A Randomized, Open-label, Comparative Study to Evaluate an Intermittent Dosing Regimen of Fluticasone Propionate 0.05% Cream (Twice Per Week) in Reducing the Risk of Relapse When Added to Regular Daily Moisturization Using PHYSIOGEL Lotion in Paediatric Subjects With Stabilized Atopic Dermatitis: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Randomized, Open-label, Comparative Study to Evaluate an Intermittent Dosing Regimen of Fluticasone Propionate 0.05% Cream (Twice Per Week) in Reducing the Risk of Relapse When Added to Regular Daily Moisturization Using PHYSIOGEL Lotion in Paediatric Subjects With Stabilized Atopic Dermatitis (NCT01915914), a randomized phase 4 study. The primary outcome is Time to the First Relapse of AD During the Maintenance Phase.

2. Clinical Question

The registry describes an acute phase before maintenance: eligible subjects received fluticasone propionate 0.05% cream twice daily for up to 4 weeks, with assessments every 2 weeks or until treatment success. Randomization occurred for maintenance, after which participants received one of the two regimens for up to 20 weeks. The posted primary result therefore addresses relapse after stabilization and randomization, not initial response during acute treatment.

The comparison is clinically specific to the study’s paediatric population, the stabilization process, the stated emollient regimen and the maintenance period. The record does not report a broader population or a comparison with other treatment regimens. It also does not report the original design assumptions or a sample-size justification.

3. Trial Design

NCT01915914 is recorded as a completed Phase 4 randomized, open-label, parallel-group study sponsored by GlaxoSmithKline. The registry lists enrollment as 107. Its design included screening, acute treatment, maintenance and follow-up. The primary comparison was between emollient plus intermittent fluticasone propionate cream and emollient alone during maintenance.

In the acute phase, eligible subjects received fluticasone propionate 0.05% cream twice daily for up to 4 weeks. Efficacy and safety assessments were planned every 2 weeks during acute treatment, up to 4 weeks or until treatment success, whichever came first. The record defines treatment success as a PSGA score of 1 or less and improvement of at least 2 compared with baseline.

Participants reaching maintenance were randomized in a 1:1 allocation. Group A received emollient twice daily plus fluticasone propionate cream once daily twice a week; Group B received emollient twice daily. Maintenance lasted up to 20 weeks, with assessments every 4 weeks or until relapse. Participants without relapse during maintenance could enter follow-up, using emollient twice daily for up to another 12 weeks. The registry describes the overall study duration as up to 37 weeks.

The primary endpoint was time from the start of maintenance fluticasone treatment to the first atopic dermatitis relapse, assessed through Week 20. The primary analysis population is described as participants randomized into maintenance, with only those available at the specified time point analyzed. The registry lists 54 participants in the combination arm and 53 in the emollient arm for the primary endpoint. The record does not report the original sample-size justification, detailed allocation-concealment procedures, or a more complete statistical analysis plan.

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: Eligible participants (par.) received twice daily (BID) fluticasone propionate 0.05% (FP) cream up to 4 weeks (wk) (in Acute Phase), then par. received (1:1) either emollient BID plus FP cream once daily (OD) twice a wk, or emollient BID up to 20 wk (Maintenance Phase). Par. who did not relapse received emollient BID up to 12 wk (Follow-up Phase).

Acute Phase

Reasons for not completing this period:

Maintenance Phase

Reasons for not completing this period:

Follow-up Phase

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: ITT Population: all participants who were randomized into the Maintenance Phase. Only participants available at the specified time point were analyzed.

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 Time to the First Relapse of AD During the Maintenance Phase, measured over From the start of treatment up to Week 20 during the Maintenance Phase.

The primary endpoint is time to the first atopic dermatitis relapse during maintenance, measured in days from the start of treatment in that phase until relapse. The registry defines relapse as a PSGA exacerbation score of at least 2 compared with the participant’s treatment-success score in the acute phase. The PSGA scale ranges from 0, clear, to 5, very severe; treatment success is defined as a score of 1 or less together with improvement of at least 2 compared with baseline.

Time-to-event endpoints preserve information about when an event occurs, rather than recording only whether it occurred by a single assessment point. They can also account for participants who have not experienced the event by the end of observation, provided the analysis handles those observations appropriately. The registry reports a log-rank analysis and a hazard ratio for the primary contrast.

This endpoint is suited to the stated question because the comparison concerns delaying or preventing a first relapse over maintenance follow-up. It distinguishes the first relapse from recurrent events and focuses on the period after participants have achieved the study’s acute-phase treatment-success criteria. The median time is a summary of the event-time distribution, while the hazard ratio summarizes the relative event rates over the analyzed period under the model’s interpretation.

The record reports a median of 142 days for the emollient arm, with a 95% confidence interval of 50 to 150 days. A median and confidence interval are not reported for the combination arm. The registry record does not explain why that median is unavailable, so the missing value should not be interpreted as a particular event pattern without additional information.

7. Results Posted in the Registry

• Emollient Plus FP 0.05% Cream vs Emollient: hazard ratio 13.50 (95% CI 4.11–44.33), reported p < 0.001 (as reported in the registry).

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.

Median Time to the First Relapse of AD During the Maintenance Phase and Follow-up Phase

• Endpoint type: Time to event: compared as a hazard ratio.
• Test: Hazard ratio as reported by the sponsor (Log Rank); it was not recalculated.
• Result:
  • Emollient Plus FP 0.05% Cream vs Emollient: hazard ratio 4.95 (95% CI 2.43–10.11), reported p < 0.001 (as reported in the registry).
• Interpretation:
  • The 95% confidence interval excludes no difference (a ratio of 1), so the hazard of the event is statistically higher for Emollient Plus FP 0.05% Cream vs Emollient than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Numbers of Recurrent Participants at the End of the Maintenance Phase (Week 20)

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Emollient: 56.6% (30/53) versus 5.6% (3/54) on control; risk difference 51.0 percentage points (95% CI 36.4 to 65.7), risk ratio 10.19 (3.31–31.37); unadjusted two-proportion z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Emollient than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Numbers of Recurrent Participants at the End of the Follow-up Phase (Week 32)

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Emollient: 60.4% (32/53) versus 18.5% (10/54) on control; risk difference 41.9 percentage points (95% CI 25.1 to 58.6), risk ratio 3.26 (1.79–5.94); unadjusted two-proportion z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Emollient than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The posted secondary outcomes extend the primary question in two directions: they describe relapse timing through follow-up and count recurrent participants at the ends of the maintenance and follow-up periods. Across the posted recurrence counts, the combination arm has fewer recorded recurrences than the emollient arm. The secondary time-to-event analysis also reports a hazard ratio for the named comparison. Taken together, these outcomes are directionally consistent as recorded in the registry, but each measures a different aspect or period of disease recurrence.

These results complement rather than replace the primary endpoint. The primary outcome concerns time to first relapse through Week 20. The follow-up time-to-event outcome extends observation through Week 32, while the recurrence counts use their own specified time windows. Because the endpoints and periods differ, they should not be treated as interchangeable repetitions of the primary result. The registry record does not report the full analysis plan linking these outcomes or the handling of all repeated and secondary comparisons.

Secondary analyses commonly have less evidential weight than a prespecified primary analysis because multiple outcomes and comparisons increase the opportunity for chance findings. Unless a multiplicity procedure is specified, individual p-values do not control the overall false-positive risk across the set of tests. The registry record does not report a multiplicity adjustment. The secondary p-values should therefore be regarded as unadjusted, and the pattern should be interpreted as supportive description rather than as equivalent to an independently confirmed primary finding.

The registry lists additional secondary outcomes, including treatment success during acute treatment, quality-of-life changes, and questionnaire assessments. Numerical results for those outcomes are not reported in the registry record in the registry record for this analysis. They cannot be used to strengthen or qualify the recurrence pattern without those results.

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

• Median Time to the First Relapse of AD During the Maintenance Phase and Follow-up Phase
• Numbers of Recurrent Participants at the End of the Maintenance Phase (Week 20)
• Numbers of Recurrent Participants at the End of the Follow-up Phase (Week 32)
• Number of Participants With "Treatment Success" During the Acute Phase
• Change From Baseline in Quality of Life (QoL) at the End of the Maintenance Phase
• Change From Baseline in QoL at the End of the Follow-up Phase
• Number of Participants With Post-study Assessment of Skin Emollients Using Questionnaire
• Number of Participants With Post-study Assessment of Lotion Qualities (1) Using Questionnaire

9. Statistical Interpretation

10. Statistical Methods Explained

The registry reports a log-rank analysis for the primary time-to-first-relapse endpoint. The log-rank test compares event-time experience between groups over follow-up, taking account of when relapses occur rather than reducing the outcome to a simple count at one time point. It relies on the observed event and censoring information. The record identifies this method but does not report the full analysis specification, such as any stratification, covariate adjustment or detailed handling of censoring.

What does the hazard ratio describe?

A hazard ratio compares the instantaneous event rates between groups over the observed follow-up, conditional on participants having remained event-free up to that point. It is not a ratio of medians, a difference in days, or a direct probability that an individual participant will relapse. Its interpretation depends on the outcome definition, group order, reference group and event coding. The registry gives the contrast label and hazard-ratio estimate, but does not fully specify the direction convention. The estimate should therefore be quoted with that limitation rather than translated into an unsupported claim about which group had the higher event rate.

A single hazard ratio also compresses the time pattern into one summary. If the relative event rates vary over time, one estimate may not describe every part of follow-up equally well. The registry record does not report an assessment of the proportional-hazards assumption or a time-varying analysis. This does not establish that the assumption was violated; it means that the relevant diagnostic and alternative-analysis details are not reported in the registry record.

Why report a median and a confidence interval?

The median event time is the point by which half of the event-time distribution has occurred, subject to the analysis of censored observations. It offers a time-scale summary that is distinct from the hazard ratio. A confidence interval describes uncertainty around an estimate under the analysis method: a narrower interval indicates greater precision, while a wider interval allows a broader range of values. It is not a guarantee that the true value lies within the interval, and it does not account for every possible source of bias or uncertainty.

For the primary outcome, the median and its 95% confidence interval are reported for the emollient arm, but not for the combination arm. The hazard-ratio result is reported for the comparison. These summaries answer related but different questions, and the unavailable arm median cannot be filled in by interpreting the hazard ratio. The registry does not explain the missing median or provide enough detail to determine why it could not be estimated.

How do the recurrence comparisons differ?

The recurrence endpoints count participants with a recorded recurrence within a stated period. For maintenance, the sponsor reports a Fisher exact analysis; for follow-up, the sponsor reports a chi-squared analysis. These methods assess differences in categorical outcome counts. A simple unadjusted re-calculation from posted group summaries can describe the observed group contrast, but it does not substitute for the sponsor’s analysis or account for factors not contained in those summaries. The registry’s reported analyses take precedence where results or methods differ; no discrepancy is flagged in this record.

Fisher exact testing is based on the categorical count arrangement and is often used when count data are sparse. A chi-squared test uses an approximation for categorical data. Both produce a test of association or group difference, but neither by itself describes the size or clinical importance of an effect. The group counts and the reported test results should be read together, with attention to the endpoint period and analysis population.

Why can over-dispersion matter in other trial outcomes?

As a general statistical point, over-dispersion occurs when count outcomes vary more than a chosen model expects. If it is ignored, estimated uncertainty can be too small and tests can appear more conclusive than warranted. That issue is relevant when a count model is used; it is not a reason to assume over-dispersion in this trial. The primary endpoint is time-to-event, and the registry identifies log-rank testing rather than a count model for that outcome.

For all these methods, randomization supports comparison of groups at baseline in expectation, but it does not remove the consequences of missing observations, unblinded assessment or incomplete reporting. Open-label design can matter when relapse assessment includes judgment, as the PSGA-based definition does. The registry record does not report assessor masking, adjudication procedures or sensitivity analyses addressing these concerns.

11. Confidence Intervals and Precision

The primary log-rank hazard-ratio estimate is 13.499, with a 95% confidence interval from 4.111 to 44.325. The interval spans a broad range of values, indicating substantial uncertainty about the magnitude of the reported estimate. The p-value, p < 0.001, is small, but statistical evidence against the test’s null hypothesis does not make the effect estimate precise. Significance and precision are different properties.

The emollient-arm median time to first relapse is 142 days, with a 95% confidence interval from 50 to 150 days. The range of plausible median values represented by that interval is also broad. A median and a hazard ratio summarize different aspects of the event-time data, so the interval around the median should not be used to infer an interval for the hazard ratio. No median or confidence interval is reported for the combination arm.

For the secondary time-to-event result through Week 32, the reported hazard ratio is 4.952, with a 95% confidence interval from 2.426 to 10.111. That interval is also not a measure of clinical importance by itself. Both time-to-event estimates depend on the event definition, analysis population, follow-up and censoring; the registry’s available-case statement and limited description of handling missing observations are relevant to how precisely they can be interpreted.

Confidence intervals quantify sampling uncertainty under the analysis method. They do not automatically capture uncertainty from open-label assessment, departures from model assumptions, missing data, multiplicity or limits on generalizability. The registry record does not report enough analysis detail to evaluate all of those sources. Accordingly, the point estimates and intervals should be read as summaries of the posted analysis, not as complete measures of uncertainty.

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 effect, not the trial’s original design assumptions. The stated planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, a normal approximation, and an effect to detect set to the stated fraction of the observed effect. For time-to-event outcomes, the figures are total events; for the other outcomes, the figures are participants per arm.

The 80% power column describes the number of events or participants needed for a planned study to have an 80% probability of rejecting its null hypothesis if the assumed effect is present, under the calculation assumptions. The 90% power column describes the corresponding target for a 90% probability. These are design targets, not guarantees about what a future study will observe. The event totals in the time-to-event illustrations are not per-arm counts.

Each row considers planning for a different fraction of the observed effect. Planning for a smaller effect requires more participants or events to distinguish that effect from random variation under otherwise unchanged assumptions. The displayed values are illustrations based on this trial’s observed results; they do not show that the original trial was designed with those targets or that it achieved a particular power. This is not observed power.

The registry record does not report the trial’s own sample-size justification or original design assumptions. The planning illustrations also do not account for design features not included in their stated assumptions, such as losses to follow-up or any adjustment for multiple comparisons. Their role is to show how projected study size changes with the assumed effect and power target, not to revise the interpretation of the completed trial.

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

d is the total number of events needed (Schoenfeld's formula) with 1:1 allocation and proportional hazards.

Worked calculation for this trial

1. Hazard ratio HR = 13.499, so ln(HR) = 2.6026.
2. d = 4 × (1.96 + 0.842)² / [ln(13.499)]² = 4.6, rounded up to 5 events in total 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
hr <- 13.4993                              # observed hazard ratio
events_needed <- function(power, frac = 1) ceiling(round(4 * (z_a + qnorm(power))^2 / log(hr^frac)^2, 6))   # Schoenfeld, 1:1 allocation
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: events = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(events_needed(0.80, f), big.mark = ","), format(events_needed(0.90, f), big.mark = ",")))

# Simulation: given d total events under proportional hazards, events in the treatment arm are Binomial(d, hr / (1 + hr)).
# The score (log-rank type) statistic is (D1 - d / 2) / sqrt(d / 4).
set.seed(1915914)
d    <- events_needed(0.80)
reps <- 20000
D1 <- rbinom(reps, d, hr / (1 + hr))
z  <- (D1 - d / 2) / sqrt(d / 4)
cat(sprintf("Simulated power with %s events: %.3f (%d replicates)\n", format(d, big.mark = ","), mean(abs(z) > z_a), reps))

Output from running the code

100% of observed effect: events = 5 (80% power), 7 (90% power)
 75% of observed effect: events = 9 (80% power), 12 (90% power)
 50% of observed effect: events = 19 (80% power), 25 (90% power)
Simulated power with 5 events: 0.703 (20000 replicates)

The simulation uses the fact that, with d events in total and proportional hazards, the number of events in the treatment arm is binomial. It is an approximation that conditions on the total number of events, and it should land close to the target power.

14. Analyses Reported by the Sponsor

The sponsor’s primary analysis is a log-rank comparison with a reported hazard ratio of 13.499, a 95% confidence interval from 4.111 to 44.325, and p < 0.001. The registry’s rounded display gives the same result as 13.5, with limits of 4.111 and 44.33. No discrepancy is flagged between the posted result and the sponsor’s analysis.

For the primary outcome, the sponsor’s reported analysis is the principal inferential result. A simple unadjusted re-calculation from summary statistics, where one is available, cannot reproduce a time-to-event analysis because it lacks the participant-level event times and censoring information used by the log-rank method. It should not be treated as a replacement for the sponsor’s analysis.

The registry also reports sponsor analyses for recurrence counts: Fisher exact testing for recurrence by the end of maintenance and chi-squared testing for recurrence by the end of follow-up. Those reported p-values are distinct from any unadjusted re-calculation based solely on the displayed counts. The record does not provide a detailed analysis plan or additional modeling information that would permit a full independent reconstruction.

The sponsor’s analyses describe the results under the stated statistical methods. They do not, on their own, resolve limitations in endpoint assessment, missing-data handling, multiplicity or generalizability. The estimates should be interpreted alongside those limitations rather than as a complete account of the trial’s evidential strength.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary outcome and multiple secondary outcomes, and reports analyses for the primary time-to-event comparison and selected secondary outcomes. The registry record does not report a multiplicity-adjustment procedure. When several outcomes or comparisons are tested, the chance of at least one apparently unusual result can increase. Without a stated adjustment or testing hierarchy, secondary p-values should be treated as unadjusted and interpreted cautiously.

The primary population is described as the ITT population of all participants randomized into maintenance, with only participants available at the specified time point analyzed. This wording identifies the intended randomized population but also makes clear that analyses were limited to available participants at the relevant time. The registry record does not report how many participants were missing for each endpoint, reasons for missingness, imputation methods, or sensitivity analyses. It is therefore not possible to assess from this record whether missing observations could have affected the estimates.

The study was open-label, with no masking. The record does not report whether outcome assessors were masked or whether relapse decisions were independently adjudicated. Because the endpoint uses a clinical score, knowledge of treatment allocation could potentially affect assessment; the registry does not establish whether such an influence occurred.

The trial used randomization in maintenance, but details of sequence generation and allocation concealment are not reported in the registry record. Nor are full specifications for censoring, model assumptions, covariate adjustment, interim analyses, or the original statistical plan reported. These omissions limit independent assessment of the analysis and its robustness. They should be treated as reporting limitations, not as evidence that any particular procedure was or was not used.

16. 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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: bronchitis (1.9%), hand-foot-and-mouth disease (1.9%), infectious mononucleosis (1.9%), mycoplasma infection (1.9%).

The safety figures are reported in separate groups for acute fluticasone treatment, maintenance combination treatment, maintenance emollient treatment and follow-up emollient use. Serious adverse events are recorded for 0 of 116 participants in the acute fluticasone group, 2 of 53 in the maintenance combination group, 0 of 53 in the maintenance emollient group, and 0 of 66 in the follow-up emollient group. The registry records other adverse events for 4 of 116, 9 of 53, 9 of 53 and 11 of 66, respectively.

These are counts within the registry’s named safety groups and should not be combined across phases or treated as if every group represents the same treatment period. The record does not explain the relationship between all safety denominators and the maintenance randomization totals. Deaths are not reported in the registry record.

The listed serious adverse event terms include bronchitis, hand-foot-and-mouth disease, infectious mononucleosis and mycoplasma infection, each reported as 1.9%. Event counts alone do not establish causation or comparative safety. The registry record does not provide further detail needed to assess event timing, severity beyond the serious classification, relatedness, or statistical comparison between groups.

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

The results describe outcomes among paediatric subjects with stabilized atopic dermatitis who entered the randomized maintenance comparison. The primary endpoint concerns time to first relapse, and the sponsor reports a statistically significant log-rank result for the named treatment contrast. The recurrence counts and extended time-to-event analysis provide additional descriptions of relapse during maintenance and follow-up.

These findings do not establish that either regimen will produce the same results in people outside the trial population, under different treatment schedules, or in settings with different monitoring. The study was open-label, the primary endpoint relies on a clinical score, and the registry record does not report masking of assessors or adjudication. The reported results therefore need to be considered in light of how relapse was assessed.

A small p-value does not establish the size or practical importance of a difference. The primary hazard-ratio confidence interval is broad, and the combination-arm median is not reported. The registry also does not provide complete details about missing data, multiplicity, model assumptions or the original statistical plan. Those gaps limit conclusions about precision and robustness.

The findings are not medical advice and do not by themselves determine a treatment choice. They are a statistical account of the registry’s reported comparison and outcomes. Any clinical interpretation would need to consider the full study record and other relevant evidence, neither of which can be reconstructed from these posted results alone.

18. Why This Trial Matters Statistically

This trial illustrates how a randomized maintenance comparison can focus on time until a clinically defined recurrence rather than only the presence or absence of relapse at one endpoint. Time-to-event methods use the event timing and are appropriate to a question framed around when relapse first occurs. The study also reports recurrence counts, making clear that different endpoint forms can provide complementary but non-identical summaries.

It also demonstrates why an effect estimate should be read together with its uncertainty. The sponsor reports a small p-value, yet the primary hazard-ratio confidence interval spans a broad range, and a median is not reported for one arm. Statistical evidence against a null hypothesis and precision about the magnitude are not the same thing.

The maintenance population is selected through an acute treatment and stabilization process. That design makes the question specific: it concerns maintenance strategies after treatment success, not treatment of all eligible subjects from the outset. Randomization supports the maintenance comparison, while the open-label design and incomplete reporting of assessment and missing-data procedures remain relevant qualifications.

Finally, the secondary results demonstrate the importance of multiplicity and endpoint hierarchy. A broadly consistent pattern across related outcomes can be informative, but multiple unadjusted tests do not carry the same evidential weight as a clearly specified primary analysis. The registry record is useful for understanding the design and posted results, while its reporting gaps mark where interpretation must remain limited.

19. Limitations

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

The primary time-to-event result has a wide confidence interval, and a median is not reported for the combination arm. The registry does not explain the reason for that missing median, so the time-to-event summaries are incomplete for direct descriptive comparison. It also does not fully specify the hazard-ratio reference and event-coding convention, limiting interpretation of the estimate’s direction from the contrast label alone.

The study was open-label. The registry record does not report whether outcome assessors were masked, how relapse assessments were standardized beyond the PSGA definition, or whether suspected events were adjudicated. Those details matter because the endpoint depends on a clinical score.

The analysis population is described as randomized maintenance participants available at the specified time point. The record does not report endpoint-specific missingness, reasons for missing observations, methods for handling missing data, or sensitivity analyses. It also does not report a multiplicity adjustment for the secondary outcomes. These omissions limit evaluation of the robustness and combined evidential strength of the results.

The registry record does not report the original sample-size justification, detailed allocation-concealment methods, or a complete statistical analysis plan. It also does not provide all listed secondary outcome results in the registry record for this analysis. Finally, the findings concern a specific paediatric stabilized-atopic-dermatitis maintenance setting and should not be generalized beyond that setting on the basis of this record alone.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Randomized, Open-label, Comparative Study to Evaluate an Intermittent Dosing Regimen of Fluticasone Propionate 0.05% Cream (Twice Per Week) in Reducing the Risk of Relapse When Added to Regular Daily Moisturization Using PHYSIOGEL Lotion in Paediatric Subjects With Stabilized Atopic Dermatitis — NCT01915914.
• Registry record last updated 2018-10-12; status last verified 2018-09.
• Results first posted 2015-12-23.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

23. Record Summary

NCT01915914 is a completed, randomized, open-label, parallel-group Phase 4 study sponsored by GlaxoSmithKline. It compared emollient plus intermittent fluticasone propionate 0.05% cream with emollient alone during maintenance after an acute treatment and stabilization phase.

The primary outcome was time to first atopic dermatitis relapse through Week 20. The registry reports a log-rank hazard ratio of 13.499, with a 95% confidence interval from 4.111 to 44.325 and p < 0.001. The emollient-arm median was 142 days; a median for the combination arm is not reported.

Secondary recurrence counts and the extended time-to-event comparison provide related results, but the registry does not report a multiplicity adjustment. Open-label assessment, available-case analysis language, incomplete missing-data details and limited reporting of the statistical plan qualify interpretation. The posted estimates describe the trial record and do not by themselves establish broader clinical conclusions.