A Study to Evaluate the Safety and Efficacy of Ustekinumab Maintenance Therapy in Patients With Moderately to Severely Active Crohn's Disease (IM-UNITI): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study to Evaluate the Safety and Efficacy of Ustekinumab Maintenance Therapy in Patients With Moderately to Severely Active Crohn's Disease (IM-UNITI) (NCT01369355), a randomized phase 3 study. The primary outcome is Number of Participants With Clinical Remission at Week 44.

2. Clinical Question

This is a maintenance-treatment question conditioned on prior response to intravenous ustekinumab. It is not a comparison of initial induction treatment, and the primary result should not be read as an estimate for all people with Crohn's disease. The registry also describes other treatment paths for participants who did not respond to induction or who received placebo induction. Those paths are part of the broader study record, but the primary efficacy analysis population is specifically described as randomized ustekinumab induction responders after restart.

3. Trial Design

The registry describes IM-UNITI as a Phase 3, randomized, double-blind, placebo-controlled, parallel-group, multicenter study. Allocation was randomized, masking was recorded as quadruple, and the study status is completed. Enrollment was 1282. The recorded study dates are 2011-09-13 for study start and 2015-06-10 for primary completion.

The study evaluated subcutaneous ustekinumab maintenance after intravenous induction. In the primary efficacy population, participants who responded to intravenous ustekinumab induction were randomized to subcutaneous placebo, ustekinumab 90 mg q12w, or ustekinumab 90 mg q8w. The posted primary results contain 131 participants in the placebo group, 129 in the q12w group, and 128 in the q8w group. These group sizes describe the primary outcome analysis and should not be confused with total enrollment.

The registry also records pathways for participants who did not respond to intravenous ustekinumab induction, participants who received placebo induction, and participants who responded to placebo induction. Those pathways include further intravenous or subcutaneous treatment and placebo maintenance. They make the full trial structure broader than the three-arm primary efficacy comparison. The exact randomization procedures beyond the recorded randomized allocation, including any stratification factors, are not reported in the registry record.

The primary outcome was clinical remission at Week 44. The registry lists four secondary outcomes, including clinical response, remission in a specified subset, corticosteroid-free remission, and remission among participants refractory or intolerant to TNF antagonist therapy. The precise protocol details and the full statistical plan are not reported in the registry record.

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: Due to a stability issue with the batch of Intravenous (IV) drug (130 mg ustekinumab), in November 2011 sponsor temporarily suspended dosing in induction studies (CRD3001 and CRD3002) and this maintenance study (CRD3003). All 3 studies were restarted with a 90 milligram per milliliter (mg/mL) formulation for IV administration on 17 February 2012.

Only the first 6 of 9 groups are shown.

Week 0 - Week 44 (Maintenance)

Reasons for not completing this period:

Long-term Extension (Week 44-272)

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The primary efficacy analysis population in this study was randomized participants (i.e., participants who were in clinical response to ustekinumab induction dosing at Week 8 from one of the induction studies CRD3001 and CRD3002) after study restart.

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.

Only the first 6 of 7 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Number of Participants With Clinical Remission at Week 44, measured over Week 44.

The primary endpoint counts participants with clinical remission at Week 44. The registry defines remission as a Crohn's Disease Activity Index (CDAI) score below 150 points. CDAI combines information on eight Crohn's disease-related variables: extra-intestinal manifestations, abdominal mass, weight, hematocrit, total number of liquid stools, abdominal pain or cramping, use of antidiarrheal drugs or opiates, and general well-being. In general, CDAI scores range from 0 to approximately 600, with higher scores indicating greater disease activity. The registry notes that a decrease in CDAI over time indicates improvement in disease activity.

A remission endpoint gives a clinically interpretable count: whether each participant met a defined disease-activity threshold at the specified assessment. Comparing the proportions meeting that threshold can describe how remission frequency differed between the maintenance groups. It does not, by itself, describe the full distribution of CDAI scores, the size of score changes among individual participants, or the course of disease between assessments.

The endpoint is assessed at Week 44, so it captures status at that time point rather than a time-to-remission outcome. The primary analysis population is also important to its interpretation: it consists of randomized participants who had responded to ustekinumab induction, after study restart. The result therefore describes maintenance outcomes in that selected population and does not directly answer an induction-treatment question or a question about all enrolled participants.

7. Results Posted in the Registry

• Ustekinumab 90 Milligram (mg) SC Every 12 Weeks (q12w): 48.8% (63/129) versus 35.9% (47/131) on control; risk difference 13.0 percentage points (95% CI 1.0 to 24.9), risk ratio 1.36 (1.02–1.82); unadjusted two-proportion z-test p = 0.034.
• Ustekinumab 90 mg SC q8w: 53.1% (68/128) versus 35.9% (47/131) on control; risk difference 17.2 percentage points (95% CI 5.3 to 29.2), risk ratio 1.48 (1.12–1.96); unadjusted two-proportion z-test p = 0.005.

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.

Number of Participants With Clinical Response at Week 44

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Ustekinumab 90 Milligram (mg) SC Every 12 Weeks (q12w): 58.1% (75/129) versus 44.3% (58/131) on control; risk difference 13.9 percentage points (95% CI 1.8 to 25.9), risk ratio 1.31 (1.03–1.67); unadjusted two-proportion z-test p = 0.025.
  • Ustekinumab 90 mg SC q8w: 59.4% (76/128) versus 44.3% (58/131) on control; risk difference 15.1 percentage points (95% CI 3.1 to 27.1), risk ratio 1.34 (1.06–1.70); unadjusted two-proportion z-test p = 0.015.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Ustekinumab 90 Milligram (mg) SC Every 12 Weeks (q12w) 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 proportion with the event is statistically higher for Ustekinumab 90 mg SC q8w than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Number of Participants in Clinical Remission at Week 44 Among Participants in Clinical Remission to Ustekinumab at Week 0 of Maintenance Study

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Ustekinumab 90 Milligram (mg) SC Every 12 Weeks (q12w): 56.4% (44/78) versus 45.6% (36/79) on control; risk difference 10.8 percentage points (95% CI -4.7 to 26.4), risk ratio 1.24 (0.91–1.69); unadjusted two-proportion z-test p = 0.174.
  • Ustekinumab 90 mg SC q8w: 66.7% (52/78) versus 45.6% (36/79) on control; risk difference 21.1 percentage points (95% CI 5.9 to 36.3), risk ratio 1.46 (1.10–1.95); unadjusted two-proportion z-test p = 0.008.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Ustekinumab 90 Milligram (mg) SC Every 12 Weeks (q12w) and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Ustekinumab 90 mg SC q8w than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Number of Participants With Corticosteroid-free Remission at Week 44

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Ustekinumab 90 Milligram (mg) SC Every 12 Weeks (q12w): 42.6% (55/129) versus 29.8% (39/131) on control; risk difference 12.9 percentage points (95% CI 1.3 to 24.4), risk ratio 1.43 (1.03–1.99); unadjusted two-proportion z-test p = 0.031.
  • Ustekinumab 90 mg SC q8w: 46.9% (60/128) versus 29.8% (39/131) on control; risk difference 17.1 percentage points (95% CI 5.4 to 28.8), risk ratio 1.57 (1.14–2.17); unadjusted two-proportion z-test p = 0.005.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Ustekinumab 90 Milligram (mg) SC Every 12 Weeks (q12w) 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 proportion with the event is statistically higher for Ustekinumab 90 mg SC q8w than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Across the four posted secondary outcomes, the direction of the displayed point estimates generally favors the two ustekinumab maintenance groups over subcutaneous placebo, but the strength and precision of the comparisons vary by endpoint and analysis population. The registry-reported tests and the unadjusted recalculations are not interchangeable, and the endpoint-specific estimates should be interpreted in the context of who was included in each analysis. One secondary endpoint is restricted to participants in clinical remission to ustekinumab at the start of maintenance, unlike the broader primary efficacy population.

The secondary outcomes are related to the primary Week 44 remission question, but they do not measure exactly the same thing. Clinical response is distinct from clinical remission; corticosteroid-free remission adds a treatment-use condition; and the subset outcomes address more specific participant groups. A pattern across related endpoints can provide context for the primary result, but it does not make the endpoints equivalent or establish that the same effect applies to each population.

Unadjusted secondary results carry less confirmatory weight when the multiplicity plan is not reported. In general, considering several outcomes and comparisons creates more opportunities for apparently notable results to arise by chance. A p-value from one secondary comparison does not show that an overall family of tests has been controlled for that possibility. The registry record lists four secondary outcomes and ten analyses overall, but does not report an adjustment strategy or the prespecified hierarchy, if any. The secondary findings are therefore best treated as supportive or exploratory evidence rather than as independent definitive claims.

These outcomes also do not resolve limitations in the primary result’s scope. They are assessed at Week 44 and include populations defined in different ways. The registry does not report a full account of missing-data handling or the complete statistical plan for these endpoints. Taken together, the secondary results offer related observations, but conclusions should remain bounded by their analysis populations, endpoint definitions, and unreported multiplicity and missing-data procedures.

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

• Number of Participants With Clinical Response at Week 44
• Number of Participants in Clinical Remission at Week 44 Among Participants in Clinical Remission to Ustekinumab at Week 0 of Maintenance Study
• Number of Participants With Corticosteroid-free Remission at Week 44
• Number of Participants in Clinical Remission at Week 44 in the Subset of Participants Who Were Refractory or Intolerant to Tumor Necrosis Factor (TNF) Antagonist Therapy

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is binary: each participant either met the clinical-remission definition at Week 44 or did not. For a binary outcome, the observed group proportion is the number meeting the definition divided by the number assessed in that group. The registry posts the counts and group sizes, while the recalculated comparisons express the difference between group proportions as a risk difference. A risk difference is an absolute contrast, rather than a relative measure. For example, the q12w comparison is reported as a risk difference of 0.130, and the q8w comparison as 0.172. These are unadjusted recalculations from the posted summary counts.

What does an unadjusted risk difference describe?

An unadjusted risk difference compares the observed proportions in two groups without accounting for additional covariates or strata. It is directly tied to the proportions in the posted outcome data. An adjusted analysis can differ if it accounts for features of the design or analysis that an unadjusted calculation does not. The registry reports Cochran-Mantel-Haenszel chi-square tests for the primary comparisons, whereas the recalculated p-values use unadjusted comparisons. The exact strata used for the registry tests are not reported in the registry record.

Other common binary-outcome measures include the risk ratio and odds ratio. A risk ratio compares proportions by division; an odds ratio compares the odds of the outcome. These measures answer related but distinct questions, and their numerical values need not resemble a risk difference. The posted recalculated results include relative measures as well as risk differences, but the absolute differences are useful for describing how far apart the observed group proportions were. No single effect measure captures every aspect of the comparison.

Why can the reported p-values differ?

A p-value is a measure of how compatible the observed data are with a specified null hypothesis under a chosen statistical method. It is not the probability that the null hypothesis is true, and it is not a measure of effect size. The registry reports p=0.040 for the q12w comparison and p=0.005 for the q8w comparison using a Cochran-Mantel-Haenszel chi-square test. The unadjusted recalculated p-values are 0.034 and 0.005. Differences between the q12w values are consistent with using different methods; the q8w p-values are the same as posted to the reported precision. A test result should be interpreted alongside the estimated difference and its interval, not in isolation.

The Cochran-Mantel-Haenszel approach is generally used to assess an association while accounting for defined strata. In general, accounting for strata can produce a result that differs from a simple unadjusted comparison. The registry record does not specify the strata or provide a full description of the statistical model in the posted result. Therefore, the reported test should be described by its recorded name rather than by assuming the details of its implementation.

What does an interval add?

A confidence interval describes uncertainty around an estimate under the assumptions of the method used. A narrower interval indicates greater precision than a wider interval, while an interval spanning the null value for a difference indicates that values on both sides of no difference remain compatible with that interval. The recalculated primary risk-difference intervals are 0.011 to 0.249 for q12w and 0.053 to 0.292 for q8w. The confidence level and exact interval procedure are not reported in the registry record, so the recalculated intervals should be interpreted as the intervals posted with these unadjusted calculations, not as a description of the sponsor's interval procedure.

For binary outcomes, the amount of information depends partly on the number of participants and how many meet the endpoint. When event counts are limited, estimates can be less precise. The primary groups have similar sizes, but the interval still expresses uncertainty in the comparison. General statistical theory also distinguishes statistical evidence from practical importance: a small p-value does not establish that an effect is large or important, and an interval does not establish causation beyond the design and population studied.

The analysis population and endpoint definition are essential parts of the method. Here, the primary efficacy population is defined as randomized participants who responded to ustekinumab induction at Week 8 in the earlier induction studies, after study restart. The endpoint is a Week 44 CDAI threshold. A different population, endpoint definition, or treatment of participants without observed Week 44 outcomes could yield different estimates. The registry record does not report the missing-data method, so the handling of participants without an observed endpoint cannot be reconstructed from the posted summary.

The number of analyses listed in the registry and the existence of several outcomes raise an important interpretation issue: statistical tests are often considered within a broader analysis plan, including any adjustment for multiple comparisons. The registry record does not report whether or how multiplicity was addressed. General statistical reasoning is that making many comparisons can increase the chance of finding at least one apparently unusual result by chance if each is interpreted separately. Without the prespecified plan and adjustment details, the posted p-values should not be treated as if they necessarily account for all outcome and treatment comparisons.

11. Confidence Intervals and Precision

The recalculated primary risk-difference estimate for ustekinumab 90 mg q12w compared with subcutaneous placebo is 0.130, with an interval from 0.011 to 0.249. The estimate for ustekinumab 90 mg q8w compared with placebo is 0.172, with an interval from 0.053 to 0.292. These intervals show that the posted data are compatible with a range of possible absolute differences, rather than identifying one exact value. The q12w interval approaches zero at its lower end, while the q8w interval is also broad relative to its point estimate.

In general, a confidence interval expresses uncertainty about an estimate using a particular method and assumptions. A narrower interval indicates greater precision; a wider one indicates less precision. The interval limits are not a range containing a fixed proportion of individual participants, and they do not give the probability that the true effect lies within the displayed limits. The confidence level and calculation procedure for these recalculated intervals are not reported in the registry record.

The group sizes for the primary comparison were 131 for placebo, 129 for q12w, and 128 for q8w. Those counts, together with the number meeting the remission definition, inform the precision of the estimated proportions and differences. Similar group sizes do not remove uncertainty, especially when the outcome is a participant count and the observed event counts are finite.

Precision should also be considered in relation to study scope. These estimates apply to the registry-defined randomized induction responders after restart and to the Week 44 remission threshold. The posted intervals do not capture uncertainty about generalizing to other populations, treatment schedules, endpoints, or follow-up periods. They also do not address uncertainty arising from missing outcomes or multiplicity, for which relevant analysis details are not reported in the registry record.

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 give participants per arm for the binary-outcome comparisons. They are not total sample sizes for the whole study and are not total event counts. The registry’s sample-size assumptions for these illustrations are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. Each illustration uses a stated fraction of the observed effect as the effect to detect. The original trial’s own design assumptions and sample-size justification are not reported in the registry record.

The 80% power column gives the illustrative number of participants per arm for a design intended to have an 80% chance of detecting the specified effect under the stated assumptions, if that effect were the true one and the assumptions held. The 90% power column gives the corresponding per-arm planning number for a higher chance of detection. These are planning concepts, not guarantees about a particular study’s outcome. They depend on the effect size and assumptions used.

The rows using 100%, 75%, and 50% of the observed effect show how the illustrative sample size changes when the target effect is made smaller. Smaller effects require more participants to distinguish them from no difference with the same assumed significance level and power. For example, the primary q12w comparison’s illustration for 100% of the observed effect gives 228 participants per arm for 80% power and 304 for 90% power; for 50% of the observed effect, it gives 890 and 1,190, respectively. These values are planning illustrations based on the observed effect, not the trial’s documented design plan.

The same logic applies to the q8w comparison and the secondary comparisons: the effect fraction and comparison determine the displayed planning values. The illustrations should not be read as evidence that the completed trial had a specified probability of detecting its result. This is not observed power. They are prospective-style calculations that show the sample sizes implied by chosen assumptions and effect targets.

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 = 0.359; treatment proportion p₁ = 0.488; average p̄ = 0.424.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6988 = 1.3696.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6928 = 0.5830.
4. n = (1.3696 + 0.5830)² / (0.1296)² = 227.0, rounded up to 228 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 <- 0.358778625954198                          # control proportion
p1 <- 0.488372093023256                          # 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(1369355)
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 = 228 (80% power), 304 (90% power)
 75% of observed effect: n per arm = 400 (80% power), 535 (90% power)
 50% of observed effect: n per arm = 890 (80% power), 1,190 (90% power)
power.prop.test: 227 
Simulated power at n = 228 per arm: 0.806 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Analyses Reported by the Sponsor

The registry reports the sponsor’s primary comparisons using a Cochran-Mantel-Haenszel chi-square test. For q12w compared with subcutaneous placebo, the reported p-value is 0.040; for q8w compared with placebo, it is 0.005. The unadjusted recalculation from the posted group counts gives p=0.034 for q12w and p=0.005 for q8w.

The distinction matters because the two sets of p-values do not necessarily use the same analysis. The registry names the sponsor’s method, while the recalculation is a simple unadjusted comparison. The q12w values differ, whereas the q8w values match at the precision posted. A difference in p-values does not by itself imply that the underlying event counts are inconsistent; it can arise from applying different statistical procedures.

The recalculated risk differences are also unadjusted and provide a direct contrast between the observed proportions. The sponsor’s reported p-values should be understood as the registry’s own analysis results, and the recalculations as supplementary summaries of the posted counts. The registry record does not provide enough detail to reconstruct every feature of the sponsor’s analysis, including the exact strata used for the named test or a full statistical analysis plan. The posted data therefore support describing both results and their methods, without treating the recalculation as a replacement for the registry-reported analysis.

15. Multiplicity, Missing Data and Analysis Populations

The registry record does not report whether the primary and secondary comparisons were subject to a prespecified multiplicity procedure. It lists five outcome measures and ten analyses, and it presents several treatment comparisons and secondary endpoints. In general, multiplicity refers to the problem of interpreting several statistical tests together: when many tests are considered, the chance of at least one apparently unusual result can increase if each is judged separately. Adjustments or testing hierarchies can address that issue, but the registry does not describe whether such methods were used. The posted p-values should therefore be read as reported for their respective comparisons, not assumed to control an overall error rate across all outcomes and tests.

The primary efficacy population is described as randomized participants who were in clinical response to ustekinumab induction dosing at Week 8 in induction studies CRD3001 and CRD3002, after study restart. This population definition is useful, but it does not fully describe every analysis decision. The registry record does not report the missing-data method for participants without an observed Week 44 endpoint. It also does not report whether participants were analyzed according to a particular intention-to-treat convention, how intercurrent events were handled, or whether additional exclusions were applied beyond the stated population definition.

Missing outcome data can affect estimates if the participants with observed results differ systematically from those without them. General statistical approaches may use assumptions, imputation, or other rules, and different choices can lead to different results. Without a reported method and supporting details, it is not possible to assess how sensitive the posted results are to missingness or to recreate the sponsor’s handling of unavailable outcomes.

The full statistical analysis plan, sample-size justification, interim-analysis plan, and details of any prespecified testing sequence are not reported in the registry record. Their absence from the registry record means those aspects cannot be evaluated from the posted information. These limitations do not change the counts or reported tests, but they constrain interpretation of how confirmatory the comparisons were intended to be and how completely the analyses can be reproduced.

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: crohn's disease (10.3%), abdominal hernia (3.4%), gastrointestinal haemorrhage (3.4%), non-cardiac chest pain (3.4%), cholecystitis (3.4%).

The safety record lists serious adverse events and other adverse events by groups defined by induction history and maintenance or long-term-extension treatment. These are posted counts for those safety groups, not the primary efficacy population sizes. For the groups labeled ustekinumab induction responders receiving placebo maintenance, q12w maintenance, and q8w maintenance, serious adverse events affected 22 of 133, 16 of 132, and 13 of 131 participants, respectively. The corresponding posted counts of participants with other adverse events are 94 of 133, 88 of 132, and 84 of 131. No deaths are listed in those three groups.

The safety groups and denominators should be read according to their registry labels and not assumed to match the primary efficacy analysis sets exactly. The record also lists long-term-extension groups, where deaths are recorded for some groups; these should not be combined with the maintenance-period groups without a defined analysis. The most frequently listed serious-event terms include Crohn's Disease at 10.3%, with several other terms at 3.4% each. The registry record does not provide enough context in these summary fields to establish causal attribution or compare safety formally between all treatment groups. Counts alone do not show exposure-adjusted rates or the timing and duration of events.

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

The results describe Week 44 clinical remission in randomized participants who responded to intravenous ustekinumab induction, after study restart. In that population, the posted remission counts were 47 of 131 with subcutaneous placebo, 63 of 129 with ustekinumab 90 mg q12w, and 68 of 128 with ustekinumab 90 mg q8w. The recalculated absolute differences versus placebo were 0.130 for q12w and 0.172 for q8w. These estimates summarize the observed endpoint comparison and are accompanied by uncertainty intervals.

The results do not establish that every participant benefits, or that the same findings apply to people who did not respond to induction, people outside the analysis population, or other treatment schedules. The study record includes additional treatment paths, but the primary efficacy population is specifically defined by prior response. The result also does not answer whether participants remained in remission throughout the maintenance period. It records whether the endpoint threshold was met at Week 44.

A difference in remission proportions is not the same as a difference in overall disease experience. The endpoint uses a CDAI threshold, which summarizes several disease-related variables, but the posted primary result does not provide the full score distribution or individual trajectories. It does not describe all potential benefits, harms, or outcomes beyond those reported in the registry. The safety summaries are also group-based counts and do not, on their own, establish that an event was caused by a treatment.

The p-values do not convey the size of the difference or the probability that a treatment effect is true. The registry’s named test and the unadjusted recalculation use different methods for the q12w comparison. Secondary results provide related observations but carry limitations from their distinct populations and the unreported multiplicity approach. Missing-data procedures are also not reported in the registry record.

Accordingly, the posted results support a description of the observed Week 44 comparisons in this study population. They do not, by themselves, settle questions about generalizability, long-term outcomes, or the interpretation of all outcomes and safety events. Those questions require the broader trial report and analysis details, which are not reported in the registry record.

18. Why This Trial Matters Statistically

This trial illustrates why a clinical result needs to be read together with its population, endpoint, effect estimate, uncertainty, and analysis method. The primary efficacy population was not every participant enrolled: it was randomized participants who had responded to intravenous ustekinumab induction, after study restart. That distinction defines the scope of the comparison and helps prevent a maintenance result in responders from being misread as an induction result or as applying to all people with Crohn's disease.

The endpoint is a participant count based on a specified CDAI threshold at Week 44. A binary endpoint is easy to interpret at the group level, but it compresses a broader measure into whether a participant crossed a particular threshold at a particular assessment. The risk differences make the absolute contrasts visible, while intervals show that the estimates remain uncertain. Looking only at a p-value would lose information about both magnitude and precision.

The registry also places sponsor-reported tests alongside recalculated unadjusted comparisons. The difference in methods matters: a p-value belongs to the test that generated it, and results should not be treated as identical simply because they concern the same group counts. The comparison demonstrates the value of identifying the analysis method before interpreting differences in statistical output.

Finally, the multiple secondary outcomes, analysis populations, and safety groups underline how trial records can contain several distinct evidence streams. Multiplicity, missing-data procedures, and the original sample-size rationale are not reported in the registry record. Recognizing these gaps is part of a statistically responsible reading, not a reason to substitute unsupported assumptions. The trial matters statistically because it shows how design, selection, endpoint definition, estimation, and reporting detail jointly shape what a result can say.

19. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The primary efficacy analysis is restricted to participants who responded to intravenous ustekinumab induction and were randomized after study restart. This responder-based population differs from the broader enrollment of 1282 and limits generalization to participants who did not meet that response condition. The primary comparison addresses two ustekinumab maintenance schedules against subcutaneous placebo in that defined population, not every treatment pathway represented in the study.

The endpoint is remission at Week 44, defined by a CDAI score below 150 points. A single assessment does not show the timing or persistence of remission, and the posted participant counts do not show the full distribution of CDAI scores or changes from baseline. The record does not provide additional detail about missing Week 44 outcomes or their handling, which limits assessment of possible bias related to unavailable endpoint data.

The registry reports a Cochran-Mantel-Haenszel chi-square test, but does not report the exact strata or the complete analysis plan. The unadjusted recalculations use a different approach, so their p-values should not be conflated with the sponsor-reported tests. The multiplicity procedure, if any, is not reported, despite multiple outcomes and analyses. This limits how confidently the individual p-values can be interpreted as confirmatory across the full set of comparisons.

The safety summaries contain groups with different labels and denominators, including long-term-extension groups. They should not be treated as if all participants had the same exposure duration or belonged to the primary efficacy population. The posted summaries do not provide enough detail for exposure-adjusted comparisons or causal attribution of events. The original trial sample-size justification, interim analyses, and further details of the statistical plan are not reported in the registry record.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Study to Evaluate the Safety and Efficacy of Ustekinumab Maintenance Therapy in Patients With Moderately to Severely Active Crohn's Disease (IM-UNITI) — NCT01369355.
• Registry record last updated 2025-02-04; status last verified 2025-01.
• Results first posted 2017-02-23.
• 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

IM-UNITI is recorded as a completed Phase 3, randomized, double-blind, placebo-controlled, parallel-group, multicenter study with enrollment of 1282. Its primary outcome was clinical remission at Week 44, defined as a CDAI score below 150 points. The primary efficacy analysis population comprised randomized participants who responded to intravenous ustekinumab induction, after study restart.

In that analysis, remission was recorded for 47 of 131 participants receiving subcutaneous placebo, 63 of 129 receiving ustekinumab 90 mg q12w, and 68 of 128 receiving ustekinumab 90 mg q8w. The registry reports Cochran-Mantel-Haenszel p-values of 0.040 and 0.005 for the respective comparisons with placebo. Unadjusted recalculations provide risk differences of 0.130 and 0.172, with their corresponding intervals.

The record also includes four secondary outcomes and safety counts across maintenance and long-term-extension groups. Multiplicity handling, missing-data methods, the original sample-size justification, and the full statistical plan are not reported in the registry record.