A Study of LY3375880 in Adults With Moderate-to-Severe Atopic Dermatitis: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of LY3375880 in Adults With Moderate-to-Severe Atopic Dermatitis (NCT03831191), a randomized phase 2 study. The primary outcome is Percentage of Participants Achieving Validated Investigator's Global Assessment for AD (vIGA-AD) of 0 or 1 With a ≥2 Point Improvement.

2. Clinical Question

3. Trial Design

The registry describes NCT03831191 as a multicenter, randomized, double-blind, placebo-controlled, parallel-group phase 2 study of LY3375880 in adults with moderate-to-severe atopic dermatitis. The study's official title identifies it as the ADmIRe Study. The sponsor was Eli Lilly and Company. The registry lists a start date of 2019-02-12, a primary completion date of 2020-02-27, and a status of terminated. The reason for termination is not reported in the registry record.

During the induction period, participants received placebo or one of three LY3375880 doses: 50 mg, 150 mg, or 600 mg. Each was administered subcutaneously every 4 weeks. The posted primary endpoint is assessed at Week 16. The registry lists 136 participants enrolled, while the primary endpoint denominators are smaller and vary by group. The registry's definition of the primary endpoint population is all randomized participants who completed or discontinued the study before sponsor termination. Why those endpoint denominators differ from the overall enrollment is not reported in the registry record.

Random allocation and double masking are stated in the registry, but the allocation procedure, masking details, and any stratification factors are not reported in the registry record. The original target enrollment and the study's own sample-size justification are also not reported in the registry record. The registry lists eight outcome measures and 18 analyses. It does not report the full statistical analysis plan, the timing or rationale for termination, or whether interim analyses were conducted. These omissions limit how fully the planned study design can be reconstructed from 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: Responders are participants who achieved a 50% reduction in the Eczema Area and Severity Index score \[EASI-50\] response, regardless of whether rescue therapy had been initiated during induction period. Participants did not receive 300 mg because the trial was stopped early.

Only the first 6 of 15 groups are shown.

Induction Period (16 Weeks)

Reasons for not completing this period:

Maintenance Period (36 Weeks)

Reasons for not completing this period:

Post-Treatment Follow-Up Period(8 Weeks)

Analysis population for the primary outcome, as defined in the registry: Induction Period: All randomized participants who completed or discontinued the study prior to study termination by the sponsor.

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: All randomized participants.

6. Primary Endpoint

The registry lists the primary outcome as Percentage of Participants Achieving Validated Investigator's Global Assessment for AD (vIGA-AD) of 0 or 1 With a ≥2 Point Improvement, measured over Week 16.

The primary endpoint is the percentage of participants who achieved a vIGA-AD score of 0 or 1 with at least a 2-point improvement at Week 16. The vIGA-AD is a static, numeric assessment of overall atopic dermatitis severity. Its scale ranges from 0, clear, to 4, severe; higher scores indicate greater severity. The assessment considers erythema, papulation or induration, oozing or crusting, and lichenification.

A response defined by both a low final score and a minimum improvement represents two related aspects of change: the severity recorded at the assessment and improvement relative to baseline. A responder threshold gives a clinically interpretable way to describe the share of participants meeting a specified criterion. It also turns an individual assessment into a binary outcome: each participant either meets the response definition or does not.

The registry reports the endpoint as a percentage, with a 95% confidence interval for each group. It states that non-responder imputation was used for missing data. Under that approach, missing endpoint values are treated as non-response for the primary endpoint. The registry record does not report further details about how baseline values were handled or whether other missing-data sensitivity analyses were performed.

7. Results Posted in the Registry

• 50 mg LY3375880: 5.0% (1/20) versus 9.5% (2/21) on control; risk difference -4.5 percentage points (95% CI -20.3 to 11.3), risk ratio 0.53 (0.05–5.35); unadjusted two-proportion z-test p = 0.578.
• 150 mg LY3375880: 5.9% (1/17) versus 9.5% (2/21) on control; risk difference -3.6 percentage points (95% CI -20.5 to 13.2), risk ratio 0.62 (0.06–6.25); unadjusted two-proportion z-test p = 0.679.
• 600 mg LY3375880: 4.8% (1/21) versus 9.5% (2/21) on control; risk difference -4.8 percentage points (95% CI -20.3 to 10.7), risk ratio 0.50 (0.05–5.10); unadjusted two-proportion z-test p = 0.549.

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 Eczema Area and Severity Index 75 (EASI-75)

• 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:
  • 50 mg LY3375880: 15.0% (3/20) versus 19.0% (4/21) on control; risk difference -4.0 percentage points (95% CI -27.0 to 18.9), risk ratio 0.79 (0.20–3.09); unadjusted two-proportion z-test p = 0.731.
  • 150 mg LY3375880: 23.5% (4/17) versus 19.0% (4/21) on control; risk difference 4.5 percentage points (95% CI -21.8 to 30.7), risk ratio 1.24 (0.36–4.22); unadjusted two-proportion z-test p = 0.736.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between 50 mg LY3375880 and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between 150 mg LY3375880 and the control. The interval gives the range of effects that remain compatible with the data.

Percentage of Participants Achieving SCORing Atopic Dermatitis 75 (SCORAD-75)

• 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:
  • 50 mg LY3375880: 5.0% (1/20) versus 4.8% (1/21) on control; risk difference 0.2 percentage points (95% CI -13.0 to 13.4), risk ratio 1.05 (0.07–15.68); unadjusted two-proportion z-test p = 0.972.
  • 150 mg LY3375880: 0.0% (0/17) versus 4.8% (1/21) on control; risk difference -4.8 percentage points (95% CI -13.9 to 4.3); unadjusted two-proportion z-test p = 0.362.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between 50 mg LY3375880 and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between 150 mg LY3375880 and the control. The interval gives the range of effects that remain compatible with the data.

The posted secondary binary outcomes provide context alongside the primary response definition, but they do not remove the uncertainty in the primary comparison. The reported Week 16 measures include EASI-75, SCORAD-75, and achieving vIGA-AD of 0. The registry also lists change-from-baseline measures and other outcomes, including an assessment at Week 52 and a pharmacokinetic measure. The full results for all listed secondary outcomes are not reported in the registry record for this analysis.

For the secondary outcomes with posted Week 16 data, the observed response percentages vary by endpoint and dose group. The confidence intervals are broad, reflecting the small group denominators and limited event counts. The displayed comparisons do not show a consistent pattern that can be read as a clear, precise separation of the LY3375880 groups from placebo across these outcomes. Secondary measures can help describe whether the primary pattern is echoed by other definitions of response, but each outcome measures a different aspect of disease severity or response and should be interpreted on its own terms.

Unadjusted secondary comparisons carry less weight when many outcomes and comparisons are examined. In general, the more comparisons considered, the greater the chance that some results appear notable by chance alone. The registry lists multiple outcome measures and analyses, but whether and how multiplicity was controlled is not reported in the registry record. Accordingly, the secondary findings are best treated as exploratory descriptions rather than independent confirmation of the primary result. They do not establish a treatment effect or resolve the uncertainty evident in the primary endpoint estimates.

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

• Percentage of Participants Achieving Eczema Area and Severity Index 75 (EASI-75)
• Percentage of Participants Achieving SCORing Atopic Dermatitis 75 (SCORAD-75)
• Percentage of Participants Achieving vIGA-AD of 0
• Mean Change From Baseline in Eczema Area and Severity Index (EASI) Score
• Mean Change From Baseline in SCORAD
• Percentage of Participants Achieving vIGA-AD of 0 or 1 With a >=2-point Improvement at Week 52
• Pharmacokinetics (PK): Area Under the Concentration Versus Time Curve at a Steady State (AUCτ,ss) of LY3375880

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is binary at the participant level: a participant either achieved the specified vIGA-AD response at Week 16 or did not. The registry presents the percentage meeting that definition in each group. This percentage is a direct descriptive summary of the observed outcome in the stated analysis population. It is not a measure of the average change in severity, and it does not show how far each individual participant's score changed.

A confidence interval describes uncertainty around an estimated quantity. In general, a 95% confidence interval is produced by a procedure that, under its assumptions, would capture the quantity of interest in 95% of repeated applications. It is not a guarantee that a particular interval contains the true value. For the primary endpoint, the reported intervals accompany the group percentages. Their breadth indicates that the estimates are imprecise. The comparison intervals for the recalculated risk differences also cross zero, so the posted summaries are consistent with differences favoring either group.

The risk difference is the responder proportion in a LY3375880 group minus the responder proportion in the placebo group. It expresses the contrast on the percentage-point scale when proportions are written as percentages, or on the proportion scale when expressed as decimals. The recalculated risk differences are simple, unadjusted calculations from the posted summary statistics. They do not replace the sponsor's analysis. An unadjusted comparison does not account for any covariates or other model terms that may have been used in the registry's analysis.

The registry reports that its primary comparisons used logistic regression and gives odds ratios. An odds ratio compares the odds of response in one group with the odds in another. It is not the same quantity as a risk difference: the odds ratio is on a relative odds scale, while the risk difference is on an absolute scale. Logistic regression can incorporate model terms, but the registry record does not report the full model specification, any covariates, or the method used to construct the reported confidence intervals. The sponsor's odds-ratio results and the simple risk-difference recalculations therefore need not have identical numerical estimates or p-values.

The primary endpoint uses a response threshold, which can make interpretation straightforward but discards information about the size of score changes within each response category. Two participants with different degrees of improvement can be classified alike if both meet the response rule; a participant just short of the threshold is classified as a non-responder. This is a general feature of dichotomizing a continuous or ordinal measure, not a finding specific to this study. The registry also lists continuous change outcomes among its secondary measures, but their posted results are not included in the registry record for this analysis.

For the primary endpoint, the registry states that non-responder imputation (NRI) was used for missing data. In general, NRI assigns missing outcomes to the non-response category. This provides a defined rule for including participants without an observed endpoint value, but its impact depends on how many values are missing and why they are missing. Those details are not reported in the registry record. The endpoint population is described as all randomized participants who completed or discontinued the study before sponsor termination; this definition should be kept in mind when interpreting the denominators.

The recalculated p-values are unadjusted. A p-value measures how unusual data at least as incompatible with a specified null hypothesis would be under that hypothesis and the statistical model; it is not the probability that the null hypothesis is true. The registry's primary reported p-values are from its logistic-regression analyses. The registry record does not provide enough detail to determine how the sponsor handled multiple comparisons or whether the reported primary p-values were adjusted.

Small samples and sparse binary outcomes can make estimates unstable and intervals wide. General statistical methods for sparse outcomes may use exact or other specialized approaches, but the registry identifies logistic regression for the primary sponsor comparisons. The registry record does not describe any corrections for small-sample behavior or provide enough detail to evaluate the model's assumptions. The recalculated risk differences should therefore be read as descriptive, unadjusted comparisons based on the posted counts and percentages, alongside—not in place of—the reported sponsor analysis.

11. Confidence Intervals and Precision

The primary endpoint estimates are based on group denominators of 21 for placebo, 20 for 50 mg LY3375880, 17 for 150 mg LY3375880, and 21 for 600 mg LY3375880. The corresponding 95% confidence intervals are wide: 2.7% to 28.9% for placebo, 0.9% to 23.6% for 50 mg, 1.0% to 27.0% for 150 mg, and 0.8% to 22.7% for 600 mg. Each interval spans values well below and above the reported percentage, showing that the group response proportions are estimated with limited precision.

The recalculated 95% confidence intervals for the risk differences are also broad: -0.203 to 0.113 for 50 mg versus placebo, -0.205 to 0.132 for 150 mg versus placebo, and -0.203 to 0.107 for 600 mg versus placebo. Every interval includes zero and includes values on both sides of zero. Thus, the observed point estimates alone do not tightly constrain the possible difference between groups. This uncertainty matters when interpreting apparent differences in small groups, where a small change in the number of responders can shift the percentage noticeably.

In general, a confidence interval gives a range of values reasonably compatible with the data and the method used to calculate it; it is not a range containing most individual participant outcomes. A broad interval can reflect limited information, including a small denominator and few observed events. The registry does not report a separate precision target for the trial's original design. The intervals therefore describe uncertainty in the posted estimates, not proof that the groups are equivalent or that a specific difference is absent.

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 figures are participants per arm, not total enrollment or total participants across the compared groups. They are planning illustrations based on the observed effect, not the trial's own original design assumptions. The listed assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation; each scenario targets the stated fraction of the observed effect. The trial's own sample-size justification and original design assumptions are not reported in the registry record.

The 80% power column gives the illustrative number of participants per arm for a design with an 80% chance of detecting the assumed effect under the stated planning assumptions. The 90% power column gives the corresponding number for a 90% chance. These are design probabilities under the assumed effect and method, not guarantees about results. For example, planning for 75% of the observed primary risk difference for 50 mg versus placebo gives 983 participants per arm for 80% power and 1,316 for 90% power. Planning for 75% of the observed primary risk difference for 600 mg versus placebo gives 878 participants per arm for 80% power and 1,175 for 90% power.

Across the planning scenarios, a smaller effect to detect requires more participants than a larger effect under the same assumptions. The values based on the observed effects should not be treated as fixed recommendations: they depend on the observed effect estimates, and those estimates are uncertain. These calculations are not observed power. The registry record does not report that these planning calculations were used to select the study's enrollment.

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 = 0.095; treatment proportion p₁ = 0.050; average p̄ = 0.073.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.3670 = 0.7193.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.3656 = 0.3077.
4. n = (0.7193 + 0.3077)² / (-0.0452)² = 515.4, rounded up to 516 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 <- 0.0952380952380952                          # control proportion
p1 <- 0.05                          # 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(3831191)
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 = 516 (80% power), 690 (90% power)
 75% of observed effect: n per arm = 983 (80% power), 1,316 (90% power)
 50% of observed effect: n per arm = 2,358 (80% power), 3,156 (90% power)
power.prop.test: 515.4 
Simulated power at n = 516 per arm: 0.802 (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 reported primary analyses compare each LY3375880 dose group with placebo using logistic regression and report odds ratios. For 50 mg, the odds ratio was 0.55 with a 95% confidence interval from 0.04 to 6.81 and p = 0.638. For 150 mg, it was 0.69 with a 95% confidence interval from 0.05 to 8.76 and p = 0.773. For 600 mg, it was 0.58 with a 95% confidence interval from 0.05 to 7.26 and p = 0.671.

The recalculated comparisons use risk differences from posted summary statistics, without covariate adjustment. Their estimates were -0.045, -0.036, and -0.048 for 50 mg, 150 mg, and 600 mg, respectively. Their unadjusted p-values were 0.578, 0.679, and 0.549. Odds ratios and risk differences are different measures, and the sponsor's logistic-regression method may include model features not available in the registry record. The estimates and p-values are therefore not expected to match exactly.

The registry reports no discrepancy between the posted sponsor analysis and the recalculation. The sponsor's analysis is the primary reported analysis; the simple recalculations offer a complementary, unadjusted description rather than a replacement. Both sets of intervals are broad, so neither supports a precise estimate of the comparison.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists eight outcome measures and 18 analyses, including a primary endpoint, multiple secondary endpoints, and comparisons across dose groups. Considering many outcomes or comparisons raises a multiplicity issue: in general, the chance of obtaining at least one apparently unusual result can increase as the number of analyses increases. Statistical plans may address this using prespecified testing sequences, adjusted significance thresholds, or other procedures. Whether any such strategy was planned or applied is not reported in the registry record.

The sponsor's primary analyses are reported as logistic regression, while the posted secondary analyses include Fisher exact tests. The registry record does not provide the complete statistical analysis plan, the model specifications, or a full account of which analyses were prespecified. It also does not report whether the reported p-values account for the several dose-versus-placebo comparisons. The p-values should therefore be read as reported for their individual comparisons, without assuming a multiplicity adjustment.

Missing-data handling is reported specifically for the primary outcome: non-responder imputation was used. This means missing primary endpoint data were assigned to the non-response category under the stated method. The registry does not report the amount or timing of missing data, reasons for missingness, or sensitivity analyses under alternative assumptions. The missing-data procedures for the secondary measures are not reported in the registry record.

The primary endpoint population includes all randomized participants who completed or discontinued the study before sponsor termination. The record does not provide further detail about exclusions, analysis-set construction, or how the endpoint population relates to the larger enrollment. Details of interim analyses, stopping rules, and whether termination affected the analysis plan are also not reported in the registry record. These missing details limit the ability to assess how the reported analyses relate to the full prespecified plan.

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: cardiac arrest (20.0%), lung adenocarcinoma (3.0%), dermatitis atopic (3.0%), cellulitis (0.0%).

For the main induction groups in the adverse-event record, serious adverse events were reported in 0 of 33 participants at risk in the placebo group, 0 of 35 at risk with 50 mg LY3375880, 0 of 34 at risk with 150 mg, and 2 of 33 at risk with 600 mg. The registry lists no deaths in any of the reported groups. These counts describe the posted safety record; they do not by themselves establish whether an event was caused by study treatment.

Other adverse events were reported in 7 of 33 participants at risk with placebo, 9 of 35 with 50 mg, 12 of 34 with 150 mg, and 12 of 33 with 600 mg. The adverse-event listing also includes later response and follow-up groups with different denominators. Those groups should not be casually combined with the main induction groups because they represent distinct registry categories and may reflect different periods or treatment histories.

Safety counts need to be read with their denominators and the study's termination status in mind. The registry record does not provide enough detail to assess event duration, severity beyond the serious classification, relatedness to treatment, or comparative risk with precision. These figures are descriptive and are not a substitute for a complete safety analysis.

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

The primary result describes the percentage of participants in each group who met a specific clinician-rated response definition at Week 16. The observed percentages were 9.5% with placebo, 5.0% with 50 mg LY3375880, 5.9% with 150 mg, and 4.8% with 600 mg. The corresponding uncertainty intervals are wide, and the recalculated risk-difference intervals include zero. The registry's reported odds-ratio intervals are also broad.

These findings do not establish that LY3375880 is effective or ineffective for atopic dermatitis. They also do not establish that the groups are equivalent. A nonsignificant p-value is not proof of no difference, and an estimate in one direction is not proof that the underlying effect has that direction. With limited endpoint denominators and few participants meeting the response definition, the data allow a broad range of possible differences.

The result applies to the recorded endpoint, the specified assessment time, and the analysis population described in the registry. It does not directly answer every question about symptom change, longer-term response, or safety. The registry lists several other outcomes, but their full results are not reported in the registry record for this analysis. The study was terminated, and the reason is not reported in the registry record, which further limits interpretation of the study's intended scope and completion.

In general, randomized and masked comparisons can help reduce some sources of bias, but the design labels alone do not resolve uncertainty caused by small analysis groups, missing data, or incomplete reporting of analysis methods. The most defensible reading is descriptive: the posted data show low response percentages in all four groups, with imprecise between-group estimates. Stronger conclusions about the size, direction, or reliability of a difference are not supported by these results alone.

18. Why This Trial Matters Statistically

This trial is statistically informative as an example of the distinction between a point estimate and the precision of an estimate. The group percentages and comparison estimates are easy to state, but their confidence intervals are broad. A numerical difference between groups can look suggestive when considered alone; the interval shows how much uncertainty remains around that difference.

The binary response definition also illustrates a common trade-off. A responder percentage is straightforward to communicate, but a threshold reduces the information in the underlying severity assessments. The registry's other listed outcomes include different response definitions and changes in scores, which can offer complementary perspectives. However, the available secondary results do not remove the uncertainty in the primary endpoint, and the registry does not report how multiplicity was handled.

The study also demonstrates why sample-size planning depends on the effect size one wants to detect. The planning illustrations based on fractions of observed effects require more participants for smaller target effects. They are not the original design justification and are not observed power. Keeping those distinctions clear prevents a post hoc planning calculation from being mistaken for a feature of the trial's original plan.

Finally, reporting details shape statistical interpretation. The registry supplies the primary endpoint definition, the missing-data method, and the sponsor's logistic-regression results, but not the complete analysis plan, model specifications, or reason for termination. Recognizing what is and is not documented helps readers distinguish the posted evidence from assumptions that cannot be verified.

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 primary endpoint results are based on small group denominators: 21 for placebo, 20 for 50 mg LY3375880, 17 for 150 mg, and 21 for 600 mg. The resulting confidence intervals are broad, and the observed percentages depend on relatively few response events. This limits precision and makes the comparisons sensitive to small changes in event counts.

The registry reports an enrollment of 136, but the primary endpoint population is defined as randomized participants who completed or discontinued before sponsor termination, and the endpoint denominators are smaller. The reasons for the difference between enrollment and endpoint denominators are not reported in the registry record. The study status is terminated, but the reason, its effect on follow-up, and its relationship to the analysis population are not reported.

Non-responder imputation is specified for missing primary endpoint data, but the amount of missingness, reasons for missingness, and results under alternative assumptions are not reported. The complete statistical analysis plan, logistic-regression model details, and any multiplicity-control procedure are not reported. These gaps make it difficult to evaluate robustness and to determine how the reported analyses relate to prespecified methods.

The registry provides only a subset of the secondary outcome results considered in the registry entries. It lists other measures, including Week 52 and pharmacokinetic outcomes, without full results. The posted adverse-event counts are descriptive and do not provide a complete account of event relatedness, duration, or severity. These limitations are specific to the available registry record and should temper conclusions drawn from the observed efficacy and safety summaries.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Study of LY3375880 in Adults With Moderate-to-Severe Atopic Dermatitis — NCT03831191.
• Registry record last updated 2021-04-20; status last verified 2020-06.
• Results first posted 2021-04-20.
• 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

NCT03831191 was a multicenter, randomized, double-blind, placebo-controlled, parallel-group phase 2 study of LY3375880 in adults with moderate-to-severe atopic dermatitis. The induction period compared subcutaneous placebo with 50 mg, 150 mg, or 600 mg LY3375880 administered every 4 weeks. The registry lists enrollment of 136 and a terminated study status.

The primary outcome was the percentage achieving vIGA-AD 0 or 1 with at least a 2-point improvement at Week 16. Observed percentages ranged from 4.8% to 9.5% across the four groups, and the reported confidence intervals were broad. Sponsor-reported logistic-regression comparisons and unadjusted risk-difference recalculations both show substantial uncertainty. The registry specifies non-responder imputation for the primary endpoint, but does not report the full analysis plan, the reason for termination, or the study's original sample-size justification.