Loteprednol Etabonate Ophthalmic Gel for the Treatment of Ocular Inflammation and Pain Following Cataract Surgery: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Loteprednol Etabonate Ophthalmic Gel for the Treatment of Ocular Inflammation and Pain Following Cataract Surgery (NCT02208297), a randomized phase 3 study. The primary outcome is Percentage of Participants With Complete Resolution of Anterior Chamber (AC) Cells.

2. Clinical Question

The clinical question concerns a short-term sign of ocular inflammation after surgery. Anterior chamber cells are white blood cells in the anterior aqueous humor, assessed during slit lamp examination. The endpoint focuses on whether that finding had resolved completely, rather than on a change in average score or a broader measure of symptoms.

The registry also lists pain and other measures of anterior chamber inflammation as secondary outcomes, including resolution of flare and combined resolution of cells and flare. Those outcomes add context to the trial’s stated focus, but the registry record does not report their results in the record summarized for this analysis. The primary result should therefore be read as evidence about the specified cell-resolution outcome at the stated assessment time, not as a complete account of every outcome in the study.

3. Trial Design

The registry identifies this as a phase 3, randomized, double-masked, parallel-group study. Participants were assigned to loteprednol etabonate gel or vehicle gel, with both treatments administered BID. The registry lists 163 participants in each arm for the primary outcome. Enrollment is recorded as 326, and the study status is completed.

The study title and brief objective describe evaluation of safety and efficacy of loteprednol etabonate ophthalmic gel, 0.38%, following cataract surgery. Its primary outcome was the percentage of participants with complete resolution of anterior chamber cells at 8 days. The assessment used a slit lamp examination and a graded cell scale, with grade 0 representing no cells seen. The registry states that a lower score indicated a better outcome.

Random allocation helps make treatment groups comparable on average at the start of a trial, while masking is intended to reduce the influence of treatment knowledge on care, assessment, or reporting. A parallel design means the arms are separate groups rather than participants crossing between treatments. These are general design principles; details of how allocation was generated, concealed, or implemented are not reported in the registry record.

The study began in 2014-09, and primary completion is recorded as 2015-03. The sponsor is Bausch & Lomb Incorporated. Further information about the full protocol, recruitment procedures, eligibility criteria, follow-up beyond the stated primary timeframe, and the trial’s original sample-size justification 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.

Overall Study

Analysis population for the primary outcome, as defined in the registry: The analysis population included all randomized participants, with missing values imputed as non-responders.

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 Percentage of Participants With Complete Resolution of Anterior Chamber (AC) Cells, measured over 8 days.

The primary endpoint is the percentage of participants with complete resolution of anterior chamber (AC) cells in the study eye at 8 days. AC cells are white blood cells in the anterior aqueous humor. An investigator assessed them during slit lamp examination using a 5-point scale: grade 0 means no cells seen; grade 1 means 1 to 5 cells; grade 2 means 6 to 15 cells; grade 3 means 16 to 30 cells; and grade 4 means > 30 cells. Complete resolution was defined as a cell score of 0.

This endpoint turns an examination finding into a binary outcome: a participant either met the specified definition of complete resolution or did not. That makes the reported result a count of participants meeting a threshold, rather than a mean cell grade or the size of a change from baseline. The registry says lower scores indicate better outcomes.

Complete resolution is one way to describe whether inflammation observed in the anterior chamber has cleared by the assessment time. It does not, by itself, convey the full distribution of cell grades, the timing of resolution, or whether other symptoms or signs changed. Those details are not reported in the registry record. The primary analysis included all randomized participants, and missing values were imputed as non-responders. This approach counts an unobserved outcome as not meeting the endpoint, which is a general method for handling missing binary response data; the record does not describe additional sensitivity analyses.

7. Results Posted in the Registry

• Loteprednol Etabonate Gel (BID): 23.3% (38/163) versus 17.8% (29/163) on control; risk difference 5.5 percentage points (95% CI -3.2 to 14.3), risk ratio 1.31 (0.85–2.02); unadjusted two-proportion z-test p = 0.217.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is a binary response: complete resolution of AC cells, defined as a score of 0, versus not meeting that definition. The arm results are presented as counts of participants with the event and the number assessed for the primary outcome. A risk is the proportion of a group experiencing an event. A risk difference compares the two group risks by subtracting one from the other. General interpretation: a risk difference of zero represents no difference in the event proportions, and positive or negative values indicate direction according to the order of the comparison.

What does the confidence interval describe?

A confidence interval expresses uncertainty around an estimate using a method with a stated long-run coverage property. In repeated use under its assumptions, intervals constructed by that method would contain the target value at the specified coverage rate at the corresponding frequency. It does not assign a probability to the fixed underlying effect after the data have been observed. Here, the posted interval for the risk difference extends across zero. That means the interval includes no difference as well as values favoring either direction. The registry does not report the method used to construct this interval, so details such as its exact assumptions cannot be assessed from the record.

How should the p-value be read?

The posted unadjusted p-value is 0.217. General explanation: a p-value measures how unusual data at least as incompatible with a stated null model would be if that model and the test assumptions held. It is not an effect-size measure, does not quantify the probability that the null hypothesis is true, and does not show whether an effect is important in practice. The registry record does not report the primary analysis’s prespecified significance threshold or enough detail to verify the test procedure. The sample-size planning assumptions are not evidence of the original analysis threshold.

What does randomization contribute?

Random assignment is intended to reduce systematic differences between treatment groups at baseline, including differences that are not measured. Double masking can reduce the opportunity for expectations to influence treatment delivery or outcome assessment. These are general advantages of randomized, masked designs, not proof that bias was absent. The registry record does not describe the random sequence method, concealment procedures, who was masked, or whether masking was maintained.

What does the missing-data rule imply?

The record says the primary analysis included all randomized participants and imputed missing values as non-responders. For a binary response endpoint, that assigns a non-response status when the outcome is missing. This can be a conservative rule in some circumstances, but whether it is conservative depends on why values are missing and how missingness differs between arms. Those details and the number of missing outcomes are not reported in the registry record. No alternative imputation approach or sensitivity analysis is reported in the record.

Why can count outcomes need careful modeling?

General explanation: a binary outcome can be analyzed using methods for proportions, while count outcomes and continuous scores may require different models. If count data show greater variability than a simple model expects, over-dispersion can make standard errors too small and confidence intervals too narrow unless the analysis accounts for it. That concern is not directly applicable to the primary endpoint as reported, because it is a binary participant-level response rather than a count of recurrent events. The registry record does not describe the statistical model used for the primary analysis.

The comparison also reports relative measures, but their estimates and interval endpoints include values written to more than three decimal places in the record. They are not reproduced in this explanation. Regardless of scale, relative measures and absolute risk differences answer different questions: a relative measure compares event risks proportionally, whereas a risk difference expresses their separation on the risk scale. The primary interpretation rests on the posted counts, risk-difference interval, and unadjusted p-value, without treating any one measure as a substitute for the others.

The registry lists six outcome measures, including four secondary outcomes. The record does not report a multiplicity adjustment, an ordered testing procedure, or a complete statistical analysis plan. Without those details, the handling of multiple comparisons cannot be evaluated. The posted primary result should therefore be distinguished from any claim that all possible outcome comparisons were prespecified or controlled for false-positive findings.

10. Confidence Intervals and Precision

The posted risk-difference estimate is 0.055, and its confidence interval runs from -0.032 to 0.143. Because the interval includes zero, the result is compatible with no difference in the response proportions. It also includes values on either side of zero, so the direction is not determined precisely by this interval. The posted unadjusted p-value is 0.217.

Precision concerns the range of effects that remain reasonably compatible with the observed data under the interval’s method and assumptions. A relatively wide interval means the estimate leaves considerable uncertainty about the magnitude of the difference. The result’s point estimate alone should not be treated as the exact effect that would occur in a broader population. The counts—38 participants meeting the endpoint in the loteprednol etabonate arm and 29 in the vehicle arm, with 163 in each arm—are the observed data underlying this binary comparison.

The record does not report the confidence-interval construction method or the analysis details needed to evaluate its assumptions. Nor does it report sensitivity analyses for the missing-as-non-responder rule. Thus, the interval describes uncertainty as calculated for the posted analysis, while uncertainty due to alternate analysis choices cannot be assessed from the registry record. The confidence interval and p-value are not separate measures of clinical importance; neither establishes how meaningful a difference would be to patients or clinicians.

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 effect and are not the trial’s original sample-size justification. The assumptions for these illustrations are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is set at a stated fraction of the observed effect.

For planning at 100% of the observed effect, the estimates are 840 participants per arm for 80% power and 1,124 participants per arm for 90% power. At 75% of the observed effect, the estimates are 1,456 participants per arm for 80% power and 1,949 participants per arm for 90% power. At 50% of the observed effect, they are 3,191 participants per arm for 80% power and 4,271 participants per arm for 90% power.

In plain language, 80% power is the chance, under the stated assumptions and if the chosen effect were the true effect, of detecting that effect using the specified test. The 90% power column uses a higher target chance of detection and therefore requires a larger planned group size under the same framework. Smaller effects are harder to distinguish from random variation, so planning for a smaller fraction of the observed effect requires more participants per arm.

These are planning illustrations, not observed power. They do not show the probability that the trial’s result is true or that the treatment has a particular effect. The registry record does not report the trial’s original design assumptions or sample-size justification, and the illustrations should not be mistaken for those original decisions.

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, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.178; treatment proportion p₁ = 0.233; average p̄ = 0.206.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.5715 = 1.1200.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.5701 = 0.4798.
4. n = (1.1200 + 0.4798)² / (0.0552)² = 839.6, rounded up to 840 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
p0 <- 0.177914110429448                          # control proportion
p1 <- 0.233128834355828                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^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 (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(2208297)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 840 (80% power), 1,124 (90% power)
 75% of observed effect: n per arm = 1,456 (80% power), 1,949 (90% power)
 50% of observed effect: n per arm = 3,191 (80% power), 4,271 (90% power)
power.prop.test: 839.6 
Simulated power at n = 840 per arm: 0.804 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-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

The primary analysis population is described as all randomized participants, with missing values imputed as non-responders. This defines how missing primary endpoint values entered the analysis, but the registry record does not state how many values were missing, why they were missing, or whether the pattern differed between arms. It also does not report alternative assumptions or sensitivity analyses. Consequently, the effect of the imputation rule on the result cannot be assessed from the record.

The registry lists four secondary outcomes: complete resolution of AC cells by visit, grade 0 pain in the study eye by visit, complete resolution of AC flare by visit, and complete resolution of both AC cells and AC flare by visit. It lists six outcome measures overall. The record addressed does not report results for those secondary outcomes. No secondary endpoint conclusions should be inferred from the primary result.

When a study evaluates multiple outcomes or repeated assessments, the chance of at least one apparently notable finding can increase unless the analysis plan addresses that multiplicity. General approaches include adjusting testing thresholds or specifying an order for tests, but the record does not report whether either approach was used. The statistical analysis plan, prespecified hierarchy, and multiplicity strategy are not reported in the registry record.

The record does not provide details on analysis exclusions, protocol deviations, or the handling of other intercurrent events. It also does not report the planned primary analysis method beyond the endpoint and missing-data rule. These gaps limit the ability to reproduce the analysis or judge whether the posted result follows a prespecified plan. They do not change the reported counts, but they constrain interpretation of the result’s robustness and of any unreported comparisons.

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 5% reporting threshold.

Most frequent serious adverse event terms

Serious events with the highest arm-level rate: endophthalmitis (0.6%).

The registry reports serious adverse events for participants at risk in each arm. In the loteprednol etabonate gel arm, 162 participants were listed as at risk for serious adverse events, with 0 affected. In the vehicle gel arm, 163 were listed as at risk, with 1 affected. Deaths were reported as 0 in each arm. For other adverse events, the at-risk counts were 162 and 163, respectively, and 0 participants were reported as affected in either arm.

These figures describe the safety entries in the registry record; they do not establish that no other adverse events occurred outside the reported categories or time period. The record’s list of top serious terms includes Endophthalmitis at 0.6%. The group attribution and further context for that term are not reported in the registry record. The time window, event definitions, severity grading, and relatedness assessment are also not reported in the record. The counts should therefore be read as the posted safety figures, not as a complete characterization of risk.

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

The posted primary result describes complete resolution of AC cells at 8 days among participants assigned to loteprednol etabonate gel or vehicle gel. The observed event counts were 38 and 29, respectively, with 163 participants in each arm. The estimated risk difference was 0.055, with a confidence interval from -0.032 to 0.143, and the unadjusted p-value was 0.217. The interval crosses zero, leaving uncertainty about whether the underlying difference favors either arm or is close to no difference.

This result does not establish that either treatment works or does not work. It is one estimate for one defined endpoint at one stated assessment time, with missing values handled as non-responders. It does not describe partial resolution, longer-term outcomes, pain results, or the other secondary endpoints whose results are not reported in the registry record. A statistical comparison also does not, by itself, establish clinical importance.

The randomized, double-masked design is relevant context, but the registry record does not report enough detail to evaluate all potential sources of bias or reproduce the analysis. In particular, the allocation method, primary statistical model, missing-data sensitivity analyses, and multiplicity plan are not reported. These omissions limit how much can be concluded about robustness beyond the posted analysis.

Safety entries show 0 serious adverse events in the loteprednol etabonate arm and 1 in the vehicle arm, with 0 deaths in each arm. Those figures are limited to the registry’s reported categories and at-risk counts. They do not settle the comparative safety profile, especially because the record does not provide further detail about follow-up, event attribution, or the serious-term listing.

16. Why This Trial Matters Statistically

This trial illustrates why a binary endpoint should be interpreted with both its definition and its uncertainty in view. Complete resolution means an AC cell score of 0, not merely improvement from a higher score. The reported counts therefore answer a threshold question: how many participants met that specific criterion by 8 days. They do not summarize the full range of scores or describe the trajectory of inflammation.

It also illustrates the difference between an estimated effect and the precision of that estimate. The point estimate favors a higher response proportion in the loteprednol etabonate arm, while the confidence interval includes no difference and values in both directions. The p-value is not a measure of clinical importance and cannot resolve the uncertainty by itself. Reading these statistics together avoids treating a single estimate as definitive.

The stated missing-data rule is another important part of the result. Imputing missing values as non-responders affects the definition of the analyzed response outcome, and the registry does not report the extent or pattern of missingness. Without sensitivity analyses, readers cannot assess how other reasonable assumptions might affect the estimate.

Finally, the registry distinguishes a posted primary result from a full account of trial evidence. Several secondary outcomes are listed, but their results and the multiplicity approach are not reported. The sample-size illustrations are planning calculations based on the observed effect, not the original trial design or observed power. Keeping these distinctions clear helps prevent the primary result, planning figures, and safety counts from being asked to answer questions they do not address.

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 unavailable. It does not report the eligibility criteria, baseline characteristics, random-sequence method, allocation concealment, or the roles included in double masking. Without those details, readers cannot fully assess how participants entered the study or how treatment assignment and outcome assessment were protected from bias.

The primary result is reported for complete resolution at 8 days, but the record does not provide the distribution of AC cell grades, baseline values, or changes from baseline. The endpoint therefore gives a threshold-based outcome rather than a fuller account of the inflammation findings. Although secondary outcomes are listed, their results are not reported in the registry record addressed.

The missing-data approach is stated, but the amount and reasons for missingness are not reported. No sensitivity analyses are described. The statistical analysis plan, model details, confidence-interval method, and multiplicity strategy are also not reported in the registry record. These limitations make it difficult to evaluate the robustness of the estimated difference and to establish how secondary comparisons were handled.

The safety information is limited to the posted adverse-event counts and top serious term. The record does not provide event narratives, group attribution for the listed Endophthalmitis term, follow-up duration for safety reporting, or other details needed for a broader safety assessment. The original sample-size justification is not reported; the planning figures are illustrations based on the observed effect, not the study’s stated design rationale.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Loteprednol Etabonate Ophthalmic Gel for the Treatment of Ocular Inflammation and Pain Following Cataract Surgery — NCT02208297.
• Registry record last updated 2020-09-25; status last verified 2020-09.
• Results first posted 2020-09-03.
• 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

The registry record describes a completed phase 3, randomized, double-masked, parallel-group study sponsored by Bausch & Lomb Incorporated. It compares loteprednol etabonate ophthalmic gel, 0.38%, administered BID with vehicle gel administered BID after cataract surgery. Enrollment is recorded as 326; the primary outcome analysis lists 163 participants in each arm.

The primary outcome was complete resolution of AC cells at 8 days, with missing values treated as non-responders. The posted counts were 38 in the loteprednol etabonate arm and 29 in the vehicle arm. The risk-difference confidence interval includes zero, and the unadjusted p-value is 0.217. The registry also lists secondary inflammation and pain outcomes, but their results are not reported in the registry record addressed. Safety figures include 0 serious adverse events in the loteprednol etabonate arm, 1 in the vehicle arm, and 0 deaths in each arm.