A Study of Tocilizumab in Combination With DMARDs in Patients With Moderate to Severe Rheumatoid Arthritis: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of Tocilizumab in Combination With DMARDs in Patients With Moderate to Severe Rheumatoid Arthritis (NCT00531817), a randomized phase 3 study. The primary outcome is Percentage of Patients With an Improvement of at Least 50% in American College of Rheumatology (ACR) Score (ACR50) From Baseline at Week 24.

2. Clinical Question

The comparison concerns the difference between groups in the proportion meeting the ACR50 response definition. That definition requires improvement in both tender and swollen joint counts and improvement in at least three of five additional measures. Those measures cover patient and physician assessments of disease activity, patient-assessed pain, physical function measured by the Health Assessment Questionnaire-Disability Index, and an inflammation marker: C-reactive protein or, if that value is missing, erythrocyte sedimentation rate.

This endpoint captures a defined level of improvement across multiple dimensions, rather than change in a single symptom or laboratory measure. It is a binary outcome: each participant is classified as meeting or not meeting the response definition. The primary comparison therefore focuses on the percentage of participants in each group classified as responders by Week 24. The study also listed outcomes covering other response thresholds, disease activity, function, fatigue, sleep, health status, and early component-level changes. Those outcomes provide additional perspectives, but they do not change what the primary ACR50 question asks.

3. Trial Design

The registry describes a completed, randomized, double-blind, parallel-group Phase 3 study sponsored by Hoffmann-La Roche. It enrolled 619 participants with rheumatoid arthritis. The study compared tocilizumab with placebo, with both groups continuing stable antirheumatic therapy that could include permitted DMARDs. The planned allocation ratio was 2:1, with participants assigned to tocilizumab or placebo. The treatment description specifies tocilizumab 8 mg/kg intravenously or placebo intravenously every 4 weeks.

The study began in 2007-10, and primary completion was recorded as 2009-06. The anticipated time on study treatment was 3-12 months. These are registry-record descriptions; the reported primary endpoint itself covers baseline to Week 24. The primary analysis population included randomized patients who received at least one dose of study medication. The endpoint record lists 409 participants in the tocilizumab plus DMARDs group and 205 in the placebo plus DMARDs group.

Randomization is intended to distribute participant characteristics across treatment groups by chance, while double masking is intended to limit the influence of treatment knowledge on participants and study personnel. A parallel-group design assigns participants to a group and evaluates the groups over the study period, rather than having each participant receive both study treatments. These are general descriptions of the design terms; the registry record does not give further operational details about how allocation concealment or masking was implemented.

The registry gives the overall enrollment and target description, but the trial's original sample-size calculation, its assumed treatment effect, and its planned power are not reported in the registry record. Nor does the record describe interim analyses or provide a fuller protocol-level account of treatment exposure and follow-up. Those omissions limit what can be reconstructed about the original design decisions from the registry entry alone.

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.

Randomized

Entered Extended Treatment Period

Analysis population for the primary outcome, as defined in the registry: Intent-to-treat population: All randomized patients who received at least 1 dose of study medication. In determining ACR status, a last observation carried forward (LOCF) approach was used for missing joint count data. Patients with missing data or who escaped were classified as non-responders.

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 Percentage of Patients With an Improvement of at Least 50% in American College of Rheumatology (ACR) Score (ACR50) From Baseline at Week 24, measured over Baseline to Week 24.

The primary endpoint is the percentage of participants with an improvement of at least 50% in ACR score from baseline at Week 24. ACR50 is a response classification, not a continuous average change. To meet it, a participant must improve in tender and swollen joint counts and in at least three of the additional specified measures: patient and physician assessment of disease activity, patient assessment of pain, physical function on the Health Assessment Questionnaire-Disability Index, and C-reactive protein or erythrocyte sedimentation rate if C-reactive protein is missing.

A composite response rule is used to represent improvement across several aspects of the condition. Requiring improvement in joint counts as well as other measures makes the endpoint broader than one based on a single measurement. The threshold also gives a clear classification for a group comparison: the analysis counts how many participants in each group met the rule, then expresses the result as a percentage.

The endpoint does not describe every possible aspect of rheumatoid arthritis, and a response classification does not convey the exact amount of change for each individual measure. Nor does the result alone establish how participants felt on outcomes not included in the response definition. Secondary outcomes listed in the registry address additional domains, but the primary question remains whether participants met the ACR50 rule by Week 24.

7. Results Posted in the Registry

• Tocilizumab 8 mg/kg + DMARDs: 30.1% (123/409) versus 11.2% (23/205) on control; risk difference 18.9 percentage points (95% CI 12.7 to 25.1), risk ratio 2.68 (1.77–4.05); unadjusted two-proportion z-test p = <0.001.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is binary: a participant either met the ACR50 response definition by Week 24 or did not. For a binary outcome, a group percentage is the proportion of participants classified as having the event of interest—in this case, a response. The response percentage is readily interpretable, but the uncertainty in a comparison depends on both the observed proportions and the numbers of participants contributing to them.

What does the risk difference describe?

A risk difference is the response proportion in one group minus the response proportion in the other group. It is an absolute-scale contrast. If expressed as a percentage-point difference, it describes how many more or fewer participants per hundred met the endpoint in one group, based on the observed proportions. It is distinct from a relative comparison. The recalculated result is an unadjusted comparison based on the posted summary counts and does not account for any covariates.

Other measures may also be used for binary outcomes. A risk ratio divides one group’s response proportion by the other’s, while an odds ratio compares the odds of response. The ratio measures answer relative-scale questions, whereas the risk difference retains an absolute scale. The same dataset can therefore yield values that look quite different across these measures without being contradictory. Interpretation should begin with the endpoint definition and the response percentages, then use the contrast that matches the question of interest.

How should the p-value be understood?

A p-value describes how compatible the observed data are with a specified null model, under the assumptions of the test. A small p-value indicates that data such as these would be unusual under that null model; it is not the probability that the null hypothesis is true. It does not measure the size of the group difference, and it does not measure the probability that a result will reproduce in another study. The effect estimate and confidence interval are needed to understand magnitude and precision alongside the p-value.

The recalculated p-value is an unadjusted result for the primary binary comparison. The registry also reports a Fisher Exact analysis. Fisher’s exact test is a method for testing an association in a two-by-two table of group and outcome counts. The registry labels the reported parameter as Mean Difference (Final Values), despite the endpoint being a percentage and the analysis method being Fisher Exact. This wording is retained as a feature of the posted report; the registry record does not explain that parameter label further. The reported result and the recalculated unadjusted contrast should therefore be distinguished rather than silently treated as the same statistical output.

Why does the endpoint's missing-data rule matter?

The primary endpoint population is described as all randomized patients who received at least one dose of study medication. For missing joint-count data, the registry says a last observation carried forward approach was used. In general, last observation carried forward fills a missing later measurement with an earlier recorded measurement. This can retain participants with incomplete follow-up in an analysis, but it relies on an assumption about how the outcome would have evolved after the last available value. The registry record does not report sensitivity analyses that test alternative assumptions.

The record further states that patients with missing data or who escaped were classified as nonresponders. A nonresponder classification makes the endpoint rule explicit for these cases, but it can influence response proportions. Its impact depends on the pattern and amount of missingness and escape, details not reported in the registry record. The analysis should therefore be read as an estimate under the stated endpoint and missing-data rules, not as an assumption-free count of observed complete responses.

What does statistical power mean in the planning illustration?

Power is the probability that a specified statistical procedure will reject its null hypothesis when a particular effect is truly present, under the assumptions used for planning. The sample-size illustration uses the observed effect as its reference and considers stated fractions of that effect. It is not the original trial design calculation. The registry record does not report the assumptions used for the original design, including its target effect or planned power.

Over-dispersion is a general concern when observed variability exceeds what a simple statistical model assumes. In count outcomes, for example, over-dispersion can make standard errors based on a simpler model too small and confidence intervals too narrow. This primary endpoint is binary, so that count-data explanation is not a claim about this analysis. More generally, all model-based uncertainty estimates depend on whether their assumptions are suitable for the data and design.

How do multiple outcomes affect interpretation?

The registry lists 11 secondary outcomes and reports 10 secondary outcomes as posted. When many outcomes or time points are examined, the chance of at least one apparently notable result can increase if each is tested separately. Statistical methods can address this through prespecified testing hierarchies or multiplicity adjustments, but the approach used is not reported in the registry record. The existence of multiple registered outcomes does not by itself establish that the primary comparison was adjusted, nor does the registry record establish that it was not. The safest interpretation is to distinguish the primary endpoint result from the other listed outcome tests and avoid treating multiple unadjusted findings as independent confirmation.

10. Confidence Intervals and Precision

A confidence interval gives a range of values that are compatible with the observed data under the statistical method and its assumptions. In repeated use of the same method, confidence intervals at a stated confidence level would cover the target parameter at that rate, provided the assumptions hold. A 95% interval is not a statement that there is a 95% probability that this particular fixed parameter lies within the calculated limits.

The registry's own primary analysis reports an estimate of 18.85 and a 95% confidence interval from 12.29 to 25.42. The interval is expressed on the outcome scale used in that report. The recalculated unadjusted risk difference has its own interval estimate, based on a different calculation from the posted group summaries. The registry record does not flag a discrepancy between the recalculation and the sponsor analysis. The reported estimate and interval should nevertheless be read in the context of their stated methods and outcome scale, rather than assumed to be identical outputs.

Interval width is one way to assess precision: a narrower interval indicates less statistical uncertainty under the method, while a wider one indicates more. Precision is not the same as clinical importance. An interval can help show which effect sizes remain compatible with the data, but it cannot determine whether an effect is meaningful for every patient or setting. The observed percentages also matter: 30.1% met the response definition in the tocilizumab group and 11.2% in the placebo group.

These estimates are subject to the endpoint definition, analysis population, and missing-data rules. The confidence interval does not account for unreported sources of uncertainty such as deviations from the protocol or an unspecified multiplicity approach. It quantifies sampling uncertainty under the analysis method, not every limitation in the conduct or reporting of a trial.

11. 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 planning figures are participants per arm for 80% power and participants per arm for 90% power. They are not total event counts. The rows consider planning for 100%, 75%, and 50% of the observed effect. For 100% of the observed effect, the illustration gives 72 participants per arm for 80% power and 95 participants per arm for 90% power. For 75% of the observed effect, it gives 117 participants per arm for 80% power and 155 participants per arm for 90% power. For 50% of the observed effect, it gives 236 participants per arm for 80% power and 315 participants per arm for 90% power.

The assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is the stated fraction of the observed effect. These are planning assumptions for the illustration, not the trial's original design assumptions. The original sample-size justification and its assumptions are not reported in the registry record. The illustration uses equal allocation even though the registry describes planned randomization in a 2:1 ratio.

The 80% power column means the calculated design would have an 80% chance of detecting the specified effect, under the stated assumptions, if that effect were the true one. The 90% power column expresses the corresponding planning target of a 90% chance. For either target, planning for a smaller effect requires more participants per arm. These counts are planning illustrations based on the observed effect, not observed power for the completed trial.

12. 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, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.112; treatment proportion p₁ = 0.301; average p̄ = 0.206.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.5724 = 1.1219.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.5567 = 0.4685.
4. n = (1.1219 + 0.4685)² / (0.1885)² = 71.2, rounded up to 72 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
p0 <- 0.11219512195122                          # control proportion
p1 <- 0.300733496332518                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^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 (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(531817)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 72 (80% power), 95 (90% power)
 75% of observed effect: n per arm = 117 (80% power), 155 (90% power)
 50% of observed effect: n per arm = 236 (80% power), 315 (90% power)
power.prop.test: 71.2 
Simulated power at n = 72 per arm: 0.822 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-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.

13. Analyses Reported by the Sponsor

The registry's own analysis reports Fisher Exact as the method, labels the parameter Mean Difference (Final Values), and gives an estimate of 18.85 with a 95% confidence interval from 12.29 to 25.42. Its p-value is reported as less than one ten-thousandth. The recalculation instead uses the posted response summaries to produce an unadjusted comparison of the two binary response proportions, with a recalculated p-value of <0.001. These are distinct presentations and should not be conflated.

The registry record does not flag a discrepancy between the sponsor's result and the recalculation. The sponsor's reported result is the trial's own analysis; the recalculation is a simple, unadjusted reconstruction from the posted summary statistics. The report's label of Mean Difference (Final Values) is unusual alongside a binary percentage endpoint and the listed Fisher Exact method, and no further explanation of that label is reported in the registry record. Accordingly, the sponsor's estimate and interval should be read as reported, while the recalculated contrast is useful as a transparent comparison of the posted group summaries. Neither result should be interpreted without the endpoint definition and analysis population.

14. Multiplicity, Missing Data and Analysis Populations

The primary outcome is defined as ACR50 response from baseline at Week 24. The registry also lists 11 secondary outcomes, including ACR20 and ACR70 at specified visits, disease activity, EULAR response, function, health status, fatigue, sleep, and early changes in individual score components. The record indicates that 10 secondary outcomes were posted. It does not explain why the listed and posted counts differ or identify a particular missing secondary result.

Multiplicity is a concern in general when researchers test several hypotheses or examine several time points. If each test is judged separately without adjustment, the probability of at least one false positive across the set can be higher than the nominal level for an individual test. A study may address this with a prespecified hierarchy, adjusted significance thresholds, or other procedures. The approach for secondary endpoints and repeated assessments is not reported in the registry record. The record also does not report whether interim analyses were planned or conducted.

The primary population is described as randomized patients who received at least one dose of study medication. The record says last observation carried forward was used for missing joint-count data, and that patients with missing data or who escaped were classified as nonresponders. This makes the missing-data handling part of the response definition and analysis. The registry does not report how many endpoint values were missing, how many patients escaped, or sensitivity analyses under alternative missing-data assumptions.

These details matter because conclusions can depend on which patients are included and how missing outcomes are classified. The stated rule is informative, but it does not show whether missingness was balanced between groups or whether alternate methods would change the estimate. The registry record also does not report a broader statistical analysis plan or enough detail to reconstruct all analysis decisions. Interpretation should therefore remain specific to the posted primary analysis, its stated population, and its stated handling of missing data.

15. 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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: pneumonia (2.0%), cellulitis (1.7%), deep vein thrombosis (1.7%), chest pain (1.0%), pancreatitis (0.7%).

The serious adverse event figures are reported separately for three groups: Tocilizumab + DMARDs, Placebo/Tocilizumab + DMARDs, and Placebo + DMARDs. The registry lists 79 serious adverse events among 409 at risk in the first group, 27 among 173 at risk in the second, and 11 among 205 at risk in the third. These counts should be read with their respective group labels and denominators; the middle group is explicitly labelled as a combined placebo/tocilizumab group in the safety record.

The record also lists other adverse events: 259 among 409 at risk for Tocilizumab + DMARDs, 98 among 173 at risk for Placebo/Tocilizumab + DMARDs, and 69 among 205 at risk for Placebo + DMARDs. These are counts, not adjusted comparisons. Serious adverse event terms with listed percentages include Pneumonia (2.0%), Cellulitis (1.7%), Deep Vein Thrombosis (1.7%), Chest Pain (1.0%), Pancreatitis (0.7%), and Pulmonary embolism (0.7%). Death counts are not reported in the registry record.

Adverse event counts alone do not establish that an event was caused by a treatment, and comparisons can be affected by exposure duration, group definitions, and how events were collected. The registry record does not provide enough detail to assess those factors or to make a complete comparative safety interpretation. These figures describe the safety entries posted, rather than offering individualized guidance.

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

The posted primary result describes the share of participants who met a specified response threshold by Week 24. The ACR50 percentages were 30.1% with tocilizumab plus DMARDs and 11.2% with placebo plus DMARDs. The registry's analysis reports an estimate with a 95% confidence interval from 12.29 to 25.42, while the recalculation of the unadjusted binary comparison reports p <0.001. Together, these statistics describe a group-level difference in the registered endpoint and its statistical uncertainty under the stated analysis methods.

The result does not mean every participant in one group responded or every participant in the other group failed to respond. It does not indicate the response of a particular individual, nor does it capture all changes in symptoms, function, disease activity, or quality of life. ACR50 is a composite classification with specific requirements; a participant who does not meet it could still have changes in one or more component measures that do not satisfy the full response rule.

The study compared treatment groups in the setting of stable antirheumatic therapy, which could include permitted DMARDs. The primary comparison therefore concerns the regimens as described, not tocilizumab isolated from background therapy. Its findings are tied to the enrolled study population, the endpoint population, the Week 24 assessment, and the treatment and missing-data procedures recorded in the registry. They should not automatically be generalized to different populations, treatment contexts, or time periods.

Statistical significance does not by itself establish clinical importance, certainty about every outcome, or a favorable overall balance of benefits and harms. The endpoint is one part of the evidence; the listed secondary outcomes and safety information are also relevant to a full interpretation, while the completeness of those registry results is limited. The result is best understood as evidence about a specific response comparison, not as a universal conclusion about rheumatoid arthritis or an individual treatment decision.

17. Why This Trial Matters Statistically

This trial illustrates how the choice of endpoint shapes a clinical comparison. ACR50 is not a single symptom measure; it requires improvement in joint counts and multiple additional assessments. Explaining the response rule is essential because the percentage is meaningful only when readers know what qualifies as a response. The binary format makes the result accessible as a proportion, while the composite definition means that it should not be mistaken for a complete account of every component outcome.

The study also shows why absolute and relative measures should be distinguished. The posted group percentages communicate how often the endpoint occurred in each arm. A risk difference expresses their separation on an absolute scale, while ratios express a relative comparison. Reporting the endpoint rates alongside an interval and a test result gives readers a fuller basis for judging the result than a p-value alone.

There is also a practical lesson in registry interpretation. The sponsor's own result and an unadjusted reconstruction can use different labels or scales, even when both concern the same posted endpoint. Keeping the original analysis distinct from a recalculation makes the source and meaning of each estimate clearer. The reported parameter label in this record does not align intuitively with the binary endpoint, so careful description is more useful than assuming the label explains itself.

Finally, missing-data rules and multiple outcomes are not peripheral details. Last observation carried forward and nonresponder classification influence how response status is assigned, while the listed outcomes create questions about multiplicity that the registry does not resolve. These considerations show why statistical interpretation should combine the result with the endpoint definition, population, analysis rules, and limits of registry reporting.

18. 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.
• Participant counts were reconstructed from posted percentages and arm sizes, so small rounding differences are possible.

The registry record gives a concise outcome and design description but not the full protocol or analysis plan. It does not report the original sample-size assumptions or justification, including the effect size and planned power, and it does not describe interim analyses. Thus the planning illustration cannot be used to infer the assumptions that guided the trial's original enrollment target.

The primary analysis is described for randomized participants who received at least one dose. The record reports last observation carried forward for missing joint-count data and classifies patients with missing data or who escaped as nonresponders. It does not report the number of missing endpoint values, the extent of escape, or sensitivity analyses using alternative assumptions. As a result, the robustness of the primary estimate to missing-data choices cannot be assessed from this registry entry alone.

The record lists multiple secondary outcomes, while 10 are reported as posted. The reason for that difference is not reported in the registry record, and the multiplicity strategy is also not reported. The registry does not provide enough information to determine how secondary comparisons or repeated visits were handled statistically. The reported sponsor parameter is labelled Mean Difference (Final Values), a label whose relation to the binary percentage endpoint is not explained in the record.

Safety entries include distinct group labels, including Placebo/Tocilizumab + DMARDs, and death counts are not reported in the registry record. The registry summary does not provide enough detail to infer event attribution or account for exposure duration in the groups. These limitations concern what can be concluded from the posted record; they do not establish that the trial did or did not collect information beyond what the registry reports.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: A Study of Tocilizumab in Combination With DMARDs in Patients With Moderate to Severe Rheumatoid Arthritis — NCT00531817.
• Registry record last updated 2012-08-20; status last verified 2012-08.
• Results first posted 2010-08-04.
• 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.

22. Record Summary

NCT00531817 is a completed, randomized, double-blind, parallel-group Phase 3 study sponsored by Hoffmann-La Roche. It enrolled 619 participants with moderate to severe active rheumatoid arthritis and compared tocilizumab 8 mg/kg intravenously with placebo, both alongside stable antirheumatic therapy including permitted DMARDs. The registry records a 2:1 planned allocation ratio.

The primary endpoint was the percentage meeting ACR50 from baseline at Week 24. The posted response percentages were 30.1% for tocilizumab plus DMARDs and 11.2% for placebo plus DMARDs. The registry reports its own estimate and confidence interval, and an unadjusted recalculation is also available. Missing-data rules, multiple listed outcomes, and limited registry detail constrain interpretation beyond this defined comparison.