Reduction in the Occurrence of Center-Involved Diabetic Macular Edema: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Reduction in the Occurrence of Center-Involved Diabetic Macular Edema (NCT00090519), a randomized phase 3 study. The primary outcome is Mean Duration of Definite Center of Macula-involved Diabetic Macular Edema (DME).

2. Clinical Question

3. Trial Design

This was a randomized, phase 3, parallel-group trial with quadruple masking. The registry record lists 731 participants enrolled and the study status as completed. The study started in 2004-02, and primary completion was recorded in 2010-04. Chromaderm, Inc. is listed as the sponsor.

Participants were assigned to ruboxistaurin or placebo. The ruboxistaurin arm received 32 milligrams orally once daily for up to 36 months; the placebo arm received placebo orally once daily for up to 36 months. Random allocation supports a comparison between groups by reducing systematic differences at assignment, while masking is intended to limit the influence of knowledge about treatment assignment. The registry record does not describe the allocation procedure or the specific people covered by quadruple masking.

The primary outcome was assessed over 6 Months through 36 Months. It was the mean duration of definite center-involved diabetic macular edema, determined through central grading of stereoscopic fundus photographs. The primary analysis population is described as intent-to-treat, meaning participants were analyzed in their randomized groups, regardless of whether they took or received the assigned treatment.

The registry lists 10 outcome measures and 9 analyses. Eight secondary outcomes are posted, covering visual acuity, photocoagulation, contrast sensitivity, diabetic retinopathy progression, kidney-related measures, visual function, and adverse events. The registry record does not report the original sample-size justification, detailed analysis plan, or how the recorded number of analyses relates to all planned outcomes.

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

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Intent-to-treat (ITT) population including all randomized participants analyzed according to the treatment group to which they were originally assigned by random allocation even if the participant does not take the assigned treatment, does not receive the correct treatment, or otherwise does not follow the protocol.

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 Duration of Definite Center of Macula-involved Diabetic Macular Edema (DME), measured over 6 Months through 36 Months.

The primary endpoint was mean duration of definite center-involved diabetic macular edema, measured in months per participant. The registry describes center involvement as determined by central grading of stereoscopic fundus photographs. When involvement was identified at a visit, a participant was considered to have definite center involvement for a specified duration between adjacent visits; the durations were then totaled for each participant. The group mean was calculated from participant durations.

This endpoint represents time accumulated with center involvement over the study’s outcome window, rather than a single measurement of edema at one visit. A mean offers a concise summary of the average duration across participants, while the standard deviation describes how spread out participant durations were around that average. The posted means were 1.72 months per participant for ruboxistaurin and 1.69 for placebo, with standard deviations of 4.95 and 4.41, respectively.

In general, a duration measure can capture differences in how long a condition is present, which is distinct from whether it is present at a particular assessment. The registry record does not report additional details about the grading procedure, handling of uncertain assessments, or the rationale for choosing this endpoint over other measures.

7. Results Posted in the Registry

• Ruboxistaurin: mean 1.72 versus 1.69 on control; difference 0.03 (95% CI -0.67 to 0.73), standardized effect d = 0.01; z-test p = 0.933.

8. Statistical Interpretation

9. Statistical Methods Explained

What does a mean difference represent?

General statistical explanation: A mean is the arithmetic average of the measured values in a group. A mean difference compares two group averages; its sign depends on the order of subtraction. For this unadjusted recalculation, the comparison is ruboxistaurin minus placebo, so a positive value means the observed average was higher in the ruboxistaurin group. The posted group means are 1.72 and 1.69 months per participant, and the recalculated difference is 0.03.

The mean summarizes the center of the observed distribution but does not describe how similar individual participants were to one another. The posted standard deviations are 4.95 for ruboxistaurin and 4.41 for placebo. In general, a standard deviation describes the spread of observations around the group mean. It is not the uncertainty in the mean itself; uncertainty in a comparison also depends on the amount of data and the analysis method.

What does the recalculation represent?

The unadjusted recalculation uses the posted group summaries to estimate a difference between group means and an interval around that estimate. It is not the sponsor’s reported ANOVA analysis. The recalculated estimate is 0.03, while the registry’s reported ANOVA estimate is -0.01. A difference between estimates can arise when analyses use different methods or target different quantities, but the reason for the difference is not reported in the registry record. The registry record does not identify the recalculation as the trial’s prespecified primary analysis.

General statistical explanation: An unadjusted comparison uses the group summaries without adjustment for other variables. An adjusted model can account for specified covariates or other features, depending on its design. Whether an adjustment is suitable depends on the prespecified analysis and data structure; those details cannot be inferred from the summary statistics. The registry lists ANOVA for the sponsor analysis but does not report its model specification or diagnostics.

How should a confidence interval be read?

General statistical explanation: A confidence interval gives a range of estimates compatible with the observed data under a statistical procedure and its assumptions. It communicates precision as well as direction. A narrow interval indicates greater precision than a wider interval on the same scale; an interval containing zero does not establish that the true difference is zero. The sponsor’s reported 95% interval runs from -0.714 to 0.686. It includes both negative and positive differences, so the posted analysis does not pinpoint the direction of the group difference.

What does ANOVA contribute?

General statistical explanation: Analysis of variance, or ANOVA, is a method for comparing means within a statistical model. The registry identifies the sponsor’s analysis as ANOVA and labels its parameter as mean difference in final values. The reported estimate is -0.01, with p = 0.969. The record does not report whether the model included covariates, how assumptions were checked, or how missing observations were handled. ANOVA is a method label, not a complete account of the analysis.

Why might spread and distribution matter?

General statistical explanation: Statistical models often rely on assumptions about the distribution of measurements or model errors. If observations are highly variable or the distribution is strongly skewed, the mean and its uncertainty may be sensitive to the analysis method. For duration outcomes, values may be concentrated differently across participants, and the mean alone cannot reveal that pattern. The posted standard deviations indicate substantial spread relative to the group means, but summary statistics alone do not establish skewness, over-dispersion, or a model violation. Individual-level distributions and diagnostics are not reported in the registry record.

Over-dispersion is a general concern when observed variability exceeds what a model assumes. In such cases, a model that understates variability can produce confidence intervals that are too narrow. This is a general statistical consideration, not a finding about this trial: the registry record does not report an over-dispersion assessment or enough distributional information to evaluate one.

What does intent-to-treat mean?

General statistical explanation: An intent-to-treat analysis keeps participants in the groups to which randomization assigned them, regardless of treatment received or adherence. This approach preserves the original randomized comparison. The registry describes the primary population in these terms. The posted text does not fully describe missing-data handling, and intent-to-treat labeling alone does not reveal how unavailable outcome measurements were addressed.

What is not specified?

The registry record does not report the original sample-size justification, planned interim analyses, detailed statistical analysis plan, multiplicity adjustment, model diagnostics, or the method used for missing primary-outcome data. These omissions limit the extent to which a reader can assess whether the reported analysis followed a prespecified plan or how robust it is to alternative assumptions. The absence of a reported detail is not evidence that a particular procedure was or was not used.

10. Confidence Intervals and Precision

Precision concerns how tightly an estimate is determined by the data and analysis. A point estimate alone does not show precision; the interval around it provides information about the range of values compatible with the analysis. The sponsor’s ANOVA estimate was -0.01, with a 95% confidence interval from -0.714 to 0.686. Because the interval extends in both negative and positive directions, it does not settle the direction of the difference.

In general, a 95% confidence interval is produced by a procedure that, across repeated comparable samples under its assumptions, would contain the target parameter in 95% of intervals. It is not a statement that there is a 95% probability that this particular interval contains the true value. The interpretation depends on the analysis model and its assumptions.

The unadjusted recalculation also reports an interval spanning values on both sides of zero. That interval accompanies a recalculated mean difference of 0.03. The sponsor estimate and recalculated estimate differ slightly, and their intervals arise from distinct analyses. The registry record does not explain the methodological reasons for the difference.

The posted standard deviations, 4.95 for ruboxistaurin and 4.41 for placebo, describe variation among participant durations, not the precision of the group means. The registry record does not provide further distributional detail or a full account of missing outcome data. Those factors, along with the analysis method, affect how an interval should be understood.

Accordingly, the estimates are best read with their intervals rather than as exact values. The intervals do not establish equivalence, and the registry record does not report an equivalence margin. They also do not support a confident claim about which group had a larger underlying mean duration.

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 sample-size figures are planning illustrations based on the observed effect, not the original trial design. The registry record does not report the trial’s own sample-size assumptions or justification. The illustration assumes a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. It considers effects equal to 100%, 75%, and 50% of the observed effect.

For this continuous outcome, the entries mean participants per arm for 80% power and participants per arm for 90% power. In general, power is the chance that a planned study will detect an effect of a specified size using a specified procedure, if that effect is present and the assumptions hold. The 80% and 90% columns therefore describe different planning targets, not results observed in this trial. These figures are not observed power.

For planning at 100% of the observed effect, the illustration gives 383,484 participants per arm for 80% power and 513,375 participants per arm for 90% power. At 75% of the observed effect, it gives 681,748 and 912,667 participants per arm, respectively. At 50% of the observed effect, it gives 1,533,933 and 2,053,500 participants per arm, respectively. Smaller effects require more participants in this illustration because they are harder to distinguish from random variation.

The general pattern is that a higher power target or a smaller effect to detect calls for a larger planned sample under the stated assumptions. These calculations use the observed effect, which is itself uncertain, so they should be read as planning illustrations rather than as evidence about the trial’s design or its achieved power. The figures are participants per arm, not total event counts.

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 = 0.030; pooled standard deviation σ = 4.689.
2. n = 2 × (1.96 + 0.842)² × 4.689² / 0.030² = 383,483.0, rounded up to 383,484 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 <- 0.03                      # observed difference in means
sigma <- 4.68894873716276                     # 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(90519)
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 = 383,484 (80% power), 513,375 (90% power)
 75% of observed effect: n per arm = 681,748 (80% power), 912,667 (90% power)
 50% of observed effect: n per arm = 1,533,933 (80% power), 2,053,500 (90% power)
power.t.test: 383484 
Simulated power at n = 383,484 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. Analyses Reported by the Sponsor

The registry reports the sponsor’s ANOVA analysis of final values as a mean difference of -0.01, with a 95% confidence interval from -0.714 to 0.686 and p = 0.969. The separate unadjusted recalculation from the posted summaries gives a mean difference of 0.03. The point estimates are close to zero but differ in sign.

These are not identical analyses: the registry labels the sponsor result as ANOVA, while the recalculation is an unadjusted comparison derived from the posted group summaries. The registry record does not explain the difference in estimates or provide enough model detail to reconcile the analyses fully. The discrepancy flag is false, so the posted information does not identify a discrepancy requiring the sponsor analysis to take precedence over the recalculation.

The sponsor’s confidence interval includes negative and positive values, and its p-value is 0.969. In general, these statistics describe uncertainty under the sponsor analysis; they do not prove equivalence or show that the true difference is exactly zero. The sponsor result should be read as the registry’s reported analysis, while the recalculation is a separate descriptive check based on the posted summaries.

14. Multiplicity, Missing Data and Analysis Populations

The registry lists 10 outcome measures, 8 posted secondary outcomes, and 9 analyses. The secondary outcomes include visual acuity, first occurrence of focal/grid photocoagulation, contrast sensitivity, progression of nonproliferative diabetic retinopathy, estimated glomerular filtration rate, albumin/creatinine ratio, visual function, and adverse events. These outcomes address different aspects of vision, disease progression, kidney-related measures, and safety.

General statistical explanation: When many outcomes or comparisons are tested, the chance of at least one apparently unusual result can increase. Multiplicity procedures adjust interpretation or testing to account for a defined family of analyses. The registry record does not report whether a multiplicity adjustment was planned or used, how outcomes were prioritized beyond identifying the primary outcome, or how the listed analyses correspond to the outcome measures.

The primary population is described as intent-to-treat, with participants analyzed according to randomized assignment even if they did not take or receive the assigned treatment. The registry record does not report details of the missing-data method, the amount or pattern of missing primary-outcome data, or whether sensitivity analyses were conducted. Intent-to-treat terminology does not, by itself, specify how missing measurements are handled.

The registry record also does not report planned interim analyses, stopping rules, a complete statistical analysis plan, or whether analyses were prespecified before access to outcomes. These details matter when assessing the strength of evidence across multiple outcomes and analysis choices. Their absence from the record should not be interpreted as proof that no such procedures existed.

15. 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.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: diabetes mellitus inadequate control (1.7%), myocardial infarction (1.6%), acute myocardial infarction (1.4%), coronary artery disease (1.4%), angina pectoris (1.1%).

The serious adverse event figures list 371 participants at risk in the ruboxistaurin group, with 93 affected, and 360 at risk in the placebo group, with 82 affected. These counts describe participants with serious adverse events, not the number of distinct event episodes. The registry also lists 298 participants affected by other adverse events in the ruboxistaurin group and 299 in the placebo group, each among the same group-specific at-risk counts.

The listed serious adverse-event terms include Diabetes mellitus inadequate control at 1.7%, Myocardial infarction at 1.6%, Acute myocardial infarction at 1.4%, Coronary artery disease at 1.4%, Angina pectoris at 1.1%, and Cerebrovascular accident at 1.1%. The registry record does not specify the denominator or additional context for these term-level percentages, so they should not be interpreted as rates calculated from the at-risk counts without that information.

Deaths are not reported in the registry record. Adverse-event counts are descriptive safety information; by themselves, they do not establish that an event was caused by a study treatment or support a comparative causal conclusion.

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

The posted primary outcome describes mean duration of definite center-involved diabetic macular edema from 6 Months through 36 Months. The group means were 1.72 months per participant for ruboxistaurin and 1.69 for placebo. The unadjusted recalculation gives a mean difference of 0.03, while the sponsor’s ANOVA analysis reports -0.01. Both estimates are close to zero, but they come from distinct analyses.

The sponsor’s 95% confidence interval, from -0.714 to 0.686, includes differences in either direction. The unadjusted recalculation interval also spans both sides of zero. Together, these summaries do not precisely determine the direction of the group difference. They do not demonstrate that the groups are equivalent, since an equivalence margin and suitable equivalence analysis are not reported in the registry record.

The findings should not be turned into a claim that ruboxistaurin works or does not work. The posted data describe a particular duration outcome and the uncertainty in its comparison. Other outcomes are listed separately, and the adverse-event counts describe safety observations; none can be substituted for the primary comparison without considering its own analysis and uncertainty.

The registry record identifies randomization and an intent-to-treat primary population, but important analytic details are not reported, including missing-data handling, the full statistical plan, and the original sample-size justification. These limitations constrain interpretation of how the analysis was conducted and how robust it may be to alternative assumptions. The record’s results support a careful description of the observed group summaries, not a broader conclusion beyond the outcomes and analyses reported.

17. Why This Trial Matters Statistically

This trial illustrates why a group mean should be read together with its variability and the uncertainty around a comparison. The posted means, 1.72 and 1.69 months per participant, are close, but the means alone do not establish how precisely the underlying difference is estimated. The sponsor’s confidence interval extends across negative and positive values, and the unadjusted recalculation likewise does not isolate a direction.

It also demonstrates that two analyses of the same trial can produce point estimates that differ slightly. The sponsor’s ANOVA estimate is -0.01, while the unadjusted recalculation is 0.03. Because the registry record does not give full model details or explain this difference, the appropriate statistical reading is to identify the analyses separately rather than treating them as interchangeable.

The endpoint is a duration summary based on assessments between visits, not a single-visit measurement. That makes the definition of the outcome and the handling of intervals between assessments important to understanding what the mean represents. The registry describes the general calculation but does not report further procedural detail.

Finally, the planning illustration makes clear that sample-size calculations depend on the effect chosen for detection and the assumptions used. The trial’s original design assumptions are not reported, so the illustration based on the observed effect should not be mistaken for the basis of the completed trial. These distinctions help separate observed summaries, inferential uncertainty, and hypothetical planning from one another.

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 primary-outcome group sizes are 349 for ruboxistaurin and 346 for placebo, while overall enrollment is listed as 731. The registry record does not explain the relationship between these counts, such as the availability of primary-outcome measurements. The primary population is described as intent-to-treat, but missing-data handling and the number of participants contributing measurements over time are not reported.

The primary outcome depends on identifying definite center involvement from centrally graded stereoscopic fundus photographs and assigning duration between adjacent visits. The registry record does not provide further details about grading reliability, assessment timing, or treatment of uncertain or incomplete observations. These omissions make it difficult to assess measurement uncertainty from the record alone.

The sponsor analysis is reported as ANOVA of final values, but model specifications, diagnostics, and any covariate adjustment are not reported. The unadjusted recalculation and sponsor analysis have slightly different estimates; the registry record does not explain why. The record also does not report a multiplicity strategy, missing-data sensitivity analyses, interim analyses, or the original sample-size justification.

These limitations concern the completeness of the registry record and the scope of interpretation possible from its posted summaries. They do not establish that particular methods were absent from the conduct of the trial.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Reduction in the Occurrence of Center-Involved Diabetic Macular Edema — NCT00090519.
• Registry record last updated 2016-10-06; status last verified 2016-08.
• Results first posted 2016-05-16.
• 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

The registry record describes a completed, randomized, phase 3, parallel-group trial with quadruple masking and 731 participants enrolled. Ruboxistaurin was given orally once daily at 32 milligrams for up to 36 months; placebo was also taken orally once daily for up to 36 months. The primary endpoint was mean duration of definite center-involved diabetic macular edema from 6 Months through 36 Months, analyzed in an intent-to-treat population.

Posted primary-outcome means were 1.72 months per participant for ruboxistaurin and 1.69 for placebo. The sponsor’s ANOVA estimate was -0.01, with a 95% confidence interval from -0.714 to 0.686 and p = 0.969. The unadjusted recalculation gives a mean difference of 0.03. The analyses are distinct, and the registry record does not explain the difference in their estimates.

The record lists secondary outcomes and adverse-event counts, but does not report important details including the original sample-size justification, missing-data methods, or multiplicity procedures.