Epiduo Pediatric Acne Study: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Epiduo Pediatric Acne Study (NCT01138735), a randomized phase 4 study. The primary outcome is Success Rate.

2. Clinical Question

The registry describes a randomized, vehicle-controlled comparison in a parallel-group trial. The primary measure is a participant-level success rate, rather than a mean change in an acne count. In the posted results, 47.2% of participants in the adapalene/benzoyl peroxide arm met the success definition, compared with 15.4% in the vehicle arm. The sponsor's reported comparison used the Cochran-Mantel-Haenszel method and gave p < 0.001.

This question concerns the recorded clinical assessment at the stated follow-up and the defined threshold for success. It does not, by itself, describe every participant's change, the timing of improvement, or longer-term outcomes. The primary result is based on last observation carried forward, so it also reflects the registry's specified approach to using available observations for participants without a Week 12 value. Further details about missing observations and the analysis plan are not reported in the registry record.

3. Trial Design

NCT01138735 is listed as a completed Phase 4 study sponsored by Galderma R&D. The registry record describes randomized allocation, a parallel-group model, and triple masking. Its official title uses the term double-blind, while the design record lists triple masking; the specific roles included in that masking classification are not reported in the registry record. The study enrolled 285 participants.

The two arms were adapalene/benzoyl peroxide and topical gel vehicle. The active gel was described as adapalene and benzoyl peroxide 0.1%/2.5%, applied topically once daily for 12 weeks. The vehicle was also applied topically once daily for 12 weeks. The stated study purpose was to evaluate safety and efficacy in subjects 9 to 11 years of age with acne vulgaris. The registry lists Acne as the condition.

The recorded start was 2010-06 and primary completion was 2011-08. These dates describe the registry's study timeline; they do not establish when individual participants entered or completed treatment. The primary endpoint timeframe was baseline to Week 12, using last observation carried forward. The primary analysis population is identified as intent to treat.

The arm-level primary outcome records 142 participants for adapalene/benzoyl peroxide and 143 for vehicle. These are the denominators attached to the posted endpoint values, and they sum to the recorded enrollment of 285. The registry also lists four outcome measures and four analyses, while two secondary outcomes are named in the record. Further details about allocation procedures, recruitment, adherence, protocol deviations, or the reasons for the count of analyses are not reported in the registry record.

The randomized vehicle comparison supports a between-arm interpretation of the recorded outcomes, subject to the way the study was conducted and analyzed. It does not make the two topical gels identical except for their active ingredients; details of vehicle composition 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.

Overall Study

Analysis population for the primary outcome, as defined in the registry: Intent to treat (ITT) population, LOCF

5. Baseline Characteristics

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

6. Primary Endpoint

The registry lists the primary outcome as Success Rate, measured over Baseline to Week 12 (Last Observation Carried Forward [LOCF]).

The primary endpoint, Success Rate, is the percentage of participants rated Clear or Almost Clear on the Investigator's Global Assessment, with at least 2 grades of reduction from baseline. The reported timeframe is baseline to Week 12, with last observation carried forward. The analysis population is identified as intent to treat.

This endpoint turns a clinician's global assessment into a defined binary classification: each participant either meets the stated success criterion or does not. Reporting the percentage meeting that criterion makes the result interpretable as the proportion of participants reaching a specified level of improvement, rather than as an average score alone. The threshold also means that the endpoint does not distinguish between participants who meet it by a small margin and those with greater improvement.

As a general statistical point, a binary endpoint can be useful when a trial question is framed around the proportion of participants attaining a clinically defined state. Its meaning depends on the assessment scale, the threshold for success, and the timing of assessment. The registry gives the success definition and assessment time but does not report details such as assessor training, assessment reliability, or the handling of any borderline ratings.

The observed success rates were 47.2% with adapalene/benzoyl peroxide and 15.4% with vehicle. The registry's sponsor-reported comparison gives p < 0.001, but no confidence interval or comparative effect estimate for this primary outcome is reported in the registry record. The p-value indicates evidence against the equality tested by that analysis; it does not convey the precision of the size of the difference. Interpretation should therefore distinguish the observed arm rates from the uncertainty around their contrast, which is not quantified in the posted primary analysis.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary comparison concerns a binary outcome: whether each participant met the registry's definition of success. The endpoint is recorded as a percentage of participants, and the sponsor reports a Cochran-Mantel-Haenszel comparison between the two arms. The registry does not provide a numerical comparative estimate or confidence interval for this analysis, so the arm rates and reported p-value are the available primary results.

What does the reported p-value mean?

In general, a p-value measures how unusual data at least as incompatible with a specified null hypothesis would be if that hypothesis and the analysis assumptions were true. It is not a measure of the probability that the treatments are equivalent, and it does not quantify the size of a contrast. The primary p-value is reported as p < 0.001. Because the registry does not give the test statistic, a confidence interval, or a comparative estimate for the primary analysis, the precision of the primary contrast is not directly reported.

The Cochran-Mantel-Haenszel method is a general approach for comparing categorical outcomes while accounting for strata. The registry names this method for the primary outcome but does not report the stratification variables or other implementation details. It is therefore not possible to determine from the record which strata were used, or how any stratification affected the result. The registry also lists this method for the total-lesion secondary outcome, without a numerical sponsor estimate.

How should the secondary mean differences be read?

The two secondary outcomes are continuous changes from baseline. For each, the posted arm summaries include a mean and standard deviation. The recalculated contrast is the active-arm mean minus the vehicle-arm mean, as reported in the accompanying estimate. A negative mean difference corresponds to a more negative change in the active arm, given the outcome's change-from-baseline scale. These are simple unadjusted recalculations from the posted summary statistics; they are not adjusted estimates from the trial's own models.

For total lesion-count percentage change, the recalculated mean difference was -46.2, with a 95% confidence interval from -56.35 to -36.05. For inflammatory lesion-count change, it was -6.7, with a 95% confidence interval from -9.593 to -3.807. Each recalculated p-value was < 0.001. In general, a confidence interval conveys uncertainty about an estimated contrast: narrower intervals indicate greater precision under the method's assumptions, while wider intervals indicate less precision. The posted intervals do not incorporate adjustment for covariates or multiplicity.

The registry lists ANCOVA for the inflammatory-lesion sponsor analysis. ANCOVA is a general method for comparing group outcomes while accounting for one or more covariates. The registry record does not identify the covariates, the model specification, or the sponsor's estimated contrast and interval. The recalculated unadjusted mean difference should not be described as the ANCOVA result.

What do the planning illustrations represent?

The planning illustrations are based on observed secondary-outcome effects and are not the study's own design calculations. They assume a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation. Their purpose is to show how the number of participants per arm would change if a future study were planned to detect a stated fraction of an observed effect, under those assumptions. For total lesion-count percentage change, the illustrations give 15 participants per arm for 80% power and 19 for 90% power when planning for 100% of the observed effect; 26 and 34 when planning for 75%; and 57 and 76 when planning for 50%. For inflammatory lesion-count change, the corresponding values are 55 and 73; 97 and 130; and 218 and 291.

These are planning illustrations, not observed power and not evidence that the trial was adequately or inadequately designed. A planning calculation tied to an observed effect can be optimistic if that effect is larger than a plausible future effect. The registry does not report the trial's original sample-size justification, target effect, or design assumptions. General sample-size calculations depend on the outcome scale, variability, significance level, desired power, allocation, and the effect worth detecting; changing those assumptions changes the planning result.

Why can over-dispersion matter?

As a general concept, count outcomes can vary more than a simple model assumes. When the observed variability exceeds the model's assumed variability, standard errors may be understated and tests or intervals may appear more precise than warranted. This is called over-dispersion. The posted lesion-count outcomes are analyzed as mean changes with standard deviations for the recalculations, while the registry names ANCOVA for one sponsor analysis. It does not report a count-regression model or an over-dispersion assessment. No conclusion about whether over-dispersion was present or relevant can be drawn from the registry record.

More broadly, a statistical method is only one part of interpretation. The outcome definition, analysis population, missing-data handling, model assumptions, and multiplicity strategy all matter. The record identifies intent to treat and LOCF for the primary outcome, and names sponsor methods for the comparisons, but does not report many implementation details. Those absences limit the ability to reproduce or fully assess the sponsor's analyses from the registry record alone.

10. Analyses Reported by the Sponsor

The sponsor-reported primary analysis compares the success rates using the Cochran-Mantel-Haenszel method and reports p < 0.001. The registry lists no sponsor-reported effect estimate or confidence interval for that primary comparison. The arm-specific rates are 47.2% for adapalene/benzoyl peroxide and 15.4% for vehicle. These rates describe the observed groups; they are not a substitute for a comparative interval.

For the secondary outcomes, the registry names Cochran-Mantel-Haenszel for percent change in total lesion counts and ANCOVA for change in inflammatory lesion counts. Each listed sponsor analysis reports p < 0.001 but does not include a numerical sponsor estimate or interval. The recalculated mean differences and 95% confidence intervals are simple unadjusted calculations from the posted group summaries. They may not match estimates from an adjusted analysis because the model, covariates, and other implementation details are not reported.

Accordingly, the recalculated results are useful as transparent descriptions of the posted summaries, but they should be kept distinct from the sponsor's own analyses. The record does not flag a discrepancy between sponsor and recalculated results. The sponsor's analysis is the appropriate reference for the trial's reported inferential testing; the recalculations do not replace it or establish that the analyses used identical assumptions.

11. Multiplicity, Missing Data and Analysis Populations

The primary endpoint is identified as the success rate in the intent-to-treat population, assessed from baseline to Week 12 with last observation carried forward. The record also names two secondary outcomes and reports four outcome measures and four analyses. It does not describe a testing hierarchy, a family-wise error strategy, or any adjustment for multiple comparisons. Whether the sponsor controlled multiplicity, and how any such procedure was specified, is not reported in the registry record.

As a general statistical principle, testing several outcomes can increase the chance that at least one result appears statistically notable by chance if each is interpreted without adjustment. A prespecified hierarchy or adjustment can address that issue, but the record does not state whether either was used. The reported p-values for the secondary outcomes should therefore be treated as nominal unless the analysis plan establishes otherwise. The primary comparison remains the designated primary endpoint, but its interpretation still depends on the full planned analysis and its implementation.

The primary analysis specifies last observation carried forward. In general, LOCF substitutes a participant's last recorded value for a later missing assessment. This can retain participants with incomplete follow-up, but it assumes that the carried-forward value is an appropriate representation of the later outcome. That assumption may be unrealistic when outcomes change over time. The registry does not report the number or reasons for missing Week 12 observations, the timing of last observations, sensitivity analyses, or whether other missing-data methods were used.

The primary analysis population is identified as intent to treat. The registry does not provide a detailed definition of that population, such as how randomized participants with no post-baseline observations were handled. Nor does it report adherence, withdrawals, protocol deviations, or the analysis populations used for the secondary outcomes. These omissions make it difficult to assess the extent to which missing data or departures from assigned treatment could influence the results.

In general, missing-data assumptions and multiplicity decisions are part of the statistical plan, not details that can be inferred from p-values. The registry record does not report the original statistical analysis plan, interim analyses, sample-size justification, or sensitivity analyses. Their absence limits assessment of whether the posted analyses correspond to a prespecified strategy and how robust the results are to alternative reasonable assumptions.

12. Safety Results

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

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: right frontal arteriovenous malformation (0.7%).

The adverse-event summary records serious events for 0 of 142 participants in the adapalene/benzoyl peroxide arm and 1 of 143 participants in the topical gel vehicle arm. The registry lists RIGHT FRONTAL ARTERIOVENOUS MALFORMATION at 0.7% among the top serious terms. The record does not establish from this listing whether the term was considered related to study treatment. Death counts are not reported in the registry record.

Other adverse events are recorded for 37 of 142 participants in the adapalene/benzoyl peroxide arm and 18 of 143 participants in the vehicle arm. These counts are descriptive. The registry record does not provide event-level detail, severity grading, exposure-adjusted rates, or a comparative statistical analysis for these other events. The figures should not be interpreted as proof that any event was caused by an assigned gel or as a complete account of safety.

In general, serious adverse-event counts can be sparse, and the occurrence of an event in one arm does not alone establish a treatment-related difference. The absence of a recorded serious event in an arm does not establish absence of risk. These posted figures are limited to the registry's reported categories and denominators; additional safety details are not reported in the registry record.

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

The registry results describe a randomized comparison of adapalene/benzoyl peroxide gel and topical gel vehicle in children 9 to 11 years of age with acne vulgaris. For the primary outcome, the recorded success rate at Week 12 was 47.2% in the adapalene/benzoyl peroxide arm and 15.4% in the vehicle arm. The sponsor-reported comparison had p < 0.001. The secondary outcomes also show arm-level mean changes and reported comparisons, with recalculated unadjusted differences and confidence intervals available for those summaries.

These results do not establish the exact magnitude or precision of the primary contrast: a primary comparative estimate and confidence interval are not reported in the registry record. The p-value does not describe the probability that the null hypothesis is true, and statistical evidence alone does not tell how important the observed difference is. The endpoint's threshold also compresses a range of participant outcomes into success or not success.

The study addresses a defined age range, treatment schedule, comparator, and follow-up period. The record does not establish outcomes beyond the stated 12-week treatment period or for populations outside those described. It does not provide enough information to assess all aspects of missing-data handling, multiplicity, adherence, or model implementation. The safety counts are likewise descriptive and do not establish causation.

In general, randomized comparisons can help distinguish outcome patterns between assigned groups, but their interpretation depends on trial conduct, endpoint definitions, analysis choices, and uncertainty. The reported results should be read as evidence about the outcomes and time frame recorded, not as a guarantee of an individual outcome or as a complete account of long-term effects. Further protocol and analysis details are not reported in the registry record.

14. Why This Trial Matters Statistically

Statistically, this trial illustrates the distinction between a binary primary endpoint and continuous secondary measures. The primary endpoint asks what proportion reached a prespecified global-assessment threshold. The secondary endpoints describe mean changes in lesion counts. Each scale answers a different question: a success percentage focuses on crossing a defined clinical category, while a mean change summarizes the distribution of measured changes.

The registry also illustrates why statistical significance and precision are separate. The primary comparison reports p < 0.001, but no comparative confidence interval or effect estimate. The secondary outcomes have recalculated estimates and intervals from arm summaries, yet these are unadjusted and not the sponsor's model-based results. Readers therefore have different levels of detail for the primary and secondary findings.

The missing-data specification matters as well. LOCF is stated for the primary analysis, but the record does not report the amount or pattern of missing follow-up or sensitivity analyses. Likewise, multiple outcomes are reported without a documented multiplicity strategy. These limitations do not erase the recorded comparisons, but they constrain how completely their inferential reliability can be assessed.

The planning illustrations provide another statistical lesson: sample-size calculations depend on the effect chosen for planning and the assumptions used. Illustrations tied to observed effects are not the trial's design rationale and are not observed power. A future design would require an independently justified target effect and assumptions, none of which are reported as the original design basis in the registry record.

For clinical researchers and biostatisticians, the record is a useful example of how outcome definitions, analysis populations, missing-data methods, model descriptions, and multiplicity plans jointly shape interpretation. It also shows the value of separating the sponsor's stated analyses from transparent recalculations of posted summary statistics.

15. 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 registry record contains limited information for judging the conduct and analysis in detail. It identifies randomized allocation, a parallel model, and triple masking, while the official title describes the study as double-blind. The particular masking roles, randomization sequence method, allocation concealment, and details of implementation are not reported in the registry record. The relationship between the title wording and the masking classification is not explained.

The primary result lacks a reported comparative effect estimate and confidence interval. This leaves the precision of the primary contrast unclear, despite the small reported p-value. For the secondary outcomes, the effect estimates and intervals are recalculations from summary statistics, not the sponsor's adjusted estimates. Model covariates and implementation details are not reported.

Missing-data detail is also limited. LOCF is specified for the primary endpoint, but counts and reasons for missing observations, alternative analyses, and sensitivity analyses are not reported. The primary population is called intent to treat without a more detailed definition in the record. Adherence, withdrawals, and protocol deviations are not reported.

The registry lists multiple outcome measures and analyses but does not report a multiplicity strategy or hierarchy. It also does not report the trial's sample-size justification, interim analyses, or original statistical analysis plan. The planning illustrations based on observed secondary effects do not fill that gap and should not be interpreted as the rationale for the completed study.

Finally, the outcomes are reported for baseline to Week 12, and the record does not provide longer-term follow-up. The available safety summary has limited event detail and no reported deaths count. These limitations mean the registry can support a description of the posted results, but not a complete appraisal of every design, analysis, or safety question.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Epiduo Pediatric Acne Study — NCT01138735.
• Registry record last updated 2021-02-18; status last verified 2013-04.
• Results first posted 2013-03-26.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

19. Record Summary

NCT01138735 is a completed Phase 4 randomized, parallel-group study of adapalene/benzoyl peroxide gel versus topical gel vehicle in children 9 to 11 years of age with acne vulgaris. The registry records 285 participants, triple masking, and once-daily topical use for 12 weeks. The primary endpoint was the Week 12 percentage rated Clear or Almost Clear with at least 2 grades of improvement from baseline on the Investigator's Global Assessment, using an intent-to-treat population and LOCF.

Primary success rates were 47.2% and 15.4%, with sponsor-reported p < 0.001. Two posted secondary outcomes concern total and inflammatory lesion-count changes. Their sponsor-listed analyses also report p < 0.001; the recalculated estimates are unadjusted. Confidence intervals for the primary contrast, multiplicity procedures, detailed missing-data information, and the original sample-size justification are not reported in the registry record.