Treatment of Meibomian Gland Dysfunction Prior to Cataract Surgery: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Treatment of Meibomian Gland Dysfunction Prior to Cataract Surgery (NCT01808560), a randomized na study. The primary outcome is Mean Change in Total Meibomian Gland Score.

2. Clinical Question

3. Trial Design

This was a randomized, unmasked, crossover-model pilot study. The registry reports enrollment of 34 and gives a status of completed. The study began in 2013-03, with primary completion in 2015-01. Its phase is listed as NA. The sponsor was TearScience, Inc.

The primary comparison was between people assigned to LipiFlow pre-treatment and people assigned to an untreated control group. Pre-treatment consisted of a 12-minute LipiFlow System treatment in both eyes one month before cataract surgery. The control group received no MGD treatment before surgery. The registry says control-group participants received a 12-minute crossover treatment in both eyes three months after surgery.

The crossover arrangement matters when interpreting which treatment contrast the posted result represents. The primary endpoint is explicitly described as a comparison of the pre-treatment and untreated groups from baseline to three months after surgery. The record posts no primary endpoint result for the LipiFlow post-treatment group. It therefore does not establish a result for the later crossover treatment from the values reported for the primary comparison.

Masking is listed as none. The registry does not report details of sequence generation, allocation concealment, adherence, or whether outcome assessors were masked. The primary endpoint population is identified as the intent-to-treat population of all randomized subjects. The analysis-specific group counts in the posted results differ from the safety denominators and the enrollment figure; the registry record does not explain these differences. The study is identified as a pilot feasibility study, but its original sample-size rationale 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.

Randomization and Treatment Allocation

1-Month After Cataract Surgery

3-Month After Cataract Surgery

Crossover Treatment

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 Total Meibomian Gland Score, measured over Baseline and 3 Months post-surgery.

The primary endpoint is the mean change in total meibomian gland score from baseline to three months after cataract surgery. It compares the LipiFlow pre-treatment group with the untreated control group. The endpoint is intended to capture change in the function of the meibomian glands, assessed through the characteristics of secretions from gland openings along the lower eyelid.

The examination used a handheld Meibomian Gland Evaluator under a slit-lamp biomicroscope. The device applies a standardized amount of pressure, which the registry describes as a way to support consistent measurement. Fifteen glands were evaluated across the temporal, central, and nasal regions of the lower eyelid. Each gland was graded from 0 to 3 according to secretion characteristics: 3 for clear liquid, 2 for cloudy liquid, 1 for inspissated or toothpaste-like secretion, and 0 for no secretion. The total score is the sum of those grades and ranges from 0 to 45.

As a continuous measure, the endpoint summarizes the average change across the group rather than the number of people who crossed a threshold. It is used to assess the stated gland-function question, but it is not itself a direct measure of visual quality, ocular comfort, or cataract-surgery success. The registry does not report a clinically important threshold for interpreting a change in score.

7. Results Posted in the Registry

• LipiFlow Pre-treatment: mean 12.80 versus 1.90 on control; difference 10.90 (95% CI 5.14 to 16.66), standardized effect d = 1.28; z-test p = <0.001.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Mean Change in Dry Eye Questionnaire Score From Baseline at 3 Months

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • LipiFlow Pre-treatment: mean -3.40 versus 1.20 on control; difference -4.60 (95% CI -8.59 to -0.61), standardized effect d = -0.83; z-test p = 0.024.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for LipiFlow Pre-treatment than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The registry posts one secondary outcome: mean change in the dry-eye questionnaire score from baseline to three months after surgery. The direction of its group contrast differs from the primary gland-score contrast: the pre-treatment group had a negative mean change, while the untreated group had a positive mean change. The unadjusted comparison is reported with a p-value of 0.024 and a confidence interval that does not include zero. These findings describe the posted questionnaire result; they do not establish that the symptom measure and gland score track the same underlying experience or mechanism.

The secondary outcome complements the primary endpoint because it concerns questionnaire score rather than the gland examination. It is relevant to the study's stated interest in ocular comfort, but the registry does not report a threshold for meaningful change or detail on questionnaire interpretation. A difference in an objective-style gland assessment and a difference in questionnaire score are distinct kinds of evidence and should not be collapsed into a single conclusion.

Unadjusted secondary results generally carry less weight when the testing strategy and multiplicity handling are not documented. Considering multiple outcomes can create opportunities for chance findings, and the record lists six outcome measures while only one secondary outcome is posted in the registry record for this analysis. The registry record does not report a prespecified hierarchy, correction, or comprehensive outcome analysis. The secondary result is therefore best treated as supportive and exploratory rather than as independent confirmation of the primary result. No additional secondary outcome results should be inferred from the fact that the record lists other outcome measures.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Mean Change in Dry Eye Questionnaire Score From Baseline at 3 Months

9. Statistical Interpretation

10. Statistical Methods Explained

What does a mean change represent?

The endpoint is continuous: each person has a score at baseline and a score at follow-up, and the reported group result is the average change. A mean difference compares the average change in the pre-treatment group with the average change in the untreated group. In this record, the mean changes are 12.8 and 1.9, and the reported between-group mean difference is 10.9. A mean summarizes the group but can be influenced by unusually high or low observations. The registry does not post individual-level observations or distribution plots in the registry record for interpretation.

The gland score is built by adding grades for assessed glands. Because the total score is a sum of ordinal grades, treating its change as a continuous outcome assumes that an average difference is a useful summary of the scale. The registry identifies the endpoint as a mean and reports standard deviations, but it does not explain the rationale for the statistical model or discuss the scale's measurement properties.

What does the confidence interval contribute?

A confidence interval expresses uncertainty around an estimated effect under the model and sampling assumptions used to calculate it. The primary interval is 5.143 to 16.66 for the mean difference. It describes a range of values compatible with the observed data and method; it is not a range containing a known proportion of individual participants' responses. The interval is above zero, while its width also makes clear that the point estimate is not exact. General statistical interpretation should not be confused with certainty about the true effect.

The record reports standard deviations for each group and a pooled standard deviation used in the recalculated comparison. Standard deviation describes variation among observations within a group; it is not the uncertainty of the group mean. The uncertainty of a mean difference also depends on the amount of information in each group. The endpoint group counts are 17 and 13. The registry's posted interval and p-value are unadjusted, and the record does not describe a model with covariate adjustment.

How should the p-value be read?

The primary unadjusted p-value is <0.001. In general, a p-value measures how incompatible the observed data are with a specified null model, assuming the model and analysis conditions hold. It is not the probability that the null hypothesis is true, nor does it measure the probability that the result is due to chance in a broad, unconditional sense. It also does not tell whether a difference is important to patients. The confidence interval and the outcome's meaning provide complementary information about effect size and precision.

The record gives no multiplicity adjustment for the primary or secondary comparisons. If several outcomes or comparisons are examined, repeated testing can increase the chance of at least one apparently unusual result under null effects. The registry does not report a prespecified hierarchy or analysis plan to show how the primary and secondary outcomes were prioritized. This makes the unadjusted secondary finding less conclusive than a result from a clearly documented, prespecified testing strategy.

What is the role of randomization and masking?

Randomization is intended to make treatment groups comparable on average with respect to measured and unmeasured baseline characteristics. It does not guarantee balance in a small study, and it cannot prevent bias arising from how participants are retained, outcomes are measured, or analyses are selected. The registry identifies the allocation as randomized but does not report the sequence-generation or concealment methods.

Masking is listed as none. In general, lack of masking can affect behavior, co-interventions, reporting of subjective outcomes, and assessment of outcomes, depending on how those outcomes are collected. The gland score has a described examination procedure, but the record does not say whether the person assessing the glands knew group assignment. It also does not report whether the dry-eye questionnaire was self-administered or how missing questionnaire responses were handled.

What does recalculation mean?

The posted mean difference and interval are presented as a simple unadjusted comparison based on the posted group summaries. This kind of recalculation is not a substitute for the sponsor's full analysis if that analysis is available; the registry record lists no reported analyses. The discrepancy indicator is false, so the record does not flag a conflict requiring preference for a sponsor analysis. Still, summary-data calculations cannot account for features that are absent from the record, such as covariate adjustment, repeated-measures modeling, or imputation.

Because the study used a crossover model, the later LipiFlow treatment for the control group is relevant to the study design. The posted primary comparison, however, is explicitly between pre-treatment and untreated control at the three-month post-surgery timeframe. The record does not post an outcome analysis of the post-treatment period, nor does it explain how any later crossover observations were modeled. The primary result should therefore be interpreted as the stated pre-treatment versus untreated comparison, not as a crossover estimate across treatment periods.

11. Confidence Intervals and Precision

The primary mean difference is 10.9 points, with a 95% confidence interval from 5.143 to 16.66. This interval is entirely above zero and indicates that the estimated difference is positive under the analysis used. Its span also signals uncertainty around the point estimate: the data do not identify the mean difference with exact precision. The confidence interval is more informative about plausible effect sizes than the p-value alone, while still depending on the assumptions underlying the calculation.

Group variability was substantial relative to the scale's possible range: the reported standard deviations were 9.9 for the pre-treatment group and 6.1 for the untreated group. The analysis counts were 17 and 13, respectively. In general, greater variability and limited information make estimates less precise. The registry does not report a clinically important difference threshold, so the interval cannot be classified as excluding effects that would be unimportant to patients.

The posted confidence interval is not a prediction interval for an individual's response, and it does not show how consistently participants improved. It concerns the estimated group mean difference. The registry record does not report individual-level distributions, subgroup intervals, or adjusted estimates that could clarify variation in response or the effect of baseline characteristics. The result is therefore reasonably precise about the direction of the estimated mean contrast in this dataset, but uncertainty remains about its magnitude and generalizability.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning figures are participants per arm, not total events. They are illustrations based on the observed effect, rather than the trial's original design calculation. The stated planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The assumed effect to detect is expressed as a stated fraction of the observed effect. The original trial's own sample-size assumptions and justification are not reported in the registry record.

For each assumed effect, the 80% power column gives the planned participants per arm under those assumptions to provide an 80% chance of detecting that effect at the stated significance level, if the assumptions hold. The 90% power column gives the corresponding planning count for a 90% chance. These are design calculations, not guarantees about a future study and not a measure of what this completed study achieved.

For the primary outcome, the planning entries are 10 participants per arm for 80% power and 13 for 90% power when planning for 100% of the observed effect. For 75% of the observed effect, they are 17 and 23. For 50% of the observed effect, they are 39 and 51. The corresponding entries for the secondary outcome are 24 and 31 at 100%, 41 and 55 at 75%, and 93 and 124 at 50%.

Smaller effects require more participants to distinguish them from random variation under the same planning assumptions. These calculations use the observed effect as their starting point, so they should be treated as planning illustrations rather than a claim that the original design was based on that effect. This is not observed power.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = 10.900; pooled standard deviation σ = 8.483.
2. n = 2 × (1.96 + 0.842)² × 8.483² / 10.900² = 9.5, rounded up to 10 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
delta <- 10.9                      # observed difference in means
sigma <- 8.48250299987316                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(1808560)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 10 (80% power), 13 (90% power)
 75% of observed effect: n per arm = 17 (80% power), 23 (90% power)
 50% of observed effect: n per arm = 39 (80% power), 51 (90% power)
power.t.test: 10.6 
Simulated power at n = 10 per arm: 0.778 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Multiplicity, Missing Data and Analysis Populations

The registry identifies an intent-to-treat population of all randomized subjects for the primary endpoint. This indicates the stated target population for that endpoint, but the posted analysis counts are 17 for pre-treatment and 13 for untreated control, while enrollment is listed as 34. Safety denominators are different again. The record does not explain why these counts differ or give a participant flow account that would show the timing and reasons for missing observations.

Handling of missing data is not reported in the registry record. It does not describe whether participants without a three-month assessment were excluded, whether missing scores were imputed, or whether sensitivity analyses were conducted. These details matter because missingness can bias a treatment comparison if the probability of having an observed outcome relates to treatment assignment or the outcome itself. Without the reasons and methods, the effect of missing data on the reported means cannot be assessed.

The registry lists six outcome measures, posts one secondary outcome in addition to the primary result, and reports no analyses in its analysis listing. It does not report how outcome multiplicity was addressed, whether the primary endpoint was selected before results were examined, or whether other analyses were conducted but not posted. The p-values are unadjusted. Consequently, the registry record does not establish a familywise testing strategy or identify which comparisons, if any, were considered confirmatory.

Randomization is reported, but sequence generation and allocation concealment are not reported in the registry record. Masking is listed as none, and the record does not clarify masking of outcome assessors. The crossover treatment is described, but the primary comparison is the pre-treatment group versus untreated control at three months after surgery. The handling of post-crossover data is not reported. These reporting gaps limit assessment of analysis populations, missingness, and the risk of bias.

15. 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 adverse-event denominators separately from the primary endpoint analysis counts. For LipiFlow pre-treatment, 18 were listed at risk for serious adverse events and 18 for other adverse events; the affected counts were zero in both categories, and deaths were zero. For untreated control, the corresponding at-risk counts were 16 and 16, with zero affected in both categories and zero deaths. For LipiFlow post-treatment, the at-risk counts were 13 and 13, with zero affected in both categories and zero deaths.

These figures mean that the registry lists no serious or other adverse events as affecting participants in the reported categories, and no deaths, during the recorded adverse-event reporting. They do not establish that risk is absent. The record does not report event terms, event-collection details, follow-up completeness, or an assessment of comparative safety. The absence of listed events in this pilot should therefore be read as limited descriptive safety information, not as evidence that uncommon harms cannot occur.

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

The posted primary result is a between-group difference in average change in a gland-function score over the stated period. It shows a larger mean increase in the pre-treatment group than in the untreated control group, with a 95% confidence interval above zero. The secondary questionnaire result also shows a group contrast, but it is a separate, unadjusted outcome and should not be treated as interchangeable with the primary endpoint.

The result does not establish the effect of LipiFlow treatment after surgery, because the record does not post a result for the later crossover period. Nor does a gland-score difference by itself demonstrate a change in visual quality, ocular comfort, cataract-surgery outcomes, or any other unreported endpoint. The registry's study purpose mentions several aims, but the primary result addresses only the specified gland score.

The estimates come from a pilot study with limited endpoint group counts and incomplete reporting of analysis details. The record does not report a clinically important threshold, missing-data methods, assessor masking, or a multiplicity plan. The confidence interval describes uncertainty in the estimated group mean difference under the analysis assumptions; it does not show the range of individual responses or establish how well the result applies beyond this study.

Accordingly, the posted data support a descriptive conclusion about the observed group contrast in this study, while leaving important questions about precision, bias, clinical relevance, and reproducibility. They should not be interpreted as a definitive statement about treatment benefit or lack of benefit. A statistical difference and a clinically meaningful difference are not the same thing, and the registry does not provide the threshold needed to judge that distinction.

17. Why This Trial Matters Statistically

This trial raises statistical issues common to small randomized pilot studies of procedural interventions. The randomized comparison is useful for organizing an estimate of a between-group contrast, while the modest analysis counts mean that the estimate can remain sensitive to sampling variability and missing observations. The posted confidence interval helps convey that uncertainty more fully than a significance test alone.

The endpoint also illustrates how measurement choices shape interpretation. A standardized gland examination produces a score intended to represent gland function, whereas a questionnaire measures reported dry-eye symptoms. These endpoints can contribute different information. Their results should not be combined conceptually without evidence that they represent the same construct or move together in a clinically meaningful way.

The crossover structure adds an important design consideration. The control group was scheduled to receive treatment after surgery, but the primary result compares pre-surgery treatment with no pre-surgery treatment at the three-month post-surgery assessment. Keeping the estimand clear prevents the primary contrast from being misread as a comparison of treatment timing across all periods. The lack of posted post-treatment results also limits what can be learned about the crossover phase.

Finally, the record highlights the value of transparent reporting: participant flow, missing-data methods, allocation procedures, assessor masking, multiplicity plans, and original sample-size assumptions all affect how confidently a result can be interpreted. The posted estimates can be described, but the missing methodological details constrain evaluation of robustness and generalizability.

18. 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 trial is identified as a pilot feasibility study, and the endpoint analysis counts are limited: 17 in the pre-treatment group and 13 in the untreated control group. The registry does not report the original sample-size calculation or its assumptions, so the adequacy of the planned enrollment for a targeted effect cannot be judged from the record.

The record lists no masking. It does not say whether gland-score assessors were masked to treatment assignment, how the randomized sequence was generated or concealed, or whether treatment delivery and follow-up were consistent. These omissions matter for an outcome based on an examination and for a study in which participants may know whether they received treatment.

There is no reported participant flow or explanation of the difference between enrollment, endpoint analysis counts, and adverse-event denominators. Missing-data methods and sensitivity analyses are not reported. The treatment crossover is described, but outcomes for the post-treatment period are not posted, and the record does not explain how later observations were handled.

The registry also does not report a multiplicity strategy, clinically important thresholds for the continuous measures, or a detailed statistical analysis plan. Only one secondary outcome is posted among six outcome measures listed. These gaps limit interpretation of the secondary finding and prevent assessment of whether results were selected from a broader set of analyses. Finally, the primary endpoint is a gland score; the record does not establish how strongly it relates to visual quality, ocular comfort, or other patient-relevant outcomes.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Treatment of Meibomian Gland Dysfunction Prior to Cataract Surgery — NCT01808560.
• Registry record last updated 2018-09-04; status last verified 2018-08.
• Results first posted 2018-05-14.
• 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.

22. Record Summary

NCT01808560 is a completed randomized, unmasked pilot study with a crossover model. It enrolled 34 and compared LipiFlow treatment one month before cataract surgery with no MGD treatment before surgery; the control group was scheduled for crossover treatment three months after surgery.

The primary outcome was mean change in total meibomian gland score from baseline to three months after surgery. The posted comparison favored a larger mean increase in the pre-treatment group, with a mean difference of 10.9 and a 95% confidence interval of 5.143 to 16.66. One secondary dry-eye questionnaire outcome and adverse-event counts are also reported. The registry does not report the original sample-size rationale, missing-data methods, multiplicity handling, or results for the post-treatment period.