Study Comparing Etanercept in Combination With Methotrexate in Subjects With Rheumatoid Arthritis: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Study Comparing Etanercept in Combination With Methotrexate in Subjects With Rheumatoid Arthritis (NCT00565409), a randomized phase 4 study. The primary outcome is Percentage of Participants Achieving 28 Joint Disease Activity Score (DAS28) Less Than or Equal to (≤) 3.2 at Week 88.

2. Clinical Question

3. Trial Design

The registry identifies this as a randomized, triple-masked, parallel-group phase 4 study of etanercept and methotrexate in people with rheumatoid arthritis. The record lists enrollment as 834 and study status as completed. The study began in 2008-03, and primary completion was recorded in 2011-05. Pfizer is listed as sponsor. The meaning of each masking role, the allocation procedure, and the detailed protocol are not reported in the registry record.

The primary Week 88 endpoint lists three groups: etanercept 50 mg plus methotrexate, etanercept 25 mg plus methotrexate, and placebo plus methotrexate. The endpoint population is modified intent-to-treat, defined as participants who took at least one dose of double-blind study treatment and had at least one post-randomization DAS28 evaluation. The endpoint-group counts are 201, 201, and 197, respectively. These analysis counts need not represent every person enrolled, because they correspond to the endpoint population and record-specific group definitions.

The registry summary frames the principal comparison as etanercept 50 mg once weekly plus methotrexate versus methotrexate monotherapy over 88 weeks. The posted endpoint group labels instead specify placebo plus methotrexate as the comparator. The registry lists 42 outcome measures and 385 analyses; 41 secondary outcomes are posted. The original design’s sample-size assumptions, the detailed statistical analysis plan, and any interim analyses are not reported in the registry record. Those omissions limit what can be concluded about how the study was planned or how the posted analyses were selected.

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.

Period 1 Open Label

Reasons for not completing this period:

After Period 1, Prior to Period 2

Reasons for not completing this period:

Period 2 Double-Blind

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Modified Intent to Treat population (mITT): all participants who took at least 1 dose of double-blind test article and had at least 1 post-randomization DAS28 evaluation

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 Participants Achieving 28 Joint Disease Activity Score (DAS28) Less Than or Equal to (≤) 3.2 at Week 88, measured over Week 88.

The primary endpoint is the percentage of participants achieving a DAS28 score of 3.2 or less at Week 88. The registry defines a score of 3.2 or less as low disease activity. DAS28 combines the number of swollen and painful joints assessed with a 28-joint count, erythrocyte sedimentation rate, and the participant’s general health assessment on a visual analogue scale. The joint counts could be prorated when less than 20% of the swollen- or painful-joint data were missing.

This endpoint turns a multi-component measure into a binary outcome: each participant either meets the stated threshold at the scheduled assessment or does not. Reporting the percentage meeting a threshold offers a direct group-level description of how often the defined state was observed. It is not the same as reporting average change in the score, and it does not show how long a participant remained below the threshold.

Threshold outcomes can be useful for clinical interpretation, but they discard information about how far each score lies above or below the cutoff. A small difference around the threshold can change a participant’s classification, while a larger score change that stays on one side of the threshold may not change it. The endpoint also describes status at Week 88 rather than the full course of disease activity during the 88-week period. The primary endpoint’s population restriction and the requirement for a post-randomization DAS28 assessment are important when interpreting the reported percentages.

7. Results Posted in the Registry

• Etanercept 50 mg Plus MTX -Period 2: 82.6% (166/201) versus 42.6% (84/197) on control; risk difference 39.9 percentage points (95% CI 31.3 to 48.6), risk ratio 1.94 (1.63–2.30); unadjusted two-proportion z-test p = <0.001.
• Etanercept 25 mg Plus MTX-Period 2: 79.1% (159/201) versus 42.6% (84/197) on control; risk difference 36.5 percentage points (95% CI 27.6 to 45.4), risk ratio 1.86 (1.55–2.21); 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 Time Participants Had Low Disease Activity DAS28 Week 36 to Week 88

• 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 errors.
• Result:
  • Etanercept 50 mg Plus MTX -Period 2: mean 0.85 versus 0.58 on control; difference 0.27 (95% CI 0.21 to 0.33), standardized effect d = 0.96; z-test p = <0.001.
  • Etanercept 25 mg Plus MTX-Period 2: mean 0.84 versus 0.58 on control; difference 0.26 (95% CI 0.20 to 0.32), standardized effect d = 0.92; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Etanercept 50 mg Plus MTX -Period 2 than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Etanercept 25 mg Plus MTX-Period 2 than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Prorated Swollen Joint Count at Week 36

• 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 errors.
• Result:
  • Etanercept 50 mg Plus MTX -Period 2: mean 0.63 versus 0.59 on control; difference 0.04 (95% CI -0.23 to 0.31), standardized effect d = 0.03; z-test p = 0.769.
  • Etanercept 25 mg Plus MTX-Period 2: mean 0.57 versus 0.59 on control; difference -0.02 (95% CI -0.26 to 0.22), standardized effect d = -0.02; z-test p = 0.868.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Etanercept 50 mg Plus MTX -Period 2 and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Etanercept 25 mg Plus MTX-Period 2 and the control. The interval gives the range of effects that remain compatible with the data.

Painful Joint Count at Week 36

• 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 errors.
• Result:
  • Etanercept 50 mg Plus MTX -Period 2: mean 0.64 versus 0.74 on control; difference -0.10 (95% CI -0.34 to 0.14), standardized effect d = -0.08; z-test p = 0.406.
  • Etanercept 25 mg Plus MTX-Period 2: mean 0.69 versus 0.74 on control; difference -0.05 (95% CI -0.30 to 0.20), standardized effect d = -0.04; z-test p = 0.694.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Etanercept 50 mg Plus MTX -Period 2 and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Etanercept 25 mg Plus MTX-Period 2 and the control. The interval gives the range of effects that remain compatible with the data.

The posted secondary outcomes cover several related views of disease activity, including DAS28 status and change over time, time to loss of low disease activity, the proportion of follow-up time with low disease activity, and joint-count outcomes. Across the secondary results with group summaries, the broad pattern is that the etanercept combination groups have higher reported low-disease-activity summaries than the placebo combination group, while the Week 36 swollen- and painful-joint-count summaries are relatively similar across groups. These measures provide context around the primary Week 88 threshold result rather than an independent replacement for it.

The secondary outcomes differ in their measurement scales, timeframes, and interpretation. A threshold or proportion of time outcome addresses the frequency or duration of a defined state; a change score or joint count describes a different aspect of disease activity. Agreement across related measures can provide a coherent descriptive picture, but those measures are correlated and should not be treated as separate, unrelated confirmations.

The registry lists many outcome measures and analyses. When multiple outcomes or time points are examined, the chance of obtaining at least one apparently notable result by chance can increase. The registry record does not report a multiplicity adjustment for the secondary analyses, and no sponsor-reported inferential analyses are posted for the secondary measures described in these data. Their unadjusted recalculations therefore carry less evidential weight than a prespecified, multiplicity-controlled analysis would. They are best read as supportive descriptive context, with attention to their intervals and the analysis details available for each outcome.

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

• Percentage of Participants Achieving DAS28 Low Disease Activity or Remission at Baseline, Weeks 4, 8, 12, 20, 28 and 36
• Percentage of Participants Achieving DAS28 Low Disease Activity or Remission
• Change From Baseline in DAS28 at Weeks 4, 8, 12, 20, 28 and 36
• Change From Week 36 in DAS28 at Weeks 40, 48, 56, 64, 72, 80 and 88
• Time to Loss of Low Disease Activity DAS28 and a Change of ≥ 0.6 Units in the DAS28
• Time to Loss of Low Disease Activity DAS28
• Proportion of Time Participants Had Low Disease Activity DAS28 Week 36 to Week 88
• Change From Baseline in Prorated Swollen Joint Count at Weeks 4, 8, 12, 20, 28 and 36

9. Statistical Interpretation

10. Statistical Methods Explained

The primary endpoint is binary because the registry reports the percentage meeting a DAS28 threshold. A binary outcome reduces each participant’s endpoint status to whether the specified criterion was met. The group percentages describe the observed frequency of that status in the endpoint population. They do not retain all the detail in the underlying DAS28 score, such as the degree of improvement or the distance from the threshold.

What does an odds ratio describe?

The sponsor-reported comparisons use the Cochran-Mantel-Haenszel method and report odds ratios. An odds ratio compares the odds of the endpoint between groups; odds are the probability of an event relative to the probability of no event. An odds ratio above 1 indicates higher odds in the first named group relative to the second, as the groups are ordered in the comparison. It is not a risk difference, and it should not be read as the ratio of the two observed percentages. When an outcome is common, odds ratios can appear more pronounced than comparisons expressed as relative probabilities.

The Cochran-Mantel-Haenszel method is a stratified approach to comparing categorical outcomes. In general, it can combine information across strata while accounting for the stratification used in an analysis. The strata and any covariates used in this trial’s sponsor analysis are not reported in the registry record, so the detailed implementation cannot be reconstructed from the posted result alone.

How should the p-value be read?

A p-value describes how unusual data at least as incompatible with a specified null hypothesis would be under the assumptions of the statistical test. It is not the probability that the null hypothesis is true, and it does not measure the size or clinical importance of an effect. The sponsor-reported p-values below < 0.001 for the two placebo comparisons indicate strong statistical evidence against the corresponding no-association null under the analysis used. The p-value of 0.381 for the comparison between dose groups does not provide strong evidence against its null. A non-small p-value does not prove that the groups are equivalent.

The posted comparisons also include simple recalculations based on the observed group summaries. These use an unadjusted risk-difference approach and are not the same method as the sponsor’s Cochran-Mantel-Haenszel odds-ratio analysis. Such recalculations can help readers understand the data on a different scale, but they are not a substitute for the trial’s reported analysis. In this record, the discrepancy flag means the sponsor’s analysis takes precedence. The recalculations should therefore be treated as illustrative rather than as an alternative definitive result.

Why use an interval as well as a point estimate?

A point estimate gives one summary of the observed contrast. A confidence interval describes a range of values compatible with the data and method under repeated-sampling assumptions. Wider intervals indicate less precision, while narrower intervals indicate greater precision. A confidence interval is not a guarantee that the true value lies in the interval, nor does it account automatically for bias, incomplete reporting, or multiplicity.

For an odds ratio, a confidence interval that includes 1 is compatible with no difference on the odds-ratio scale at the stated confidence level. The sponsor-reported interval for the comparison between the etanercept dose groups includes 1. The intervals reported for each etanercept combination versus placebo do not include 1. These statements concern the particular analyses and do not establish that every possible analysis or outcome would have the same conclusion.

What does modified intent-to-treat mean?

The primary endpoint population includes participants who took at least one dose of double-blind study treatment and had at least one post-randomization DAS28 evaluation. This is the registry’s definition, not a claim that every randomized participant contributed to this endpoint. Requiring a post-randomization assessment means participants without such an evaluation are outside the stated endpoint population. The record does not report the number excluded for that reason, how missing Week 88 endpoint values were handled, or whether the analysis used imputation or another missing-data method.

In general, excluding participants based on post-randomization data availability can affect comparability if availability is related to treatment or outcome. Randomization supports fair comparisons at assignment, but the analysis population and missing-data handling still matter for interpreting the estimates. The registry record does not provide enough detail to evaluate the consequences of missing endpoint assessments.

How should over-dispersion be understood?

Over-dispersion is a general issue for count data: observed variability may exceed what a simple probability model assumes. It can make standard errors too small if the model does not account for that extra variability. This concern is relevant in general when analyzing counts, but the primary endpoint is a binary threshold outcome, and the registry does not report a model-based count analysis for it. No claim about over-dispersion or a correction for it can be made for this primary analysis from the registry record.

11. Confidence Intervals and Precision

The sponsor-reported primary comparison of etanercept 50 mg plus methotrexate with placebo plus methotrexate has an odds ratio of 5.36 and a 95% confidence interval from 3.2 to 9.0. The interval is entirely above 1, the value corresponding to equal odds. The reported comparison of etanercept 25 mg plus methotrexate with placebo plus methotrexate has an odds ratio of 4.81 and a 95% confidence interval from 3.0 to 7.6, also entirely above 1. These intervals convey uncertainty about the odds-ratio estimates under the reported method.

For the comparison between the etanercept combination groups, the odds ratio is 1.15 and the 95% confidence interval ranges from 0.7 to 2.0. This interval includes 1 and spans values on either side of equal odds. The point estimate alone should not be treated as a settled difference between groups. The reported p-value for this comparison is 0.381.

Confidence intervals concern sampling uncertainty under the analysis assumptions. They do not measure uncertainty from incomplete registry reporting, possible bias, unreported analytic choices, or the interpretation of a threshold endpoint. The registry record does not give detailed information on missing Week 88 observations or the sponsor analysis’s strata, so the intervals cannot resolve those questions.

The sponsor-reported intervals are the relevant intervals for the trial’s reported comparisons. The registry also flags a discrepancy with simple recalculations from the posted summaries; no recalculated interval should replace the sponsor’s analysis. Precision should therefore be read together with the analysis method and the record’s limitations, not as a standalone guarantee of reproducibility or certainty.

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 illustrations translate an observed effect into hypothetical planning quantities. For these binary comparisons, the figures are participants per arm, not total participants or total events. The stated assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The illustration considers detecting the full observed effect, 75% of the observed effect, or 50% of the observed effect.

The 80% power column gives the planned participants per arm under those assumptions for a hypothetical study to have an 80% chance of detecting the specified effect if that effect were the true one. The 90% power column gives the corresponding per-arm planning quantity for a 90% chance. These are design calculations, not guarantees about a future study’s results. The original trial’s own sample-size justification and planning assumptions are not reported in the registry record.

The rows using smaller fractions of the observed effect require more participants per arm in these illustrations. That is a general feature of power calculations: smaller effects are harder to distinguish from random variation, so more information is needed under otherwise fixed assumptions. The calculations are based on the observed trial effect, which may not be the effect a future study would find. They are not observed power: observed power is not a useful substitute for interpreting the estimated effect, its confidence interval, and the study’s analysis.

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 = 0.426; treatment proportion p₁ = 0.826; average p̄ = 0.626.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6842 = 1.3411.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6232 = 0.5245.
4. n = (1.3411 + 0.5245)² / (0.3995)² = 21.8, rounded up to 22 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 <- 0.426395939086294                          # control proportion
p1 <- 0.825870646766169                          # 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(565409)
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 = 22 (80% power), 29 (90% power)
 75% of observed effect: n per arm = 42 (80% power), 56 (90% power)
 50% of observed effect: n per arm = 97 (80% power), 130 (90% power)
power.prop.test: 21.8 
Simulated power at n = 22 per arm: 0.805 (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 sponsor-reported analyses compare categorical outcomes using the Cochran-Mantel-Haenszel method and express results as odds ratios. The posted recalculations instead use simple unadjusted comparisons of group summaries on a risk-difference scale. Different estimands and methods need not produce identical estimates or p-values, particularly when the outcome is common or when an analysis accounts for features not represented in a simple summary comparison.

The registry record explicitly flags a discrepancy between the recalculations and the sponsor’s reported analysis. Under that circumstance, the sponsor’s own analysis takes precedence when describing the trial’s inferential result. For the primary placebo comparisons, the sponsor reports odds ratios of 5.36 and 4.81, each with p < 0.001. The sponsor reports an odds ratio of 1.15 and p = 0.381 for the comparison between the two etanercept combination groups.

The registry does not report enough detail to reproduce the sponsor’s analysis fully, including the strata used in the Cochran-Mantel-Haenszel method. The simple recalculations remain descriptive checks based on posted summaries; they should not be presented as corrections to, or replacements for, the sponsor’s results.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists 42 outcome measures and 385 analyses, alongside 41 posted secondary outcomes. With multiple endpoints, groups, and assessment times, repeated testing can increase the chance of observing small p-values even when there is no underlying difference for some comparisons. Statistical methods can address this through a prespecified hierarchy or multiplicity adjustment, but the registry record does not report whether such procedures were planned or applied to the secondary analyses. The reported p-values should therefore be interpreted in light of that missing information rather than as if each test were the only one conducted.

The record does not report the original statistical analysis plan, endpoint hierarchy, handling of interim analyses, or details of any adjustment across the primary and secondary outcomes. It also does not report a complete account of missing endpoint observations or the method used to handle them. For the primary endpoint, the modified intent-to-treat definition requires at least one post-randomization DAS28 evaluation, but this definition alone does not explain how missing Week 88 data were treated.

Analysis populations matter because the endpoint results apply to participants meeting the stated modified intent-to-treat criteria, not necessarily every randomized participant. The registry does not provide enough information to assess whether exclusions or unavailable measurements could have affected group comparability. No assumptions about imputation, sensitivity analyses, or the reasons for missing data can be made from the record. These gaps do not erase the reported results, but they limit independent evaluation of robustness and of how well the estimates represent all randomized participants.

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: urinary tract infection (1.0%), pneumonia (0.6%), myocardial infarction (0.5%), gastric ulcer (0.5%), gastritis (0.5%).

The serious adverse event listings use separate risk populations for Period 1 and Period 2. In Period 1, the record lists 38 serious adverse events among 834 at risk. For Period 2, it lists 13 among 202 at risk in the etanercept 50 mg plus methotrexate group, 7 among 202 at risk in the etanercept 25 mg plus methotrexate group, and 15 among 200 at risk in the placebo plus methotrexate group. These are posted counts, not comparative safety estimates adjusted for exposure or other factors.

The registry also lists urinary tract infection at 1.0%, pneumonia at 0.6%, myocardial infarction at 0.5%, gastric ulcer at 0.5%, gastritis at 0.5%, and umbilical hernia at 0.5% among the top serious-event terms. The record does not report deaths in the listed fields. Period labels and risk populations should be kept distinct when interpreting the counts; the figures should not be combined across periods or treated as rates without additional exposure information.

These listings do not establish that an event was caused by a particular treatment, and they do not by themselves support a comparative safety conclusion. Detailed event definitions, exposure-adjusted analyses, and further safety context are not reported in the registry record.

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

The primary result describes the percentage of participants in the defined modified intent-to-treat population who met the DAS28 low-disease-activity threshold at Week 88. The posted percentages are higher in both etanercept plus methotrexate groups than in the placebo plus methotrexate group. Sponsor-reported analyses likewise show odds ratios above 1 for each combination compared with placebo, with reported p-values below < 0.001. These statements summarize the posted endpoint and its reported analysis; they are not medical advice.

The result does not mean that every participant in a group achieved low disease activity, or that group membership determines an individual’s outcome. It does not describe the full trajectory of disease activity between assessments, and it does not show how far participants’ scores moved relative to the threshold. A single threshold assessment cannot convey all dimensions of symptoms, function, quality of life, or disease course.

The comparison between the two etanercept combination groups was not statistically persuasive in the sponsor’s reported analysis: its 95% confidence interval includes 1 and its p-value is 0.381. That finding does not establish equivalence or prove that the doses have the same effect. Establishing equivalence would require a suitable design and prespecified margin, neither of which is reported in the registry record.

The findings also do not by themselves settle the balance of benefits and harms. Serious adverse event counts are posted, but they are not a complete causal or exposure-adjusted safety comparison. Missing details on the original analysis plan, multiplicity, and endpoint missing-data handling further limit interpretation. The strongest reading is therefore bounded: the record reports a clear difference in the primary endpoint comparisons with placebo under the sponsor’s analysis, while leaving important questions about other outcomes, durability, and safety to the wider evidence base.

18. Why This Trial Matters Statistically

This trial illustrates why the endpoint definition, analysis population, and comparison method all matter in clinical research. The primary endpoint is not a raw disease-activity score but a threshold classification at Week 88. That choice makes the result easy to express as a percentage, yet it also loses information about the score’s magnitude and about changes at other times. Statistical interpretation should stay tied to what the endpoint actually measured.

The record also illustrates that an unadjusted recalculation need not reproduce a sponsor’s stratified analysis. The discrepancy flag makes the distinction explicit: a straightforward comparison of posted group summaries is useful as a check, but it is not automatically the trial’s inferential result. Readers need to know which estimate is being discussed, which scale it uses, and whether the source reports an adjusted or stratified method.

Finally, the number of outcomes and analyses makes multiplicity relevant. A collection of secondary findings can add context, but it does not carry the same weight as a primary endpoint analyzed under a clearly reported plan. The registry’s missing detail on multiplicity, missing data, and the original sample-size rationale shows why statistical significance alone is not a complete account of evidence. Precision, analysis population, outcome definition, and transparent reporting all shape what a trial result can support.

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 provides limited detail about eligibility, treatment administration beyond the named groups, masking roles, and the full protocol. Although it identifies the study as randomized, triple-masked, and parallel, it does not report the allocation procedure or the original statistical analysis plan. The discrepancy between recalculated and sponsor-reported results is flagged, but the information needed to reproduce the sponsor’s stratified analysis is not included.

The primary endpoint population requires a post-randomization DAS28 evaluation. The record does not report how many otherwise eligible participants lacked the required assessment, how missing Week 88 data were handled, or whether sensitivity analyses were performed. The record also does not report the original sample-size assumptions or justification, so the planning illustrations should not be confused with the trial’s own design rationale.

Many outcomes and analyses are listed, but multiplicity procedures are not reported. This makes secondary p-values and exploratory comparisons harder to interpret as confirmatory evidence. Secondary outcomes also span different constructs and timeframes, so their results should not be treated as interchangeable with the primary Week 88 threshold endpoint.

Safety reporting includes event counts for named periods and selected serious-event terms, but the registry does not provide exposure-adjusted comparisons or enough detail to attribute events to treatment. The listed fields do not report deaths. These limitations restrict what can be concluded from the registry record alone about robustness, generalizability, and comparative safety.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Study Comparing Etanercept in Combination With Methotrexate in Subjects With Rheumatoid Arthritis — NCT00565409.
• Registry record last updated 2015-08-10; status last verified 2015-08.
• Results first posted 2015-08-10.
• 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

This completed phase 4 randomized, triple-masked, parallel-group study examined etanercept and methotrexate in rheumatoid arthritis. The primary endpoint was the percentage achieving DAS28 of 3.2 or less at Week 88 in a modified intent-to-treat population. The posted percentages were 82.6% for etanercept 50 mg plus methotrexate, 79.1% for etanercept 25 mg plus methotrexate, and 42.6% for placebo plus methotrexate.

The sponsor-reported analyses found odds ratios of 5.36 and 4.81 for the two etanercept combination groups versus placebo plus methotrexate, each with p < 0.001. The comparison between the etanercept groups had an odds ratio of 1.15 and p = 0.381. Because the record flags a discrepancy with simple recalculations, the sponsor’s analysis takes precedence. Multiplicity procedures, missing Week 88 data handling, and the original sample-size assumptions are not reported in the registry record.