This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the results posted in the ClinicalTrials.gov record; calculations added by Clinical Biostats are labelled as such.
1. Trial at a Glance
This page covers the registry results for Azelaic Acid on Demodex Counts in Rosacea (NCT03035955), a randomized phase 2, phase 3 study. The primary outcome is Demodex Count.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT03035955 |
| Phase | Phase 2, Phase 3 |
| Condition | Rosacea |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | Single |
| Enrollment | 20 |
| Arms in the primary outcome | 2 |
| Lead sponsor | Wake Forest University |
| Results posted | Yes |
2. Clinical Question
The study asks whether applying azelaic acid to one side of the face is associated with a different Demodex folliculorum count at Week 4 than leaving the other side untreated. It also describes lesion count and redness as outcomes of interest in the study rationale, but the registry lists only Demodex Count as a primary outcome measure.
Population
Adults with mild to moderate papulopustular rosacea. The registry lists an enrollment of 20.
Intervention
Azelaic acid, identified in the arm description as Finacea Gel 15%, applied twice daily to either the left or right side of the face.
Comparator
No treatment on the opposite side of the face. The comparison is within a split-face arrangement, rather than between separate active-treatment and untreated participant groups.
Primary question
What Demodex count was reported at Week 4 for each randomized side-allocation arrangement? The registry does not provide a treatment-versus-control estimate based on paired side-level results.
The study description says that Demodex counts were assessed at baseline and Week 4, while also noting that only Week 4 was reported. The central statistical question is therefore not just whether the posted values differ, but whether the record gives enough information to attribute any difference to treatment. It does not provide the paired counts, a treatment-effect estimate, or an uncertainty interval. The two posted values correspond to the two arrangements of azelaic acid and no treatment across the face; they should not be read as results from independent treatment and control groups.
3. Trial Design
The registry describes a randomized, single-masked study and labels its model as parallel. Its narrative and arm descriptions, however, specify a split-face design: each participant used azelaic acid on one side of the face and had no treatment on the other side. The side receiving azelaic acid could be either the left or right side. The operational details explaining the registry's parallel-model label alongside the split-face description are not reported in the registry record.
The listed arm arrangements are azelaic acid on the left with no treatment on the right, and azelaic acid on the right with no treatment on the left. Each arrangement is listed with 10.0 participants. These are side-allocation groups, not independent groups in which one set of participants received treatment and another set received no treatment. The split-face design means that both treatment conditions are represented within each participant, while the arm labels distinguish which side received azelaic acid.
The study lists Phase 2 and Phase 3, a total enrollment of 20, and a completed status. The listed condition is rosacea, and the sponsor is Wake Forest University. The record identifies azelaic acid as the drug intervention and names Finacea Gel 15% in the arm description. The brief summary describes a twice-daily application and a study population with mild to moderate rosacea.
For the primary outcome, the timeframe is Week 4. The outcome description mentions counts at baseline and Week 4 but specifies that only Week 4 was reported. The registry record does not describe the masking procedure, the method used to assign the treated side, treatment adherence, or the exact schedule of assessments beyond the stated timeframe. Those missing details limit how fully the conduct of the study can be reconstructed from the 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
| Stage | Azelaic Acid Left/No Treatment Right | Azelaic Acid Right/No Treatment Left |
|---|---|---|
| STARTED | 10 | 10 |
| COMPLETED | 9 | 9 |
| NOT COMPLETED | 1 | 1 |
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.
| Characteristic | Unit | Azelaic Acid Left/No Treatment Right | Azelaic Acid Right/No Treatment Left | Total |
|---|---|---|---|---|
| Participants analysed | 10 | 10 | 20 | |
| Age, Categorical: <=18 years | Participants | 0 | 0 | 0 |
| Age, Categorical: Between 18 and 65 years | Participants | 10 | 10 | 20 |
| Age, Categorical: >=65 years | Participants | 0 | 0 | 0 |
| Sex: Female, Male: Female | Participants | 8 | 9 | 17 |
| Sex: Female, Male: Male | Participants | 2 | 1 | 3 |
6. Primary Endpoint
The registry lists the primary outcome as Demodex Count, measured over Week 4.
| Element | Registry entry |
|---|---|
| Result type | Number |
| Measure of spread | not applicable |
| Unit | number of demodex |
| Endpoint classification | not reduced to a single comparison |
The primary endpoint is Demodex Count, expressed as the number of demodex. It is a count outcome: it records how many organisms were identified under the study's measurement procedure, rather than measuring a continuous quantity such as weight or time. The registry's outcome description refers to counts at baseline and Week 4 but says only Week 4 was reported.
A count is relevant to the stated research question because the investigators hypothesized that azelaic acid treatment would lead to a decreased Demodex folliculorum count. Count data can describe the level observed under each condition, and a treatment comparison can assess whether counts tend to differ between treated and untreated sides. In a split-face study, that comparison should preserve the within-participant pairing: the treated side and untreated side of a participant are related observations, not independent observations.
The record reports 40.0 for the arrangement with azelaic acid on the left and no treatment on the right, and 18.0 for the reverse arrangement. The registry does not identify whether these values are means, medians, or another summary, and it does not provide a measure of spread or a paired treatment contrast. It also does not report the sampling or counting procedure. As a result, the endpoint values can be described as posted, but their precise statistical interpretation and measurement context are limited.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The registry posts a Week 4 Demodex count value of 40.0 for the arrangement with azelaic acid on the left side and no treatment on the right. It posts 18.0 for the arrangement with azelaic acid on the right and no treatment on the left. These are the reported values for the two side-allocation arrangements. The record does not establish that they are averages, nor does it state how the values were calculated.
Those values alone do not provide a direct estimate of the effect of azelaic acid compared with no treatment. In this split-face design, the relevant treatment comparison would ordinarily use the treated-side and untreated-side results within the same participants. The registry does not post those paired side-level values or a paired contrast. The arrangement-level values could reflect differences between the participants in the two allocation groups, differences in how the values were summarized, or other features of the data; the record does not allow those possibilities to be separated.
No confidence interval, test statistic, p-value, or sponsor analysis is posted. The registry lists zero analyses. Accordingly, the record does not establish the precision of a treatment-effect estimate or whether a statistical comparison was made. A numerical difference between the posted arrangement values is not, by itself, evidence of a treatment effect. The outcome description also says that baseline and Week 4 were assessed but only Week 4 was reported, so no change from baseline can be evaluated from the posted results.
The endpoint is a count, but no dispersion measure or distributional description is available. Without these details, it is not possible to judge how variable the observations were or whether a simple mean-based comparison would describe them adequately. The record also does not provide the number of paired observations contributing to an endpoint analysis or information about missing outcome data.
The appropriate reading is therefore descriptive and limited: two Week 4 values are recorded for the two side-allocation arrangements, but the registry does not supply the participant-level comparison needed to estimate the within-person treatment contrast or its uncertainty. These data do not support a firm conclusion about whether counts differed between treated and untreated sides. The sponsor's own analysis is not posted in the registry record, and there is no recalculated analysis to compare with it.
9. Statistical Methods Explained
The design and outcome structure determine which statistical comparisons would be meaningful. The registry describes randomized side allocation and a split-face arrangement, with azelaic acid applied to one side and no treatment on the other. The side allocation matters because it is the unit assigned to treatment, while each participant contributes observations under both conditions. Any analysis that treats the treated and untreated sides as unrelated observations would fail to reflect that pairing.
Why does the split-face structure matter?
In general, a split-face design can control for characteristics shared by both sides of a participant, such as overall participant-level features. The treatment contrast is then based on the within-participant difference between sides. This can be more informative than comparing unrelated groups when the paired observations are appropriately measured and analyzed. Pairing does not remove all sources of bias: side-specific differences, measurement procedures, adherence, and other design features may still matter.
For a paired analysis, the key information is the set of corresponding treated-side and untreated-side counts for each participant, or an appropriate summary of those paired differences. The registry does not report these data. Its two posted values are organized by which side received treatment, and the record does not say that the values are paired contrasts. It therefore does not support reconstruction of a within-participant effect estimate.
How should count outcomes be analyzed?
In general, counts are non-negative observations and may have distributions that differ from those commonly assumed for continuous measurements. Their spread can increase as their typical level increases, and counts can be more variable than a simple model expects. Over-dispersion describes variation greater than that expected under a chosen count model; ignoring it can make uncertainty estimates too small. A suitable analysis depends on the observed distribution, the measurement process, and the design.
For paired count outcomes, the analysis also needs to account for the relationship between the two sides of the same participant. A count model or another method may be appropriate depending on the data and assumptions, but the registry does not report a statistical model, an estimate of dispersion, or an analysis plan. It would not be appropriate to infer which method the investigators used.
What can the posted values establish?
A point estimate describes a result in the observed data, while a confidence interval conveys the uncertainty around an estimate under the assumptions of the analysis. A hypothesis test provides a test statistic or p-value for a specified comparison and model. These tools are most interpretable when the estimand, analysis population, and handling of paired observations are clear. The registry gives two Week 4 values but no inferential statistic or interval, and it does not identify the summary measure used for the values. The precision and statistical evidence for a treatment contrast therefore cannot be assessed from the registry results.
Baseline adjustment is another possible analytical consideration when baseline measurements are available. Here, the outcome description mentions baseline measurement, but the registry says only Week 4 was reported. Baseline values and any adjustment for baseline are not reported in the registry record. A change-from-baseline comparison or baseline-adjusted estimate cannot be reconstructed.
What do planning calculations mean?
In general, sample-size planning uses assumptions about the effect to detect, the variability of the outcome, the significance level, and the design. A planning illustration is not an observed result and does not show the study's achieved power. The registry describes a planning approach based on fractions of an observed effect, a two-sided significance level of 0.05, equal allocation, and a normal approximation. Those are planning assumptions, not evidence that the trial used that approach to choose its enrollment. The trial's own sample-size justification is not reported in the registry record. For a split-face count outcome, planning would also need to reflect the paired design and relevant count variability; the record does not give the information needed to evaluate those assumptions.
There is no sponsor analysis posted for the primary endpoint, and no recalculated comparison is available. Thus, statistical interpretation should remain at the level of the recorded Week 4 values, not an inferred test or model-based treatment effect.
10. Multiplicity, Missing Data and Analysis Populations
The registry lists one primary outcome measure, Demodex Count, and no secondary outcomes are posted. It lists zero analyses. These registry entries indicate that the registry record contains no posted set of inferential analyses for which multiplicity handling can be assessed. The record does not report whether additional outcomes were collected, whether multiple comparisons were planned, or whether any adjustment for multiplicity was used.
The outcome description refers to baseline and Week 4, but only Week 4 is reported. Baseline values, participant-level paired results, and reasons for any unavailable measurements are not reported in the registry record. The record does not describe an approach to missing data, such as a model-based method, imputation, or a complete-case analysis. It also does not identify an analysis population or explain whether all enrolled participants contributed to the endpoint summary.
For a split-face study, it would be useful to know whether both sides were measured for each participant and how incomplete pairs were handled. That information is not reported in the registry record. The listed participant counts for the two side-allocation arrangements do not establish how many participants had usable Week 4 measurements for both sides.
Without a prespecified analysis description, baseline results, missing-data details, or posted analyses, it is not possible to assess selective outcome reporting, the treatment of incomplete observations, or the robustness of any inferential result. These are limits on interpretation, not evidence that a particular analysis or reporting practice occurred.
11. Safety Results
The registry reports serious adverse events as affected participants divided by participants at risk.
| Arm | Participants with serious adverse events | Participants at risk | Percentage |
|---|---|---|---|
| Azelaic Acid | 0 | 20 | 0.0% |
| no Treatment | 0 | 20 | 0.0% |
Deaths
| Arm | Deaths (participants affected of at risk) |
|---|---|
| Azelaic Acid | 0 of 20 |
| no Treatment | 0 of 20 |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 5% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| Azelaic Acid | 0 of 20 |
| no Treatment | 0 of 20 |
The safety figures are listed separately for azelaic acid and no treatment. For each condition, the registry lists 20 at risk for serious adverse events and 0 affected; it also lists 20 at risk for other adverse events and 0 affected. Deaths are listed as 0 in each group. No serious adverse event terms are listed.
These figures describe the safety entries in the registry record. Because the study used a split-face arrangement, the azelaic-acid and no-treatment groups are not necessarily separate sets of participants. The same participant could contribute to the at-risk count for both conditions, so the entries should not be added to create a participant total. The record does not provide event-level details, definitions, observation periods, or further explanations of how at-risk counts were assigned.
Zero affected is the posted count for the listed adverse-event categories; it does not establish that no event of any kind occurred outside the categories or reporting framework represented in the record. The safety figures should be read as registry reporting, not as a comparison with a calculated rate or as proof of safety.
12. What the Result Does, and Does Not, Mean
The result establishes that the registry contains two Week 4 Demodex count values: 40.0 for the side-allocation arrangement in which azelaic acid was applied to the left, and 18.0 for the arrangement in which azelaic acid was applied to the right. It also establishes that the record describes a randomized split-face study in adults with mild to moderate papulopustular rosacea and lists no treatment on the opposite side.
The record does not establish a within-person treatment effect. The posted values are arranged by treatment-side assignment, and the registry does not identify their summary statistic or provide the treated-side and untreated-side counts paired within each participant. It supplies no confidence interval, p-value, or analysis result for a direct comparison. The values therefore cannot show how precisely a treatment contrast was estimated or whether a statistical test supported one.
The absence of baseline results also means that a change from baseline cannot be evaluated. Although the outcome description mentions baseline and Week 4, the record says that only Week 4 was reported. It does not explain how missing measurements were handled or state which participants were included in an endpoint analysis.
These limitations do not show that azelaic acid had no effect, nor do they establish that it reduced Demodex counts. They mean that the posted registry results are insufficient to determine the direction and magnitude of a paired treatment effect with statistical precision. The stated hypotheses about Demodex counts, lesion counts, and redness should not be confused with observed results: the registry lists only Demodex Count as a primary outcome, and it does not post results for lesion count or redness.
The safest statistical conclusion is narrow: the registry reports Week 4 values for two side-allocation arrangements, but it does not report the analysis needed to interpret them as a treatment-versus-no-treatment effect. No clinical recommendation follows from these registry entries.
13. Why This Trial Matters Statistically
This trial is statistically notable because its split-face design makes the comparison structure central to interpretation. Each participant is described as receiving azelaic acid on one side and no treatment on the other. That design is different from a conventional comparison of separate treatment and control groups: the two observations from a participant are linked, and the analysis should respect that relationship.
The registry's result format does not preserve that paired comparison. It gives a value for each side-allocation arrangement, rather than a reported treated-side versus untreated-side contrast within participants. This highlights a practical reporting principle: summaries should match the randomization and measurement structure. Without paired results, a reader cannot distinguish the treatment contrast from differences between the participants assigned to each side arrangement.
The endpoint also illustrates why the type of outcome matters. Demodex Count is a count measure, so its variability and distribution are relevant to the choice and interpretation of an analysis. The registry does not report a spread measure, distribution, or model. A point value without those details gives little information about precision, especially when the paired structure is also unavailable.
Finally, the record distinguishes between an outcome being named and an outcome being analyzed. Demodex Count is listed as the primary measure, but no inferential analysis is posted. The entry offers a useful example of how registry results can document outcome values while leaving the treatment estimate and its uncertainty unresolved. For clinical researchers, it underscores the importance of reporting paired data or a clear paired effect estimate, along with the model, uncertainty, and handling of missing observations.
14. 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 main limitation is that the registry does not provide a paired treatment comparison, although the study description specifies a split-face design. The arm-level values are reported by side-allocation arrangement, not as a within-participant contrast. The record does not explain whether the values are means, medians, or another summary, and no spread measure is given.
The outcome description mentions baseline and Week 4, but only Week 4 results are reported. No baseline values, changes from baseline, or baseline-adjusted analyses appear in the record. The sampling and counting methods for Demodex, and the rules used to derive the posted values, are not reported in the registry record.
The enrollment is 20, with 10.0 listed for each side-allocation arrangement. The record does not provide a sample-size justification or explain whether the enrollment was intended to support a particular level of precision. No inferential analysis, confidence interval, p-value, or analysis population is posted. Missing-data methods, adherence information, and reasons for any unavailable measurements are also not reported in the registry record.
The design metadata labels the study parallel while the narrative describes a split-face comparison. The registry does not reconcile those descriptions or give the masking procedure and allocation details. These limitations constrain evaluation of the design and analysis. They should not be treated as evidence that a specific protocol feature was absent; they are details not reported in the registry record.
15. Related Tutorials
16. Related Statistical Calculators
17. Sources
- ClinicalTrials.gov: Azelaic Acid on Demodex Counts in Rosacea — NCT03035955.
- Registry record last updated 2018-09-07; status last verified 2018-08.
- Results first posted 2017-05-30.
- 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.
Continue through the Clinical Biostats statistical pathway
Use this trial's endpoint as a starting point for deeper study of effect measures, confidence intervals, and sample-size planning.
18. Record Summary
The registry record describes a completed randomized, single-masked study of azelaic acid and Demodex counts in adults with mild to moderate papulopustular rosacea. It lists Phase 2 and Phase 3, an enrollment of 20, and Wake Forest University as sponsor. The study narrative describes a split-face arrangement, with azelaic acid applied twice daily to one side and no treatment on the other.
The primary outcome is Demodex Count at Week 4. The record reports 40.0 for azelaic acid on the left and no treatment on the right, and 18.0 for the reverse arrangement. Although the outcome description mentions baseline and Week 4, only Week 4 is reported. No inferential analysis, paired treatment contrast, uncertainty interval, or secondary outcome result is posted. The safety entries list 0 affected by serious adverse events, 0 affected by other adverse events, and 0 deaths for each condition.