A Study of the Efficacy and Safety of Upadacitinib (ABT-494) in Participants With Moderately to Severely Active Ulcerative Colitis: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of the Efficacy and Safety of Upadacitinib (ABT-494) in Participants With Moderately to Severely Active Ulcerative Colitis (NCT03653026), a randomized phase 3 study. The primary outcome is Percentage of Participants Who Achieved Clinical Remission Per Adapted Mayo Score at Week 8.

2. Clinical Question

3. Trial Design

This was a Phase 3, randomized, parallel-group, placebo-controlled induction study with quadruple masking. The registry lists ulcerative colitis as the condition and describes the study objective as evaluating upadacitinib’s efficacy and safety, compared with placebo, for inducing clinical remission. The study is completed. The registry reports enrollment of 522 participants, a start date of 2018-12-06, and a primary completion date of 2021-01-14.

The Part 1 intervention period was eight weeks. Participants assigned to upadacitinib received 45 mg once daily; those assigned to placebo received matching placebo once daily. According to the arm descriptions, participants who did not achieve clinical response at Week 8 entered an open-label extension and received upadacitinib for an additional eight weeks. The posted primary endpoint is measured at Week 8, so it addresses the induction period rather than the later extension.

The primary endpoint population is described as the Part 1 intent-to-treat population: randomized participants who received at least one dose of study drug in Part 1. The record also states that participants from one site were excluded for non-compliance. Its description of the missing-data approach is incomplete in the registry record, so the full handling of missing observations cannot be established from that description.

The registry lists 12 outcome measures and 12 analyses, while the results record includes primary and secondary endpoints. The planned allocation ratio, original sample-size justification, interim-analysis arrangements, and full statistical plan are not reported in the registry record. The planning illustrations accompanying the posted results are based on the observed effect and should not be mistaken for the study’s original design assumptions.

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 Part 1 participants were randomized in a 2:1 ratio to upadacitinib or placebo. Randomization was stratified by biologic-inadequate responder (Bio-IR) status (bio-IR vs non-bio-IR), corticosteroid use (yes or no), and Adapted Mayo score (≤ 7 or \> 7) at Baseline. Within bio-IR, randomization was further stratified by number of prior biologic treatments (≤ 1 or \> 1). Within non-bio-IR, randomization was further stratified by previous biologic use (yes or no).

Part 1: Placebo-controlled Period

Reasons for not completing this period:

Part 2: Open-label Extension

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The Part 1 intent-to-treat population (ITT1) includes randomized participants who received at least 1 dose of study drug in Part 1. The ITT1 population excludes 6 participants from 1 site with non-compliance. Non-responder imputation incorporating multiple imputation to handle missing data due to Coronavirus Disease - 2019 (COVID-19) was used.

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 Who Achieved Clinical Remission Per Adapted Mayo Score at Week 8, measured over Week 8.

The primary endpoint is the percentage of participants who achieved clinical remission per the Adapted Mayo score at Week 8. The score combines stool frequency, rectal bleeding, and endoscopic findings into a measure of ulcerative-colitis disease activity. The registry describes the overall score as ranging from 0 to 9, with higher scores representing more severe disease. Clinical remission requires an overall score of 2 or less, a stool-frequency subscore of 1 or less and no higher than baseline, a rectal-bleeding subscore of 0, and an endoscopic subscore of 1 or less.

A remission definition combines symptoms with an endoscopic assessment, rather than relying on a single symptom alone. That makes the endpoint a clinically interpretable summary of several dimensions of disease activity, while also requiring each stated component criterion to be met. The result is reported as a percentage of participants, with a 95% confidence interval.

Interpretation depends on the defined population and the handling of missing outcomes. The registry identifies the Part 1 intent-to-treat population and notes a site-level exclusion, but the full missing-data description is not reported in the registry record. The endpoint therefore should be read together with the population definition and the reported analysis method, not as a simple count of all enrolled participants.

7. Results Posted in the Registry

• Upadacitinib 45 mg: 33.4% (114/341) versus 4.0% (7/174) on control; risk difference 29.4 percentage points (95% CI 23.6 to 35.2), risk ratio 8.31 (3.96–17.44); 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 With Endoscopic Improvement at Week 8

• 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 45 mg: 44.0% (150/341) versus 8.0% (14/174) on control; risk difference 35.9 percentage points (95% CI 29.3 to 42.6), risk ratio 5.47 (3.26–9.16); 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 45 mg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Percentage of Participants With Endoscopic Remission at Week 8

• 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 45 mg: 18.2% (62/341) versus 1.7% (3/174) on control; risk difference 16.5 percentage points (95% CI 11.9 to 21.0), risk ratio 10.55 (3.36–33.11); 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 45 mg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Percentage of Participants Who Achieved Clinical Response Per Adapted Mayo Score at Week 8

• 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 45 mg: 74.5% (254/341) versus 25.3% (44/174) on control; risk difference 49.2 percentage points (95% CI 41.3 to 57.1), risk ratio 2.95 (2.26–3.83); 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 45 mg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The detailed secondary endpoint results follow a consistent numerical direction: the reported percentages for the upadacitinib group are higher than those for placebo for endoscopic improvement, endoscopic remission, and clinical response at Week 8. The registry reports sponsor analyses for these endpoints using the Cochran-Mantel-Haenszel method, with superiority labels and p < 0.001. These outcomes offer related views of disease activity, including endoscopic findings and clinical response, and thus provide context alongside the primary remission endpoint.

The secondary outcomes are not identical to the primary endpoint. Each uses its own definition and can capture a different aspect or threshold of response. A pattern aligned with the primary result can add descriptive coherence, but it does not turn the secondary outcomes into independent replications: they come from the same trial population and may be statistically related. The registry lists additional outcome measures, but detailed result entries are not present for every listed secondary measure in the record summarized for analysis. Their results should not be inferred from the endpoints with posted values.

Unadjusted secondary p-values carry less weight when the plan for handling multiple comparisons is unknown. Testing several outcomes can raise the chance of at least one apparently strong finding under a null hypothesis. A prespecified hierarchy or multiplicity adjustment can address that issue, but the multiplicity procedure is not reported in the registry record. The secondary findings therefore support a descriptive account of the results, while confirmatory interpretation depends on a statistical plan that is not available in the record.

Secondary results should be read with their confidence intervals and endpoint definitions, rather than ranked only by p-value. They help describe the breadth of the Week 8 outcome pattern but do not expand the primary endpoint’s scope to longer-term outcomes or the open-label extension.

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

• Percentage of Participants With Endoscopic Improvement at Week 8
• Percentage of Participants With Endoscopic Remission at Week 8
• Percentage of Participants Who Achieved Clinical Response Per Adapted Mayo Score at Week 8
• Percentage of Participants Who Achieved Clinical Response Per Partial Adapted Mayo Score at Week 2
• Percentage Of Participants Who Achieved Histologic-Endoscopic Mucosal Improvement at Week 8
• Percentage of Participants Who Reported No Bowel Urgency at Week 8
• Percentage of Participants Who Reported No Abdominal Pain at Week 8
• Percentage of Participants Who Achieved Histologic Improvement at Week 8

9. Statistical Interpretation

10. Statistical Methods Explained

Several statistical ideas help interpret these results. The explanations below describe general concepts; the study-specific details are those stated in the registry record and posted results.

What does a percentage endpoint represent?

A binary endpoint classifies each participant as meeting or not meeting the remission definition at the specified assessment. The group percentage is the estimated proportion meeting that definition. It is not a measure of how much each participant improved, and it does not show the distribution of score changes among participants who did not meet remission criteria. The registry reports the primary endpoint as a percentage of participants at Week 8.

What is a risk difference?

A risk difference compares the event proportions in two groups by subtracting one group proportion from the other. It expresses the absolute difference in the probability of the recorded event, in this case clinical remission. A positive difference in the direction used for the posted comparison means the event percentage is higher in the upadacitinib group. The registry’s sponsor-reported adjusted response-rate difference is 29.0 percentage points. The recalculated comparison is unadjusted and uses posted summary statistics; it should not be confused with the sponsor’s adjusted estimate.

What does the Cochran-Mantel-Haenszel method do?

As a general method, the Cochran-Mantel-Haenszel procedure can estimate or test an association between treatment group and a binary outcome while accounting for defined strata. The sponsor’s reported analysis names this method and reports an adjusted response-rate difference. The registry record does not specify the strata or provide the full analysis details needed to reproduce that adjustment. The reported adjusted estimate should therefore be distinguished from a simple calculation using only the arm-level percentages.

What does a p-value convey?

A p-value is a measure of how compatible the observed data are with a specified null hypothesis under the assumptions of the analysis. It is not the probability that the null hypothesis is true, and it does not measure the probability that a result is due to chance in an everyday sense. The sponsor reports p < 0.001 for the primary comparison. That result indicates strong statistical evidence against the null hypothesis used in the reported test, but the p-value does not convey the effect’s magnitude or resolve uncertainty about bias, missing data, or generalizability.

What does an adjustment change?

An adjusted analysis can account for features of the analysis plan, such as stratification, that a crude comparison of pooled percentages does not incorporate. Adjustment may alter the estimate and its uncertainty. The sponsor’s adjusted estimate and the simple unadjusted recalculation need not match exactly, even when they use the same underlying trial. The registry record reports the sponsor’s method and adjusted result; its available summary does not provide every detail necessary to recreate the adjustment.

Why does the analysis population matter?

An analysis population defines which participants contribute to the endpoint estimate. Here, the primary result is described for the Part 1 intent-to-treat population, comprising randomized participants who received at least one dose of study drug in Part 1, with participants from one site excluded for non-compliance. That description is important when interpreting the percentages. The registry’s wording about missing data is incomplete, so the exact imputation procedure and its assumptions are not reported in the registry record.

What does superiority mean in the record?

A superiority analysis tests whether the results support a difference in a specified direction under the analysis plan. The sponsor’s primary result is labelled superiority, with the adjusted difference favoring a higher remission percentage for upadacitinib and p < 0.001. This is a statistical classification of the comparison, not a guarantee about the effect in other settings or a statement about every participant. A confidence interval remains important because it describes the range of effect sizes compatible with the data and method.

Finally, the primary endpoint is one of several outcomes. When multiple outcomes are examined, the chance of obtaining at least one apparently persuasive result can increase if no adjustment or hierarchy is used. The registry record does not report the multiplicity strategy, so the inferential status of the secondary findings should be considered less definitive than a result interpreted under a fully documented multiplicity plan.

11. Confidence Intervals and Precision

The primary arm percentages are estimates, not exact population values. Their 95% confidence intervals are 28.5 to 38.5 for upadacitinib 45 mg and 1.1 to 7.1 for placebo. The sponsor-reported adjusted difference is 29.0 percentage points, with a 95% confidence interval from 23.2 to 34.7. These intervals communicate uncertainty around the estimates under the methods used. In general, a confidence interval is a range generated by a procedure that, over repeated similar studies under its assumptions, would cover the true value at the stated rate.

The adjusted interval for the difference lies above zero, consistent with the reported superiority test and p < 0.001. Its span also shows that the exact size of the adjusted difference is not pinned down to a single value. The interval describes statistical uncertainty; it does not account for every possible source of error, such as bias from design or conduct, incomplete reporting, or departures from the analysis assumptions.

The recalculated unadjusted comparison has its own confidence interval, derived from the posted group summaries. It is not interchangeable with the sponsor’s adjusted interval because the estimates use different analysis approaches. The registry record does not report enough detail to reconstruct the adjustment fully.

Confidence intervals for secondary endpoints also indicate varying degrees of precision, and estimates should be interpreted in light of their endpoint definitions and analysis populations. A narrow interval can indicate greater statistical precision under the model, but does not ensure that an estimate is free of bias or clinically important. Conversely, a wider interval signals greater uncertainty about the magnitude. Precision should be considered alongside the endpoint, analysis plan, and incomplete reporting of missing-data and multiplicity procedures.

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 show planning scenarios based on the observed effect, rather than documenting how the original trial was designed. For these binary outcomes, the displayed quantities are participants per arm, not total events. The planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. Each scenario uses the stated fraction of the observed effect as the effect to detect. The original study’s own sample-size assumptions and justification are not reported in the registry record.

The 80% power column gives the illustrative number of participants per arm for a design targeting an 80% chance of detecting the specified effect under the planning assumptions, if that effect and those assumptions were correct. The 90% power column describes the corresponding target for a higher chance of detection. These are prospective planning calculations, not a measure of what the completed trial achieved.

Rows that plan for a smaller fraction of the observed effect require more participants per arm in the illustrations. This is a general feature of power calculations: smaller differences are harder to distinguish from random variation, so larger sample sizes are generally needed when other assumptions remain fixed. The displayed scenarios are illustrative rather than guarantees; actual power depends on the true effect, event rates, analysis method, and other design features.

These figures are not observed power. They are not evidence that the trial had a particular probability of success after the results were known, and they should not be used as a substitute for interpreting the reported effect estimate and confidence interval. The registry does not report the original design’s sample-size justification, and the planning illustrations should be read only under their stated 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 = 0.040; treatment proportion p₁ = 0.334; average p̄ = 0.187.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.5517 = 1.0814.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.5110 = 0.4301.
4. n = (1.0814 + 0.4301)² / (0.2941)² = 26.4, rounded up to 27 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 <- 0.0402298850574713                          # control proportion
p1 <- 0.334310850439883                          # 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(3653026)
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 = 27 (80% power), 35 (90% power)
 75% of observed effect: n per arm = 41 (80% power), 54 (90% power)
 50% of observed effect: n per arm = 73 (80% power), 96 (90% power)
power.prop.test: 26.4 
Simulated power at n = 27 per arm: 0.868 (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’s primary analysis is a Cochran-Mantel-Haenszel adjusted response-rate difference of 29.0 percentage points, with a 95% confidence interval from 23.2 to 34.7 and p < 0.001. It is labelled as a superiority analysis. The posted arm-level percentages are 33.5% for upadacitinib 45 mg and 4.1% for placebo.

The recalculation uses the posted summary statistics to form a simple, unadjusted comparison. It is a useful check on the direction and approximate scale of the group difference, but it is not a reproduction of the sponsor’s adjusted analysis. The adjusted and unadjusted estimates can differ because the methods do not necessarily use the same information or assumptions. The registry record does not give sufficient detail about the sponsor’s adjustment to independently recreate it from the summarized results.

The registry marks no discrepancy between the posted sponsor analysis and the recalculation. That does not mean the estimates are identical: they are distinct analyses, and the sponsor-reported result remains the study’s reported inferential analysis. The unadjusted recalculation should be understood as a descriptive comparison from group summaries, not as a replacement for the prespecified or sponsor-reported analysis.

15. Multiplicity, Missing Data and Analysis Populations

The registry record describes a primary endpoint and multiple secondary outcomes, and lists 12 outcome measures and 12 analyses. When multiple hypotheses are tested, the chance of at least one small p-value can increase even if none of the underlying effects is real. Statistical plans commonly address this through procedures such as a testing hierarchy or an adjustment to the significance threshold. Which procedure was planned for this study is not reported in the registry record.

This gap matters most when interpreting secondary results. Their reported p-values may describe the individual comparisons, but without a documented multiplicity strategy it is not possible to determine from the registry record whether they were intended as confirmatory tests or how family-wise error was controlled. The secondary pattern can still be described, but it should not be treated as if every endpoint had the same evidential status as the primary analysis.

The primary population is described as the Part 1 intent-to-treat population, including randomized participants who received at least one dose of study drug in Part 1, with participants from one site excluded for non-compliance. The registry’s population description also begins to describe a non-responder imputation approach incorporating multiple imputation for missing data, but the wording is incomplete. The remaining details, including the full missing-data procedure and assumptions, are not reported in the registry record.

Missing-data assumptions can affect estimated response rates because participants without an observed endpoint may differ from those with observed outcomes. Different imputation approaches can make different assumptions about those unobserved outcomes. Without the complete method, sensitivity analyses, and details of analysis populations for each endpoint, the robustness of the estimates to missing data cannot be assessed from this record alone. The registry also does not report the full statistical analysis plan or interim-analysis details.

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: colitis ulcerative (1.7%), abdominal discomfort (0.9%), abdominal pain lower (0.9%), covid-19 pneumonia (0.9%), chronic obstructive pulmonary disease (0.9%).

Safety results are reported separately for Part 1 and Part 2, and the denominators differ across groups and periods. In Part 1, the registry lists 11 serious adverse events among 344 participants at risk in the upadacitinib group and 8 among 177 at risk in the placebo group. It reports zero deaths in each Part 1 group. Part 2 lists participants who received upadacitinib after either initial assignment; the registry reports one serious adverse event among 68 at risk in the upadacitinib/upadacitinib group and 4 among 116 at risk in the placebo/upadacitinib group, with zero deaths in each.

These are descriptive counts and should be read with their stated group and period. Part 2 follows a different treatment history from the randomized Part 1 comparison, so those groups should not be treated as a continuation of the original placebo-controlled contrast. The registry does not provide enough detail in these summary fields to characterize event duration, relatedness to treatment, or comparative risk-adjusted rates.

The record also lists serious-event terms and their posted percentages, including ulcerative colitis at 1.7%; abdominal discomfort, lower abdominal pain, COVID-19 pneumonia, and chronic obstructive pulmonary disease at 0.9% each; and anaemia at 0.6%. These terms are descriptive listings, not evidence by themselves of causation. Safety counts do not determine whether a treatment is appropriate for an individual.

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

The primary result describes a Week 8 comparison of clinical remission under the trial’s induction assignments. The registry reports a greater percentage meeting the remission definition in the upadacitinib group than in the placebo group, and the sponsor’s adjusted analysis reports a positive difference with a confidence interval above zero and p < 0.001. This is evidence about the trial’s stated endpoint, population, and analysis—not a guarantee about outcomes for every participant.

The result does not establish the precise effect that would occur in every clinical setting. Trial participants, study procedures, outcome assessment, missing-data assumptions, and follow-up all shape the estimate. The available record does not provide the complete missing-data method, multiplicity strategy, or full analysis plan. The posted estimate also addresses the induction endpoint at Week 8; it does not establish outcomes after the extension period or long-term comparative effects.

A statistically persuasive comparison is not the same as a complete assessment of clinical importance. The size of the adjusted difference and its confidence interval convey magnitude and uncertainty, while the p-value addresses compatibility with a null hypothesis under the analysis assumptions. Neither resolves all questions about durability, adverse effects, or applicability to people who were not represented in the analyzed population.

Safety counts are also limited summaries. They report serious adverse events and deaths by listed group and period, but do not provide the context needed to infer causality or assess an individual’s risks. The registry’s serious-event terms are descriptive and should not be treated as a complete account of every possible benefit or harm.

In short, the posted results support a Week 8 statistical difference in the measured remission endpoint under the sponsor’s reported analysis. They do not answer every clinical question, establish long-term outcomes, or replace individualized medical judgment.

18. Why This Trial Matters Statistically

This trial illustrates why an endpoint definition matters. Clinical remission is not merely a change in one symptom: the Adapted Mayo definition combines stool frequency, rectal bleeding, and endoscopic findings with specified component thresholds. The reported percentage therefore has a precise interpretation, and results for related endpoints such as clinical response or endoscopic improvement should not be treated as interchangeable with remission.

It also shows why an adjusted analysis and a simple group comparison can both be informative while remaining distinct. The sponsor reports a Cochran-Mantel-Haenszel adjusted difference, whereas the recalculation uses posted arm-level summaries without that adjustment. Their estimates are close in direction and scale, but the adjusted result is the sponsor’s inferential analysis and cannot be reconstructed fully from the available summary.

Confidence intervals add information beyond a significance label. They show a range of effect sizes compatible with the data and method, making the estimated magnitude and its uncertainty visible. The posted primary interval for the adjusted difference excludes zero, while still allowing a range of plausible effects.

The record also highlights the importance of transparent reporting. The analysis population is identified, but the missing-data description is incomplete, and the multiplicity strategy and original sample-size justification are not reported. Those omissions constrain how fully readers can evaluate assumptions and the status of secondary outcomes. Statistical interpretation depends not only on estimates and p-values, but also on understanding who was analyzed, how missing outcomes were handled, and which comparisons were planned.

19. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.
• Participant counts were reconstructed from posted percentages and arm sizes, so small rounding differences are possible.

The registry record leaves several trial-specific details unclear. The primary population description identifies participants who were randomized and received at least one dose in Part 1, and notes the exclusion of participants from one site for non-compliance. The description of the missing-data approach is cut off, so its full imputation method and assumptions are not reported in the registry record. This limits independent assessment of how missing Week 8 outcomes may have influenced the estimates.

The sponsor’s adjusted analysis is named as Cochran-Mantel-Haenszel, but the summary does not state the stratification variables or give enough detail to reproduce the adjusted result. The registry also does not report the multiplicity procedure for the numerous outcomes, the original sample-size justification, or interim-analysis arrangements. Consequently, secondary p-values cannot be placed within a fully documented family-wise testing plan based on this record alone.

The endpoint comparison is specifically at Week 8 during induction. Participants who did not achieve clinical response were described as entering an open-label extension and receiving upadacitinib, so later observations do not retain the same placebo-controlled treatment contrast. The primary result should not be generalized to that extension or to long-term outcomes.

Finally, the summarized safety figures give serious-event counts and listed terms by period, but do not include enough information to assess event relatedness, duration, or a comprehensive comparative safety profile. These limitations do not negate the posted results; they define what can and cannot be concluded from the registry record.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Study of the Efficacy and Safety of Upadacitinib (ABT-494) in Participants With Moderately to Severely Active Ulcerative Colitis — NCT03653026.
• Registry record last updated 2022-03-02; status last verified 2022-02.
• Results first posted 2021-11-26.
• 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

NCT03653026 is a completed Phase 3, randomized, quadruple-masked, parallel-group, placebo-controlled induction study in participants with moderately to severely active ulcerative colitis. The study compared upadacitinib 45 mg once daily with matching placebo during the eight-week induction period.

The primary endpoint was clinical remission per the Adapted Mayo score at Week 8. The registry reports remission in 33.5% of the upadacitinib group and 4.1% of the placebo group. The sponsor’s adjusted response-rate difference was 29.0 percentage points, with a 95% confidence interval from 23.2 to 34.7 and p < 0.001.

The record also includes secondary outcome results and safety summaries for Part 1 and Part 2. Important interpretive details—including the complete missing-data method, multiplicity strategy, and original sample-size justification—are not reported in the registry record.