An Effectiveness and Safety Study of Intravenous Golimumab in Patients With Active Rheumatoid Arthritis Despite Treatment With Methotrexate Therapy: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for An Effectiveness and Safety Study of Intravenous Golimumab in Patients With Active Rheumatoid Arthritis Despite Treatment With Methotrexate Therapy (NCT00973479), a randomized phase 3 study. The primary outcome is Proportion of Participants With an American College of Rheumatology (ACR) 20 Response at Week 14.

2. Clinical Question

The comparison is between treatment strategies in the trial’s specified population, not between golimumab and no treatment: both groups continued methotrexate. The primary outcome is a binary response, so each participant is classified as meeting or not meeting the ACR 20 definition at the stated assessment. The registry record reports the proportions, a recalculated unadjusted risk difference with a confidence interval, and a sponsor-reported analysis. These describe different aspects of the comparison and should be read in light of their respective methods.

3. Trial Design

This was a Phase 3, randomized, parallel-group, placebo-controlled trial. The official title describes a multicenter, double-blind study, and the registry lists quadruple masking. The trial was completed and enrolled 592 participants. The listed enrollment period began in 2009-09, and primary completion was recorded as 2011-03. The sponsor was Centocor, Inc.

Participants were assigned to placebo plus methotrexate or golimumab plus methotrexate. The listed arm sizes were 197 in the placebo-plus-methotrexate group and 395 in the golimumab-plus-methotrexate group. Participants in both groups continued a stable dose of commercial methotrexate. The placebo group was scheduled to cross over to golimumab at Week 24; the golimumab group received placebo infusions at Week 16 and Week 24 to maintain the blind. These schedules matter when interpreting later assessments, particularly an outcome measured at Week 24.

The primary outcome was assessed at Week 14, and the primary population was intent-to-treat, with participants analyzed according to randomized treatment. The registry does not report the trial’s original sample-size justification, the detailed randomization procedure, or the full statistical analysis plan. It also does not report enough detail to characterize all operational features of the masking or allocation process. The planning illustrations accompanying the results are calculations based on observed outcomes and do not establish the original design rationale.

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: 592 participants were randomly allocated to 2 treatment arms. At Week 16 some participants in Placebo arm switched to Golimumab earlier than the rest at Week 24. Thus, the Adverse Events are presented based upon the time when participants began to receive Golimumab in the following manner: Golimumab from Day 0, from Week 16 and from Week 24.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Intent-to-treat: all patients analyzed according to the treatment for which they were randomized.

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 Proportion of Participants With an American College of Rheumatology (ACR) 20 Response at Week 14, measured over Week 14.

The primary endpoint is the proportion of participants who met the ACR 20 response definition at Week 14. It is a binary outcome: a participant either met the response criteria or did not. The definition combines joint counts with several other assessments, requiring at least 20 percent improvement in swollen and tender joint counts and at least 20 percent improvement in at least three of five additional measures. Those additional measures cover pain, participant and physician assessments of disease activity, physical function measured by the Health Assessment Questionnaire, and C-reactive protein.

This composite endpoint is used to describe response across more than one aspect of rheumatoid arthritis. Requiring improvement in both joint counts and a specified number of additional assessments means that the endpoint is not simply a measure of one symptom or laboratory result. The percentage summarizes how many participants met the complete definition in each group at the assessment time.

For the primary outcome, the registry lists 24.9% for placebo plus methotrexate and 58.5% for golimumab plus methotrexate. These percentages describe the observed response proportions in the listed arms. They do not, by themselves, describe precision or account for the analysis method; the risk-difference estimate, confidence interval, and reported analysis provide complementary information. ACR 20 is the endpoint definition used in this trial record, and the result should be interpreted as a Week 14 comparison under that definition rather than as a complete account of every clinical outcome.

7. Results Posted in the Registry

• Group II: Golimumab 2 mg/kg + Methotrexate (MTX): 58.5% (231/395) versus 24.9% (49/197) on control; risk difference 33.6 percentage points (95% CI 25.9 to 41.4), risk ratio 2.35 (1.82–3.04); unadjusted two-proportion z-test p = <0.001.

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.

Proportion of Participants With Moderate or Good Response in Disease Activity Index Score 28 (DAS28) Using C-reactive Protein (CRP) at Week 14

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Group II: Golimumab 2 mg/kg + Methotrexate (MTX): 81.3% (321/395) versus 40.1% (79/197) on control; risk difference 41.2 percentage points (95% CI 33.3 to 49.0), risk ratio 2.03 (1.70–2.42); unadjusted two-proportion z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Group II: Golimumab 2 mg/kg + Methotrexate (MTX) than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Change From Baseline in Health Assessment Questionnaire (HAQ) Score at Week 14

• 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 II: Golimumab 2 mg/kg + Methotrexate (MTX): mean 0.50 versus 0.12 on control; difference 0.38 (95% CI 0.28 to 0.47), standardized effect d = 0.66; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Group II: Golimumab 2 mg/kg + Methotrexate (MTX) than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Proportion of Participants Who Achieved American College of Rheumatology (ACR) 50 Response at Week 24

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Group II: Golimumab 2 mg/kg + Methotrexate (MTX): 34.9% (138/395) versus 13.2% (26/197) on control; risk difference 21.7 percentage points (95% CI 15.1 to 28.4), risk ratio 2.65 (1.81–3.88); unadjusted two-proportion z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Group II: Golimumab 2 mg/kg + Methotrexate (MTX) than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The secondary outcomes cover several domains beyond the primary ACR 20 response. They include a Week 14 DAS28 response based on C-reactive protein, change in HAQ score at Week 14, an ACR 50 response at Week 24, and change in the total van der Heijde modified Sharp score at Week 24. The registry provides outcome results for the first three of these listed secondary measures. Results for the radiographic score change are not reported in the registry record.

Across the posted secondary results, the reported group summaries and registry-reported p-values point in a direction consistent with the primary comparison. They describe a Week 14 disease activity response, a functional score change, and a more stringent response threshold at Week 24. These measures are related to the primary endpoint but are not duplicate versions of it: they use different definitions, scales, and, in one case, a later assessment time. The Week 24 result should be read with the stated crossover schedule in mind, and the record does not describe how the timing of crossover was handled in that outcome analysis.

Each reported secondary analysis has p < 0.001 in the registry record. These are unadjusted secondary results in the sense that the record does not report a multiplicity correction across the family of outcomes. When many outcomes or comparisons are examined, the chance of at least one small p-value under null effects can increase. Without a reported multiplicity strategy, secondary p-values carry less confirmatory weight than a prespecified primary result and are better viewed as supportive or exploratory evidence. The registry record does not report which secondary outcomes, if any, were designated as key confirmatory endpoints or how conclusions were adjusted for multiple testing.

The secondary outcomes therefore broaden the description of measured effects, but they do not establish an independent replication of the primary finding. Their different endpoint definitions, time points, and analysis methods matter, as does the absence of posted results for the fourth listed secondary measure.

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

• Proportion of Participants With Moderate or Good Response in Disease Activity Index Score 28 (DAS28) Using C-reactive Protein (CRP) at Week 14
• Change From Baseline in Health Assessment Questionnaire (HAQ) Score at Week 14
• Proportion of Participants Who Achieved American College of Rheumatology (ACR) 50 Response at Week 24
• Change From Baseline in Total Van Der Heijde Modified Sharp (vdH-S) Score at Week 24.

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is binary because each participant either met the ACR 20 criteria at Week 14 or did not. A proportion is the number of participants meeting the endpoint divided by the number assessed or analyzed. The posted arm percentages provide a direct summary of response frequency. The primary registry analysis is reported as Cochran-Mantel-Haenszel, while the accompanying recalculation uses posted summary statistics for a simple, unadjusted comparison.

What does the risk difference describe?

The risk difference compares the response proportions by subtraction on the proportion scale. The recalculated estimate is 0.336, with a 95% confidence interval from 0.259 to 0.414. This measure describes an absolute difference in response frequency, rather than a relative comparison. A value on the proportion scale can be read as the difference between the two response proportions; it should not be confused with a risk ratio or odds ratio, which express relative contrasts in different ways. The risk difference is unadjusted and is not the same as an estimate from a stratified or covariate-adjusted analysis.

Why are the analysis methods distinct?

The registry reports a Cochran-Mantel-Haenszel analysis with p < 0.001. This method is commonly used to compare categorical outcomes while accounting for strata. The registry record does not identify the strata or provide an adjusted effect estimate and confidence interval for this analysis. The recalculation, by contrast, is based on the posted group summaries and does not reproduce any unreported stratification or adjustment. The similar reported p-value does not make the two methods identical. The registry’s reported analysis and the unadjusted recalculation answer related, but not necessarily identical, statistical questions.

A p-value describes how incompatible the observed data are with a specified null hypothesis under the assumptions of the test. It is not the probability that the null hypothesis is true, nor does it measure the size or importance of an effect. The reported value, p < 0.001, indicates that the reported analysis returned a result below that threshold. The effect estimate and its confidence interval are needed to understand the estimated magnitude and precision; a p-value alone cannot supply either.

How should the confidence interval be understood?

A confidence interval gives a range of estimates compatible with the observed data under the method used to construct it. In repeated use of the same procedure under its assumptions, the stated confidence level describes the long-run coverage property of the procedure. It is not a probability statement that this particular interval contains the true effect. The recalculated risk-difference interval is unadjusted and should be interpreted as such. The registry-reported analysis does not include an effect confidence interval in the posted results summarized in the record.

What do the secondary analyses add?

The registry reports Cochran-Mantel-Haenszel analyses for the binary secondary outcomes and an ANOVA of van der Waerden scores for change in HAQ score. The latter method ranks observations and analyzes transformed rank information rather than relying on a standard comparison of raw means alone. The registry record does not provide the underlying rank-based effect estimate or its confidence interval. The posted HAQ group means and standard deviations support a separate unadjusted mean-difference recalculation, but that calculation should not be mistaken for the sponsor’s rank-based analysis.

Binary outcomes and continuous outcomes summarize different data structures. A response percentage describes the frequency of meeting a threshold. A mean change describes an average change on a scale, while the standard deviation describes the spread of participant values around that mean. A mean-difference confidence interval describes uncertainty in the difference of averages under the recalculation’s assumptions. These estimands are not interchangeable: one cannot infer a response proportion directly from a mean change, or vice versa.

Why can over-dispersion matter?

General statistical explanation: for count outcomes, a simple model may assume that variation is determined by the mean. If actual variation is greater, the data are over-dispersed relative to that model, and standard errors can be too small unless the analysis accounts for the extra variation. The primary endpoint is a binary response rather than a count outcome, and the registry does not report an over-dispersion analysis. This general issue should not be mistaken for a stated feature of this trial’s primary analysis.

All recalculated comparisons are simple, unadjusted calculations from posted summary statistics. They do not add information about the trial’s original statistical plan, missing-data approach, or covariate adjustment. The registry record does not report those details, so they cannot be inferred from the recalculated estimates or from the reported p-values.

11. Confidence Intervals and Precision

The recalculated primary risk difference is 0.336, with a 95% confidence interval from 0.259 to 0.414. The interval is a measure of statistical precision: it displays a range of values compatible with the observed comparison under the calculation’s assumptions. It is not a range of individual participant outcomes, and it does not describe how much responses vary from person to person. The interval pertains to the unadjusted risk-difference calculation, not to the registry’s Cochran-Mantel-Haenszel result.

The interval does not include zero. In the context of a risk difference, zero corresponds to equal response proportions. This observation is consistent with the small reported p-value, but the interval contributes information that the p-value does not: it shows the range of estimated differences compatible with the data and method. Neither the interval nor the p-value establishes that the observed estimate is free of bias or that the same estimate would apply beyond the trial population and setting.

The registry reports p < 0.001 for its Cochran-Mantel-Haenszel primary analysis. It does not report a confidence interval for that sponsor analysis in the posted record. The recalculated interval therefore should not be presented as if it were the sponsor’s interval. Its use is to describe the precision of the unadjusted estimate calculated from the posted summaries.

Precision depends on the number of observations and the observed distribution, among other features. The registry lists 197 participants in the placebo group and 395 in the golimumab group for the primary outcome. Those arm sizes differ, and the unadjusted interval reflects the posted comparison rather than an equal-allocation design. The record does not report a separate interval for each group’s percentage, nor does it provide details about missing observations at the assessment. Thus the displayed interval quantifies uncertainty under the recalculation, while leaving questions about the sponsor’s model and missing-data handling unanswered.

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, not the trial’s original design calculation. For the primary binary outcome, they give participants per arm for 80% power and participants per arm for 90% power. The calculations assume a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. Each planning scenario uses a stated fraction of the observed effect as the effect to detect. The trial’s own original sample-size assumptions and justification are not reported in the registry record.

For the primary endpoint, planning for 100% of the observed effect gives 33 participants per arm for 80% power and 44 participants per arm for 90% power. Planning for 75% of the observed effect gives 57 participants per arm for 80% power and 76 participants per arm for 90% power. Planning for 50% of the observed effect gives 123 participants per arm for 80% power and 164 participants per arm for 90% power. These are illustrative calculations based on the observed effect, not the enrolled group sizes.

In plain language, 80% power means that, under the calculation’s assumptions and for the effect size specified, a study of that planned size would have an 80% chance of detecting an effect at the stated significance level if that effect were real. The 90% column uses a higher probability of detection under the same kind of assumptions. Smaller effects are harder to distinguish from random variation, so planning for a smaller fraction of the observed effect requires more participants per arm.

These figures are not observed power. They do not describe the probability that the completed trial found an effect, and they do not validate the trial’s original design. The corresponding curve should be read as a visual representation of the planning relationship between effect size and required sample size, under the stated assumptions. For non-time-to-event outcomes such as this one, the figures are participants per arm, not total events.

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

Worked calculation for this trial

1. Control proportion p₀ = 0.249; treatment proportion p₁ = 0.585; average p̄ = 0.417.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6972 = 1.3666.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6555 = 0.5517.
4. n = (1.3666 + 0.5517)² / (0.3361)² = 32.6, rounded up to 33 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 <- 0.248730964467005                          # control proportion
p1 <- 0.584810126582278                          # 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(973479)
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 = 33 (80% power), 44 (90% power)
 75% of observed effect: n per arm = 57 (80% power), 76 (90% power)
 50% of observed effect: n per arm = 123 (80% power), 164 (90% power)
power.prop.test: 32.6 
Simulated power at n = 33 per arm: 0.835 (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.

14. Analyses Reported by the Sponsor

The registry’s own primary analysis is reported as a Cochran-Mantel-Haenszel analysis with p < 0.001. The registry record does not report the corresponding sponsor effect estimate or confidence interval in the results. That reported analysis is distinct from the simple, unadjusted recalculation based on the posted response summaries.

The recalculation gives a risk difference of 0.336, with a 95% confidence interval from 0.259 to 0.414, and p < 0.001. It is a descriptive statistical reconstruction from the posted summaries, not a replacement for the sponsor’s analysis. The two analyses report the same direction of comparison in the available results, but differences in analysis method mean the estimates and uncertainty measures should not be combined as though they came from a single procedure.

The discrepancy flag is false, so there is no stated disagreement requiring the sponsor’s analysis to take precedence over a conflicting recalculation. Even so, the sponsor analysis remains the registry-reported inferential result, and the recalculated interval should be labeled unadjusted. The registry record does not report enough detail to reconstruct the sponsor’s full analysis, including any strata used in the Cochran-Mantel-Haenszel method.

15. Multiplicity, Missing Data and Analysis Populations

The registry record identifies a primary endpoint and lists secondary outcomes, but does not report a multiplicity adjustment strategy. When several outcomes are tested, the probability of obtaining at least one apparently unusual result can increase even when no real effects exist. Methods such as prespecified testing hierarchies or adjustment procedures can address that issue, but no such procedure is reported in the registry record. The secondary p-values should therefore not be interpreted as if each were an isolated, prespecified test with no multiplicity concern.

The primary analysis population is explicitly described as intent-to-treat: all patients analyzed according to the treatment for which they were randomized. This is an important feature of the stated primary analysis population. However, the registry record does not report the number of participants with observed Week 14 outcomes, the extent or pattern of missing outcome data, or how missing values were handled. It does not report sensitivity analyses assessing alternative assumptions about missing observations. The listed arm sizes should not be assumed to equal the number with complete data at every assessment.

The registry reports the sponsor’s primary analysis method as Cochran-Mantel-Haenszel but does not identify strata or give further details of the model or test implementation. It also does not report whether the primary result was adjusted for baseline covariates, the precise handling of protocol deviations, or the full analysis population for every secondary endpoint. A separate sample-size justification and details of interim analyses are not reported in the registry record.

These omissions limit how fully the posted results can be audited or independently reproduced. They do not erase the reported results, but they constrain interpretation of the inferential plan. In particular, a reader cannot determine from the registry record whether missing-data decisions, multiplicity control, and all analysis choices were prespecified and applied as planned.

16. 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: rheumatoid arthritis (2.9%), angina pectoris (1.5%), atrial fibrillation (1.5%), myocardial infarction (1.5%), sick sinus syndrome (1.5%).

The registry reports serious adverse event figures for several exposure groups. In the golimumab 2 mg/kg plus methotrexate group, 78 of 395 participants at risk were affected by a serious adverse event. The combined golimumab group lists 106 of 584 at risk. The placebo-plus-methotrexate groups that later crossed over to golimumab are listed separately: 11 of 68 at risk for the group crossing at Week 16 and 17 of 121 at risk for the group crossing at Week 24. These group labels and denominators should be retained when reading the figures; the record does not establish that they represent comparable follow-up periods.

For other adverse events, the registry lists 199 of 395 in the golimumab-plus-methotrexate group and 271 of 584 in the combined golimumab group, with corresponding figures for the crossover groups also reported. The leading listed serious terms include rheumatoid arthritis at 2.9%, and angina pectoris, atrial fibrillation, myocardial infarction, sick sinus syndrome, and device related infection at 1.5% each. These are registry-listed terms, not estimates of comparative risk between randomized groups.

Deaths are not reported in the registry record. The safety figures do not provide a randomized, time-adjusted comparison or enough detail to establish causal attribution. They should be read as counts among the stated groups at risk, not as proof that a treatment caused or prevented an event.

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

The posted primary data show that the proportion meeting the Week 14 ACR 20 response definition was 24.9% with placebo plus methotrexate and 58.5% with golimumab plus methotrexate. The unadjusted risk-difference estimate is 0.336, and its 95% confidence interval runs from 0.259 to 0.414. The registry-reported Cochran-Mantel-Haenszel analysis has p < 0.001. Together, these describe a sizeable observed separation in response proportions in this randomized comparison and the uncertainty attached to the unadjusted estimate.

The result does not state that every participant experienced the same change. It summarizes the proportion meeting a composite threshold, so it does not show the distribution of individual improvements or what happened to participants who did not meet the threshold. It also does not measure persistence of response beyond the stated assessment, and it cannot by itself describe every dimension of disease activity, function, structural change, or safety.

The comparison was golimumab plus methotrexate versus placebo plus methotrexate in participants with active rheumatoid arthritis despite methotrexate therapy. It is not a comparison against no treatment or against another active medicine. The primary population was analyzed according to randomized assignment, but the registry does not report the missing-data details needed to judge how incomplete observations affected the result.

The statistical evidence applies to the trial population, treatment regimens, outcome definition, and assessment time recorded. It does not automatically extend to other populations, schedules, comparators, or follow-up periods. The secondary outcomes provide additional measurements but have multiplicity and reporting limitations. Serious adverse event counts are not, by themselves, evidence of a causal difference in safety. Statistical results are one part of understanding a trial and do not constitute individualized medical advice.

18. Why This Trial Matters Statistically

This trial illustrates why a clinical result should be read through more than its p-value. The primary outcome is a threshold-based response, the reported sponsor test is Cochran-Mantel-Haenszel, and the unadjusted recalculation expresses the comparison as a risk difference with an interval. Each answers a related but distinct question. Separating the sponsor’s analysis from the simple recalculation avoids treating a reanalysis of summary statistics as if it reproduced an unreported full analysis.

The confidence interval adds a precision perspective to the evidence. It gives a range for the unadjusted estimated difference under the calculation’s assumptions, whereas the p-value indicates compatibility with a null comparison under the test. Neither is a substitute for considering the endpoint definition, the analysis population, missing data, or the trial’s treatment schedule.

The design also makes timing relevant. The primary endpoint was assessed at Week 14, while the placebo group’s scheduled crossover to golimumab was at Week 24. The Week 24 secondary outcome therefore sits at a different point in the treatment schedule than the primary result. Recognizing this prevents treating all reported outcomes as though they were measured under the same exposure conditions.

Finally, the record demonstrates the distinction between planning illustrations and trial design justification. The participant counts shown for hypothetical power targets are based on the observed effect and stated calculation assumptions; they do not reveal the assumptions used when the original trial was planned. Statistical interpretation depends not only on numbers but on clarity about what each number represents and which questions remain unanswered in the registry record.

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

The registry record leaves several details unavailable for independent evaluation. It does not report the original sample-size justification, a full statistical analysis plan, interim analyses, or the detailed handling of missing observations. It gives an intent-to-treat description for the primary population but does not report observed assessment counts or sensitivity analyses for missing data. These gaps limit assessment of how closely the posted analysis follows a prespecified plan.

The listed primary arm sizes are unequal: 197 participants were in the placebo-plus-methotrexate group and 395 in the golimumab-plus-methotrexate group. The planning illustrations assume equal allocation, so those illustrations are not a direct description of the completed allocation. The record also does not provide the allocation procedure or explain the reason for the arm-size imbalance.

The Week 24 ACR 50 outcome is reported at the same time as the placebo group’s scheduled crossover to golimumab. The registry does not specify the timing of outcome assessment relative to crossover or how crossover was addressed in the analysis. This makes the Week 24 between-group interpretation less straightforward than the Week 14 primary comparison, which precedes the stated crossover.

Results for the listed Week 24 radiographic score change are not reported in the registry record. The record also does not report a multiplicity adjustment across secondary outcomes, detailed strata for the primary Cochran-Mantel-Haenszel analysis, or a sponsor confidence interval for the primary effect. Safety counts are presented by listed groups, without enough detail to make time-adjusted causal comparisons. These limitations qualify what can be concluded from the registry results without negating the results that are reported.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: An Effectiveness and Safety Study of Intravenous Golimumab in Patients With Active Rheumatoid Arthritis Despite Treatment With Methotrexate Therapy — NCT00973479.
• Registry record last updated 2013-12-25; status last verified 2013-11.
• Results first posted 2013-10-14.
• 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

The registry record describes a completed Phase 3 randomized, parallel-group, placebo-controlled trial of intravenous golimumab with methotrexate in active rheumatoid arthritis despite methotrexate therapy. It lists 592 participants enrolled, quadruple masking, and primary completion in 2011-03. The primary endpoint was the Week 14 proportion achieving ACR 20 in the intent-to-treat population.

The posted primary proportions were 24.9% with placebo plus methotrexate and 58.5% with golimumab plus methotrexate. The registry reports a Cochran-Mantel-Haenszel p-value of p < 0.001. The unadjusted recalculation gives a risk difference of 0.336, with a 95% confidence interval from 0.259 to 0.414.

Three secondary outcomes have posted results; the listed radiographic score outcome does not. The registry also reports serious adverse event counts by exposure group. Details on multiplicity, missing-data handling, the original sample-size rationale, and the full statistical analysis plan are not reported in the registry record.