Cirrus HD-OCT Measurement of Area of Increased Light Penetration Under the Retinal Pigment Epithelium (RPE): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Cirrus HD-OCT Measurement of Area of Increased Light Penetration Under the Retinal Pigment Epithelium (RPE) (NCT01272076), a registered phase not stated study. The primary outcome is Difference in mm Squared of Cirrus HD-OCT Automated Measurements of the Illumination Areaa Under the RPE to Expert Manual Measurement of Areas of Hypofluorescence Typical of Geographic Atrophy (GA)..

2. Clinical Question

The study addresses a measurement question: whether an area identified through increased light penetration beneath the RPE on OCT corresponds in size to an area of hypofluorescence identified manually in fundus images. The objective is not framed as a comparison of treatments. The registry describes the relationship of increased illumination to areas of geographic atrophy and proposes assessing the difference between the two measurement approaches.

The record identifies a single clinical group, described as patients with dry AMD and geographic atrophy. It does not give separate participant counts for the imaging methods, report whether both measurements were made on each participant, or detail how images or regions were selected. Those design details are not reported in the registry record. The comparison should therefore be read as a comparison of measurement approaches, not as evidence of a randomized or like-for-like treatment control comparison.

3. Trial Design

The registry record identifies NCT01272076 as a completed study sponsored by Carl Zeiss Meditec, Inc. It lists an enrollment of 85 and describes the population as patients with dry AMD and geographic atrophy. The recorded study start is 2011-01, and primary completion is listed as 2011-02. The primary outcome timeframe is recorded as August 2011. The record does not explain the relationship between the listed timeframe and the primary completion date.

The stated objective is to compare automated Cirrus HD-OCT measurements of the area of increased illumination beneath the RPE with expert manual measurements of hypofluorescent areas typical of geographic atrophy in fundus autofluorescence images. The endpoint is continuous and is expressed as a difference in mm^2. The registry describes reporting a mean and standard deviation.

The registry lists a GA Group and describes it as patients diagnosed with dry AMD and geographic atrophy. This is not a randomized treatment comparison. Allocation is not stated, and the registry record indicates that the arms are not a like-for-like control comparison. The methods used to obtain and match the two types of image measurement, including whether measurements were paired within participants, are not reported in the registry record.

The record lists one outcome measure and no analyses. It does not report a statistical analysis plan, an inferential test, an adjustment strategy, or a sample-size justification beyond noting that the required sample size was calculated based on previous experience with the device and its variability. Further details of that calculation are not reported in the registry record. The record is marked completed, but completion status alone does not supply the missing outcome estimates or analysis.

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: After enrollment a screening image of the retina is done to ensure that the area of Geographic Atrophy is consistent with the inclusion criteria. We enrolled 85 subjects, however, 19 were determined not to be eligible per study protocol. In addition, images of 14 subjects did not meet predetermined qualification criteria per study protocol.

Overall Study

Analysis population for the primary outcome, as defined in the registry: Subjects with dry AMD and Geographic Atrophy. The required sample size was calculated based on previous experience with the device and its variability.

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 Difference in mm Squared of Cirrus HD-OCT Automated Measurements of the Illumination Areaa Under the RPE to Expert Manual Measurement of Areas of Hypofluorescence Typical of Geographic Atrophy (GA)., measured over August 2011.

The primary endpoint is the difference in mm^2 between Cirrus HD-OCT automated measurements of increased illumination beneath the RPE and expert manual measurements of hypofluorescent areas typical of geographic atrophy. It is a continuous measurement endpoint: the quantity of interest is an area, rather than an event count or a time until an event. The record says that the mean and standard deviation are the intended summary measures.

The biological rationale described in the registry is that, in retinal areas with atrophy, light from the Cirrus penetrates the sclera and choroid, which are described as more reflective than the RPE. The resulting areas of greater illumination are associated with geographic atrophy and may be used to quantify its area. The comparator is a manual assessment of hypofluorescence on fundus images by retina specialists.

Comparing the measured areas can indicate whether the automated imaging measure tends to differ in size from the manual reference measure. A mean difference describes average directional offset, while the standard deviation describes variation in the measured differences or values, depending on the analysis definition. The registry does not specify the precise calculation of the difference, its direction convention, or whether observations are paired. It also reports no numerical mean difference or standard deviation. Consequently, the stated endpoint identifies the intended measurement question but does not establish the size, direction, or precision of any observed discrepancy.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary measure is continuous because it is an area expressed in mm^2. The registry states that the endpoint is a mean and that dispersion is to be described with a standard deviation. A mean difference can describe average offset between two measurement approaches, provided the difference is defined consistently. The record does not state which method is subtracted from which, so the sign of a hypothetical difference cannot be interpreted from the endpoint title alone.

What does a mean difference tell us?

In general, the mean difference is an estimate of average systematic separation between methods. If calculated for matched measurements, each difference would represent the contrast between the automated and manual measurement of the same unit, such as a retinal area. If measurements are not matched, the statistical interpretation is different. The registry does not report whether observations were paired or how the comparison was calculated, so neither structure should be assumed.

A mean is useful for describing average offset, but it does not show the full distribution of differences. A standard deviation describes spread around a mean under the stated summary approach. It is not the same as a confidence interval: the standard deviation describes variability among observations, whereas a confidence interval expresses uncertainty about an estimated quantity. General methods for comparing measurements may also examine the range of differences across the measurement scale, because average offset alone can conceal larger discrepancies for some observations. No such detailed results are reported in the registry record.

How should uncertainty be assessed?

In general, a confidence interval gives a range of values compatible with the observed data and the assumptions of the analysis. Its width reflects both the variability in the data and the amount of information available for the estimate. A narrow interval indicates greater statistical precision than a wide interval, but precision alone does not establish that a measurement difference is unimportant or acceptable. Clinical or technical criteria for acceptable disagreement would also be needed; those criteria are not reported in the registry record.

The registry does not report a confidence interval, a test statistic, or a p-value for the primary comparison. It is therefore not possible to judge inferential uncertainty or whether the observed difference was distinguishable from a reference value using a reported test. A statistically significant difference, if one had been reported, would not by itself establish practical importance. Conversely, an imprecise estimate would not demonstrate that the methods agree.

What is the role of the sample-size illustration?

The record describes a planning illustration based on the observed effect, with a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. It says the effect to detect is a stated fraction of the observed effect. These are assumptions for planning illustrations, not a report of the study's original design or an observed-power calculation. The registry separately says the required sample size was calculated based on previous experience with the device and its variability, but the details of that original calculation are not reported in the registry record.

Sample-size illustrations describe how many participants per arm might be needed under specified assumptions. Their interpretation depends on the effect size chosen, variability, allocation, significance level, and statistical model. If the assumed effect is based on an observed result, the illustration is conditional on that estimate and should not be treated as independent evidence that the study had adequate power. The record does not supply the numerical illustration results or clarify how a two-arm planning setup relates to the listed measurement-comparison group. The figures therefore cannot be used to infer the study's actual allocation or analysis.

Why can variability matter beyond the mean?

General measurement-comparison methods distinguish average bias from agreement across the range of observations. Two methods could have similar averages but differ substantially for individual measurements; they could also show a consistent offset while tracking changes similarly. The distribution and scale of the measurement can matter as well. For example, variability may increase as measured areas increase, a pattern sometimes described generally as heteroscedasticity. The registry provides no individual observations or stratified summaries with which to assess such patterns.

Over-dispersion is a general concern when observed variability is greater than assumed by a model. In a continuous measurement comparison, an incorrect variance assumption can make uncertainty estimates too small or too large. The registry does not identify a variance model or report diagnostic checks. No model choice should be inferred from the stated mean and standard deviation alone.

More broadly, interpreting a measurement study requires clarity about the unit of analysis, whether observations are paired, how image regions are selected, and how repeated measurements or multiple images are handled. The record does not describe these details, nor does it report an analysis population or handling of missing measurements. Thus the methods information defines the intended endpoint at a high level but does not permit reconstruction or evaluation of the statistical analysis.

10. Confidence Intervals for Each Arm

No two-arm comparison is shown for this trial, but each arm's mean can be given with a t-based 95% confidence interval computed from the posted mean, spread and number of participants.

R code and output

arm <- c("GA Group")
n   <- c(52)
m   <- c(0.1)
sp  <- c(1.9)
sd  <- sp      # posted standard deviations
for (i in seq_along(n)) {
  se <- sd[i] / sqrt(n[i]); h <- qt(0.975, n[i] - 1) * se
  cat(sprintf("CARM|%s|%d|%.3f|%.3f|%.3f|%.3f\n", arm[i], n[i], m[i], sd[i], m[i] - h, m[i] + h))
}

CARM  GA Group  52  0.100  1.900  -0.429  0.629

11. Multiplicity, Missing Data and Analysis Populations

The registry lists one outcome measure and no analyses. It does not report whether multiple hypotheses, image regions, time points, or measurement comparisons were tested, or whether any multiplicity adjustment was planned. With no analysis results or statistical plan in the registry record, the consequences of testing multiple outcomes cannot be assessed. The single listed outcome does not establish that only one analysis was conducted.

Missing-data handling is not reported in the registry record. There is no account of missing images, incomplete measurements, exclusions, or the number of observations included in the endpoint analysis. Enrollment of 85 does not establish how many participants or measurements contributed to the primary result. The record also does not say whether the two measurement approaches were available for the same observations, which would matter for a paired comparison.

The analysis population is not described. No intention-to-treat or per-protocol population is defined, and neither label should be inferred for this measurement study. The registry does not describe eligibility for the analysis, how unusable scans were handled, or whether image-quality criteria were applied. These omissions limit evaluation of selection effects and the representativeness of any potential estimate.

For a continuous endpoint, missing measurements can affect both the estimated mean and its precision, particularly if missingness is related to the underlying area or image quality. This is a general statistical concern, not a finding about this study. Without counts, reasons, and a prespecified handling approach, the direction or magnitude of any such effect cannot be assessed. The registry record also does not report sensitivity analyses to examine how alternative assumptions about missing measurements might affect conclusions.

12. 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.

The registry reports serious adverse events for the GA Group with 52 at risk and 0 affected. For other adverse events, it also reports 52 at risk and 0 affected. These figures are event summaries in the registry record; they are not the primary measurement outcome and do not describe the difference between the OCT and manual measurements.

The at-risk figure is the denominator recorded for each adverse-event category. A zero affected count means that no events are recorded in that category in this entry; it does not establish that no event could occur, nor does it explain how events were collected or classified. The record does not provide a list of serious-event terms, and it does not report deaths. The relationship between the adverse-event denominator and the enrollment figure is not explained in the registry record, so they should not be reconciled by assumption.

Because the study is a measurement comparison and the registry does not describe a treatment control comparison, these safety figures should not be interpreted as comparative evidence about treatment safety. They are limited to the recorded group and categories in this registry entry. Details such as event definitions, observation procedures, and completeness of adverse-event reporting are not reported in the registry record.

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

The record describes a study intended to compare an automated OCT-based area measurement with expert manual assessment of hypofluorescent areas associated with geographic atrophy. It identifies the population, measurement approaches, continuous endpoint, unit, and planned summary statistics. This establishes what the investigators sought to measure.

It does not establish the result of that comparison. The registry contains no numerical estimate of the mean difference, no standard deviation, and no confidence interval or inferential analysis. It therefore cannot show whether the automated measurements were generally larger or smaller, how much measurements differed on average, or how much variation occurred across observations. The record cannot support a conclusion that the methods agree, are equivalent, or are interchangeable.

Even if an average difference were available, its interpretation would require attention to the distribution of individual differences and the intended use of the measurements. An average offset alone does not describe the range of disagreement. A precision estimate would also be needed, and a criterion for acceptable measurement differences would be needed to judge practical importance. Such criteria and results are not reported in the registry record.

The enrollment figure and completed status do not fill this evidentiary gap. Neither tells how many observations were analyzed or what the measured differences were. The sample-size planning assumptions are illustrative rather than observed results, and the adverse-event counts address a separate question. The record also does not give enough design detail to determine whether measurements were paired, how images were selected, or how missing measurements were treated.

Accordingly, the registry record supports a description of the intended measurement comparison but not a conclusion about its outcome. Any broader claim about the performance or clinical value of either measurement approach would go beyond the information recorded.

14. Why This Trial Matters Statistically

Measurement studies answer a different statistical question from treatment trials. Rather than asking whether an intervention changes an outcome, this study asks how an automated imaging measurement compares with a manual reference measurement. That distinction matters because a comparison of averages is not necessarily a demonstration of agreement for individual observations.

The record highlights an important methodological issue: a new or automated measure may be associated with the same underlying feature as a reference method, yet still differ systematically or variably in the areas it assigns. Evaluating such a measure requires estimates of offset and spread, along with enough methodological detail to understand what was compared. In general, reporting only an average can obscure the variation relevant to individual measurements.

Here, the registry makes the intended endpoint clear but does not report the result or analytical details needed to evaluate it. That gap illustrates why registry completion status and an endpoint description should not be mistaken for evidence of measurement agreement. Numerical results, uncertainty, the relationship between measurements, and handling of missing observations are essential to statistical interpretation.

The record also notes that the original required sample size was based on previous experience with the device and its variability, while further details are absent. This underscores that sample-size planning depends on assumptions about variability and the effect of interest. Planning illustrations based on an observed effect do not replace transparent reporting of the actual analysis and its uncertainty.

For clinical researchers and biostatisticians, the main statistical lesson is to separate the question being asked from the evidence available to answer it. The comparison is potentially informative as a measurement question, but the posted record contains no estimate with which to assess magnitude, precision, or agreement.

15. 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 most consequential limitation of the registry record is the absence of primary outcome results. Although the endpoint is defined as a difference in mm^2 and a mean and standard deviation are specified, the numerical summaries and any uncertainty estimates are not reported. The record therefore cannot support an assessment of the actual measurement comparison.

Important design and analysis details are also not reported. The record does not clarify whether the OCT and manual measurements were paired on the same retinal areas, how images or areas were selected, how measurements were matched, or whether there were repeated measurements. It does not specify the sign convention for the difference. These details are necessary to understand the estimand and the appropriate statistical model.

The record does not give the number of measurements included in the analysis, missing-data counts, reasons for missingness, exclusions, or the analysis population. It lists enrollment of 85, but that figure cannot establish the analyzed sample size. Criteria for image quality and rules for handling incomplete or unusable measurements are not reported in the registry record.

The registry identifies one outcome measure but no analyses. The statistical test, confidence-interval method, model assumptions, and any multiplicity procedures are not reported. The record says sample-size planning drew on previous device experience and variability, but the original calculation details are absent. The planning illustration assumptions do not resolve that limitation and do not indicate the precision achieved in the study.

Finally, the record gives limited safety information and no detail about adverse-event ascertainment. The serious and other adverse-event counts are separate from the primary imaging endpoint. These reporting limitations mean the registry entry describes the study question more fully than it documents the evidence needed to answer it.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Cirrus HD-OCT Measurement of Area of Increased Light Penetration Under the Retinal Pigment Epithelium (RPE) — NCT01272076.
• Registry record last updated 2014-01-20; status last verified 2013-12.
• Results first posted 2013-11-14.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

19. Record Summary

NCT01272076 is a completed study sponsored by Carl Zeiss Meditec, Inc., with recorded enrollment of 85. It concerns subjects with dry AMD and geographic atrophy and compares automated Cirrus HD-OCT measurement of increased illumination beneath the RPE with expert manual measurement of hypofluorescent areas in fundus autofluorescence images. The primary continuous endpoint is a difference in mm^2, with the mean and standard deviation identified as summary measures.

The registry lists one outcome measure and no analyses. It does not report the numerical primary result, uncertainty estimates, or key details of the analysis and missing-data handling. The adverse-event entry records 0 serious events and 0 other events among 52 at risk in the GA Group; deaths are not reported in the registry record. The study's measurement question is clear, but its comparative result cannot be judged from the registry record.