Evaluation of AZR-MD-001 in Patients With Meibomian Gland Dysfunction (MGD): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Evaluation of AZR-MD-001 in Patients With Meibomian Gland Dysfunction (MGD) (NCT04391959), a randomized phase 2 study. The primary outcome is Meibum Gland Secretion Score (MGS).

2. Clinical Question

3. Trial Design

This was a randomized, triple-masked, parallel-group Phase 2 study comparing AZR-MD-001 Active with AZR-MD-001 Vehicle in people with Meibomian Gland Dysfunction. The study is listed as completed, with enrollment of 31 participants. The registry record gives a start date of 2020-07-17 and a primary completion date of 2021-01-06. It describes both interventions as dosed up to twice weekly.

The primary endpoint was change from baseline in MGS at Month 3, reported as a mean with standard deviation. The posted endpoint summaries list 3 participants for vehicle and 4 for active treatment. These are the counts attached to the posted arm-level outcome values; why they are smaller than the overall enrollment is not reported in the registry record. The record does not provide an account of participant flow, withdrawals, exclusions, or missing endpoint measurements.

Randomization and masking are important design features because they can reduce the influence of treatment-selection and outcome-assessment expectations. The registry record does not report the randomization method, allocation concealment procedures, or operational details of masking. It also does not report the protocol’s sample-size justification or a detailed statistical analysis plan. Accordingly, the posted design information supports a description of the trial structure, but not a reconstruction of its procedures beyond what is explicitly recorded.

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: Target Population

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 Meibum Gland Secretion Score (MGS), measured over Month 3.

MGS is a clinical score intended to characterize the secretions expressed from meibomian glands. The registry record says that the scale ranges from 0, described as highly abnormal, to 45, described as Normal. Assessment uses a handheld Meibomian Gland Evaluator along the eyelid margin. The lower eyelid assessment covers 15 glands: 5 in each of the temporal, central, and nasal regions. Each gland’s expressed secretion characteristics receive a grade on a 0-3 scale.

The primary endpoint is change from baseline in MGS at Month 3. A change score focuses on how a participant’s measurement differs from their own baseline rather than only reporting the Month 3 score. The registry record summarizes this continuous endpoint with an arm-level mean and standard deviation. A higher MGS value is closer to the scale description of Normal, but interpreting a group mean still requires attention to variability, the number of measured participants, and the comparison between randomized groups.

The endpoint is relevant to the study’s stated focus on MGD and secretion characteristics. Its description does not establish how much change would be important to patients or how a given change relates to other clinical outcomes; such interpretation is not reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The posted primary outcome is continuous: it is a score that can take values across a stated scale, and the registry record summarizes it using a mean and standard deviation. For a continuous endpoint, the mean describes the center of the observed values, while the standard deviation describes how spread out those values are around the mean. These answer different questions. A standard deviation is not a confidence interval and does not, by itself, quantify how precisely the population mean or a between-group contrast has been estimated.

What would a treatment comparison estimate?

In a randomized, parallel-group study, a direct treatment comparison would ordinarily express how the outcome differs between the active and comparator groups, with an uncertainty measure such as a confidence interval. A confidence interval is a range of values compatible with the data and the assumptions of the statistical method; it helps show the precision of an estimate. It is not reported for the primary comparison in the registry record. The record also does not provide a p-value or a formal analysis result, so neither statistical significance nor the range of plausible group differences can be evaluated from a reported inferential comparison.

The posted values are mean changes from baseline at Month 3, not final scores alone. Analysis of change scores is one way to express how measurements shift from baseline. Other approaches can model the follow-up outcome while accounting for baseline measurement, but the registry record does not state which approach was planned or used for an inferential analysis. The posted arm summaries do not permit the analyst to reconstruct that choice or a model-adjusted result.

Why do the small outcome counts matter?

The endpoint summaries include 3 vehicle participants and 4 active participants. With such small groups, each observation can materially affect an arm mean and standard deviation. Estimates from small groups are generally less stable than estimates from larger groups, and uncertainty about a treatment contrast can be substantial. The standard deviations reported for the arms describe observed within-arm spread; they do not remove this uncertainty. Because no confidence interval for the comparison is reported, the precision of the apparent difference cannot be read directly from the record.

The overall enrollment is 31, whereas the posted outcome summaries represent fewer participants. The reason for this difference is not reported in the registry record. It would be inappropriate to infer whether the difference reflects missing measurements, exclusions, a particular analysis population, or some other process. Participant flow and the number randomized to each group are also not provided in the registry record analyzed for this result.

What do planning illustrations mean?

Sample-size planning illustrations use the observed effect as a reference and describe hypothetical planning under stated assumptions. In this case, the stated assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is defined as a stated fraction of the observed effect. Such illustrations are planning aids, not evidence that the completed trial had a particular power, and they are not observed power. They do not establish that the original trial was designed using those assumptions; the protocol’s sample-size justification is not reported in the registry record.

In general, sample-size calculations for a continuous endpoint depend on the target effect, variability, allocation, and statistical assumptions. A smaller target effect usually calls for a larger study to distinguish it from random variation under otherwise similar assumptions. That is a general principle, not a reconstruction of the design of this trial. No sample-size illustration values are stated in the registry record details considered for this interpretation, so no numerical planning estimate can be reported.

Why can over-dispersion matter?

Over-dispersion is a general concern when the observed variability is greater than a statistical model expects. If a model understates variability, its uncertainty estimates can be too narrow, making results appear more precise than they are. The MGS result is reported as a continuous measure with a standard deviation, and the registry record does not report a model-based comparison or a variance assumption to assess for over-dispersion. The concept is useful when evaluating statistical models, but it does not establish that over-dispersion occurred in this study.

In general, randomized allocation supports a fair comparison when outcomes and participants are analyzed according to an appropriate plan. But randomization alone does not supply missing outcome measurements, explain differences between enrollment and endpoint counts, or determine the best statistical model. Those details, along with an analysis plan and inferential results, are not reported in the registry record. The most defensible reading is therefore descriptive: the record reports arm means, standard deviations, and outcome counts, without a formal primary comparison.

10. Confidence Intervals for Each Arm

No two-arm comparison is shown for this trial, but each arm's mean can be given with a t-based 95% confidence interval computed from the posted mean, spread and number of participants.

R code and output

arm <- c("AZR-MD-001 Vehicle", "AZR-MD-001 Active")
n   <- c(3, 4)
m   <- c(0.5, 7.4)
sp  <- c(1.83, 1.53)
sd  <- sp      # posted standard deviations
for (i in seq_along(n)) {
  se <- sd[i] / sqrt(n[i]); h <- qt(0.975, n[i] - 1) * se
  cat(sprintf("CARM|%s|%d|%.3f|%.3f|%.3f|%.3f\n", arm[i], n[i], m[i], sd[i], m[i] - h, m[i] + h))
}

CARM  AZR-MD-001 Vehicle  3  0.500  1.830  -4.046  5.046
CARM  AZR-MD-001 Active  4  7.400  1.530  4.965  9.835

11. Multiplicity, Missing Data and Analysis Populations

The registry record lists Total OSDI as a secondary outcome and reports two outcome measures, but a secondary outcome estimate is not reported in the registry record. A multiplicity strategy—how to interpret several endpoints or comparisons while controlling the risk of false-positive findings—is not reported in the registry record. Without that information, it is not possible to determine whether any planned adjustment or hierarchy among outcomes was intended.

Missing-data handling is also not reported in the registry record. The posted MGS summaries include 3 participants in the vehicle group and 4 in the active group, while overall enrollment was 31. The record does not explain the difference, identify missing measurements, or state whether participants were excluded from an analysis. It does not describe imputation, sensitivity analyses, or assumptions about missing outcomes. No reason for the small outcome counts should be inferred from the summaries alone.

The analysis population is not reported in the registry record. In particular, the record does not say whether the outcome summaries represent all randomized participants, participants with available measurements, or another defined group. The statistical analysis plan and handling of protocol deviations are also not reported. These omissions matter because analysis populations and missing-data assumptions can affect both the estimated treatment contrast and the uncertainty around it. In the absence of those details, the arm summaries can be described, but their relation to the randomized population and any intended confirmatory analysis remains unclear.

12. 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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: dysuria (6.7%).

The registry record lists serious adverse events for 0 of 16 participants at risk in the vehicle group and 1 of 15 participants at risk in the active group. It lists other adverse events for 2 of 16 participants at risk in the vehicle group and 8 of 15 participants at risk in the active group. Deaths are listed for 0 participants in the vehicle group and 1 participant in the active group. These are descriptive counts, not estimates adjusted for follow-up time or other differences.

The registry record identifies dysuria at 6.7% among the top serious adverse event terms. It does not specify the group for that term in the registry record analyzed here, so it should not be assigned to an arm. The adverse-event summaries do not provide event timing, duration, severity beyond the serious-event classification, relatedness assessments, or a narrative about the reported death. Those details are not reported in the registry record.

Small numbers make safety comparisons imprecise, and the event counts alone do not establish whether an intervention caused an event. The at-risk counts are the relevant denominators attached to these safety summaries; they differ from the small counts in the posted primary outcome summaries. No causal or medical interpretation should be drawn from these descriptive figures alone.

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

The posted results establish that the registry record contains Month 3 MGS change summaries for both randomized arms. The active-arm mean change was 7.4, with a standard deviation of 1.53 and 4 participants; the vehicle-arm mean change was 0.5, with a standard deviation of 1.83 and 3 participants. The score is described as ranging from 0, highly abnormal, to 45, Normal. These values give a limited descriptive view of the measured outcome among the participants represented in those summaries.

They do not establish the size or precision of a treatment effect. The record does not report a formal between-group estimate, confidence interval, or p-value. Nor does it explain the relationship between the endpoint counts and the overall enrollment of 31. Without a clear account of missing data and analysis populations, readers cannot determine how representative the posted outcome summaries are of all randomized participants.

The summaries also do not establish whether the observed change is clinically important. The registry record does not report a threshold for meaningful change in MGS or a link between a change in this score and other outcomes. It lists Total OSDI as a secondary outcome, but a result for that measure is not reported in the registry record details analyzed.

Statistical interpretation is further limited because no analysis plan, multiplicity approach, or inferential outcome analysis is reported. The descriptive values should not be treated as proof that the intervention produces a particular effect, or proof that it does not. They are a small set of arm-level observations requiring cautious interpretation, not a basis for medical advice or a treatment recommendation.

14. Why This Trial Matters Statistically

This trial illustrates why enrollment and endpoint-analysis counts need to be distinguished. The registry record gives overall enrollment of 31, while the Month 3 MGS summaries represent 3 vehicle participants and 4 active participants. Without a participant-flow account or missing-data explanation, readers cannot tell how the endpoint summaries relate to the enrolled and randomized population. That gap is important for both interpretation and reproducibility.

It also shows why arm-level means are not a substitute for a treatment comparison. Means and standard deviations describe observed outcomes within each group, but they do not by themselves quantify the uncertainty in a between-group effect. A formal estimate and confidence interval would help readers assess precision; neither is reported for the primary endpoint. Small outcome counts make that absence especially consequential, because individual observations may strongly influence the summaries.

The study record includes a continuous endpoint, a vehicle comparator, and randomized parallel-group design, yet the formal analysis details are not reported. The contrast between the registered design features and the sparse posted inferential information highlights a general principle: readers need clear analysis populations, handling of missing data, and methods for multiple outcomes to evaluate what a trial’s results support.

Finally, planning illustrations based on an observed effect must be kept separate from the design and conduct of the trial. They can illustrate hypothetical future planning under assumptions, but do not retrospectively demonstrate observed power or reveal the protocol’s original sample-size rationale. Keeping those distinctions clear prevents descriptive results and planning calculations from being given more certainty than the registry record supports.

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 main trial-specific limitation is the very small number of participants represented in the primary endpoint summaries: 3 for vehicle and 4 for active treatment. These counts are substantially smaller than the recorded enrollment of 31, but the reason is not reported in the registry record. The record does not provide enough participant-flow information to determine how many were randomized, how many had Month 3 measurements, or how missing outcomes were handled.

The registry record reports arm-level means and standard deviations, but no formal between-group estimate, confidence interval, or p-value. It also does not report a detailed analysis plan, the analysis population, baseline-adjustment method, or multiplicity strategy. Consequently, the posted summaries cannot be used to assess the precision of a primary treatment contrast or the influence of analysis choices.

The outcome description is specific about the MGS scale and assessment procedure, but a threshold for clinically important change is not reported in the registry record. The secondary outcome Total OSDI is listed, while its result is not reported in the registry record details analyzed for this interpretation. These limits restrict interpretation to the posted MGS descriptions and prevent a broader assessment across outcomes.

Safety counts are also descriptive, with no event narratives or details on timing, duration, or relatedness in the record details. The trial’s completed status does not resolve these reporting gaps. The randomization and masking labels describe design features, but operational details are not reported in the registry record.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Evaluation of AZR-MD-001 in Patients With Meibomian Gland Dysfunction (MGD) — NCT04391959.
• Registry record last updated 2022-05-16; status last verified 2021-01.
• Results first posted 2022-05-16.
• 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

NCT04391959 is a completed, randomized, triple-masked, parallel-group Phase 2 study of AZR-MD-001 Active versus AZR-MD-001 Vehicle in patients with Meibomian Gland Dysfunction. Enrollment was 31, and both interventions were described as dosed up to twice weekly.

The primary endpoint was change from baseline in MGS at Month 3. The posted means were 0.5 (standard deviation 1.83; 3 participants) for vehicle and 7.4 (standard deviation 1.53; 4 participants) for active treatment. The registry record does not report a formal between-group analysis or explain why the endpoint counts are smaller than enrollment. Total OSDI is listed as a secondary outcome, but its result is not reported in the registry record.