A Study With Upadacitinib (ABT-494) in Subjects From China and Selected Countries With Moderately to Severely Active Rheumatoid Arthritis Who Have Had an Inadequate Response to Conventional Synthetic Disease-Modifying Anti-Rheumatic Drugs (csDMARDs): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study With Upadacitinib (ABT-494) in Subjects From China and Selected Countries With Moderately to Severely Active Rheumatoid Arthritis Who Have Had an Inadequate Response to Conventional Synthetic Disease-Modifying Anti-Rheumatic Drugs (csDMARDs) (NCT02955212), a randomized phase 3 study. The primary outcome is Percentage of Participants With an American College of Rheumatology 20% (ACR20) Response at Week 12.

2. Clinical Question

The question concerns a short-term comparison of response proportions during the placebo-controlled period. The registry record also describes evaluation of longer-term safety, tolerability, and efficacy in Period 2. Because participants originally assigned to placebo switched to upadacitinib in that period, the Period 2 arrangement is not a continued placebo comparison.

3. Trial Design

This was a Phase 3, randomized, double-blind, placebo-controlled, parallel-group study with quadruple masking. The registry record lists 338 participants, with 169 in each Period 1 group. One group received upadacitinib 15 mg once daily and the other received placebo once daily, both for 12 weeks. Participants were taking stable doses of csDMARDs and had moderately to severely active rheumatoid arthritis despite an inadequate response to those medicines.

The trial included participants from China and selected countries, including Brazil and South Korea. The stated objectives for Period 1 were to compare efficacy, safety, and tolerability of upadacitinib and placebo. In Period 2, the objective was to evaluate long-term safety, tolerability, and efficacy among participants who completed Period 1. The placebo group switched to upadacitinib 15 mg once daily for up to 52 weeks; the upadacitinib group could continue that dose for up to an additional 52 weeks.

The registry record lists the study start as 2018-01-03, primary completion as 2019-08-14, and status as completed. The sponsor was AbbVie. The original trial sample-size calculation, its assumptions, and the planned treatment of interim analyses are not reported in the registry record. The reported primary endpoint is the Week 12 percentage meeting ACR20 response criteria. The record also lists other Week 12 outcomes, including disease activity, physical function, quality of life, and additional response criteria.

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: Participants who met eligibility criteria were randomized in a 1:1 ratio to one of two treatment groups. Randomization was stratified by country and the Chinese population was expected to comprise up to 80% of the total study population.

Period 1: Double-blind Treatment Period

Reasons for not completing this period:

Period 2: Open-label Extension

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Full analysis set; participants who prematurely discontinued from study drug prior to Week 12 or for whom ACR data were missing at Week 12 were considered non-responders.

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

Population: The Full Analysis Set (FAS) included all randomized participants who received at least 1 dose of study drug.

6. Primary Endpoint

The registry lists the primary outcome as Percentage of Participants With an American College of Rheumatology 20% (ACR20) Response at Week 12, measured over Baseline and Week 12.

The primary endpoint is the percentage of participants meeting the American College of Rheumatology 20% (ACR20) response criteria at Week 12. A participant had to meet all three parts of the definition: at least 20% improvement from baseline in the 68-tender joint count; at least 20% improvement in the 66-swollen joint count; and at least 20% improvement in at least 3 of 5 additional measures. Those measures were physician global assessment of disease activity, patient global assessment of disease activity, patient assessment of pain, the Health Assessment Questionnaire Disability Index (HAQ-DI), and high-sensitivity C-reactive protein.

ACR20 combines joint findings with other assessments of disease activity and patient experience. It is a binary endpoint: each participant either meets the response definition or does not. Reporting the proportion who meet it allows a direct comparison between randomized groups at the specified visit. It does not describe the size of each participant’s improvement beyond the threshold, nor does it by itself capture every aspect of disease status.

The primary population was the full analysis set. Participants who prematurely discontinued study drug before Week 12 or whose ACR data were missing at Week 12 were classified as nonresponders. This rule makes the handling of those particular missing or discontinued observations part of the endpoint definition and its interpretation.

7. Results Posted in the Registry

• Upadacitinib 15 mg: 71.6% (121/169) versus 31.4% (53/169) on control; risk difference 40.2 percentage points (95% CI 30.5 to 50.0), risk ratio 2.28 (1.79–2.91); 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.

Percentage of Participants Achieving Low Disease Activity (LDA) Based on DAS28 (CRP) at Week 12

• 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:
  • Upadacitinib 15 mg: 46.2% (78/169) versus 13.6% (23/169) on control; risk difference 32.5 percentage points (95% CI 23.4 to 41.7), risk ratio 3.39 (2.24–5.13); 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 Upadacitinib 15 mg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Percentage of Participants Achieving Clinical Remission Based on DAS28 (CRP) at Week 12

• 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:
  • Upadacitinib 15 mg: 29.6% (50/169) versus 5.3% (9/169) on control; risk difference 24.3 percentage points (95% CI 16.6 to 31.9), risk ratio 5.56 (2.82–10.93); 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 Upadacitinib 15 mg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Percentage of Participants Achieving Low Disease Activity Based on Clinical Disease Activity Index (CDAI) at Week 12

• 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:
  • Upadacitinib 15 mg: 35.5% (60/169) versus 11.2% (19/169) on control; risk difference 24.3 percentage points (95% CI 15.6 to 32.9), risk ratio 3.16 (1.97–5.05); 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 Upadacitinib 15 mg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The numerical secondary results reported in the registry record follow the same overall direction as the primary endpoint: the upadacitinib group had higher observed proportions for the listed favorable disease activity outcomes at Week 12. These measures cover related but distinct aspects of disease status. Their alignment with the ACR20 result gives additional context, but they do not duplicate the primary endpoint or independently establish the same conclusion.

Secondary outcomes are generally interpreted with more caution when many outcomes are assessed. Testing multiple outcomes creates more opportunities for apparently notable findings, even when no true difference exists for some outcomes. The registry record does not report a multiplicity adjustment or a hierarchy for interpreting these secondary tests. The reported p-values therefore should not be treated as if each outcome were the sole primary question with error control guaranteed across the full set.

The record lists additional secondary measures, including changes in disease activity, physical function, quality of life, and other response thresholds. Numerical findings for secondary outcomes without displayed results are not reported in the registry record, so their direction and precision cannot be assessed from the listed information. The available results do not establish whether the secondary outcome analyses were prespecified in a particular order or how missing observations were handled for each one; those details are not reported in the registry record.

In general, secondary measures can help assess whether a primary result is accompanied by changes in related domains, but they should be interpreted in light of their definitions, multiplicity, precision, and analysis populations. They add context to the primary finding rather than replacing it.

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

• Change From Baseline in Disease Activity Score Based on CRP (DAS28 [CRP]) at Week 12
• Change From Baseline in Health Assessment Questionnaire Disability Index (HAQ-DI) at Week 12
• Change From Baseline in Short-Form 36 (SF-36) Physical Component Score (PCS) at Week 12
• Percentage of Participants Achieving Low Disease Activity (LDA) Based on DAS28 (CRP) at Week 12
• Percentage of Participants Achieving Clinical Remission Based on DAS28 (CRP) at Week 12
• Percentage of Participants Achieving Low Disease Activity Based on Clinical Disease Activity Index (CDAI) at Week 12
• Percentage of Participants With an American College of Rheumatology 50% (ACR50) Response at Week 12
• Percentage of Participants With an American College of Rheumatology 70% (ACR70) Response at Week 12

9. Statistical Interpretation

10. Statistical Methods Explained

What kind of endpoint is ACR20?

ACR20 is a binary responder endpoint. The primary result is the number or percentage classified as meeting the response definition, rather than an average change in a measurement. The definition combines required improvement in tender and swollen joint counts with improvement in a minimum number of additional measures. That composite gives one outcome for analysis while retaining several clinical assessment domains in the response rule. It also means the response percentage does not show how far participants improved beyond the stated thresholds.

What does the response-rate difference describe?

The response-rate difference compares the proportion meeting ACR20 in the upadacitinib group with the proportion in the placebo group. It expresses an absolute contrast in response proportions. The registry-reported difference is 40.2 percentage points. An absolute difference is useful because it states how far apart the observed group proportions are on the percentage scale. It is distinct from a risk ratio or an odds ratio, which express the comparison on relative scales and can give a different impression of the same data.

The posted results also include a simple unadjusted re-calculation from the response counts and group sizes. Such a calculation describes the crude contrast using the displayed summaries. It does not reproduce any adjustment or stratification that may have been part of the sponsor’s analysis. The sponsor reported the Cochran-Mantel-Haenszel method; the particular strata or other implementation details are not reported in the registry record.

How should the p-value be read?

The reported p-value is p < 0.001. In general, a p-value measures how incompatible the observed data, or more extreme data, are with a specified null model when the model assumptions hold. It is not the probability that the null hypothesis is true, and it does not quantify the probability that the result will be replicated. It also does not communicate the size of the response-rate difference; the estimate and its confidence interval address magnitude and precision.

The registry lists the sponsor analysis as a superiority analysis. This describes the statistical question of whether the response rate in the upadacitinib group exceeds that in the placebo group under the analysis. A small p-value alone does not establish clinical importance, generalizability, or the appropriateness of a treatment decision. Those questions require consideration of the endpoint, study population, safety information, and limitations as well as statistical evidence.

What does a confidence interval add?

A confidence interval gives a range of estimates compatible with the data and the stated statistical procedure. In repeated use of the same procedure under its assumptions, intervals constructed in this way would cover the fixed underlying value at the stated confidence level at the corresponding frequency. It is not a range containing a specified proportion of individual participant outcomes, and it is not a guarantee that the true value lies in this particular interval.

For the primary response-rate difference, the registry reports a 95% confidence interval from 30.5 to 50.0 percentage points. The lower and upper limits indicate uncertainty around the estimated contrast. The width reflects sampling uncertainty as well as the analysis method; it does not address biases that could arise from design, conduct, outcome measurement, or reporting.

Why can the reported and unadjusted estimates differ?

The sponsor’s Cochran-Mantel-Haenszel analysis and an unadjusted calculation from displayed counts need not be numerically identical. In general, a method that accounts for strata can produce an estimate and interval that differ from a crude comparison. The registry record does not identify the strata or give all implementation details. For the primary endpoint, the posted sponsor estimate and the simple unadjusted re-calculation are in close agreement at the precision reported. The sponsor’s analysis is the registry-reported inferential result.

What is the role of the analysis population and missing-data rule?

The primary endpoint analysis uses the full analysis set. The record explicitly classifies participants who discontinued study drug before Week 12 or had missing ACR data at Week 12 as nonresponders. In general, nonresponder imputation assigns a nonresponse outcome to cases covered by that rule instead of excluding them from the responder count. The rule can affect the estimated response proportion, so it is important to interpret the result alongside the specified population and missing-data handling. Further details about other analysis populations or sensitivity analyses are not reported in the registry record.

What does over-dispersion mean in general?

For binary outcomes, a basic model may assume that variation follows the pattern expected under its probability model. Over-dispersion means observed variability is greater than that model expects. If present and not accommodated, uncertainty estimates can be too small. This is a general statistical concept; over-dispersion testing or adjustment is not reported in the registry record, and the reported comparison should not be assumed to have used such an adjustment.

11. Confidence Intervals and Precision

The primary response proportions are 31.4% for placebo and 71.6% for upadacitinib 15 mg. Their registry-reported 95% confidence intervals are 24.4 to 38.4% and 64.8 to 78.4%, respectively. These intervals convey sampling uncertainty around each group’s response proportion. They are not ranges for individual participant responses.

For the comparison, the reported response-rate difference is 40.2 percentage points, with a 95% confidence interval from 30.5 to 50.0 percentage points. The interval is entirely on the positive side of zero, and the reported p-value is p < 0.001. Together, the estimate, interval, and p-value describe the observed contrast and its statistical uncertainty. The interval does not include uncertainty from every possible source, such as limitations in study conduct or the possibility that the selected population differs from other populations.

Confidence intervals are most useful when read with the estimate and the endpoint definition. The primary estimate concerns the proportion meeting a threshold-based ACR20 definition at Week 12, with the prespecified nonresponder rule for early discontinuation and missing ACR data. It does not directly provide precision for an average change in disease activity, a different response threshold, or a later time point.

The interval’s width also depends on the amount of information and the analysis procedure. The trial had 169 participants in each Period 1 group, but a confidence interval is not a measure of whether the sample is representative of all people with rheumatoid arthritis. Nor does a narrow interval remove concerns about multiplicity, missing-data assumptions, or the limited duration of the placebo-controlled comparison.

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 use the observed primary effect as a reference and ask how many participants per arm would be needed under specified hypothetical conditions. The entries are participants per arm, not total sample sizes or total events. They are planning illustrations based on the observed effect, not the trial’s own original design calculation. The assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation; each illustration considers an effect equal to the stated fraction of the observed effect. The trial’s original sample-size assumptions are not reported in the registry record.

For example, planning for 100% of the observed primary effect gives 24 participants per arm for 80% power and 31 participants per arm for 90% power. Planning for 75% of the observed effect gives 42 participants per arm for 80% power and 56 participants per arm for 90% power. Planning for 50% of the observed effect gives 93 participants per arm for 80% power and 124 participants per arm for 90% power. The curve should be read as illustrating how the planned group size varies as the assumed effect changes, under the same stated assumptions.

Power is the probability that a planned statistical test will reject its null hypothesis if a specified effect is present and the assumptions hold. The 80% and 90% columns therefore represent different target probabilities for detecting the assumed effect. They are not guarantees that a future study will reach a particular result. Smaller assumed effects are harder to distinguish from random variation, so they require more participants under these assumptions. These figures are not observed power.

Because these are hypothetical calculations anchored to an observed result, they should not be read as evidence that the original study was designed with these group sizes or assumptions. They also do not account for design features or practical considerations that are not included in the stated calculation assumptions.

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 = 0.314; treatment proportion p₁ = 0.716; average p̄ = 0.515.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.7068 = 1.3853.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6470 = 0.5445.
4. n = (1.3853 + 0.5445)² / (0.4024)² = 23.0, rounded up to 24 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 <- 0.313609467455621                          # control proportion
p1 <- 0.715976331360947                          # 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(2955212)
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 = 24 (80% power), 31 (90% power)
 75% of observed effect: n per arm = 42 (80% power), 56 (90% power)
 50% of observed effect: n per arm = 93 (80% power), 124 (90% power)
power.prop.test: 23 
Simulated power at n = 24 per arm: 0.843 (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 a Cochran-Mantel-Haenszel analysis of the primary response-rate difference. It gives a difference of 40.2 percentage points, a 95% confidence interval from 30.5 to 50.0 percentage points, and p < 0.001. The reported analysis is identified as a superiority analysis.

A simple unadjusted re-calculation from the posted response counts and group sizes gives a result consistent with the sponsor’s reported estimate at the displayed precision. The re-calculation is based only on the displayed summaries and does not reproduce any features of the Cochran-Mantel-Haenszel procedure that may affect the estimate or interval. The specific stratification details are not reported in the registry record.

The sponsor’s analysis is the primary registry-reported inferential result. The unadjusted calculation is a check based on the available summaries, not a replacement for the sponsor’s analysis. The two approaches point to the same broad interpretation of the Week 12 contrast, while the sponsor’s method and reported confidence interval should be used when describing the official analysis.

15. Multiplicity, Missing Data and Analysis Populations

The registry record lists one primary endpoint and several secondary outcomes. It does not report a multiplicity procedure, such as an adjustment across secondary comparisons, or a testing hierarchy that would specify how evidence from multiple outcomes should be interpreted. Without such information, the secondary p-values should be considered unadjusted unless the record states otherwise. Repeated testing can increase the chance of at least one small p-value across a group of tests, so isolated secondary findings carry less confirmatory weight than the primary analysis.

The primary analysis population is the full analysis set. The registry explicitly states that participants who prematurely discontinued study drug before Week 12 or had missing ACR data at Week 12 were considered nonresponders. This is a defined approach for the primary endpoint, but it does not reveal the amount or pattern of missing data. The number of participants affected by this rule, reasons for missing observations, and sensitivity analyses evaluating alternative assumptions are not reported in the registry record.

How missing data are handled can influence a binary response estimate. Counting the specified cases as nonresponders preserves them in the endpoint classification, but the result still depends on the appropriateness of that assumption for those cases. The record does not provide enough detail to assess the impact of alternative missing-data assumptions or the handling of missing values for each secondary outcome.

The registry identifies the sponsor’s primary method as Cochran-Mantel-Haenszel, but details of any stratification factors and other model specifications are not reported in the registry record. Interim analyses, stopping rules, and the original sample-size justification are also not reported in the registry record. These omissions limit assessment of the full analysis plan and how the reported tests relate to any planned multiplicity control.

16. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: pneumonia (2.5%), herpes zoster (1.2%), tendon rupture (1.2%), cholelithiasis (0.9%), appendicitis (0.6%).

During Period 1, the registry reports serious adverse events in 5 of 169 participants in the placebo group and 12 of 169 participants in the upadacitinib 15 mg group. Other adverse events are reported for 13 of 169 participants and 25 of 169 participants, respectively. No deaths are reported in either Period 1 group.

For the group labeled Period 1+2: Upadacitinib 15 mg, the registry reports serious adverse events in 55 of 322 participants and other adverse events in 198 of 322 participants; deaths are reported as 0. This combined-period group has a different period of observation and group composition from either Period 1 arm, so these counts are not a direct comparison with the placebo-controlled Period 1 counts.

The registry lists pneumonia (2.5%), herpes zoster (1.2%), tendon rupture (1.2%), cholelithiasis (0.9%), appendicitis (0.6%), and postoperative wound infection (0.6%) among the serious adverse event terms. These are reported event terms, not evidence by themselves that treatment caused an event. The record does not provide exposure-adjusted rates or additional details needed to compare event risk over equivalent follow-up.

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

The primary result describes a difference between randomized groups in the percentage meeting a defined ACR20 response threshold at Week 12. The observed response proportion was higher in the upadacitinib 15 mg group than in the placebo group, and the registry-reported comparison had a 95% confidence interval from 30.5 to 50.0 percentage points with p < 0.001. This is a statistical description of the posted short-term result, not a guarantee about an individual participant’s outcome.

The endpoint captures a combination of joint-count and other assessment changes. It does not tell how much an individual’s condition changed beyond the threshold, whether the outcome persisted after Week 12, or how every aspect of health was affected. Other outcomes listed in the registry may address related domains, but results not reported in the record cannot be inferred from the primary result.

The trial enrolled participants from China and selected countries who had an inadequate response to csDMARDs and were on stable doses of those medicines. The findings are most directly relevant to the study population, treatment regimen, comparator, and follow-up described in the registry record. Generalizing to different populations, doses, background treatments, or settings requires evidence beyond this comparison.

The placebo-controlled comparison was conducted in Period 1. Participants in the placebo group then switched to upadacitinib for Period 2, so the combined follow-up does not maintain the original placebo comparison. Safety counts from the combined period therefore answer a different descriptive question from the Period 1 comparison.

Statistical evidence does not, on its own, establish the balance of benefits and harms for a particular person or determine a treatment choice. The posted endpoint, confidence interval, missing-data rule, secondary outcome limitations, and safety counts all contribute to understanding the evidence, while the registry record leaves some design and analysis details unreported.

18. Why This Trial Matters Statistically

This trial illustrates how a binary clinical response endpoint can provide a direct comparison between randomized groups. ACR20 turns a multi-part definition involving joint counts and additional disease assessments into a responder proportion. That makes the result easy to express as an absolute difference, while also requiring careful attention to the threshold and the classification rule for missing or discontinued cases.

The record gives both a sponsor-reported comparison and sufficient response counts and group sizes for a simple unadjusted re-calculation. Their agreement at the reported precision offers a useful consistency check, while the difference in analysis methods reminds readers that crude and stratified estimates are not necessarily identical. Statistical interpretation benefits from reading the estimate and its confidence interval alongside the p-value rather than treating significance as a complete summary.

The study also highlights the distinction between randomized short-term evidence and longer-term follow-up. The placebo group switched to upadacitinib in Period 2, meaning that the later combined-period safety counts do not preserve a parallel placebo comparison. The time period and group definition must therefore accompany any reading of those counts.

Finally, the registry record permits a planning illustration based on the observed primary effect, but that exercise is distinct from the trial’s original design. The stated assumptions make clear how target power and effect size influence hypothetical participant requirements. The original design justification is not reported, so these calculations should not be used to reconstruct it.

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 study population is defined by moderately to severely active rheumatoid arthritis, inadequate response to csDMARDs, and participation from China and selected countries, including Brazil and South Korea. The findings may not transfer directly to populations with different disease characteristics, prior treatment histories, or geographic representation. The record does not give a detailed distribution of participants across countries or other demographic characteristics.

The placebo-controlled comparison is limited to Period 1, which lasted 12 weeks. Participants assigned to placebo switched to upadacitinib in Period 2. As a result, the longer follow-up described in the registry does not offer a continued randomized placebo comparison for evaluating longer-term differences between the original groups.

The primary outcome is a threshold-based composite response. It does not capture the full distribution of changes in each component or distinguish participants who just meet the criteria from those with larger improvements. The primary missing-data rule counts early study-drug discontinuations and missing Week 12 ACR data as nonresponse; the number affected and sensitivity analyses are not reported in the registry record.

The registry record does not report the original sample-size justification, details of interim analyses, multiplicity control, or all aspects of the statistical analysis plan. Numerical findings for several listed secondary measures are not reported in the registry record. Safety data are presented as counts, and exposure-adjusted rates or equivalent follow-up comparisons are not reported. These details constrain evaluation of secondary evidence, longer-term safety comparisons, and the robustness of conclusions beyond the primary result.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Study With Upadacitinib (ABT-494) in Subjects From China and Selected Countries With Moderately to Severely Active Rheumatoid Arthritis Who Have Had an Inadequate Response to Conventional Synthetic Disease-Modifying Anti-Rheumatic Drugs (csDMARDs) — NCT02955212.
• Registry record last updated 2021-09-27; status last verified 2021-08.
• Results first posted 2020-08-06.
• 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

NCT02955212 was a completed Phase 3 randomized, double-blind, placebo-controlled parallel-group study sponsored by AbbVie, with quadruple masking. The registry record lists 338 participants. In Period 1, 169 participants received upadacitinib 15 mg once daily and 169 received placebo once daily for 12 weeks. Participants had moderately to severely active rheumatoid arthritis and an inadequate response to csDMARDs.

The primary endpoint was the percentage meeting ACR20 response criteria at Week 12. The registry reports response rates of 71.6% for upadacitinib and 31.4% for placebo, with a response-rate difference of 40.2 percentage points (95% confidence interval 30.5 to 50.0; p < 0.001). The primary analysis treated early study-drug discontinuation and missing Week 12 ACR data as nonresponse.

Period 2 evaluated longer-term outcomes, with participants in the placebo group switching to upadacitinib. The record also reports selected secondary outcomes and safety counts. Several statistical-plan details, including the original sample-size justification and multiplicity procedures, are not reported in the registry record.