Treatment of Meibomian Gland Dysfunction and Dry Eye in Contact Lens Wearers: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Treatment of Meibomian Gland Dysfunction and Dry Eye in Contact Lens Wearers (NCT02102464), a randomized na study. The primary outcome is Mean Change in Meibomian Gland Score From Baseline at 3 Months.

2. Clinical Question

The question is about a measured score over a specified period, not about every aspect of dry eye or contact lens use. The registry also lists a secondary outcome concerning mean change in dry eye questionnaire score from baseline at 3 months, but the primary result is the gland score comparison. The study is identified as a pilot study. Its record does not report details that would establish how broadly the findings apply to other contact lens wearers or other treatment settings.

3. Trial Design

This was a randomized, unmasked, crossover study sponsored by TearScience, Inc. The registry lists an enrollment of 55 and a status of completed. The study began in 2014-04, with primary completion in 2015-08. Its phase is listed as NA. The study title concerns treatment of meibomian gland dysfunction and dry eye in contact lens wearers, and the official title describes it as a pilot study of treatment for meibomian gland dysfunction and evaporative dry eye in contact lens wearers.

The main comparison was between LipiFlow and an untreated control. The LipiFlow intervention was described as a single 12-minute treatment; the control received no intervention during the initial comparison period. After 3 months, the untreated control group was scheduled to cross over to LipiFlow treatment. The primary endpoint is defined as change from baseline to 3 months, which is the time point at which the initial treatment and untreated-control groups are compared.

The endpoint results list 29 subjects in the LipiFlow group and 25 in the untreated control group. Those are the group sizes attached to the reported primary outcome estimates, not a reconciliation of the overall enrollment. The relationship between overall enrollment and the outcome group counts, including any reasons for missing outcome data, is not reported in the registry record. Safety reporting lists separate at-risk counts for the LipiFlow, untreated control, and crossover LipiFlow groups; those entries should not be treated as interchangeable with the primary endpoint group sizes.

Randomization can help make assigned groups comparable on average, while an unmasked design means that masking was not used. The registry does not report the randomization procedure, allocation concealment, or details about how outcome assessment was conducted. It also does not describe the sequence or timing beyond the stated crossover after 3 months. These omissions limit what can be concluded about safeguards against bias and about analyses of the crossover period.

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.

Randomized Study

Reasons for not completing this period:

Crossover LipiFlow Treatment of Control

Analysis population for the primary outcome, as defined in the registry: Intent to Treat (ITT) Population of all randomized subjects.

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 Mean Change in Meibomian Gland Score From Baseline at 3 Months, measured over 3 Months.

The primary endpoint was mean change in meibomian gland score from baseline to 3 months. The score was based on secretion characteristics of 15 glands along the lower eyelid: five glands each in the temporal, central, and nasal regions. Each gland was graded from 0 to 3. A grade of 3 represented clear liquid, 2 cloudy liquid, 1 inspissated or toothpaste-like secretion, and 0 no secretion. The total was the sum across the 15 glands, with a possible scale range from 0 to 45. A higher score reflects less meibomian gland dysfunction.

This measure was chosen to assess gland function, which was a stated objective of the study. A baseline-to-follow-up change can describe how the measured score shifted over the specified period, while comparison with an untreated group provides a reference for changes that also occurred without the intervention. The score is a clinical measure defined by the registry; the record does not report evidence about its measurement reliability, or how score assessments were standardized across assessors.

The reported group mean changes were 12.4 for LipiFlow, with a 95% confidence interval from 10.0 to 14.8, and 1.4 for untreated control, with a 95% confidence interval from -0.4 to 3.3. The registry’s reported between-group mean difference was 11.0, with a 95% confidence interval from 8.0 to 14.0. The record also lists a dry eye questionnaire score as a secondary outcome, but that is a different measure from the primary gland score.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Statistical methods help distinguish the size of an estimated contrast from the uncertainty around it. The registry identifies the primary endpoint as a continuous measure summarized by a mean. For this kind of outcome, a group mean describes the average score change in the group; it does not show how individual participants’ scores were distributed. The record gives group means and confidence intervals but does not report a standard deviation or other measure of individual-level spread.

What does the confidence interval describe?

A confidence interval expresses uncertainty about an estimated quantity under the assumptions of the statistical method. The 95% confidence interval for the reported mean difference extends from 8.0 to 14.0. In repeated use of a method with its assumptions met, intervals constructed at this confidence level would cover the underlying parameter at the stated rate. The interval is not a range containing 95% of individual participants’ changes, and it is not a guarantee about the true value in this particular study.

The registry also gives confidence intervals for the arm-level mean changes: 10.0 to 14.8 for LipiFlow and -0.4 to 3.3 for untreated control. Each group interval concerns its own mean. The interval for the between-group difference is the direct uncertainty measure for the comparison. In general, one should use the confidence interval for the contrast rather than infer the contrast’s precision solely from separate group intervals.

How should the p-value be read?

The reported test is a two-sided t-test, and the p-value is < 0.001. In general, a p-value describes how incompatible the observed data, or data more extreme, would be with a specified null model if that model and the test assumptions held. It is not the probability that the null hypothesis is true, and it does not measure clinical importance. The registry does not report further detail about the assumptions checked for the t-test or adjustments to the analysis.

The test is described as two-sided, meaning that departures in either direction are considered in the test. The registry labels the analysis as a superiority comparison. That label indicates the type of statistical question being tested; it does not on its own establish the clinical importance of the estimated score difference. Interpretation still depends on the outcome’s meaning, the study design, and the precision of the estimate.

What does a crossover design add?

In a crossover design, participants may receive different conditions at different times. In this study, the untreated control group was scheduled to receive LipiFlow after 3 months. The primary endpoint compares the initial groups at 3 months, before that stated crossover treatment. This timing makes the initial randomized comparison distinct from any later within-person or crossover-period comparison. The registry does not report a statistical analysis of the crossover period, period effects, treatment sequence effects, or carryover effects. Those features should not be inferred from the design label alone.

Randomization establishes the intended allocation process, while masking concerns whether participants or study personnel were unaware of treatment assignment. The record says the study was randomized and unmasked. In general, randomization can reduce systematic differences between assigned groups, but it does not ensure equal groups in every feature, especially in a pilot study. Lack of masking can matter for outcomes influenced by expectations or subjective assessment. The registry does not report who assessed the gland score or whether assessors were independent of treatment delivery.

What does the registry leave unclear about the analysis?

The primary endpoint description refers to mean change from baseline to 3 months, while the analysis parameter is labeled “Mean Difference (Final Values).” These descriptions do not unambiguously establish whether the tested comparison used final scores or changes from baseline. The registry also does not report whether any covariates were included, how missing outcome values were handled, or whether the test used a pooled or unequal-variance form. No additional analytical assumptions should be attributed to the study without a registry description.

The record describes the primary population as the intent-to-treat population of all randomized subjects. In general, an intent-to-treat approach aims to analyze participants according to their randomized groups, regardless of treatment received. That principle does not specify how missing data were handled, and the registry does not report that detail. It also does not explain the relationship between overall enrollment and the group counts attached to the primary outcome estimates.

The registry lists six outcome measures, three analyses, and one secondary outcome. Multiple outcomes or analyses can create multiplicity: the more questions tested, the greater the chance of at least one small p-value by chance if no adjustment or testing strategy is used. The registry record does not report a multiplicity adjustment, a testing hierarchy, or how these outcomes relate to the three analyses. It is therefore not possible to determine from the record whether or how multiplicity was controlled.

The record includes a sample-size planning illustration based on the observed effect, with a two-sided significance level of 0.05, equal allocation, and a normal approximation; the effect to detect is expressed as a fraction of the observed effect. Such illustrations concern hypothetical planning for a future study. They are not observed power and do not show that this pilot study had a particular probability of detecting an effect. The record does not report the trial’s own sample-size justification or planning assumptions.

10. Analyses Reported by the Sponsor

The sponsor-reported primary analysis is a two-sided t-test. It gives a mean difference in final values of 11.0, with a 95% confidence interval from 8.0 to 14.0 and p < 0.001. This is the direct reported comparison to use when describing the sponsor’s result. The registry also reports the two group mean changes and their confidence intervals, but the analysis label “Mean Difference (Final Values)” is not fully aligned with the endpoint description of change from baseline. The record does not clarify the exact relationship between those descriptions.

A separate recalculated two-arm comparison is not reported as estimable in the registry record. The posted group summaries do not include all the information needed to reconstruct the sponsor’s stated test and its uncertainty. No discrepancy is identified between the sponsor’s reported analysis and a recalculation. As a result, the sponsor’s analysis is the available inferential comparison; there is no independent recalculated estimate to place alongside it.

In general, recalculations from summary statistics can be useful checks when the necessary quantities and definitions are available. They are unadjusted checks and may not reproduce an analysis that uses participant-level data or additional modeling. Here, no claim about such an alternative method is warranted. The reported confidence interval and p-value should be attributed to the registry’s sponsor-reported analysis, not to an independently reconstructed test.

11. Multiplicity, Missing Data and Analysis Populations

The registry lists six outcome measures, three analyses, and one posted secondary outcome: mean change in dry eye questionnaire score from baseline at 3 months. The primary endpoint is mean change in meibomian gland score from baseline at 3 months. The record does not describe a hierarchy among outcomes, a prespecified sequence of tests, or an adjustment for multiple comparisons. It is therefore not reported in the registry record whether multiplicity was addressed in the analysis.

Multiplicity matters because, when several hypotheses are tested, the chance of finding at least one apparently unusual result can increase if each test is judged separately. Adjustments or a prespecified testing strategy can address that issue, but no particular approach should be assumed. The primary result can be described as the reported result for its endpoint; the record does not establish how its inferential interpretation was coordinated with other outcomes or analyses.

The registry identifies the primary analysis population as the intent-to-treat population of all randomized subjects. It does not report the detailed analysis-population rules, the number of participants with an observed primary endpoint, or the reasons for any missing outcome data. Although the primary endpoint group counts are listed as 29 and 25, the record does not explain their relationship to the overall enrollment of 55. There is no basis in the record to infer which participants were missing data or why.

Missing data can affect estimates when the probability of an outcome being unavailable is related to participant characteristics or outcomes. General statistical approaches include analyzing observed cases, imputing missing values, or modeling the missingness process, but the registry does not state which approach was used. It also does not report sensitivity analyses assessing how alternative assumptions about missing values might affect the result. Consequently, the reported estimate should be read with that uncertainty about outcome completeness in mind.

The sample-size illustration is described as planning based on the observed effect, with specified general assumptions. It does not substitute for information about how the trial’s enrollment target was set. The registry does not report the original sample-size justification, any interim analysis plan, or details of a statistical analysis plan. Those items should be treated as not reported in the registry record rather than inferred from the posted result.

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.

The registry reports zero serious adverse events among the at-risk entries for LipiFlow, untreated control, and the crossover LipiFlow group. The corresponding serious-event at-risk counts are 29, 26, and 25. It also reports zero other adverse events for each of those entries and zero deaths. These are the safety figures recorded for the listed groups.

The safety entries are distinct from the primary efficacy outcome counts. In particular, the untreated control at-risk count in the safety reporting is 26, while its primary outcome group count is 25. The registry does not explain this difference, so the figures should be quoted within their own contexts rather than reconciled by assumption. The crossover entry represents a group recorded after the control group’s scheduled crossover to LipiFlow after 3 months.

Zero reported events in these entries does not establish that an event is impossible or that the intervention is free of risk. Safety counts describe what was recorded in this study and do not by themselves characterize event rates in other settings or over longer follow-up. The registry does not report the safety observation period, event-collection procedures, or detailed definitions beyond the serious and other adverse-event categories. No serious adverse-event terms are listed.

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

The primary result addresses a specific statistical question: the difference in mean meibomian gland score change from baseline to 3 months between a LipiFlow group and an untreated control group. The registry reports a mean difference of 11.0, with a 95% confidence interval from 8.0 to 14.0 and a two-sided p-value of < 0.001. The group mean changes were 12.4 for LipiFlow and 1.4 for untreated control. Because higher scores reflect less meibomian gland dysfunction, the reported direction corresponds to a larger increase in the score in the treatment group during the comparison period.

The result does not, by itself, answer whether the change is important to patients, because the registry does not report a threshold for clinical importance. Nor does it describe the full range of individual responses; group means can conceal variation among participants. The outcome is a score derived from gland secretion assessments, and it is not a direct measure of every dimension of dry eye, contact lens comfort, or daily function.

The result also does not establish what happens after the 3-month comparison. The untreated control group was scheduled to receive LipiFlow after 3 months, but the registry record does not report a later crossover analysis or longer-term results. The secondary dry eye questionnaire outcome is listed, but a result for that outcome is not included in the posted primary comparison details.

Interpretation is further limited by reporting details. The endpoint is described as mean change from baseline, while the reported analysis parameter says “Mean Difference (Final Values),” and the registry does not explain that distinction. It does not report the method used for missing data, multiplicity handling, or detailed analysis procedures. Thus, the reported estimate is best understood as the sponsor-reported comparison for this endpoint, rather than as a complete account of all possible analyses or outcomes.

Finally, the study is identified as a pilot study and enrolled 55. Pilot results can inform understanding of a measured contrast, but this record does not establish generalizability to populations or settings not described. The statistical result is evidence about the posted comparison under its analysis assumptions; it is not medical advice and does not settle broader questions beyond the registry’s stated endpoint and follow-up.

14. Why This Trial Matters Statistically

This study illustrates why defining the estimand and outcome precisely matters. The primary measure is not a general label such as “dry eye improvement”; it is a score assembled from secretion grades across glands in specified lower-eyelid regions. The registry explains how the scale is constructed and what a higher score means. That definition helps readers understand which aspect of gland function the analysis addresses and prevents the score from being mistaken for a comprehensive measure of symptoms or treatment experience.

It also demonstrates the value of separating group estimates from the contrast of interest. Each arm has a mean change and a confidence interval, while the sponsor’s analysis reports a confidence interval for the between-group mean difference. The latter directly quantifies uncertainty around the comparison. A p-value alone would not communicate that range of plausible contrast sizes or the scale on which the endpoint is measured.

The design raises important interpretation questions. Randomization and an untreated comparison group provide a structure for comparing the groups at 3 months, while the planned crossover after that point changes the interpretation of later observations. The record identifies the study as unmasked, which is relevant to possible measurement or expectation effects, especially when assessment procedures are not described. Reporting the timing of the comparison helps distinguish the initial parallel-group contrast from later crossover experience.

Finally, the result shows why registry completeness matters to statistical interpretation. The endpoint description and analysis label do not fully clarify whether the contrast is based on changes or final values. Details on missing data, multiplicity, and the relationship between enrollment and endpoint counts are also absent. These are not merely administrative details: they affect how confidently readers can connect the numerical result to the intended question. Clear reporting of these elements is essential for evaluating precision, bias, and the limits of inference.

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.

Several trial-specific limitations follow from what is not reported in the registry record. The endpoint is defined as mean change from baseline, but the sponsor’s analysis parameter is labeled “Mean Difference (Final Values).” The record does not clarify whether the comparison was performed on final scores or changes, or whether any adjustment for baseline values was used. This leaves uncertainty about the exact quantity represented by the reported contrast.

The registry gives an intent-to-treat description but does not explain how that population was operationalized, how missing primary outcome values were handled, or whether sensitivity analyses were performed. Overall enrollment is listed as 55, while the primary endpoint results have group counts of 29 and 25. The record does not explain the relationship between these counts. The safety entries use their own at-risk counts, including a different control count, and should not be used to fill in missing details about the efficacy analysis.

The study is unmasked, and the registry does not state who assessed gland secretion or what steps were taken to standardize assessment. It does not report allocation concealment, the randomization procedure, or assessor masking. The crossover schedule is described, but later-period analyses and handling of possible period or carryover effects are not reported. These omissions constrain evaluation of potential bias and of results beyond the initial 3-month comparison.

The registry lists multiple outcomes and analyses but does not report whether multiplicity was addressed. It also does not report a sample-size justification for the trial, interim analyses, or a detailed statistical analysis plan. The primary endpoint group intervals describe uncertainty in the mean estimates, but the record does not provide measures of individual-level variability. Finally, no clinical-importance threshold is reported, so statistical evidence for a nonzero contrast cannot be translated into a judgment about the importance of the score change.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Treatment of Meibomian Gland Dysfunction and Dry Eye in Contact Lens Wearers — NCT02102464.
• Registry record last updated 2017-10-18; status last verified 2017-09.
• Results first posted 2017-09-18.
• 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

NCT02102464 is a completed randomized, unmasked crossover pilot study sponsored by TearScience, Inc. It examined a single 12-minute LipiFlow treatment in contact lens wearers with meibomian gland dysfunction and evaporative dry eye, compared with an untreated control at 3 months. The untreated group was scheduled to receive crossover LipiFlow treatment after 3 months.

The primary endpoint was mean change in meibomian gland score from baseline at 3 months. Mean change was 12.4 for LipiFlow and 1.4 for untreated control. The sponsor-reported mean difference was 11.0, with a 95% confidence interval from 8.0 to 14.0 and a two-sided p-value of < 0.001. The score’s higher values reflect less gland dysfunction.

The record lists enrollment of 55 and describes the primary analysis population as intent to treat. It does not clarify the relationship between enrollment and the outcome group counts, or report detailed missing-data, multiplicity, or analysis-plan information. The statistical result is specific to the reported endpoint, comparison, and 3-month time point.