Study of Subcutaneous Golimumab in Chinese Patients With Active Rheumatoid Arthritis Despite Methotrexate Therapy: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Study of Subcutaneous Golimumab in Chinese Patients With Active Rheumatoid Arthritis Despite Methotrexate Therapy (NCT01248780), a randomized phase 3 study. The primary outcome is American College of Rheumatology (ACR) 20 Response, Using CRP (C-reactive Protein), at Week 14.

2. Clinical Question

The clinical question is framed for people whose rheumatoid arthritis remained active despite methotrexate therapy. Because both groups received methotrexate, the contrast concerns adding golimumab rather than comparing golimumab alone with no treatment. The randomized, double-blind, placebo-controlled design is intended to support a between-group comparison while limiting some sources of bias. The registry record does not report further details of how randomization was implemented or how masking was maintained.

3. Trial Design

The registry record describes a Phase 3, multicenter, randomized, double-blind, placebo-controlled, parallel-group study. Its official title identifies Chinese subjects with active rheumatoid arthritis despite methotrexate therapy. Enrollment is reported as 264, and the record lists 132 participants in each randomized group for the primary outcome. The listed study dates are a start in 2010-09 and primary completion in 2011-10; status is completed.

Both groups were assigned methotrexate at a stable dose. One group was assigned golimumab 50 mg by subcutaneous injection every 4 weeks. The other group received placebo injections through Week 20, then golimumab beginning at Week 24 and every 4 weeks through Week 48, according to the arm description. Thus the primary Week 14 comparison takes place during the placebo-controlled period described for the comparator group. The registry record does not report the detailed protocol for treatment changes, adherence, or concomitant care.

The primary outcome is the number of participants with an ACR 20 response at Week 14, analyzed according to randomized assignment regardless of whether assigned treatment was received. This is a binary outcome: each participant either meets the stated response criterion or does not. The registry record also posts three secondary outcomes covering disease activity response, ACR 20 response at Week 24, and change in the HAQ Disability Index from baseline to Week 24.

The record identifies quadruple masking. It does not explain which roles were masked, how the random sequence was generated or concealed, or whether masking was assessed. The original sample-size justification, anticipated effect, and design assumptions are not reported in the registry record. The planning illustrations associated with the observed effects are separate from the trial's own 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: In this trial, 264 participants were randomly assigned to the 2 treatment arms.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Participants randomized (ie, intent-to-treat population) according to their assigned treatment group regardless of whether or not they received the assigned 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.

6. Primary Endpoint

The registry lists the primary outcome as American College of Rheumatology (ACR) 20 Response, Using CRP (C-reactive Protein), at Week 14, measured over Week 14.

The primary endpoint is the number of participants with an ACR 20 response at Week 14. ACR 20 is defined in the registry record as at least 20% improvement in rheumatoid arthritis symptoms and disease activity. It is a binary response measure: the count records participants who meet the criterion, while those who do not meet it are not included in that response count.

A response threshold can make a trial outcome easier to interpret than a change in a single measurement alone. It asks whether participants reached a specified degree of improvement across the criteria incorporated into the ACR response definition. In general, such responder outcomes give a clinically recognizable summary, while also simplifying a complex pattern of symptoms and disease activity into a yes-or-no classification. That simplification means the count does not show how far an individual response exceeded the threshold, nor the particular pattern of component measures for each participant.

The posted primary results are 21 participants in the placebo plus methotrexate group and 54 in the golimumab plus methotrexate group, with 132 randomized participants listed for each group. The endpoint is reported for the intent-to-treat population according to assigned group. The registry record does not report further details about how missing Week 14 outcome data were handled or how the component criteria were assessed.

7. Results Posted in the Registry

• Group II: Golimumab 50 mg + MTX: 40.9% (54/132) versus 15.9% (21/132) on control; risk difference 25.0 percentage points (95% CI 14.5 to 35.5), risk ratio 2.57 (1.65–4.00); 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.

Disease Activity Index Score (DAS 28) Response, Using CRP (C-reactive Protein)

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Group II: Golimumab 50 mg + MTX: 65.2% (86/132) versus 30.3% (40/132) on control; risk difference 34.8 percentage points (95% CI 23.6 to 46.1), risk ratio 2.15 (1.61–2.87); 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 50 mg + MTX than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

American College of Rheumatology 20 Response, Using CRP, at Week 24

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Group II: Golimumab 50 mg + MTX: 42.4% (56/132) versus 15.9% (21/132) on control; risk difference 26.5 percentage points (95% CI 16.0 to 37.0), risk ratio 2.67 (1.72–4.14); 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 50 mg + MTX than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

HAQ (Disability Index of the Health Assessment Questionnaire) Score Change From Baseline

• 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 50 mg + MTX: mean -0.26 versus 0.15 on control; difference -0.41 (95% CI -0.57 to -0.26), standardized effect d = -0.66; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Group II: Golimumab 50 mg + MTX than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The posted secondary outcomes follow a generally consistent pattern in the direction of the primary comparison: the binary response measures report more responders in the golimumab plus methotrexate group, and the continuous HAQ change measure has a lower reported mean change in that group. The sponsor-reported p-values for these secondary comparisons are each <0.001. Taken together, these results provide additional context across disease activity, response status at a later assessment, and disability score change.

These outcomes are related to, but not identical with, the primary endpoint. They address different measures or assessment times, and they can capture different aspects of disease activity and function. Agreement in direction can make the overall pattern more coherent, but it does not mean the outcomes are interchangeable or that each independently confirms the primary result. The Week 24 response is also assessed at a different point in the treatment schedule than the primary Week 14 response.

Secondary analyses generally carry less evidential weight when many outcomes or comparisons are examined without a reported multiplicity strategy. Repeated testing increases the chance of obtaining apparently strong results by chance across a family of tests. The registry record does not report whether multiplicity adjustment was used, whether the secondary outcomes were prespecified in a hierarchy, or whether any were intended as confirmatory. For that reason, their p-values should be treated as supportive descriptive evidence rather than definitive stand-alone tests.

The secondary outcomes do not resolve limitations in the primary analysis, including the lack of registry entries about missing-data handling and model specifications. Their value is in adding different outcome perspectives while remaining subordinate to the prespecified primary question. The registry record identifies three secondary outcomes, but it does not give enough information to establish how their testing was coordinated or how their results were intended to affect the overall conclusions.

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

• Disease Activity Index Score (DAS 28) Response, Using CRP (C-reactive Protein)
• American College of Rheumatology 20 Response, Using CRP, at Week 24
• HAQ (Disability Index of the Health Assessment Questionnaire) Score Change From Baseline

9. Statistical Interpretation

10. Statistical Methods Explained

The primary endpoint is binary, because each randomized participant is classified according to whether the ACR 20 response criterion was met at Week 14. The posted counts are 21 of 132 for placebo plus methotrexate and 54 of 132 for golimumab plus methotrexate. A comparison of proportions is appropriate for this kind of outcome. The risk difference expresses the absolute separation between the two response probabilities, rather than a relative comparison.

What does the risk difference describe?

The risk difference is 0.25, with a 95% confidence interval from 0.145 to 0.354. It describes the estimated difference in response probability between the randomized groups. A positive value indicates a higher response proportion in the golimumab group under the ordering used for the comparison. Unlike a relative risk, this measure retains the absolute scale of the outcome. It can therefore be read in relation to the underlying proportions in the two groups. The interval communicates uncertainty around that absolute contrast; it does not mean that the true effect is known to lie within a fixed range with certainty.

How should the p-value be read?

The unadjusted comparison gives p < 0.001. The registry record separately reports the sponsor's Cochran-Mantel-Haenszel analysis with p < 0.001. A p-value describes how compatible the observed data are with a specified null hypothesis, under the analysis model and its assumptions. It does not quantify the magnitude of the effect, the probability that a treatment is effective, or the likelihood that the findings will reproduce. The effect estimate and confidence interval are needed to understand magnitude and precision.

Why are there different comparison methods?

The posted recalculation is an unadjusted comparison based on the summary counts. The Cochran-Mantel-Haenszel method is the analysis named in the registry record. In general, this method can combine comparisons across strata, with weighting determined by the analysis. The registry record does not report the strata used, the detailed implementation, or a numerical estimate and confidence interval from that sponsor analysis. The sponsor's reported p-value is therefore not interchangeable with the recalculated risk difference and its interval; they answer related questions using different reported methods.

The other posted binary outcomes are likewise summarized as counts and compared as proportions. These include the disease activity response measure and ACR 20 response at Week 24. The HAQ outcome is continuous and is reported as a group mean change with standard deviation. For a continuous measure, a mean difference compares average change between groups; the posted interval describes uncertainty in that difference. The registry record names ANOVA for the sponsor's HAQ analysis. In general, ANOVA compares group means, although the specific model details are not reported in the registry record.

For continuous data, variability among participants matters because a mean can conceal a wide spread of individual outcomes. The standard deviation describes that spread in the group, while the standard error reflects uncertainty in an estimated mean. The two are not the same. For binary data, the number of responses and the number assessed contribute to the precision of the proportion and its difference. These general principles explain why an estimate should be read alongside its uncertainty rather than in isolation.

The primary endpoint is described as analyzed according to randomized assignment, whether or not the participant received assigned treatment. This is an intent-to-treat approach as stated in the record. In general, analyzing participants by assigned group preserves the randomized comparison and estimates the effect of assignment to the treatment strategy. The registry record does not report how missing outcomes, protocol deviations, or incomplete follow-up were addressed, so the extent to which the posted analysis maintained that population definition cannot be assessed from the record alone.

The unadjusted recalculations use posted summary statistics and should not be mistaken for the sponsor's full analysis. The registry record reports no discrepancy between the recalculation and the sponsor analysis. Nevertheless, the reported analysis method and p-value remain the sponsor's formal analysis as recorded; the recalculated estimate is a complementary, transparent summary rather than a replacement for it.

11. Confidence Intervals and Precision

The primary risk difference is 0.25, and its 95% confidence interval runs from 0.145 to 0.354. The interval is wholly on the positive side of zero, indicating that the observed data are compatible with a positive absolute difference under the method used for the recalculation. Its span also shows that the exact size of the difference is uncertain: the point estimate should not be treated as a precisely known effect.

In general, a confidence interval is a range produced by a statistical procedure intended to capture the unknown effect with a stated long-run coverage rate when the assumptions hold. It summarizes sampling uncertainty, not all sources of uncertainty. It does not account automatically for limitations in the trial population, measurement, missing data, treatment adherence, or selective reporting. Nor does it imply that every value inside the interval is equally plausible.

The interval for the primary endpoint is a recalculated, unadjusted interval from posted summary counts. The sponsor's reported Cochran-Mantel-Haenszel result provides a p-value of <0.001, but the registry record does not give that analysis's effect estimate or confidence interval. This means the sponsor analysis's precision cannot be assessed from a reported interval in the registry record. The two approaches should be kept distinct rather than combining the estimate from one with the significance test from the other as though they came from a single model.

Other outcomes have their own uncertainty intervals in the recalculations, and their widths vary with the outcome scale and observed data. In general, a narrower interval indicates greater statistical precision for the modeled contrast, while a wider interval leaves more uncertainty about its size. Precision is not the same as importance: a precisely estimated small contrast may have limited practical significance, and an imprecisely estimated contrast may include effects of differing practical importance.

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 planning illustrations report participants per arm for 80% power and participants per arm for 90% power. All posted outcomes used for these illustrations are not time-to-event outcomes, so the entries refer to participants per arm rather than total events. The calculations are planning illustrations based on the observed effect; they are not the trial's original design justification. The original sample-size assumptions are not reported in the registry record.

The listed 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 set to the stated fraction of the observed effect. The rows planning for 100%, 75%, and 50% of the observed effect therefore explore how required group size changes when the target effect is varied. For example, for the primary endpoint, the illustration for 100% of the observed effect gives 50 participants per arm for 80% power and 67 participants per arm for 90% power. Planning for 75% of the observed effect gives 84 participants per arm for 80% power and 111 participants per arm for 90% power; planning for 50% gives 173 and 230, respectively.

Power is the chance, under the planning assumptions and if the specified effect were the true effect, that a study would meet its chosen statistical significance criterion. The 80% power column and 90% power column express different target chances for detecting that assumed effect. A higher target power generally requires more participants under the same assumptions. Likewise, smaller effects are harder to distinguish from random variation and require more participants for the same target power. These figures are not observed power.

Each illustration depends on the observed effect being a suitable planning reference. Observed effects can be uncertain, and a study designed around an effect smaller than the observed one would require larger group sizes, as the listed fractions illustrate. The calculations should not be read as evidence that the original trial was adequately or inadequately planned, because its own design assumptions and sample-size rationale are not reported.

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 = 0.159; treatment proportion p₁ = 0.409; average p̄ = 0.284.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6378 = 1.2500.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6128 = 0.5157.
4. n = (1.2500 + 0.5157)² / (0.2500)² = 49.9, rounded up to 50 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 <- 0.159090909090909                          # control proportion
p1 <- 0.409090909090909                          # 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(1248780)
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 = 50 (80% power), 67 (90% power)
 75% of observed effect: n per arm = 84 (80% power), 111 (90% power)
 50% of observed effect: n per arm = 173 (80% power), 230 (90% power)
power.prop.test: 49.9 
Simulated power at n = 50 per arm: 0.813 (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 record reports a Cochran-Mantel-Haenszel analysis for the primary comparison, with p < 0.001. The posted recalculation from the group counts gives an unadjusted risk difference of 0.25, a 95% confidence interval from 0.145 to 0.354, and p < 0.001. The record indicates no discrepancy between the recalculation and the sponsor's analysis.

The recalculation is a simple unadjusted comparison based on the posted summary counts. The sponsor's named method is the formal analysis reported in the registry record. The record does not report the sponsor's numerical effect estimate or confidence interval, nor details about any strata used in the Cochran-Mantel-Haenszel analysis. Accordingly, the recalculated estimate is useful as a transparent summary of the posted counts, but it should not be described as the estimate produced by the sponsor's model.

Both reported p-values are below the same stated threshold, but p-values alone do not establish that the estimates are identical. They summarize evidence against a null hypothesis under their respective calculations. Since the record flags no discrepancy, there is no conflicting result to resolve. Where an actual discrepancy exists, the sponsor's own analysis takes precedence; that condition is not flagged for this result.

The distinction matters because a formal analysis may use information or assumptions not represented in the summary-count recalculation. Without the complete analysis details, the recalculation cannot reproduce or validate every aspect of the sponsor's procedure. It should be read as a complementary check of the reported group-level data, not as an independent replacement for the trial's reported inferential analysis.

15. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when investigators make several statistical comparisons. Even when each test uses a conventional significance threshold, conducting multiple tests can increase the chance that at least one result appears statistically significant by chance. General approaches include adjusting the threshold, controlling an error rate across tests, or specifying a testing hierarchy. The registry record does not report whether any such approach was used for the secondary outcomes, whether the secondary tests were confirmatory, or how the primary and secondary hypotheses were ordered.

The primary endpoint has a clearly identified Week 14 assessment, and the registry record reports a sponsor analysis using the Cochran-Mantel-Haenszel method. It does not report the full statistical analysis plan, the precise hypotheses, covariate adjustments, stratification factors, or whether interim analyses were conducted. These details are not reported in the registry record and should not be inferred from the study title or the reported p-values.

The primary outcome description specifies an intent-to-treat population: participants randomized, analyzed according to assigned group regardless of whether they received assigned treatment. That states the intended analysis population for the primary endpoint. The record does not describe the number or pattern of missing Week 14 outcomes, imputation methods, sensitivity analyses, exclusions, or treatment of participants with incomplete follow-up. It is therefore not possible to determine from the registry record how missing data affected the analysis or whether additional populations were examined.

The record also does not report the detailed handling of protocol deviations, crossover, or treatment discontinuation in the efficacy analysis. These issues can matter because an assigned-group analysis and an analysis based on treatment received answer different questions. The treatment descriptions indicate a later transition from placebo to golimumab in the comparator group, but the primary assessment is at Week 14. Further information about subsequent analyses and populations is not reported in the registry record.

These omissions do not negate the posted results, but they limit how completely the analytical process can be evaluated. The sponsor's p-value and the unadjusted recalculations can be described as recorded; conclusions about multiplicity control, missing-data robustness, interim monitoring, and the full analysis plan cannot be made from the registry record.

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: acute myocardial infarction (1.5%), coronary artery disease (0.8%), lung infection (0.8%), white blood cell count decreased (0.8%), abortion spontaneous (0.8%).

The safety record reports serious adverse events for 4 participants at risk in the placebo plus methotrexate group that later received golimumab, and 8 participants at risk in the group assigned golimumab plus methotrexate. The corresponding numbers at risk are 128 and 131. For other adverse events, the record reports 40 affected participants among 128 at risk in the former group and 35 among 131 at risk in the latter group. These are counts as posted; the group labels and exposure periods should be retained when interpreting them.

The placebo group safety label includes a transition to golimumab 50 mg plus methotrexate, so it should not be read as a group exposed only to placebo throughout follow-up. The registry record does not provide event-level timing, duration, severity, relatedness, recurrent-event counts, or detailed exposure-adjusted rates. It also does not report deaths in the posted safety fields, so deaths are not reported in the registry record.

The listed serious adverse event terms include acute myocardial infarction at 1.5%, and coronary artery disease, lung infection, white blood cell count decreased, abortion spontaneous, and granulocytopenia at 0.8% each. These entries describe terms and percentages as recorded, but do not establish that an event was caused by treatment. Safety counts should be considered alongside exposure, follow-up, and clinical assessment rather than used alone to compare risk.

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

The primary result shows a higher observed Week 14 ACR 20 response proportion in the group assigned golimumab plus methotrexate than in the group assigned placebo plus methotrexate. The estimated unadjusted absolute difference is 0.25, with a 95% confidence interval from 0.145 to 0.354, and the posted p-value is <0.001. These statistics describe a between-group result in the randomized trial population and indicate that the estimated difference is not highly compatible with a zero difference under the unadjusted comparison's assumptions.

The result does not mean that every participant responded, that all participants experienced the same degree of change, or that the observed effect will be identical in other settings. ACR 20 is a threshold-based response definition; a participant who meets it may have a different experience from another responder. The endpoint also does not convey every aspect of rheumatoid arthritis symptoms, function, disease activity, or quality of life.

The comparison is specifically between golimumab plus methotrexate and placebo plus methotrexate in Chinese patients with active rheumatoid arthritis despite methotrexate therapy. It does not estimate the effect of golimumab alone, and it does not answer how the treatment compares with other active therapies. The registry record's treatment description also includes a later transition from placebo to golimumab, so later observations require attention to the assigned treatment schedules and assessment times.

Statistical significance is not equivalent to clinical importance. The risk difference and interval describe the magnitude and precision of this outcome contrast, but the registry record does not specify a threshold for a clinically important group difference. The findings also do not by themselves establish an individual treatment decision, long-term outcomes, or the balance of benefits and harms. Safety information is limited to the posted counts and terms, with important event details not reported in the registry record.

The secondary outcomes provide related evidence in other measures, but the registry record does not report whether multiplicity was controlled. Missing-data methods and a complete analysis plan are also not reported. The results should therefore be interpreted as evidence about the posted trial outcomes, with those uncertainties kept distinct from what the primary comparison directly estimates.

18. Why This Trial Matters Statistically

This trial is statistically informative because its primary outcome is a readily defined binary response and the registry record posts counts for both randomized groups. That allows the absolute difference in response probability to be expressed directly, with an uncertainty interval, rather than relying only on a significance test. Reporting both group counts and a contrast helps readers distinguish the observed scale of the result from the evidence against a null hypothesis.

The design also supports an interpretable treatment contrast: both groups received methotrexate, while the randomized assignment differed in golimumab versus placebo during the primary assessment period. This focuses the question on the addition of golimumab to a stable methotrexate background in the enrolled population. The primary endpoint's intent-to-treat description further clarifies that participants are assigned analytically according to randomization, regardless of whether assigned treatment was received.

At the same time, this example illustrates why statistical significance is only one part of trial interpretation. The sponsor reports a Cochran-Mantel-Haenszel analysis, while the posted recalculation gives a simple unadjusted contrast. The results are not flagged as discrepant, but the registry record does not include enough detail to reconstruct the sponsor's model or its estimate. The secondary outcomes broaden the clinical picture, yet multiplicity handling is not reported, and the missing-data approach is also not described.

The planning illustrations show how required group size changes with a target effect and power, but they are derived from the observed effects rather than the original design rationale. That distinction helps prevent retrospective planning calculations from being mistaken for evidence about the study's original power. Overall, the trial demonstrates the value of reading an effect estimate, interval, design, analysis method, and reporting gaps together.

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.

The primary comparison is reported for a specific population: Chinese patients with active rheumatoid arthritis despite methotrexate therapy. The findings may not transfer directly to populations with different characteristics, treatment histories, or clinical settings. The registry record does not provide detailed demographic or baseline clinical characteristics in the posted material considered for this analysis, so the balance and representativeness of those characteristics cannot be assessed.

The registry record does not report the original sample-size justification, the detailed protocol, the full statistical analysis plan, or details of random sequence generation and allocation concealment. It identifies quadruple masking but does not specify which roles were masked or how masking was maintained. These omissions limit assessment of potential design and analysis biases.

Missing-data handling for the primary and secondary efficacy outcomes is not reported in the registry record. The primary population is described as intent-to-treat, but the record does not state how unavailable outcome measurements, treatment discontinuations, or protocol deviations were treated. The sponsor's primary method is named, but its detailed model specification and numerical estimate are not reported. The recalculated comparisons use posted summary statistics and are unadjusted.

The registry record does not report a multiplicity strategy for the secondary outcomes, and it does not state whether interim analyses were performed. The later transition from placebo to golimumab means that outcomes at later times need to be interpreted in light of differing treatment schedules. Safety reporting is also limited: event timing, severity, relatedness, exposure-adjusted analyses, and deaths are not reported in the posted safety fields. These are trial-specific reporting limitations, not evidence that such procedures or events were absent.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Study of Subcutaneous Golimumab in Chinese Patients With Active Rheumatoid Arthritis Despite Methotrexate Therapy — NCT01248780.
• Registry record last updated 2013-09-06; status last verified 2013-08.
• Results first posted 2013-07-01.
• 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, multicenter, randomized, double-blind, placebo-controlled, parallel-group study of golimumab in Chinese patients with active rheumatoid arthritis despite methotrexate therapy. Enrollment is reported as 264. Participants were assigned golimumab 50 mg plus stable-dose methotrexate or placebo plus stable-dose methotrexate, with the comparator group scheduled to receive golimumab later in the study.

The primary outcome was ACR 20 response at Week 14. The posted counts were 21 of 132 for placebo plus methotrexate and 54 of 132 for golimumab plus methotrexate. The unadjusted risk difference was 0.25, with a 95% confidence interval from 0.145 to 0.354 and p < 0.001. The sponsor's reported Cochran-Mantel-Haenszel analysis also gave p < 0.001.

Three secondary outcomes and safety counts are posted. The registry record does not report the original sample-size rationale, multiplicity strategy, or detailed missing-data methods. The planning illustrations are separate calculations based on observed effects, not the trial's original design assumptions.