Safety and Tolerability of Artificial Tears in Dry Eye Subjects: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Safety and Tolerability of Artificial Tears in Dry Eye Subjects (NCT00932477), a randomized na study. The primary outcome is Tolerability Questionnaire Mean Scores at 1 Week.

2. Clinical Question

The central statistical question is whether participant-reported tolerability differed among products after each product was used. Because each participant was to receive every product, the intended comparison is within participants rather than a comparison of separate groups that each received only one product. That feature can be useful: participant characteristics that remain stable across product periods can be accounted for by comparing each participant with themself.

At the same time, product order and the timing of assessment can matter in a crossover study. The registry reports random assignment of product order and one week of use per product before switching. It does not report further details about the sequence assignments, the precise timing of questionnaire completion, or how period and order effects were handled in the analysis. Those details are important for interpreting a product comparison.

The endpoint is a questionnaire with eight tolerability questions on continuous visual analog scales from 0-100. Its questions do not all use the same direction: the first four measure increasing tolerability, with 0=worst and 100=best, while the second set measures decreasing tolerability, with 0=best and 100=worst. The endpoint description therefore identifies questions with opposite score directions. The registry does not state how any combined score was constructed or report results by question.

The registry lists three secondary outcomes: patients with at least one severity-grade increase in biomicroscopy findings at one week, best-corrected visual acuity status at one week, and the number of ophthalmic adverse events at one week. Results for these outcomes are not reported in the registry record. The study is marked completed, but completion alone does not supply evidence about the size, precision, or direction of differences in tolerability or these secondary measures.

3. Trial Design

This was a randomized, double-masked, crossover study of artificial tears in participants with dry eye syndrome. Enrollment was 47. The study summary says every participant would receive all three products, in a randomly assigned order, using one product at a time for one week before switching to the next. In a crossover design, participants serve as their own comparison across treatment periods; the products are not assigned to permanently separate groups.

The three products were Artificial Tear Formulation 1, Artificial Tear Formulation 2, and a glycerin and polysorbate 80 based artificial tear. The first two formulations are described as based on carboxymethylcellulose sodium, glycerin, and polysorbate 80. The third is described as glycerin and polysorbate 80 based and as a currently available artificial tear in the study summary. The registry does not give further product details in the record summarized for this analysis.

Random assignment of product order is relevant because product comparisons in crossover studies may be affected by the order in which treatments are received. Double masking is also part of the recorded design, but details of who was masked, how masking was maintained, or whether participants and assessors could identify the products are not reported in the registry record. No washout period is described in the study summary. Its presence or absence should not be inferred from the switching schedule alone.

The primary outcome was the mean Tolerability Questionnaire score at one week. The primary population is described as the safety population, consisting of patients who started the study, were randomized, and received treatment. The record also identifies three secondary outcomes covering biomicroscopy findings, best-corrected visual acuity, and ophthalmic adverse events, each at one week. The study status is completed. The record lists Allergan as sponsor.

The study ran from 2009-08 to primary completion in 2009-09, according to the dates recorded. The record identifies the phase as NA. It does not report the planned sample-size justification, the intended statistical model, the handling of period or sequence effects, or the details of treatment allocation beyond randomization of order. These omissions limit what can be said about how the planned analysis addressed the crossover structure.

A crossover design can be statistically efficient when within-participant responses are correlated, because a comparison can remove some stable differences between participants. However, the analysis needs to respect the repeated observations contributed by each participant. A method that treats all treatment-period observations as independent could misstate uncertainty. The registry record does not establish which method was used, and it does not report analysis estimates or uncertainty intervals.

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: This is a 3-treatment, 3-period, 6-sequence crossover study. Each subject received all 3 products in a randomly assigned order. The subject used 1 product at a time for a duration of 1 week before switching to the next assigned product (e.g., with treatments A, B, and C, the sequences were ABC, ACB, BAC, BCA, CAB, and CBA).

Treatment 1

Treatment 2

Treatment 3

Analysis population for the primary outcome, as defined in the registry: Safety population, which consisted of all patients who started the study (randomized) and received treatment.

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 Tolerability Questionnaire Mean Scores at 1 Week, measured over 1 Week.

The primary endpoint was the mean score on the Tolerability Questionnaire at one week. The questionnaire contains eight questions about selected performance measures, each scored on a continuous visual analog scale from 0-100. For the first four questions, higher values represent better tolerability: 0=worst and 100=best. For the second set of four questions, lower values represent better tolerability: 0=best and 100=worst.

A participant-reported tolerability measure can capture experiences that may not be fully described by an examination or an event count. It provides a structured way to record perceptions of product performance, and continuous scores can preserve gradations in those perceptions rather than reducing them immediately to categories. This is a general rationale for such measures, not evidence that this questionnaire is validated or that a particular scoring method was used in this study; those details are not reported in the registry record.

The opposite score directions matter when interpreting means. For the first four questions, a higher mean would indicate better reported tolerability; for the second set, a lower mean would indicate better reported tolerability. A pooled or total score would require a defined scoring and direction-harmonization approach. The registry identifies mean scores as the endpoint but does not explain whether results were reported separately for each question, combined, or transformed before calculating a mean.

No endpoint means, standard deviations, treatment contrasts, or confidence intervals are reported in the registry record. Consequently, the endpoint definition identifies what was intended to be measured but does not establish the observed tolerability scores or the precision of any comparison.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Statistical interpretation depends on both the study structure and the form of the outcome. This study is recorded as randomized, double-masked, and crossover, and the primary outcome is a continuous questionnaire score. The record states that participants would use every product in randomly assigned order, one at a time for one week before switching. These design details describe the intended comparisons; they do not specify the statistical model used to analyze them.

What does a mean describe?

A mean is a measure of the central value of a set of observations. For a continuous questionnaire score, it can summarize the average recorded response for a treatment and assessment time. A mean does not show how responses vary among participants, whether responses are clustered near the center or spread out, or whether some participants report very different experiences. The endpoint lists the mean as the summary measure and the standard deviation as its dispersion measure, but neither treatment-specific means nor standard deviations are reported in the registry record.

The standard deviation describes the spread of individual values around a mean. It is not the same as the uncertainty in the estimated mean. A confidence interval, when reported, describes a range of values compatible with the data and the statistical method under stated assumptions. A narrow interval indicates greater precision than a wide interval, while neither interval width nor a p-value alone determines whether a difference is important in practice. No confidence intervals or p-values for the primary endpoint are reported in this registry record.

Why does crossover structure matter?

In a crossover study, observations from the same participant are related. Stable participant characteristics can affect that participant’s responses under every product, so within-person comparisons can remove some of that between-person variation. This can improve precision relative to treating the same observations as unrelated, provided the analysis reflects the repeated-measures structure and the design assumptions are reasonable.

Repeated observations are not independent because they come from the same person. In general, a statistical model for crossover data may account for treatment, period, sequence, and participant-level variation. Which terms are appropriate depends on the design and analysis plan. The registry does not report the model, covariates, or other analysis details, so it is not possible to state which effects were included or how uncertainty was estimated.

Period effects are changes associated with the time or order of a treatment period rather than the product itself. Carryover refers to an effect of a prior treatment that persists into a later period. Randomizing product order can help balance order-related influences across participants, but it does not, by itself, establish that period or carryover effects are absent. The record describes random assignment of order and sequential one-week product use. It does not describe a washout period or report how period and carryover were assessed.

What does masking contribute?

Masking can reduce the influence of expectations on reported experiences and on assessment. The registry labels the study double-masked. The record does not specify which individuals were masked, how masking was maintained, or whether product characteristics could have revealed treatment identity. The label should therefore be treated as a design feature recorded in the registry, without inferring details beyond that description.

How should the score scale be read?

The questionnaire has eight tolerability questions, each using a continuous visual analog scale from 0-100. The first four questions are oriented so that higher scores indicate better tolerability, with 0=worst and 100=best. The second set is oriented in the opposite direction, with 0=best and 100=worst. This means that the direction of a favorable comparison depends on which question is being considered.

When scales have opposite orientations, a combined score usually needs a clearly described transformation or other scoring rule so that values have a consistent interpretation. The registry does not report such a rule, whether questions were analyzed separately, or how the stated mean was calculated. Without that information, readers should not assume a particular aggregation or scoring procedure.

Why are uncertainty and effect size both needed?

A treatment contrast describes how the estimated outcome differs between treatments. Its confidence interval communicates uncertainty around that estimate. A p-value, if reported, addresses compatibility of the data with a specified null hypothesis under a statistical model; it does not measure the probability that a treatment is effective, nor does it describe the size of a difference. Estimates and intervals are needed to judge the magnitude and precision of results. The registry record reports no primary contrasts, uncertainty intervals, or p-values.

For a crossover comparison, uncertainty should reflect that treatment responses from a given participant are correlated. A calculation that assumes unrelated observations may give an inaccurate standard error. This is a general statistical point, not a description of the analysis used in this study. The registry record does not report an analysis model or enough outcome data to reconstruct one.

What does the available safety count represent?

The safety listing gives counts of participants at risk and affected for serious and other adverse events for each treatment group. Counts are direct descriptions of what is listed, while comparative safety estimates require a defined denominator, treatment exposure context, and an analysis that accounts for the crossover structure. The record lists 47 at risk for each product and zero affected for both serious and other adverse events. It does not report further event details in the listed safety information, and deaths are not stated in the values.

In general, zero recorded events in a study does not establish that an event cannot occur. The observation is limited to the participants, treatment exposure, and follow-up represented in the record. Here, the one-week treatment periods and the listed at-risk counts describe the posted context; the record does not provide a basis for extrapolating beyond it.

What can be said about the statistical plan?

The registry record does not report the planned sample-size justification, analysis model, treatment contrast strategy, handling of missing observations, interim analyses, or multiplicity procedure. These items should not be inferred from the endpoint description or design label. The record does state that the primary measure is a mean with standard deviation as its dispersion, but that alone does not identify the hypothesis test or model. As no primary result values are posted, no independent reanalysis of the outcome is possible from the reported information.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists four outcome measures: the primary Tolerability Questionnaire mean score at one week and three secondary outcomes at one week. When a study evaluates multiple outcomes, questions, treatment contrasts, or time points, the chance of at least one apparently notable finding can increase if each is tested separately without adjustment. A multiplicity plan may define a primary comparison, control an overall error rate, or identify analyses as exploratory. This is a general statistical consideration; the registry record does not report a multiplicity procedure.

The primary endpoint itself includes eight questionnaire questions. The record describes the direction of the first four and the second set, but does not say whether the questions were analyzed individually, combined, or treated as separate inferential tests. It also does not report which treatment contrasts were planned among the three products. Therefore, the scope of any multiplicity issue cannot be reconstructed from the registry record.

Missing-data handling is also not reported. The endpoint population is described as the safety population, consisting of all patients who started the study, were randomized, and received treatment. The record does not say how many participants completed each treatment period or questionnaire, whether any endpoint observations were missing, or how missing values would have been handled in the analysis. It also does not report whether analyses included all randomized participants or a different analysis population for the primary outcome.

The registry says that each participant would receive all products, but the available record does not describe adherence, withdrawals, exclusions, or the number of observations contributing to each product-specific estimate. These details matter in crossover studies because incomplete treatment sequences can affect within-participant comparisons. Their absence does not establish that data were missing; it means the record does not describe their status or treatment in the analysis.

No interim analyses, stopping rules, or statistical analysis plan are reported in the registry record. Their presence or absence cannot be inferred. The posted results also include no primary estimates or tests from which to assess how outcomes were analyzed.

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 5% reporting threshold.

For each of the three products, the registry lists 47 participants at risk for serious adverse events and 0 affected. It also lists 47 at risk for other adverse events and 0 affected. These figures are the recorded counts for the respective treatment groups. They should be read as counts in the posted safety information, not as proof that the risk of an event is absent.

The study is a crossover trial in which participants were to receive all three products in a randomly assigned order. Therefore, the product-specific at-risk counts should not be interpreted as three independent parallel groups. The safety record does not provide event-level descriptions in the listed entries, and the death values are not stated. It also does not give additional detail about exposure duration beyond the study summary’s one-week use periods.

Zero affected participants is a descriptive result for the recorded safety categories and follow-up. In general, the absence of recorded events in a study does not demonstrate that an event is impossible, particularly when observations cover a limited population and period. These counts do not provide comparative estimates or confidence intervals, and they do not replace the missing tolerability questionnaire results. Broader safety conclusions are not reported in the registry record.

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

The registry record supports several statements about what was planned and what was posted. It describes a randomized, double-masked crossover study in 47 participants with dry eye syndrome, with each participant intended to receive three artificial tears in randomly assigned order. Product use was planned in one-week periods. The primary endpoint was mean Tolerability Questionnaire scores at one week, and the safety listing records zero affected by serious and other adverse events among 47 at risk for each product.

Those details do not answer the primary comparative question. The record does not report the treatment-specific questionnaire means, variability, contrasts, confidence intervals, or p-values. As a result, it cannot show the direction or magnitude of any score differences, or how precisely such differences were estimated. It also does not report results for the secondary outcomes concerning biomicroscopy, best-corrected visual acuity, or ophthalmic adverse events.

The endpoint description alone does not establish how the questionnaire was scored. Since its first four questions favor higher values and its second set favors lower values, interpretation of a combined score depends on the scoring method. That method is not reported. Nor does the record state the statistical model used to account for repeated observations, treatment order, or period effects in the crossover design.

The safety counts are limited to the entries and treatment context recorded. They describe zero affected participants in the listed serious and other adverse-event categories; they do not prove that no event could occur, or establish comparative safety beyond the study’s recorded exposure and follow-up. Deaths are not stated in the listed values.

Accordingly, the posted information is useful for identifying the study structure, population definition, measures, and recorded safety counts. It does not establish that products were equivalent, that one was better tolerated, or that any result was statistically significant. Missing outcome results should not be interpreted as evidence of no difference. More broadly, the record does not support clinical recommendations or conclusions beyond the registry entries it reports.

13. Why This Trial Matters Statistically

This trial illustrates why design labels and outcome summaries need to be read together. A crossover study has a different comparison structure from a parallel-group study: participants receive each product, and their responses across periods are related. The design can make within-person comparisons informative, but valid uncertainty estimates depend on an analysis that respects this dependence and addresses relevant features such as period and order.

It also illustrates that naming a continuous endpoint is not enough to interpret results. The Tolerability Questionnaire uses a 0-100 scale, but its first four questions and second set use opposite directions. A score’s meaning therefore depends on the item and on any aggregation rule. The record explains the directions but does not report the scoring procedure or results. Statistical communication needs both a defined measure and a transparent account of how that measure was summarized.

The safety listing gives concrete counts, but those counts answer a narrower question than a comparative treatment analysis. Zero affected participants in the listed categories is not the same as evidence of zero underlying risk. The crossover design also means the at-risk counts should be understood in the context of participants receiving multiple products, rather than as three unrelated groups.

Most importantly, the record distinguishes a completed study from a result that can be evaluated statistically. The primary endpoint is identified, yet estimates and uncertainty are absent. Without them, readers cannot assess the size or precision of the treatment comparisons. This is a useful reminder that interpreting clinical evidence requires the outcome data, the analysis approach, and the uncertainty around estimates—not only the planned question and trial design.

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 limitation of the available record is that it does not report the primary endpoint results. Treatment-specific means, standard deviations, contrasts, confidence intervals, and p-values are not available in the registry record. The primary comparison therefore cannot be evaluated for direction, magnitude, or precision. The study’s completed status does not resolve this information gap.

The endpoint scoring description leaves important interpretive details open. The questionnaire contains eight questions on 0-100 visual analog scales, with opposite favorable directions across the two sets. The record does not explain whether the questions were analyzed separately, whether any were combined, or how a common direction was established for an overall mean. It also does not report validation or other measurement properties of the questionnaire.

Details needed to assess the crossover analysis are not reported. The record describes randomized product order and one-week use periods, but it does not describe a washout period, the sequence distribution, handling of period effects, or assessment of carryover. It does not identify the statistical model or the assumptions used to estimate treatment differences. These omissions prevent assessment of whether the analysis reflected the repeated observations within participants.

Information about completion and missing data is also absent. The registry defines the safety population but does not report the number of completed periods, questionnaire completion by product, withdrawals, adherence, exclusions, or missing-data methods. The planned sample-size justification and multiplicity procedure are not reported either. The registry lists three secondary outcomes but provides no results for them in the information available.

Safety interpretation is limited to the posted counts: 47 at risk and 0 affected for serious and other adverse events for each product. Further event details are not reported in the listed safety entries, and deaths are not stated. These data do not establish comparative safety or extend beyond the recorded study exposure and follow-up.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Safety and Tolerability of Artificial Tears in Dry Eye Subjects — NCT00932477.
• Registry record last updated 2015-08-26; status last verified 2015-08.
• Results first posted 2011-12-13.
• 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

NCT00932477 is a completed randomized, double-masked crossover study of artificial tears in participants with dry eye syndrome. Enrollment was 47. The study summary states that participants would receive all three products in randomly assigned order, using each for one week before switching. The products were two carboxymethylcellulose sodium, glycerin, and polysorbate 80 based formulations and a glycerin and polysorbate 80 based artificial tear.

The primary outcome was mean Tolerability Questionnaire scores at one week. The eight visual analog scale questions use opposite score directions across the two sets. Three secondary outcomes concern biomicroscopy findings, best-corrected visual acuity, and ophthalmic adverse events. The registry record does not report the primary or secondary outcome results, analysis estimates, or statistical methods.

For each product, the safety listing reports 47 at risk and 0 affected by serious adverse events, and 47 at risk and 0 affected by other adverse events. Deaths are not stated in the listed values. The available record describes the design and counts but does not permit evaluation of comparative tolerability or statistical precision.