JAK1 Inhibitor With Medicated Topical Therapy in Adolescents With Atopic Dermatitis: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for JAK1 Inhibitor With Medicated Topical Therapy in Adolescents With Atopic Dermatitis (NCT03796676), a randomized phase 3 study. The primary outcome is Percentage of Participants Achieving Investigator's Global Assessment (IGA) Response of 'Clear' (0) or 'Almost Clear' (1) and ≥2 Points Improvement From Baseline at Week 12.

2. Clinical Question

The comparison is about the proportion of participants meeting a defined response threshold, not the average change in a continuous severity score. Participants who withdrew were counted as nonresponders for this primary endpoint. That rule means withdrawal did not remove a participant from the response definition as if the outcome were unknown; instead, the primary outcome classification treated withdrawal as failure to meet the endpoint.

The registry record describes a randomized, double-blind, placebo-controlled, parallel-group phase 3 study. Its stated scope is efficacy and safety of PF-04965842 used with medicated topical therapy in adolescents with moderate-to-severe atopic dermatitis. The main comparison is each active-dose group against placebo. The registry record does not report the details of the background topical regimens or a separate comparison between the active doses.

3. Trial Design

This was a phase 3 randomized, double-blind, placebo-controlled, multicenter, parallel-group study. The registry record lists quadruple masking and an enrollment of 287 participants. Study start was 2019-02-18, primary completion was 2020-04-08, and the status is completed. The condition is atopic dermatitis, and Pfizer is listed as sponsor.

Participants were adolescents 12 to <18 years of age with moderate-to-severe atopic dermatitis. The active groups received PF-04965842 100 mg QD or 200 mg QD, alongside background medicated topical therapy. The comparator was placebo alongside that background therapy. The trial description identifies randomized allocation and a parallel model; participants were assigned to a study group rather than being described as crossing between treatments. The registry lists double blinding in its summary and quadruple masking in its design details.

The primary endpoint was assessed from baseline to Week 12. It counted participants with an IGA rating of clear or almost clear and at least a 2-point improvement from baseline. The endpoint description states that withdrawn participants counted as nonresponders. The efficacy analysis population included participants randomly assigned to study intervention who took at least one dose, while the analyzed number refers to participants evaluable at the particular time point. Those definitions matter because the enrolled total and the numbers contributing to an endpoint need not be identical.

The registry record does not report the original sample-size justification, the intended statistical analysis plan, the trial’s allocation ratio, the details of the background topical therapy, or the masking procedures. It also does not state whether interim analyses were planned or conducted. The posted result should therefore be interpreted in light of the endpoint definition and analyses actually recorded, without assuming unreported protocol details.

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: A total of 287 adolescent participants were randomized to the study, and 285 adolescent participants received study treatment, including 96 participants in the placebo group, 95 participants in the abrocitinib 100 mg QD group, and 94 participants in the abrocitinib 200 mg QD group.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The analysis population included all participants randomly assigned to study intervention and who took at least 1 dose of study intervention. Number of Participants Analyzed refers to the number of participants evaluable for this time point.

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: Baseline analysis population included all participants randomly assigned to study intervention and who took at least 1 dose of study intervention. Participants were analyzed according to the intervention they actually received.

6. Primary Endpoint

The registry lists the primary outcome as Percentage of Participants Achieving Investigator's Global Assessment (IGA) Response of 'Clear' (0) or 'Almost Clear' (1) and ≥2 Points Improvement From Baseline at Week 12, measured over Baseline to Week 12.

The primary endpoint is the percentage of participants who had an IGA response of clear or almost clear at Week 12 and improved by at least 2 points from baseline. IGA is a global assessment of atopic dermatitis severity that reflects erythema, induration and scaling. The scale runs from 0 to 4: clear, almost clear, mild, moderate and severe, respectively. The endpoint therefore requires both a low Week 12 severity category and a specified improvement from the participant’s starting assessment.

A threshold endpoint is useful when the study question is framed around the proportion of participants who reach a predefined response state. It is different from asking whether the mean score changed, because participants are classified as responders or nonresponders according to the stated criteria. The resulting percentage is straightforward to interpret as the share meeting the endpoint in each group, but it does not describe the full distribution of score changes or the degree of improvement among people who do not cross the threshold.

The registry specifies that participants who withdrew were counted as nonresponders. This rule makes withdrawal part of the endpoint classification and can affect the observed response percentage. The registry’s analysis population also specifies that participants had to be randomly assigned and take at least one dose, with the analyzed number referring to those evaluable at the time point. These definitions identify who contributes to the reported percentage and how withdrawal is handled.

7. Results Posted in the Registry

• PF-04965842 100mg QD: 41.6% (37/89) versus 24.5% (23/94) on control; risk difference 17.1 percentage points (95% CI 3.7 to 30.5), risk ratio 1.70 (1.10–2.62); unadjusted two-proportion z-test p = 0.014.
• PF-04965842 200mg QD: 46.2% (43/93) versus 24.5% (23/94) on control; risk difference 21.8 percentage points (95% CI 8.4 to 35.1), risk ratio 1.89 (1.24–2.87); unadjusted two-proportion z-test p = 0.002.

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 Percentage BSA < 5% at Week 12

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • PF-04965842 100mg QD: 38.2% (34/89) versus 24.5% (23/94) on control; risk difference 13.7 percentage points (95% CI 0.4 to 27.1), risk ratio 1.56 (1.00–2.43); unadjusted two-proportion z-test p = 0.045.
  • PF-04965842 200mg QD: 36.6% (34/93) versus 24.5% (23/94) on control; risk difference 12.1 percentage points (95% CI -1.0 to 25.2), risk ratio 1.49 (0.96–2.33); unadjusted two-proportion z-test p = 0.073.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for PF-04965842 100mg QD than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between PF-04965842 200mg QD and the control. The interval gives the range of effects that remain compatible with the data.

Plasma PF-04965842 Concentration at Week 8

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • PF-04965842 200mg QD: mean 32.33 versus 7.88 on control; difference 24.45 (95% CI 4.46 to 44.44), standardized effect d = 0.44; z-test p = 0.017.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for PF-04965842 200mg QD than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Plasma PF-04965842 Concentration at Week 12

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • PF-04965842 200mg QD: mean 1,271 versus 486.60 on control; difference 784.40 (95% CI 488.59 to 1,080), standardized effect d = 1.06; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for PF-04965842 200mg QD than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The registry includes secondary clinical outcomes that examine additional response thresholds, symptom measures, and body-surface involvement, alongside plasma concentration outcomes. Collectively, these outcomes broaden the picture beyond the primary IGA response definition. They do not measure exactly the same construct: a symptom score, a body-surface threshold, and a global severity assessment each capture different aspects of the recorded outcomes.

For the posted Percentage BSA < 5% outcome at Week 12, the percentages were 24.5% for placebo, 38.2% for 100 mg QD, and 36.6% for 200 mg QD. The unadjusted comparison with placebo had p = 0.045 for 100 mg QD and p = 0.073 for 200 mg QD. These results offer a separate view of the outcome and do not redefine the primary IGA endpoint. The concentration outcomes are pharmacokinetic measures, rather than clinical response measures, and should not be read as direct substitutes for clinical outcomes.

Overall, the secondary outcomes address related but nonidentical questions. Similar direction across some measures can add context, while variation across measures can reflect different definitions, timing, analyzed populations, and uncertainty. The posted results do not establish that every secondary measure supports the same inference as the primary endpoint. In particular, secondary findings should be considered alongside their own estimates and intervals rather than summarized as one overall treatment effect.

Multiplicity is important because numerous secondary outcomes and analyses are listed. Unless a prespecified testing strategy controls the interpretation of multiple comparisons, isolated unadjusted p-values carry less confirmatory weight than the designated primary analysis. The registry record does not report the multiplicity procedure, if any. Secondary results are therefore most useful as contextual and exploratory evidence, with attention to effect estimates, precision, outcome definitions, and the possibility of chance findings across a collection of tests.

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

• Percentage of Participants Achieving ≥4 Points Improvement From Baseline in Peak Pruritis Numeric Rating Scale (PP-NRS) for Severity of Pruritus at Weeks 2, 4 and 12
• Change From Baseline in Pruritus and Symptoms Assessment for Atopic Dermatitis (PSAAD) at Week 12
• Percentage of Participants Achieving IGA Response of 'Clear' or 'Almost Clear' and ≥2 Points Improvement From Baseline at All Scheduled Time Points Except Week 12
• Percentage of Participants Achieving EASI Response ≥ 75% Improvement From Baseline at All Scheduled Time Points Except Week 12
• Percentage of Participants Achieving EASI Response ≥ 50% Improvement From Baseline
• Percentage of Participants Achieving EASI Response ≥ 90% Improvement From Baseline
• Percentage of Participants Achieving EASI Response =100% Improvement From Baseline
• Percent Change From Baseline in EASI Score

9. Statistical Interpretation

10. Statistical Methods Explained

What does a response percentage represent?

The primary endpoint is binary for each participant: the participant either meets the Week 12 IGA response definition or does not. The arm-level percentage is the proportion classified as responders among those analyzed at that time point. This presentation is appropriate for a threshold question, but it compresses a more detailed clinical score into two categories. It does not show how far a nonresponder’s score changed, nor does it distinguish participants who nearly met the threshold from those who did not improve.

For this endpoint, clear or almost clear is not sufficient on its own. The response definition also requires at least a 2-point improvement from baseline. The registry states that withdrawal was counted as nonresponse. That is an outcome-handling rule rather than a claim that every withdrawn participant had no improvement; it assigns a classification for the analysis.

What is a risk difference?

A risk difference subtracts the response proportion in the comparison group from the response proportion in the active group. It expresses the contrast on the same percentage-point scale as the underlying proportions. The recalculated risk differences are unadjusted comparisons from the posted summaries. They are useful for interpreting the size and direction of a simple group contrast, but they are not the same as the registry’s Cochran-Mantel-Haenszel estimates.

A confidence interval for a difference describes uncertainty in the estimated contrast under the statistical method used. It gives a range of values compatible with the data and method at the stated confidence level. It is not a range containing a specified share of individual participant outcomes, and it is not a guarantee that the true difference lies within the displayed endpoints.

How does the reported method differ?

The registry reports Cochran-Mantel-Haenszel estimates of difference for the two active-versus-placebo comparisons. This method is used for comparing categorical outcomes and can combine information across strata when such strata are part of the analysis. The registry record does not report the stratification details, so no particular stratification variables should be assumed. The reported p-values and confidence intervals belong to those sponsor-reported analyses.

The recalculated values use the posted arm-level response summaries to form simple, unadjusted contrasts. Their estimates and p-values may differ from the registry’s own results because the calculations are not the same analysis procedure. The registry record does not identify the recalculations as the prespecified trial analysis. There is no discrepancy flag for this result, and the sponsor’s reported analysis remains the trial’s own recorded result.

What does a p-value convey?

A p-value measures how unusual data at least as inconsistent with a specified null hypothesis would be under the assumptions of the test. It is not the probability that the null hypothesis is true, nor does it measure the size or clinical importance of an effect. Interpretation should consider the estimated difference and its interval as well as the test result. A small p-value does not eliminate uncertainty, and a p-value alone does not describe the range of plausible effect sizes.

The registry labels both primary comparisons as superiority analyses. The reported p-values are 0.015 for 100 mg QD versus placebo and 0.003 for 200 mg QD versus placebo. The unadjusted recalculated p-values are 0.014 and 0.002, respectively. These are separate outputs from different analyses and should not be mixed as if they came from a single model.

Why can the analysis population matter?

The primary endpoint description says the analysis population included all participants randomly assigned to study intervention who took at least one dose. It also clarifies that the number analyzed refers to the number evaluable at the time point. That means the count attached to a result is not necessarily the same as the overall enrollment. The endpoint description further says that withdrawal was counted as nonresponse. The registry record does not report additional details about missing-data methods, imputation beyond that rule, or sensitivity analyses.

General statistical principle: missing outcomes can affect an estimate when the likelihood of missingness is related to response. A prespecified and fully described approach helps readers assess that possibility. Here, the stated withdrawal rule is known, but other missing-data procedures are not reported in the registry record. Conclusions should not assume additional methods were used.

Why does multiplicity matter?

The registry lists many secondary outcomes and analyses in addition to the primary endpoint. When multiple hypotheses are tested, the chance of obtaining at least one apparently notable result by chance can rise if no multiplicity procedure is used. General statistical practice may address this through a prespecified testing hierarchy or another adjustment, but the registry record does not report whether such a plan was used for these results. Therefore, secondary p-values should not automatically be interpreted as confirmatory evidence with the same status as a designated primary analysis.

The primary endpoint is clearly identified, and its reported comparisons are described as superiority analyses. However, the registry record does not provide a complete statistical analysis plan, details of interim analyses, or the original sample-size justification. Those omissions limit the ability to reconstruct the full planned inferential framework from the registry alone.

11. Confidence Intervals and Precision

The primary endpoint confidence intervals quantify uncertainty in the response percentages and in the differences between groups. The placebo response was 24.5%, with a 95% confidence interval from 15.8 to 33.2. The 100 mg QD response was 41.6%, with a 95% confidence interval from 31.3 to 51.8. The 200 mg QD response was 46.2%, with a 95% confidence interval from 36.1 to 56.4. The intervals overlap, but overlap between separate arm intervals is not itself a formal test of the difference; the direct comparison intervals and reported tests address that contrast.

For the registry’s reported comparisons, the estimated difference for 100 mg QD versus placebo was 16.7 percentage points, with a 95% confidence interval from 3.5 to 29.9. For 200 mg QD versus placebo, the estimated difference was 20.6 percentage points, with a 95% confidence interval from 7.3 to 33.9. These intervals express uncertainty in the group contrasts according to the Cochran-Mantel-Haenszel analyses. The unadjusted recalculated intervals are different because they arise from simple comparisons of posted summary statistics.

In general, a narrower confidence interval indicates greater precision under the calculation’s assumptions, while a wider one indicates more uncertainty about the estimate. Precision depends on the amount of information and the variability in the data, among other features. The registry reports evaluable participant counts of 94 for placebo, 89 for 100 mg QD, and 93 for 200 mg QD for the primary endpoint. These counts help contextualize the estimates, but the intervals should be read using the analysis method attached to each result.

A confidence interval does not capture every source of uncertainty. It does not account automatically for uncertainty about study conduct, measurement choices, generalizability, or unreported analysis decisions. Nor should the interval be read as a prediction interval for a future study. The primary comparisons show estimated positive differences with intervals that remain above zero, while the interval widths still indicate that the exact magnitude is not determined precisely.

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 are planning calculations based on the observed effects, not the trial’s original design. For binary outcomes, the figures are participants per arm, not total enrollment. They assume a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. Each planning scenario uses the stated fraction of the observed effect as the effect to detect.

The 80% power column gives the estimated participants per arm for a hypothetical study to have an 80% chance of detecting the specified effect under those assumptions if that effect is present. The 90% power column gives the corresponding per-arm estimate for a 90% chance. These are planning targets, not observed results from the completed trial. The sample-size illustrations are not observed power.

For the primary endpoint’s 100 mg QD comparison, planning for 100% of the observed effect gives 118 participants per arm for 80% power and 157 for 90% power. Planning for 75% of the observed effect gives 203 and 271 participants per arm, respectively; planning for 50% gives 439 and 587. For the 200 mg QD comparison, the corresponding figures are 75 and 100 for 100% of the observed effect, 129 and 172 for 75%, and 277 and 370 for 50%.

The pattern reflects a general planning principle: smaller effects are harder to distinguish from random variation, so they require more participants to reach a given power target under the same assumptions. These calculations use the observed effect as their starting point, which can be an uncertain basis for planning. The registry record does not report the trial’s original sample-size justification or the assumptions used in its design.

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 = 0.245; treatment proportion p₁ = 0.416; average p̄ = 0.330.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6651 = 1.3035.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6540 = 0.5504.
4. n = (1.3035 + 0.5504)² / (0.1710)² = 117.5, rounded up to 118 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 <- 0.24468085106383                          # control proportion
p1 <- 0.415730337078652                          # 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(3796676)
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 = 118 (80% power), 157 (90% power)
 75% of observed effect: n per arm = 203 (80% power), 271 (90% power)
 50% of observed effect: n per arm = 439 (80% power), 587 (90% power)
power.prop.test: 117.5 
Simulated power at n = 118 per arm: 0.803 (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’s own primary analyses compare each active-dose group with placebo using the Cochran-Mantel-Haenszel method. The reported estimated difference is 16.7 percentage points for 100 mg QD versus placebo, with a 95% confidence interval from 3.5 to 29.9 and p = 0.015. For 200 mg QD versus placebo, the estimate is 20.6 percentage points, with a 95% confidence interval from 7.3 to 33.9 and p = 0.003.

The unadjusted recalculations use the posted arm-level response summaries and produce risk differences of 0.171 and 0.218, respectively. Their intervals and p-values are also calculated from those simple comparisons. Because the registry’s reported method and the unadjusted recalculations are not identical, small differences in estimates or p-values are expected and do not mean one set has been copied incorrectly. No discrepancy is flagged for this result.

The sponsor’s reported analysis is the trial’s recorded analysis and should be used as the primary reference when describing the trial results. The recalculations offer a transparent, unadjusted way to inspect the arm-level data; they do not replace the reported analysis or reveal unreported model details. The registry record does not describe the stratification details behind the Cochran-Mantel-Haenszel analysis.

15. Multiplicity, Missing Data and Analysis Populations

The primary endpoint is distinguished from the many secondary outcomes listed in the registry. That distinction matters because a study can test more than one hypothesis. General statistical principle: when many comparisons are examined, the probability of seeing at least one result that appears notable by chance can increase unless the analysis plan addresses multiplicity. A prespecified hierarchy, adjustment procedure, or other strategy can help define how evidence should be interpreted across outcomes.

The registry record does not report whether multiplicity was controlled, how testing across the two active-dose comparisons was handled, or whether secondary outcomes followed a prespecified hierarchy. It also does not report interim analyses or the complete statistical analysis plan. The absence of these details in the registry record does not establish that no such procedures existed; it means they cannot be described from the record.

The primary endpoint population included participants randomly assigned to study intervention who took at least one dose. The reported analyzed count refers to participants evaluable at that time point. Withdrawals were counted as nonresponders. These are the stated population and withdrawal rules. The registry record does not report further missing-data procedures, such as additional imputation methods or sensitivity analyses. The effect of unreported missing-data handling cannot be assessed from the posted information alone.

The original sample-size justification is also not reported in the registry record. The planning illustrations based on observed effects are separate from the trial’s own design rationale. They cannot establish what effect the trial was originally powered to detect or what assumptions informed its enrollment target. Together, the missing analysis-plan and design details mean that the registry results can be described as posted, but the full prespecified inferential framework cannot be reconstructed from the record.

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 2% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: anxiety (1.1%), angioedema (1.0%), dermatitis stopic (1.0%).

The safety summary reports serious adverse events among participants at risk in each group. In the placebo group, 2 of 96 participants were reported with a serious adverse event; in the 100 mg QD group, 0 of 95; and in the 200 mg QD group, 1 of 94. Deaths were reported as 0 in each group. These counts describe the registry’s posted serious adverse event figures over the study’s reporting period, which is not further specified in the registry record presented.

The record also lists serious adverse event terms with their posted percentages: Anxiety, 1.1%; Angioedema, 1.0%; and Dermatitis stopic, 1.0%. The last term is reproduced as recorded. These terms and percentages should be read as reported categories, not as a complete description of all possible safety outcomes.

Serious adverse event counts are not a complete safety assessment. The groups have different numbers at risk, and the counts alone do not establish a causal relationship or quantify all potential harms. The registry also lists other adverse event counts, but the serious-event summary does not substitute for a detailed comparison of event types, timing, severity, or relatedness. No medical conclusion should be drawn from these figures alone.

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

The primary result is a comparison of Week 12 response proportions using a specific IGA definition. The posted percentages were higher in both PF-04965842 groups than in the placebo group. The registry’s own analyses report positive estimated differences for both comparisons, with 95% confidence intervals above zero. The result therefore describes the trial’s recorded evidence for this endpoint and these treatment groups.

It does not mean that every participant in an active group achieved the response, or that every participant in the placebo group did not. It does not describe the full distribution of individual improvement, because the endpoint classifies participants according to whether they crossed a threshold. It also does not establish an effect for every outcome, time point, or population. The primary endpoint is limited to the Week 12 IGA response in adolescents meeting the study’s eligibility description.

The estimates are not exact values known without uncertainty. Their confidence intervals show a range of differences compatible with the data and analysis method, while the trial’s sample and endpoint rules constrain what can be inferred. The primary results also should not be extended automatically to longer-term effects or to settings not described in the registry record. The record does not provide enough detail to reconstruct the full original statistical plan, including the original sample-size justification or multiplicity procedures.

The safety counts are descriptive and should be interpreted separately from the efficacy endpoint. Serious adverse events and deaths are reported, but those counts alone do not settle the overall safety profile. The posted results are one source of evidence about this study; they do not constitute individualized guidance or a general conclusion about treatment for people outside the trial.

18. Why This Trial Matters Statistically

This trial illustrates why statistical interpretation should begin with the endpoint definition. “Response” is not a generic label: the primary outcome requires a Week 12 IGA category of clear or almost clear plus at least a 2-point improvement from baseline. Reading that definition makes clear what the percentages do and do not summarize.

It also shows the value of separating an arm-level percentage, an estimated group difference, and the uncertainty around that difference. The observed proportions describe the groups, the contrast quantifies their separation, and the confidence interval conveys precision under the analysis method. A p-value adds a test against a null hypothesis but does not replace the estimate or its interval.

The comparison between the registry’s Cochran-Mantel-Haenszel analyses and simple unadjusted recalculations is another practical lesson. Similar results can be expressed through different methods, and the method matters when reporting the trial’s official result. Recalculation can help readers understand the posted summaries, but it should not be presented as though it were the trial’s own prespecified analysis.

Finally, the registry identifies numerous secondary outcomes while leaving the multiplicity strategy and original design justification unreported. That makes it important to distinguish the primary finding from the broader set of secondary findings and to avoid treating every nominal result as equally confirmatory. The study is therefore a useful example of reading a clinical result in context: define the population and endpoint, identify the analysis, evaluate precision, and recognize what the registry does not document.

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 depends on a thresholded global assessment. That makes the response percentage easy to state, but it omits detail about the amount and pattern of change among participants who did not meet the response criteria. The registry record also does not provide the distribution of baseline IGA scores or further detail on how the assessment was administered beyond the scale description.

The analysis population is defined as participants randomly assigned who took at least one dose, while the analyzed count refers to those evaluable at the time point. Withdrawals counted as nonresponders. The registry record does not report additional missing-data procedures or sensitivity analyses, so the robustness of the result to alternative missing-data assumptions cannot be assessed from the record.

The registry reports Cochran-Mantel-Haenszel comparisons but does not give the stratification details. It also does not report a complete statistical analysis plan, multiplicity procedures, interim-analysis information, or the original sample-size justification. These omissions limit reconstruction of the planned inferential approach and how the multiple outcome comparisons were intended to be interpreted.

The result applies to the study’s adolescent population and the specified Week 12 endpoint; the registry record does not establish how the estimates would apply outside that context. The background medicated topical therapy is part of the study question, but its details are not reported in the registry record presented. The registry’s safety summary gives serious-event counts and selected terms, not a comprehensive account of event timing, severity, relatedness, or longer-term safety. These limitations do not erase the posted findings, but they narrow the claims that can be made from them.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: JAK1 Inhibitor With Medicated Topical Therapy in Adolescents With Atopic Dermatitis — NCT03796676.
• Registry record last updated 2022-04-13; status last verified 2022-04.
• Results first posted 2021-06-04.
• 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

NCT03796676 is a completed phase 3 randomized, double-blind, placebo-controlled, multicenter, parallel-group study sponsored by Pfizer. It enrolled 287 adolescents 12 to <18 years of age with moderate-to-severe atopic dermatitis. The active groups received PF-04965842 100 mg QD or 200 mg QD with background medicated topical therapy; the comparator was placebo with the same stated background-treatment context.

The primary endpoint was the percentage achieving clear or almost clear on the IGA and at least a 2-point improvement from baseline at Week 12. The registry reports Cochran-Mantel-Haenszel superiority comparisons against placebo, with positive estimated differences for both active-dose groups. Withdrawals counted as nonresponders. Serious adverse event counts were 2 of 96 for placebo, 0 of 95 for 100 mg QD, and 1 of 94 for 200 mg QD; deaths were 0 in all groups.

The registry record does not report the original sample-size justification, full statistical analysis plan, multiplicity procedure, or additional missing-data methods. Those gaps limit reconstruction of the planned analysis framework and should inform interpretation of secondary results and planning illustrations.