from reporting only an average change.

The design also illustrates the distinction between describing outcomes after treatment and estimating a causal treatment effect. Prospective observation can document how participants’ measurements are collected over follow-up, while treatment according to standard care reflects a real-life context. Without a described comparator and randomized allocation, however, the observed before-and-after pattern cannot isolate treatment effects from other influences. That distinction is important whenever clinical outcomes are interpreted from nonrandomized data.

The endpoint definition makes the analysis choices consequential. The threshold determines which score changes count as clinically relevant, while LOCF affects how unavailable later measurements are represented. An endpoint percentage can be easy to communicate, but its interpretation depends on the analysis population, missing-data handling, and uncertainty interval. Those elements need to be considered together rather than treating a percentage as self-explanatory.

The registry lists multiple outcome measures and analyses, which makes transparency about prespecification and multiplicity relevant. The record does not report a detailed analysis plan or adjustment strategy, and it does not report the primary result. This limits the ability to assess either the magnitude of the fatigue finding or its statistical precision. The case therefore underscores that a clear endpoint definition is valuable, but a complete interpretation also requires the estimate, uncertainty, analysis details, and an appropriate account of the study design.

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 main trial-specific limitation in interpreting the registry record is that the primary endpoint result is not reported. Although the endpoint is defined as the percentage with a clinically significant FACIT-Fatigue improvement at Month 4 and a 95% confidence interval is specified, neither the percentage nor the interval limits appear in the registry information. The result’s magnitude and precision therefore cannot be assessed.

The study is described as observational, with tocilizumab used according to standard care in a real-life setting. The record does not describe a comparator group, and allocation is not stated. Consequently, the design information does not support a direct randomized contrast. Changes in fatigue from baseline, even if reported, would not alone distinguish a treatment-associated change from other changes over time or differences in participant characteristics.

The primary endpoint description combines a LOCF method with a reference to Month 4 completers, but the relationship between those approaches is not explained. The endpoint analysis denominator, amount and pattern of missingness, and sensitivity analyses are not reported in the registry record. The patient analysis population is defined, but the record does not state how many participants contributed to the primary endpoint.

The registry lists many outcomes and analyses but does not report the statistical analysis plan, multiplicity approach, or sample-size justification. These omissions make it difficult to determine the intended hierarchy of analyses or assess how multiple findings were handled. Safety counts are descriptive and lack a comparison group in the record. Further details on event timing and overlap are not reported in the registry record.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: An Observational Study of the Impact of RoActemra/Actemra on Fatigue in Patients With Rheumatoid Arthritis (PEPS) — NCT01185522.
• Registry record last updated 2016-06-23; status last verified 2016-05.
• Results first posted 2016-06-23.
• 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.

19. Record Summary

PEPS is a completed prospective observational study of tocilizumab and fatigue in people with moderate to severe rheumatoid arthritis who had an inadequate response to disease-modifying antirheumatic drugs or anti-tumor necrosis factor drugs. The registry lists enrollment of 719, Hoffmann-La Roche as sponsor, a recorded start of 2010-01, and primary completion of 2011-05. Eligible participants receiving tocilizumab according to standard care were observed for 4 months.

The primary endpoint is the percentage with a clinically significant FACIT-Fatigue improvement at Month 4, defined as a change of >= 4 points from baseline. The endpoint result and its 95% confidence interval are not reported in the registry record. The record describes LOCF and defines the patient analysis population, but does not fully explain the endpoint’s missing-data handling.

Safety listings include 27 participants affected by serious adverse events among 693 at risk and 117 affected by other adverse events among 693 at risk. The study is observational, and no comparator group is described. Its reported information supports a description of the design and outcome definitions, but not a causal treatment comparison.