Safety And Effectiveness Of The TearCare® System In The Treatment Of The Signs And Symptoms Of Dry Eye Disease (SAHARA): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Safety And Effectiveness Of The TearCare® System In The Treatment Of The Signs And Symptoms Of Dry Eye Disease (SAHARA) (NCT04795752), a randomized na study. The primary outcome is Tear Break-Up Time.

2. Clinical Question

This is a randomized, parallel-group comparison of a device procedure and a drug treatment. The registry lists 345 participants enrolled. Its summary describes the study aim as evaluating safety and effectiveness in adult patients with dry eye disease. The primary endpoint is a continuous measure, so the comparison focuses on group means and the mean difference rather than the proportion of participants reaching a yes-or-no outcome.

The question is specific to the six-month comparison. It does not describe a comparison of TearCare with no treatment, and it does not establish a comparison between the original groups after the control group crosses over. The registry record does not report the analysis population, the handling of missing endpoint values, or a prespecified threshold defining a meaningful change.

3. Trial Design

SAHARA is listed as a completed, randomized, single-masked, parallel-group trial. The registry reports enrollment of 345 participants. It identifies Sight Sciences, Inc. as the sponsor and gives the trial start as 2021-04-28 and primary completion as 2024-09-23. The phase is listed as NA.

The primary six-month comparison uses the two groups as originally randomized: 172 participants in the TearCare Group (Study Device) and 173 in the Restasis Group (Control). The registry describes the TearCare group as receiving the study device from baseline to Month 24. The control group used Restasis from baseline to Month 6. Participants originally assigned to Restasis then stopped Restasis, received a single TearCare treatment at Month 6, and were followed through Month 12.

This crossover means the original treatment groups no longer represent two distinct assigned treatments after Month 6. The six-month primary endpoint occurs at the end of the stated Restasis comparison period, before the described crossover treatment. Longer-term observations after the crossover therefore address a different treatment history from the original randomized comparison.

Although masking is recorded as single, the identity of the masked party or parties is not reported in the registry record. Further details about randomization procedures, allocation concealment, visit schedules, adherence, protocol deviations, and the statistical analysis plan are not reported in the registry record. The record also does not state whether the six-month comparison was analyzed according to assignment, treatment received, or another analysis population. These details matter when assessing how closely the posted estimate follows the original randomized design.

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.

Primary Endpoint Phase

Reasons for not completing this period:

X-over-Restasis arm; Durability-TC arm

Reasons for not completing this period:

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 Tear Break-Up Time, measured over 6 months.

The primary endpoint is change in Tear Break-Up Time (TBUT) from baseline to Month 6, measured in seconds. TBUT is the time until a tear film first breaks up after a blink. It is used as a continuous indicator of tear-film stability: a longer time represents a longer interval before observed breakup. Because the endpoint is continuous, the registry reports mean change and standard deviation for each group.

The posted mean change is 2.71 seconds for TearCare, with a standard deviation of 3.35, and 1.22 seconds for Restasis, with a standard deviation of 2.46. The reported mean difference is 1.49 seconds in favor of the TearCare group’s numerical change. These are group-level summaries; they do not show how much individual participants changed or how many had an improvement.

TBUT is one measure of dry-eye signs, not a complete description of symptoms or overall disease experience. The registry summary states that the trial aimed to evaluate signs and symptoms, but the posted primary endpoint is TBUT. The record does not report a threshold for a clinically meaningful change, nor does it provide the relationship between TBUT change and participants’ symptoms. Statistical separation in a continuous endpoint therefore should not be treated as a complete account of clinical importance.

7. Results Posted in the Registry

• TearCare Group (Study Device): mean 2.71 versus 1.22 on control; difference 1.49 (95% CI 0.87 to 2.11), standardized effect d = 0.51; z-test p = <0.001.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is a continuous variable, and the registry reports its mean and standard deviation in each randomized group. A mean gives the average change across participants included in the group summary. A standard deviation describes how spread out individual measurements are around that group mean. It is not the standard error of the mean and does not directly describe uncertainty in the estimated group average.

What does the mean difference describe?

The reported mean difference is 1.49 seconds, obtained as the difference between the posted group means of 2.71 seconds and 1.22 seconds. It expresses the contrast in average change, in the endpoint’s original unit of seconds. It is not a percentage change, a ratio, or a change expected for every participant. The interpretation depends on the endpoint’s definition and the groups being compared at the stated time point.

For a randomized trial, comparing groups according to their original assignment can preserve the benefit of randomization when participants are analyzed as assigned. However, the registry record does not identify the analysis population or say whether the posted calculation followed that approach. It also does not give enough detail to establish whether the estimate is adjusted for baseline measurements or other covariates. No particular adjustment or testing procedure should be inferred from the summary statistics alone.

What does the interval add?

A confidence interval is a range of values compatible with the observed data under a specified statistical procedure. It conveys uncertainty around an estimate more directly than a p-value alone: a narrower interval indicates greater precision, while a wider interval indicates less precision. The reported interval for the mean difference has a positive lower bound and an upper endpoint of 2.11 seconds. The record does not state the confidence level or the procedure used to construct this interval, so those details cannot be confirmed from the registry entry.

In general, the width of an interval depends on the variability of the outcome, the amount of information available, and the method used to estimate uncertainty. A larger sample can often improve precision, but it cannot remove uncertainty due to measurement error, selection, missing observations, or limitations in the analysis. The standard deviations in the posted results show that individual TBUT changes varied within both groups; the interval summarizes uncertainty about the difference in means rather than the range of individual responses.

How should the p-value be read?

The posted unadjusted p-value is <0.001. A p-value is calculated under a null model and a specified test. It describes the compatibility of the observed data with that model, rather than the probability that a treatment effect is real or the size of an effect. A small p-value does not establish that the difference is clinically important. The registry does not identify the test used, its assumptions, or whether the test was prespecified.

For continuous outcomes, common analyses can use models that compare means, potentially with adjustment for baseline values or other variables. That is general statistical context, not a description of the method used in this trial. The registry record does not report a specific model, variance assumption, covariate set, or robustness analysis. The posted means and standard deviations do not reveal those choices.

Why can over-dispersion matter?

Over-dispersion is a general concern when observed variability is greater than a model assumes. If a method underestimates variability, standard errors can be too small and intervals too narrow; if it accounts for variability appropriately, uncertainty estimates can be more credible. This concept is especially relevant to count or event data, but the primary endpoint is continuous. The registry record does not report an over-dispersion assessment, and none should be presumed.

The primary result is based on a six-month comparison, while later follow-up includes crossover from Restasis to TearCare. This timing matters because a statistical contrast depends on which groups and treatment periods are being compared. A contrast after crossover would not have the same interpretation as the original randomized six-month contrast. The record does not provide enough detail to describe an analysis of post-crossover outcomes.

Finally, the registered outcomes and analyses do not give a full account of all planned statistical work. The record lists 2 outcome measures and 0 analyses, and no secondary results are posted. Details about missing-data handling, multiplicity, protocol deviations, and analysis populations are not reported in the registry record. These are not minor labels: they can influence which observations contribute to an estimate and how uncertainty is interpreted. Accordingly, the numerical summaries support a focused reading of the posted six-month result, while a fuller audit of the analysis requires information not reported in the registry record.

10. Confidence Intervals and Precision

The estimated mean difference is 1.49 seconds, with a reported interval whose lower bound is positive and whose upper bound is 2.11 seconds. The interval therefore does not cross zero, consistent with the reported unadjusted p-value of <0.001. The registry record does not state the confidence level or interval-construction method, so the interval should be described as reported rather than assigned an unstated confidence percentage.

Precision concerns uncertainty in the group difference, not whether the endpoint captures every aspect of dry eye disease or whether the difference matters to participants. A confidence interval gives a range of estimates compatible with the data under its statistical assumptions. Its width can be considered alongside the point estimate: the result is not a single certain value, even when the interval excludes zero.

The within-group standard deviations are 3.35 seconds for TearCare and 2.46 seconds for Restasis. These values describe spread in the group outcomes, whereas the interval concerns the estimated difference between group means. They answer related but distinct questions. The record does not give individual-level distributions, the number of observations with missing endpoint values, or a description of how uncertainty was calculated.

Interpretation is also limited by the lack of a reported clinical-importance threshold. The interval helps show statistical precision around the mean contrast, but it cannot establish whether values throughout that range correspond to a meaningful change in symptoms or daily experience. The result is most directly informative about the six-month average TBUT change in the two randomized groups.

11. 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 mean difference and assume two-sided significance level 0.05, equal allocation between two arms, and a normal approximation. Each row considers detecting a stated fraction of the observed effect. The trial’s own original design assumptions, including its sample-size justification, are not reported in the registry record.

For planning for 100% of the observed effect, the displayed requirement is 62 participants per arm for 80% power and 82 participants per arm for 90% power. For planning for 75% of the observed effect, the corresponding figures are 109 and 146 participants per arm. For planning for 50% of the observed effect, the figures are 245 and 327 participants per arm.

Power is the probability, under the planning assumptions and a particular true effect, that a statistical test will reject its null hypothesis. The 80% and 90% columns indicate different target probabilities for detecting the effect used in that planning scenario. A higher target power generally requires more participants. Likewise, smaller effects are harder to distinguish from random variation and need more participants: the figures rise as the planned fraction of the observed effect becomes smaller.

These are planning illustrations, not observed power. They should not be read as proof that the completed trial had a particular probability of finding a result, nor as a reconstruction of the sponsor’s original design rationale. They also rely on the stated approximation and assumptions; the registry record does not report whether those assumptions match the trial’s original planning process.

12. 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 = 1.490; pooled standard deviation σ = 2.938.
2. n = 2 × (1.96 + 0.842)² × 2.938² / 1.490² = 61.0, rounded up to 62 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 <- 1.49                      # observed difference in means
sigma <- 2.93760314625682                     # 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(4795752)
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 = 62 (80% power), 82 (90% power)
 75% of observed effect: n per arm = 109 (80% power), 146 (90% power)
 50% of observed effect: n per arm = 245 (80% power), 327 (90% power)
power.t.test: 62 
Simulated power at n = 62 per arm: 0.801 (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.

13. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when several outcomes, time points, comparisons, or analyses are tested. If many hypotheses are examined, the chance of at least one small p-value can increase unless the analysis plan addresses that issue. The registry lists 2 outcome measures and 0 analyses, and no secondary results are posted. The primary TBUT result has an unadjusted p-value of <0.001. The record does not report whether any multiplicity adjustment was planned or used, or how the primary endpoint was prioritized among outcomes.

Missing-data methods are also important. If endpoint measurements are unavailable for some randomized participants, the analysis may depend on how those observations are handled. The registry does not report missingness at the six-month endpoint, reasons for missing measurements, imputation methods, or sensitivity analyses. It is therefore not possible to assess from the posted summary whether missing values could have affected the estimate.

The analysis population is not reported in the registry record. The record does not say whether all randomized participants were included, whether participants were analyzed according to assignment, or whether another population was used. It also does not report how protocol deviations, adherence, or treatment discontinuation were handled. These details help establish what population the estimate represents and whether the randomized comparison was maintained in analysis.

The crossover is relevant to analysis populations and time points. At six months, the stated comparator period ends and participants assigned to Restasis begin TearCare. Post-crossover measurements therefore reflect a different treatment history. The record does not describe the analysis approach for those later observations, and they should not be treated as an ongoing untreated control comparison. A complete assessment of multiplicity, missingness, and analysis populations would require information not reported in the registry record.

14. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: cholelithiasis (0.6%), pulmonary edema (0.6%), gall bladder tumor (0.6%).

The registry reports serious adverse events for the TearCare group from baseline to 24 months, the Restasis control group from baseline to Month 6, and the participants originally assigned to Restasis after crossover from Month 6 to Month 12. The corresponding numbers at risk are 172, 173, and 163. The numbers affected by serious adverse events are 2, 0, and 0, respectively. One death is listed in the TearCare group; no deaths are listed in either Restasis period.

These groups have different follow-up periods and treatment histories, so their counts should not be read as a direct comparison over identical observation time. The record does not describe event adjudication, attribution to treatment, or the detailed definitions used for these entries. It also lists Cholelithiasis, Pulmonary Edema, and Gall bladder tumor among the top serious terms, each reported as 0.6%. These entries do not establish causation. The posted figures describe reported safety counts and should be interpreted within the stated periods and limits of the registry record.

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

The six-month posted results show a larger average change in TBUT in the TearCare group than in the Restasis group. The reported mean difference is 1.49 seconds, the interval has a positive lower bound and an upper bound of 2.11 seconds, and the unadjusted p-value is <0.001. This is a statistical comparison of group averages for one continuous endpoint at one time point.

The result does not show that every participant in the TearCare group improved, nor does it describe the distribution of individual responses. It does not indicate how many participants experienced a change that they considered important. A group mean can summarize an overall pattern while concealing substantial variation among individuals. The registry record does not report a threshold for clinical importance or connect the TBUT result to a specific symptom change.

The endpoint captures a sign of dry eye disease, while the trial summary also refers to symptoms. The posted primary result alone cannot represent all signs, symptoms, or experiences relevant to dry eye disease. No secondary results are posted in the registry record, so the additional outcome evidence cannot be assessed from this entry.

The comparison is tied to the first six months. Once the Restasis group crossed over to TearCare, later follow-up no longer compared the original randomized groups receiving separate treatments. The six-month estimate should not be extended to claims about later durability or outcomes after crossover. Nor does the posted result establish comparative effects in populations or settings not described by this trial record.

Finally, statistical evidence of a difference is not the same as a determination of clinical value. The record provides a mean contrast and uncertainty information, but does not report the clinical threshold, analysis details, missing-data approach, or multiplicity strategy needed for a fuller assessment. These results describe what was posted for the registered endpoint; they do not serve as medical advice or as a recommendation for an individual.

16. Why This Trial Matters Statistically

This trial is statistically informative because it compares a continuous sign of dry eye disease between randomized groups and reports both group means and a mean difference. The endpoint is measured in seconds, which allows the estimate to be expressed in the measure’s original units. Reporting an absolute difference helps readers assess the scale of the group contrast rather than relying on a p-value alone.

The result also illustrates why several statistics should be read together. The means describe average change, the standard deviations describe variation within groups, the confidence interval conveys uncertainty around the contrast, and the p-value summarizes compatibility with a null model under the test used. None of these measures answers every question. In particular, statistical separation does not by itself define clinical importance or indicate how individual participants fared.

The timing of the crossover is another key design feature. The original randomized comparison ends at Month 6, when participants in the Restasis group move to TearCare. This makes the six-month time point central to interpreting the randomized contrast and limits direct interpretation of later comparisons between original groups.

The registry entry also demonstrates the value of complete reporting. Details about the analysis population, missing observations, multiplicity, and the original sample-size rationale are not reported in the registry record. Without those details, the posted results can be interpreted as a focused comparison, but the full analysis pathway cannot be reconstructed. Clear reporting of these elements helps readers distinguish the estimate itself from uncertainty about how it was produced and which participants contributed to it.

17. 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 registry record leaves several trial-specific details unreported. It does not identify who was masked, describe the randomization or allocation-concealment procedures, or provide the full eligibility criteria and participant characteristics. These omissions limit assessment of how the enrolled sample was formed and how the masking process operated.

For the primary result, the record does not report the analysis population, the method used to compare means, any baseline or covariate adjustment, missing-data handling, or sensitivity analyses. It also does not state the confidence level or construction method for the reported interval. These gaps make it difficult to evaluate the analysis assumptions or determine how robust the estimate may be to alternate reasonable methods.

The posted efficacy information is focused on the six-month TBUT change. Although the registry summary refers to signs and symptoms, no secondary results are posted. The record does not report a threshold for a clinically meaningful TBUT change or a link between the mean contrast and symptom outcomes. The mean difference therefore has a clear statistical unit but an unreported clinical-importance benchmark.

The crossover further limits interpretation beyond six months. Participants originally assigned to Restasis received TearCare after Month 6, so later observations do not represent an unchanged randomized comparison. The registry does not report a separate analysis that would resolve the effect of this treatment history.

Safety counts also cover different observation periods across groups, and the record does not provide detailed event definitions, attribution, or adjudication information. The planning figures are based on the observed effect under stated assumptions, not the trial’s own original sample-size rationale. Collectively, these limitations call for a focused interpretation of the posted six-month result rather than broader conclusions about all outcomes or follow-up periods.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Safety And Effectiveness Of The TearCare® System In The Treatment Of The Signs And Symptoms Of Dry Eye Disease (SAHARA) — NCT04795752.
• Registry record last updated 2025-11-24; status last verified 2025-11.
• Results first posted 2025-11-24.
• 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.

21. Record Summary

SAHARA (NCT04795752) is a completed, randomized, single-masked, parallel-group trial sponsored by Sight Sciences, Inc. It enrolled 345 participants and lists Dry Eye and Meibomian Gland Dysfunction as conditions. The registry gives a start date of 2021-04-28 and a primary completion date of 2024-09-23.

The primary endpoint is change in TBUT from baseline to Month 6. Mean change was 2.71 seconds in the TearCare group and 1.22 seconds in the Restasis group; the reported mean difference was 1.49 seconds, with an unadjusted p-value of <0.001. The Restasis group crossed over to TearCare after Month 6. The registry does not report several analysis details, including missing-data methods, the analysis population, and the original sample-size rationale.