Study Evaluating the Safety and Efficacy of NCX 4251 Ophthalmic Suspension for the Treatment of Blepharitis: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Study Evaluating the Safety and Efficacy of NCX 4251 Ophthalmic Suspension for the Treatment of Blepharitis (NCT04675242), a randomized phase 2 study. The primary outcome is Percentage of Study Eyes With Complete Cure (Score 0) of Blepharitis Signs and Symptoms at Day 15.

2. Clinical Question

3. Trial Design

This was a multicenter, randomized, double-masked, placebo-controlled, parallel-group Phase 2b study of NCX 4251 ophthalmic suspension for acute exacerbations of blepharitis. The registry lists the study as Phase 2 and completed, with enrollment of 224. The recorded start date is 2020-12-14 and the primary completion date is 2021-06-16. The official title specifies NCX 4251 at 0.1% once daily. The comparator was the vehicle of the NCX 4251 suspension.

Randomization assigns participants to treatment groups by chance, helping make groups comparable on average with respect to measured and unmeasured factors. A parallel design means the groups are followed in separate arms rather than participants crossing from one treatment to another. Double masking is intended to reduce the influence of treatment knowledge on participants’ reports, clinical care, and outcome assessment. The registry record does not describe the randomization sequence, allocation concealment, or exactly who was masked.

The primary endpoint concerned study eyes with complete cure across three scored features at Day 15. The recorded endpoint population was described as intent to treat, but included only study eyes with data for all three scales at baseline and the Day 15 visit. The arm-level result records give n values of 109 for NCX 4251 and 111 for placebo. Those analysis counts should not be treated as a complete accounting of all enrolled participants; the registry record does not explain the difference between the enrollment figure and the endpoint denominators.

The registry lists three secondary outcomes: change in eye dryness symptoms, change in an eye dryness sign assessed by fluorescein staining of the inferior cornea, and treatment-emergent adverse events. The full statistical protocol, detailed visit schedule beyond the primary endpoint timeframe, and methods for handling departures from assigned treatment 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.

Before assignment: 7 to 14 day screening period

Overall Study

Analysis population for the primary outcome, as defined in the registry: Intent to treat (for study eyes with data for all 3 scales at the Baseline and Day 15 Visit)

5. Baseline Characteristics

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

Population: Study Eye

6. Primary Endpoint

The registry lists the primary outcome as Percentage of Study Eyes With Complete Cure (Score 0) of Blepharitis Signs and Symptoms at Day 15, measured over Day 15 Visit.

The primary endpoint was the percentage of study eyes with complete cure, defined as a score of 0, for each of eyelid margin redness, eyelid debris, and eyelid discomfort at the Day 15 visit. The scales ranged from 0, none, to 3, severe. This is a stringent composite definition: an eye qualifies only when all three assessed signs and symptoms meet the complete-cure criterion. It therefore reflects the joint status of several aspects of blepharitis rather than change in any one feature alone.

A complete-cure endpoint can be useful when the clinical question concerns whether assessed features have resolved to the scale’s lowest score by a specified visit. It is straightforward to interpret as an outcome category, but it does not convey the size of improvement among eyes that do not meet the all-features criterion. It also depends on having assessments for every component at both baseline and follow-up. The registry’s population description restricts the endpoint to eyes with data for all three scales at both visits.

The posted arm results are 4.0 for NCX 4251 and 2.0 for placebo. The registry labels the endpoint as a percentage, but the record does not give additional detail on the calculation, rounding, or handling of multiple study eyes from a participant. The statistical interpretation should therefore remain tied to the posted endpoint and reported analysis rather than infer unreported details.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does the registry-reported test do?

The registry identifies Fisher Exact as the statistical method for the comparison. In general, Fisher’s exact test evaluates whether the distribution of observations across categories differs between groups, using an exact calculation rather than relying on a large-sample approximation. It is commonly used for categorical outcomes, particularly when counts are limited. The registry does not explain how the test was applied to the primary endpoint, and the analysis label “Mean Difference (Final Values)” does not clearly match the endpoint’s description as a percentage. That makes it important to distinguish the method name and estimate as reported from a fully specified, independently reproducible analysis.

The primary endpoint is defined at the eye level, with complete cure requiring a score of 0 on all three components at the Day 15 visit. The arm summaries report values of 4.0 and 2.0, alongside denominators of 109 and 111. The record does not provide a complete subject-level or eye-level dataset, nor further calculation details that would allow the displayed analysis to be reconstructed. The record also explicitly indicates that a two-arm recalculation for this count-type endpoint is not estimable from the posted summaries. There is consequently no recalculated p-value or confidence interval to compare with the registry analysis.

How should the confidence interval be read?

A confidence interval expresses uncertainty around an estimated effect under the assumptions of the analysis. The registry reports a 95% confidence interval from -11.5 to 15.3 for the mean difference of 1.9. In general, a wider interval represents less precision than a narrower interval. Because this interval includes zero and extends in both directions, it does not rule out no difference or differences favoring either group under the estimate’s stated orientation. The record does not define the subtraction order or clarify the interpretation of the analysis label, so the sign should not be assigned a treatment-favoring direction beyond the order implicit in the registry’s reported analysis.

A confidence interval is not a range that contains the true effect with a 95% probability after the data have been observed. Rather, in repeated use of the same procedure under its assumptions, intervals constructed in this way would cover the fixed effect at the stated rate. That general explanation does not resolve uncertainty about whether the method and estimate are aligned with the endpoint as described. The interval should therefore be interpreted cautiously as the interval reported in the registry.

What does the p-value add?

The reported p-value is 0.443. In general, a p-value measures the compatibility of the observed data with a specified null hypothesis and statistical model. It is not a measure of effect size, clinical importance, or the chance that the null hypothesis is true. Nor does a larger p-value prove that two groups are identical. Here, the p-value is consistent with the broad uncertainty conveyed by the interval, but neither statistic establishes equivalence or a definitive absence of a difference.

Statistical significance is a decision convention that depends on a threshold chosen in advance. Although the registry reports the p-value, the statistical analysis plan and any prespecified threshold for this endpoint are not reported in the registry record. It is therefore possible to describe the reported p-value without claiming that it reflects a documented decision rule or a complete inferential plan.

Why do missing measurements and over-dispersion matter?

In general, missing outcome data can affect both the estimated group comparison and its uncertainty. The primary population is described as intent to treat for eyes with data on all three scales at baseline and Day 15. This is a restricted observed-data set: randomization alone does not ensure that every randomized participant contributes to an endpoint requiring complete measurements. The record does not report the number or reasons for missing values, the assumptions used to handle missingness, or analyses that test how conclusions change under alternative assumptions.

Over-dispersion is a general issue for count outcomes when variability exceeds what a simple count model assumes. It can make standard errors too small if the extra variability is ignored. It is not possible to assess over-dispersion from the posted arm summaries alone, and the registry does not report a count-model specification, variance assessment, or adjustment for over-dispersion. The endpoint is described as a percentage, while the record’s result type and note refer to a count-type endpoint; this makes it especially important not to impose a model that is not documented.

What can be inferred from the design?

Randomization can reduce confounding in the comparison of assigned groups, and double masking can limit some forms of assessment and behavior bias. In general, these design features strengthen a treatment-group comparison when allocation is implemented appropriately and outcome data are analyzed according to a prespecified plan. However, the registry record does not provide enough detail to evaluate allocation concealment, masking success, adherence, protocol deviations, or the precise analysis population. Design labels alone do not answer those questions.

The study is described as Phase 2b, but the registry does not report the sample-size justification, assumptions used to plan enrollment, interim analyses, or a full statistical analysis plan. Those details should not be inferred from the observed result. Any sample-size illustrations based on observed effects are planning exercises, not observed power, and do not demonstrate that the trial was designed or powered according to those assumptions.

10. Analyses Reported by the Sponsor

The registry reports a Fisher Exact analysis with a mean difference of 1.9, a 95% confidence interval from -11.5 to 15.3, and p = 0.443. The registry labels this as the reported analysis. There is no separately available recalculated two-arm comparison for this count-type endpoint: the record identifies the values as missing or not estimable for that recalculation. It is therefore not possible to assess agreement or disagreement between the reported analysis and an independent recalculation from the arm summaries.

The endpoint is described as a percentage, while the analysis is labeled “Mean Difference (Final Values)” and names Fisher Exact. These labels leave uncertainty about the exact form of the analyzed data and the interpretation of the estimate. The registry record does not provide enough information to resolve that ambiguity or reproduce the analysis. The reported numbers can be described as recorded, but should not be presented as independently confirmed by an unadjusted recalculation.

No discrepancy between a reported analysis and a recalculation is flagged in the record. That does not create a recalculated result where none is estimable, and it does not remove the need to interpret the reported analysis cautiously. The sponsor’s analysis is the available comparison, with its method and labeling understood as reported in the registry.

11. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary endpoint and three secondary outcomes: change from baseline to Day 15 in study-eye dryness symptoms, change in a dryness sign measured by fluorescein staining of the inferior cornea, and treatment-emergent adverse events. It records four outcome measures and three analyses. Multiple outcomes create an opportunity for false-positive findings when many hypotheses are tested. In general, multiplicity procedures can control error rates or otherwise define how findings across several outcomes should be interpreted. The statistical analysis plan and any multiplicity adjustment or testing hierarchy are not reported in the registry record.

The presence of a primary endpoint does not by itself reveal whether secondary outcomes were formally tested, whether any analyses were exploratory, or how tests were ordered. No adjustment should be assumed. The results for different endpoints should not be treated as independent confirmations, especially when their measurements may be related. The registry record does not report a comprehensive list of all analyses undertaken or explain whether the posted analyses represent all planned tests.

The primary endpoint population is described as intent to treat for study eyes with data for all three scales at baseline and Day 15. Requiring complete data for these assessments means that the analyzed endpoint population is narrower than a simple statement of intent to treat might suggest. The registry record does not report how many participants or eyes were excluded from this endpoint because of missing measurements, why data were missing, or whether missingness differed by group. It also does not specify an imputation method, a model using partial measurements, or sensitivity analyses under alternative missing-data assumptions.

Those omissions matter because a complete-case comparison can be distorted if the availability of measurements is related to treatment assignment or outcome status. This is a general concern, not evidence that bias occurred in this trial. Without the missing-data details, readers cannot judge how robust the endpoint comparison is to incomplete follow-up. The record also does not report whether analysis populations were defined differently for secondary outcomes or safety outcomes.

In general, multiplicity and missing-data methods are best interpreted from a prespecified statistical plan. That plan, any interim-analysis procedure, and the detailed analysis-population rules are not reported in the registry record. Accordingly, the primary result can be described as posted, but the available information does not support a broader account of the trial-wide testing strategy or the consequences of incomplete outcome data.

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 non-serious adverse event terms

The adverse-event record lists no serious adverse events affecting participants in either group: 0 of 111 at risk for NCX 4251 and 0 of 113 at risk for placebo. It also lists 0 deaths in each group. These figures describe the serious-event and death entries recorded for the study; they do not establish that the risk of a serious event is zero. With no recorded serious events in these groups, the registry figures cannot precisely characterize the frequency of uncommon harms.

The record separately lists other adverse events affecting 12 of 111 participants at risk in the NCX 4251 group and 20 of 113 at risk in the placebo group. These are counts, not an adjusted comparison. The registry record does not provide event-level detail in the listed serious-event terms, and it does not give a statistical comparison of the other adverse-event counts in the registry record for this analysis. These figures should not be read as proving that one group has greater or lesser overall safety risk.

Safety interpretation also depends on how events were collected, classified, and followed over time. The registry lists treatment-emergent adverse events as a secondary outcome, but detailed definitions, exposure duration, event severity, relatedness, and handling of multiple events per participant are not reported in the registry record. The posted counts therefore provide a limited description, not a comprehensive assessment of comparative safety.

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

The results describe a Day 15 comparison of complete cure across eyelid margin redness, eyelid debris, and eyelid discomfort in study eyes with all required baseline and follow-up data. The posted arm values are 4.0 for NCX 4251 and 2.0 for placebo. The registry-reported mean difference is 1.9, with a 95% confidence interval from -11.5 to 15.3 and p = 0.443. The interval includes zero and spans values in either direction, indicating substantial uncertainty around the reported comparison.

These results do not establish that the groups differ, but they also do not demonstrate that they are equivalent. A failure to obtain strong evidence against no difference is not the same as evidence that the difference is negligible. An equivalence conclusion requires a design that specifies an acceptable range of differences and an analysis that evaluates that range. Such a margin and analysis are not reported in the registry record.

The endpoint is a complete-cure measure, not a general measure of every aspect of blepharitis or every possible change over time. It requires all three components to score 0 at the Day 15 visit. It does not, by itself, describe how much individual components improved among eyes that did not meet the full definition, how symptoms changed at other times, or the durability of any status beyond the specified visit.

The results also require caution because the registry’s endpoint description, result type, and analysis label do not fully clarify how the comparison was calculated. The endpoint population excludes eyes without data for all three scales at baseline and Day 15, while missing-data details are not reported in the registry record. There is no recalculated two-arm estimate available from the posted summaries to resolve those limitations.

The findings are therefore best read as a record of the observed endpoint summaries and the registry-reported inferential statistics, with uncertainty about precision and analysis details. They do not by themselves establish a clinical benefit, a lack of benefit, equivalence, or a complete safety profile.

14. Why This Trial Matters Statistically

This trial illustrates why the endpoint definition and the analysis label need to be read together. Complete cure is defined using three signs or symptoms, each scored on a scale from 0 to 3, with the endpoint requiring the lowest score on every component. That makes the outcome clinically interpretable as an all-components status at the Day 15 visit, but it also means that partial improvement is not captured by the complete-cure category alone.

The registry’s result presentation raises a statistical interpretation question: the endpoint is described as a percentage, yet the reported analysis is labeled as a mean difference in final values and names Fisher Exact. Those descriptions do not clearly identify the data transformation or estimand. In addition, the record says a two-arm recalculation is missing or not estimable for this endpoint type. This is a practical example of why reported methods, outcome scale, and available summary data must align before independent reproduction can be attempted.

The confidence interval is broad relative to the central estimate and includes zero. This highlights the difference between an estimated effect and the precision with which it is known. A single point estimate can look definite when presented without uncertainty; the interval makes clear that the posted result is compatible with a range of possible comparisons. The p-value alone cannot convey that range, and it cannot establish equivalence.

The endpoint population also illustrates the distinction between an intention-to-treat label and the actual observations contributing to an analysis. The registry describes an intent-to-treat population but limits the primary endpoint to eyes with complete data for all three measures at baseline and Day 15. Without missingness details, readers cannot judge how closely this subset represents everyone randomized.

Together, these features make the trial statistically informative as an example of interpreting a randomized comparison with an all-components endpoint, limited analysis detail, and incomplete reproducibility from registry summaries. The most defensible interpretation keeps the reported evidence distinct from questions that the record does not resolve.

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 primary limitation for statistical interpretation is that the registry’s endpoint and analysis descriptions are not fully aligned. The outcome is described as a percentage of study eyes with complete cure, but the analysis is labeled “Mean Difference (Final Values)” and identifies Fisher Exact. The record does not explain the calculation sufficiently to determine exactly how the estimate and interval map to the endpoint. A separate recalculation is not estimable from the posted summaries, so the comparison cannot be independently checked using the available arm-level information.

The primary population requires data for all three scales at baseline and Day 15. The record does not report the number of missing assessments, reasons for missingness, or methods used to handle incomplete observations. Consequently, the extent to which the analyzed population differs from all randomized participants, and whether missing data could affect the result, cannot be evaluated.

The registry reports enrollment of 224 and endpoint arm denominators of 109 and 111, but does not explain the endpoint-specific flow from enrollment to analysis. The reported counts should therefore be understood in their stated contexts rather than assumed to represent every enrolled or randomized participant. Details on protocol deviations, treatment adherence, and the analysis set for each secondary outcome are not reported in the registry record.

The statistical analysis plan is not reported in the registry record. This includes the sample-size justification, any prespecified effect size, the handling of multiple outcomes, the analysis model, interim analyses, and any sensitivity analyses. Without that information, it is not possible to determine whether the posted inferential procedure reflects a prespecified plan or how the trial’s analysis strategy addressed key design choices.

The endpoint is assessed at Day 15 and does not by itself describe longer-term outcomes or change patterns across visits. The available safety figures provide counts but not detailed event characteristics or a formal comparative safety analysis. These limitations restrict how broadly the results can be interpreted beyond the specific posted endpoint and registry entries.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Study Evaluating the Safety and Efficacy of NCX 4251 Ophthalmic Suspension for the Treatment of Blepharitis — NCT04675242.
• Registry record last updated 2023-06-18; status last verified 2023-06.
• Results first posted 2022-12-14.
• 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

The registry record describes a completed, multicenter, randomized, double-masked, placebo-controlled, parallel Phase 2b study of NCX 4251 ophthalmic suspension for acute exacerbations of blepharitis. It lists enrollment of 224, a start date of 2020-12-14, and a primary completion date of 2021-06-16. The primary endpoint was the percentage of study eyes with complete cure across eyelid margin redness, eyelid debris, and eyelid discomfort at Day 15.

The posted arm values were 4.0 for NCX 4251 and 2.0 for placebo. The reported comparison was a mean difference of 1.9, with a 95% confidence interval from -11.5 to 15.3 and p = 0.443. No two-arm recalculation is available for this endpoint from the posted summaries. Serious adverse events and deaths were recorded as 0 in both groups. Important details on missing data, multiplicity, the statistical analysis plan, and the exact calculation underlying the primary comparison are not reported in the registry record.