A Trial of Montelukast for Maintenance Therapy of Eosinophilic Esophagitis in Children: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Trial of Montelukast for Maintenance Therapy of Eosinophilic Esophagitis in Children (NCT01458418), a randomized na study. The primary outcome is Eosinophil Count.

2. Clinical Question

3. Trial Design

The registry describes a randomized, quadruple-masked, parallel trial of montelukast for maintenance therapy of eosinophilic esophagitis in children. The listed arms are montelukast 10 mg/day, montelukast 5mg/day, and placebo. In general, randomization is intended to reduce systematic differences between treatment groups at assignment, while masking is intended to limit the influence of treatment knowledge on care, assessment, or reporting. The registry does not explain how randomization or masking was carried out, nor does it identify the roles covered by quadruple masking.

The arm descriptions give the daily tablet regimens: two 5mg montelukast tablets in the 10 mg/day arm, one 5mg montelukast tablet plus one placebo tablet in the 5mg/day arm, and two placebo tablets in the placebo arm. A parallel design means that assigned groups are followed as separate groups rather than participants crossing between treatments. The registry does not report further details of treatment duration beyond the endpoint timeframe, which is 12 weeks.

Enrollment is listed as 4 and the study status is terminated. The primary outcome description says participants were not unblinded because enrollment was insufficient to create generalizable knowledge; consequently, the number assigned to each arm is not known. The registry reports zero participants in the arm-specific outcome groups and gives an outcome summary for the group whose arm is unknown. The sponsor is Children's Mercy Hospital Kansas City. The study phase is 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: There were no pre-assignment details. All inclusion/exclusion criteria were measured using existing, standard of care data. All subjects who signed informed consent were randomized (although it is not now know which arm they were randomized to as no unblinding was ever done due to low enrollment and lack of data).

Overall Study

Analysis population for the primary outcome, as defined in the registry: Subjects were not unblinded due to inability to enroll numbers to create generalizable knowledge, so we don't know how many subjects were in each arm.

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: Subjects were never unblinded as study was stopped prior to data analysis (number of subjects too low to create generalizable knowledge)

6. Primary Endpoint

The registry lists the primary outcome as Eosinophil Count, measured over 12 weeks.

The primary endpoint is eosinophil count, measured as eosinophils per high-powered field in the mid esophagus after 12 weeks of therapy. Eosinophils are immune cells, and the registry's study description identifies their presence in the esophagus as relevant to eosinophilic esophagitis. Counting them in a tissue sample provides a numerical measure of the cell finding targeted by the stated study question.

In general, a tissue count can be used to quantify a biological feature rather than relying only on a description of symptoms. It does not, by itself, describe every aspect of a person's condition or experience. The registry lists a mean and a full range for the outcome summary, but that summary is attributed to participants whose arm is unknown. There are no reported arm-specific means or ranges and no treatment contrast. As a result, the endpoint's role in the study is clear, but the posted outcome cannot show whether the count differed by assigned treatment.

The study purpose also mentions helping stop esophageal tightening. However, the registry lists only one outcome measure, eosinophil count. A measure of tightening, swallowing, symptoms, or another clinical feature is not reported as a posted outcome. Therefore, the eosinophil count should be read as the specific primary measure available in the record, not as a comprehensive measure of every goal described in the study purpose.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The record identifies a mean as the primary outcome summary and gives the full range. It does not report treatment-specific analyses, an uncertainty interval, or a test comparing groups. Understanding what those quantities would mean helps distinguish the descriptive result from a treatment-effect estimate.

What does the mean describe?

In general, the mean is the arithmetic center of a set of observed values. It can be useful for conveying an average, but it does not show whether observations cluster closely around that average or vary widely. Its interpretation also depends on knowing which participants contributed values and how those participants relate to the groups under study. Here, the posted mean is 8 eosinophils/HPF for subjects whose arm is unknown. It cannot be assigned to any montelukast or placebo arm.

What does the full range add?

The full range reports the lowest and highest values in the summarized observations. The posted range is 2 to 20 eosinophils/HPF. In general, a range conveys the span of observed values but does not describe their distribution, the typical distance from the mean, or the precision of an estimate. A range can also be influenced by extreme observations. The registry does not provide a standard deviation or another measure of spread for this endpoint.

What would a confidence interval mean?

In general, a confidence interval expresses uncertainty around an estimated quantity under a stated statistical procedure. For a treatment contrast, it would help readers assess which effect sizes are compatible with the data and how precisely the contrast is estimated. No confidence interval is reported for the primary outcome or for a comparison between treatment arms. The range from 2 to 20 is not a confidence interval: it describes observed endpoint values, not uncertainty about a treatment effect.

Why is an arm-specific contrast needed?

The clinical question compares montelukast regimens with placebo. The relevant statistical quantity would be a comparison of outcomes linked to those assigned groups, with uncertainty reported for that comparison. The registry lists three planned arms, but the outcome population is identified as having unknown assignment. Consequently, neither a difference between group means nor a test of that difference is available. No comparison can be recovered from the overall mean and range alone.

What does the design contribute?

In general, random allocation can support a fair comparison by making assignment independent of participants' measured and unmeasured characteristics on average. Masking can help limit biases arising from knowledge of assignment. These design features do not substitute for analyzable outcome data. In this record, the randomized, quadruple-masked, parallel design is described, but the number of participants in each arm and the treatment-specific endpoint results are not reported. The practical benefit of those design features for interpreting this outcome is therefore limited by the unavailable arm identities.

How should planning assumptions be read?

The registry's planning note describes sample-size illustrations based on a stated fraction of an observed effect, with a two-sided significance level of 0.05, equal allocation, and a normal approximation. Such calculations, in general, are planning illustrations: they describe sample sizes under specified assumptions and are not observed power. They do not show that a trial actually achieved a given power, and they do not replace a treatment-effect estimate with its uncertainty. No calculated planning values are reported in the registry record.

The registry lists no analyses. It also does not report an analysis model, a method for handling the outcome distribution, an adjustment strategy, or a confidence-interval procedure. Those methods are not recoverable from the randomized design or the descriptive summary. Any more detailed account of the statistical analysis would therefore go beyond what is reported in the registry record.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists one outcome measure, eosinophil count, and zero analyses. It lists no secondary outcomes. The study description refers to both reducing eosinophil counts and helping stop esophageal tightening, but an outcome measure for tightening is not reported. The record does not state whether other outcomes were planned but not posted, or whether any adjustment for multiple comparisons was planned. A multiplicity strategy is not reported in the registry record.

The outcome is summarized for subjects whose treatment arm is unknown. The record says they were not unblinded because enrollment was insufficient to create generalizable knowledge. The number of participants assigned to each arm is therefore not known from the record, and arm-specific endpoint summaries are absent. This is an assignment-identification limitation, distinct from a documented account of missing endpoint values.

The registry does not describe whether endpoint observations were missing, how missing observations would be handled, or whether an imputation or other sensitivity analysis was used. It does not identify an analysis population, such as an intention-to-treat or per-protocol population. Although enrollment is listed as 4 and the outcome summary includes 4 participants with unknown assignment, the record does not give enough detail to establish how participants were selected for any analysis beyond that summary.

These reporting gaps matter because analyses can depend on who is included, how missing observations are handled, and which comparisons are treated as primary. No assumptions about those decisions can be inferred from the available outcome summary.

11. 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.

The adverse-event record has an unknown-arm group with 4 participants at risk for serious adverse events and 0 affected. It also lists 4 at risk for other adverse events and 0 affected. The assigned arm for this group is unknown. The arm-specific adverse-event groups list 0 at risk, so they do not provide treatment-specific safety denominators or comparisons.

These entries describe what is recorded, not a general conclusion about safety. With treatment assignment unknown, the reported counts cannot establish how events were distributed among the montelukast and placebo arms. A zero affected count in the listed unknown-arm group means no affected participants are recorded in that category for this group; it should not be interpreted as proof that a treatment is free of risk.

Deaths are listed as null rather than as a count, so death information is not reported in the registry record. No serious-event terms are listed. The registry does not provide further event descriptions or treatment-specific safety summaries. The small enrollment and unavailable arm assignments further limit what these entries can establish about adverse events.

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

The record documents a randomized, quadruple-masked trial of montelukast 10 mg/day, montelukast 5mg/day, and placebo for maintenance therapy in children with eosinophilic esophagitis. Its primary endpoint was eosinophil count in the mid esophagus after 12 weeks of therapy. The posted endpoint summary is an overall mean of 8 eosinophils/HPF and a full range from 2 to 20 for subjects whose arm is unknown.

Those descriptive statistics do not show how the outcome differed by assigned treatment. Because the registry says participants were not unblinded and does not identify the number in each arm, the reported mean cannot be attributed to either montelukast regimen or placebo. The record gives no treatment-effect estimate, confidence interval, or p-value. It therefore does not establish whether the groups differed or how precisely any difference could be estimated.

The study purpose also mentions esophageal tightening, but the registry posts no outcome measure for that feature. The available endpoint is a count of eosinophils, not a direct measure of tightening or a complete account of symptoms and function. The record cannot answer the full clinical question described in its summary.

These limitations do not demonstrate that the treatments were equivalent, beneficial, or ineffective. They mean that the available registry results are insufficient for a comparative conclusion. The adverse-event counts are similarly limited by unknown assignment and the absence of treatment-specific denominators. The record is informative about the study's design and the data that were posted, but not about a comparative treatment effect.

13. Why This Trial Matters Statistically

This trial illustrates a key distinction between a randomized design and an interpretable randomized comparison. Randomization can support causal comparisons when outcomes can be linked to assigned groups. Here, the registry describes randomization and masking, but reports that participants were not unblinded and that arm membership is unknown. The primary outcome summary consequently cannot be separated into treatment groups.

The example also shows why an overall mean and range do not answer a comparative research question. They describe the values in a summarized group, but do not show what happened under each intervention. A mean without an identified arm and a treatment contrast cannot estimate the difference between montelukast and placebo. Nor does the range quantify the uncertainty around such a contrast.

There is also a difference between a biological endpoint and the broader clinical question. Eosinophil count is the posted primary measure, while the study description additionally refers to esophageal tightening. When a stated goal has no corresponding reported endpoint, the record cannot establish results for that goal, regardless of what the listed endpoint measures.

Finally, sparse reporting requires careful separation of unknown information from negative findings. The treatment-specific outcome, allocation counts, analysis methods, and death information are not available in the registry record. Those gaps should not be converted into claims of no difference, no risk, or no effect. Reading the record precisely means describing its observable statistics while keeping the unresolved comparisons explicit.

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 trial-specific limitation is that participants were not unblinded, with the registry explaining that enrollment was insufficient to create generalizable knowledge. The number in each arm is consequently unknown, and the primary endpoint is summarized only for subjects whose arm is unknown. This prevents interpretation of the planned comparison across montelukast and placebo.

Enrollment is listed as 4, and the study status is terminated. The record gives no treatment-specific outcome values, no treatment-effect estimate, and no confidence interval or p-value. It also does not report an analysis model, the planned analysis population, or procedures for missing data. These omissions prevent readers from assessing how the trial intended to estimate a comparison or how robust an estimate might have been.

The single posted outcome measure is eosinophil count. Although the study purpose mentions esophageal tightening, an outcome measure addressing tightening is not reported. The record therefore does not provide results for that part of the stated purpose. It also does not report a standard deviation for the eosinophil count; the given full range does not fill that gap or quantify precision.

Safety interpretation is constrained by unknown assignments. The record includes adverse-event counts for an unknown-arm group but no treatment-specific counts, and deaths are not reported in the registry record. The record does not describe the reasons for termination beyond its statement about insufficient enrollment to create generalizable knowledge. Further explanations of recruitment, conduct, or analysis are not reported in the registry record.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: A Trial of Montelukast for Maintenance Therapy of Eosinophilic Esophagitis in Children — NCT01458418.
• Registry record last updated 2020-10-30; status last verified 2020-10.
• Results first posted 2020-10-30.
• 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.

18. Record Summary

The registry describes a randomized, quadruple-masked, parallel trial of montelukast 10 mg/day, montelukast 5mg/day, and placebo in children with eosinophilic esophagitis. The listed sponsor is Children's Mercy Hospital Kansas City. Enrollment is 4, and the study status is terminated.

The primary endpoint is eosinophil count in the mid esophagus after 12 weeks of therapy. The posted summary for subjects whose arm is unknown reports a mean of 8 eosinophils/HPF and a full range from 2 to 20. The record says participants were not unblinded because enrollment was insufficient to create generalizable knowledge; arm-specific outcome values and a treatment comparison are not reported.

The registry lists no analyses and no secondary outcomes. Its adverse-event entries do not permit treatment-specific safety comparisons, and death information is not reported in the registry record. Overall, the record describes the trial and a limited endpoint summary, but does not establish comparative efficacy or safety.