Infrared Choroidal Reflectance Camera for the Detection of Childhood Cataract: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Infrared Choroidal Reflectance Camera for the Detection of Childhood Cataract (NCT03035292), a registered phase not stated study. The primary outcome is Results of Red-reflex and Infrared-reflex Tests Compared to Gold Standard.

2. Clinical Question

3. Trial Design

This was a completed proof-of-concept diagnostic comparison, registered as NCT03035292. It examined red-reflex and infrared-reflex assessment in children attending a paediatric ophthalmology clinic. The registry lists 110 eligible children, with or without cataracts, aged 1 month to 5 years. Children with previous intra-ocular surgery were excluded.

The record says each child’s right eye was assessed by a medical student using both red-reflex and infrared-reflex assessment. The findings were compared with a gold-standard examination by an ophthalmologist. This design is intended to evaluate classification against a reference examination; it is not a randomized treatment trial. The record does not state the allocation or assessment order, and its own note cautions that the listed arms are not a like-for-like control comparison.

Because both assessments were applied to the same eye, their results are paired rather than independent observations. This matters for a direct statistical comparison: the relationship between the two assessments on each eye should be considered. The statistical analysis used for such a comparison is not reported in the registry record.

The primary outcome was a count of eyes classified as true positives, true negatives, false positives, and false negatives, with a timeframe of 1 day. Secondary outcomes concerned results and sensitivity and specificity stratified by ethnicity. The registry lists the study status as completed, but does not report the primary outcome counts, accuracy estimates, subgroup results, or an account of participant flow.

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: 110 children were eligible for the study, each had their right eye assessed by a medical student using 1) red-reflex assessment 2) Infrared-reflex assessment. Findings were compared to a gold standard examination by an ophthalmologist.

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.

Population: Children having red-reflex, infrared-reflex and gold standard examination. Data for the right eye only was analysed.

6. Primary Endpoint

The registry lists the primary outcome as Results of Red-reflex and Infrared-reflex Tests Compared to Gold Standard, measured over 1 day.

The primary endpoint records the number of eyes in four diagnostic categories when each screening assessment is compared with the ophthalmologist’s examination. A true positive is an eye with cataract correctly identified by the screening test. A true negative is an eye without cataract correctly classified as not having cataract. A false positive is an eye without cataract incorrectly classified as having cataract; a false negative is an eye with cataract incorrectly classified as not having cataract. The registry description contains the wording “False negatives positives” for the last category, but its definition describes the false-negative classification.

These counts are used because they show the types of correct classifications and errors, rather than reducing diagnostic performance to a single undifferentiated total. Sensitivity describes how often a test identifies cataract among eyes that have cataract according to the reference examination. Specificity describes how often it correctly classifies eyes without cataract. General statistical context: these measures answer different questions, so neither alone fully describes the balance of missed cases and false alarms.

The outcome timeframe is 1 day. The record gives the endpoint definition but does not report the counts or derived sensitivity and specificity estimates. The endpoint therefore cannot establish from the registry record alone how accurately either assessment classified the eyes.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

General statistical context: diagnostic accuracy is evaluated by comparing a test’s classifications with a reference standard. In this study, the ophthalmologist’s examination is identified as the gold standard, while red-reflex and infrared-reflex assessment are the tests being evaluated. The strength of that comparison depends not only on the test results but also on how the reference examination was conducted and whether it was applied consistently. Further procedural detail is not reported in the registry record.

What do the four outcome categories mean?

A true positive is a cataract eye identified as having cataract by the test. A false negative is a cataract eye the test classifies as not having cataract. A true negative is an eye without cataract correctly classified as such. A false positive is an eye without cataract incorrectly classified as having cataract. These categories are defined relative to the reference examination; they are not independent judgements about the true disease state.

Sensitivity is the proportion of reference-positive eyes that the test identifies as positive. It is calculated from true positives relative to true positives and false negatives. Specificity is the proportion of reference-negative eyes that the test identifies as negative, calculated from true negatives relative to true negatives and false positives. These definitions explain why the four counts are useful: they allow the two accuracy measures to be derived while preserving the distinction between missed cataracts and false alarms. The registry record does not report the counts or derived estimates.

Why does the paired design matter?

The record states that each child’s right eye was assessed using both methods. As a result, the two test results are linked within an eye. General statistical context: a paired comparison focuses on cases where the methods agree or disagree, rather than treating every result from one method as unrelated to every result from the other. The extent and direction of disagreement can affect the evidence for a difference in diagnostic performance. The registry record does not report the paired cross-classification or the analysis used to compare the tests.

The stated primary endpoint is a count of eyes in diagnostic categories, not a continuous measurement. Counts can be presented directly and can also be used to calculate proportions such as sensitivity and specificity. Statistical uncertainty around these estimates is important, particularly when the number of eyes in a relevant reference category is limited. A confidence interval conveys uncertainty in an estimate; it does not guarantee that the true value lies within the interval. No confidence intervals are reported in the registry record.

What role does the reference examination play?

Accuracy estimates measure agreement with the reference standard, not accuracy against an unknowable state without qualification. If the reference examination misclassifies an eye, or if its interpretation is affected by knowledge of the screening result, the estimated performance of the screening tests may also be affected. General statistical context: blinding, consistent procedures, and complete application of the reference standard help reduce these risks. Whether the ophthalmologist was blinded, and details of the reference examination, are not reported in the registry record.

The record says results were obtained in a clinic population that included children with and without cataracts. Accuracy estimates can depend on the mix and characteristics of the population tested, as well as on how cases are defined. Clinic-based performance should therefore be interpreted in its stated setting. The record does not describe how children were recruited or provide a detailed account of the distribution of cataract status.

How should uncertainty and subgroup analysis be considered?

A point estimate alone does not show how much sampling uncertainty surrounds diagnostic performance. Confidence intervals help readers judge precision, while the underlying counts show how many classifications support the estimate. Neither is reported for the primary outcome. The registry lists ethnicity-stratified diagnostic outcomes, but no subgroup estimates, counts, or uncertainty measures are reported. General statistical context: subgroup estimates may be less precise than overall estimates, and apparent differences across subgroups need cautious interpretation.

Multiple outcomes or subgroup comparisons can increase the chance of findings that appear notable by chance when many tests are conducted. Methods for controlling or interpreting multiplicity, including any prespecified hierarchy, are not reported in the registry record. The record lists 4 outcome measures and 2 analyses, but it does not provide enough detail to establish how these were organized or whether the reported analyses followed a prespecified plan.

The registry describes a planned sample-size illustration based on the observed effect, with a two-sided significance level of 0.05, equal allocation, and a normal approximation; it says the effect to detect is a stated fraction of the observed effect. These are planning assumptions, not evidence of the trial’s original design or of observed power. The registry record does not report a sample-size justification for the study’s own design. Planning calculations describe hypothetical future study requirements; they do not add evidence about the accuracy observed in this study.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary outcome of diagnostic classification against the gold standard and secondary outcomes concerning results, sensitivity, and specificity stratified by ethnicity. It lists 4 outcome measures and 2 analyses. The record does not report a prespecified hierarchy among outcomes, a multiplicity adjustment, or another strategy for interpreting multiple analyses. It is therefore not possible to determine how the analysis plan addressed the chance of apparently notable results arising across several outcomes or subgroup comparisons.

Missing-data reporting is also limited. The registry states that 110 children were eligible and describes assessment of the right eye using both screening methods, followed by an ophthalmologist’s examination. It does not report the number with completed results for each assessment, whether any examinations were missing or uninterpretable, how missing results were handled, or whether exclusions occurred after eligibility was established. The primary outcome counts are not reported in the registry record, so completeness cannot be checked against the stated study population.

The analysis population is not defined beyond the description of eligible children and their eye assessments. The record does not say whether analyses included every eligible child, only those with complete paired tests, or only eyes with a completed reference examination. This matters because exclusions or incomplete verification can affect diagnostic accuracy estimates if they are related to test findings or cataract status.

The record lists a planning illustration based on the observed effect, but it does not report the study’s own sample-size justification or provide an analysis plan. No interim analyses or other statistical planning details are reported in the registry record. These gaps prevent a full assessment of whether the analyses and handling of missing information followed a prospectively specified approach.

11. 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.

The registry’s adverse-event record lists 110 children at risk, 0 serious adverse events, 0 other adverse events, and 0 deaths. These figures describe what was recorded for the group titled “Children Recruited to Diagnostic Comparison Study.” They are safety-reporting counts, not measures of diagnostic accuracy, and they do not provide the missing primary outcome results.

A zero count in a registry entry means that no events in that category are reported for the stated group; it does not by itself establish that no events occurred outside the record’s collection or reporting procedures. The registry record does not describe adverse-event definitions, surveillance methods, the duration of safety follow-up, or how events were attributed. The primary outcome has a timeframe of 1 day, but the safety follow-up timeframe is not reported in the registry record.

These counts should therefore be read narrowly as the posted adverse-event figures. They do not establish the completeness of safety monitoring or provide information about the precision of an event-rate estimate. No rate should be inferred from the record’s counts without a defined follow-up period and appropriate reporting details.

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

The registry record describes a diagnostic study comparing red-reflex and infrared-reflex assessments with an ophthalmologist’s examination in children attending a paediatric ophthalmology clinic. Its primary outcome was designed to count correct and incorrect classifications. That design addresses a clinically relevant statistical question: how the screening methods classify eyes relative to a reference examination.

However, the registry record does not report the primary outcome counts or the resulting sensitivity and specificity estimates. It consequently does not show which assessment had more true positives, true negatives, false positives, or false negatives, nor how precisely either method’s performance was estimated. The record cannot support a conclusion about comparative diagnostic accuracy from the posted primary outcome data.

Even when accuracy estimates are available, they describe performance in the population and setting studied, in relation to the chosen reference standard. They do not automatically establish how a test performs in other settings or populations. Here, the stated population was children aged 1 month to 5 years attending a paediatric ophthalmology clinic, and children with previous intra-ocular surgery were excluded. The record does not establish performance in a broader screening programme.

The study’s completed status indicates its registry status, not that all results are reported or that a particular interpretation is warranted. The record also does not describe blinding, the analysis of paired results, missing-data handling, or the statistical plan. Those omissions limit assessment of how robust any potential accuracy comparison would be.

The posted safety figures report zero serious adverse events, zero other adverse events, and zero deaths among 110 listed at risk. Those figures should be interpreted as registry-reported counts, with the safety follow-up and collection methods not reported. They do not change the central limitation: without the diagnostic outcome results, comparative performance cannot be evaluated from this record.

13. Why This Trial Matters Statistically

Diagnostic tests can make two kinds of errors: failing to identify an eye with cataract and incorrectly flagging an eye without cataract. The study’s four-category primary outcome was structured to keep these errors separate. That is statistically important because sensitivity and specificity describe different aspects of test performance, and an overall count of correct classifications could conceal an imbalance between missed cases and false positives.

The study also illustrates why the design of a comparison matters. Both assessments were applied to the same right eye, according to the registry record. This paired structure can support a direct comparison, but the analysis needs to account for the linkage between the findings. The registry does not report how that comparison was analysed, so readers cannot assess whether the statistical method reflected the design.

Reference-standard procedures and reporting are also central to diagnostic research. The ophthalmologist’s examination is identified as the gold standard, but details of its conduct and blinding are not reported. Without the outcome counts, estimates, and uncertainty intervals, the intended comparison cannot be interpreted quantitatively. The record therefore demonstrates a gap between a clearly framed diagnostic question and the information needed to assess its answer.

The ethnicity-stratified outcomes raise another general statistical issue: subgroup analyses require enough information to judge both the estimates and their uncertainty. No subgroup results are reported in the registry record. Overall, this study matters statistically because it highlights the need for complete classification counts, paired-analysis methods, transparent reference-standard procedures, and clear reporting of uncertainty when evaluating screening tests.

14. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The most consequential limitation in the registry record is the absence of primary outcome results. Although the endpoint and its categories are described, no counts, sensitivity estimates, specificity estimates, or confidence intervals are reported. The record therefore does not permit evaluation of either test’s accuracy or a direct comparison of the methods.

Several design and conduct details are also missing. The record does not report the order of the two assessments, whether the ophthalmologist was blinded to their results, how the reference examination was standardized, or how potentially uninterpretable findings were handled. These details matter in diagnostic studies because test order and knowledge of prior results can influence classification.

The study population was clinic-based and included children with and without cataracts, aged 1 month to 5 years; children with previous intra-ocular surgery were excluded. The record does not provide detailed recruitment procedures, the distribution of cataract status, or a participant-flow account. Generalizability beyond this stated setting cannot be assessed from the available record.

Although ethnicity-stratified outcomes are listed, subgroup results are not reported. The analysis plan, missing-data approach, analysis population, multiplicity strategy, and methods for comparing the paired assessments are also not reported in the registry record. Finally, the safety entry gives event counts but not the definitions, monitoring procedures, or follow-up duration. These reporting gaps constrain interpretation beyond the missing diagnostic results themselves.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Infrared Choroidal Reflectance Camera for the Detection of Childhood Cataract — NCT03035292.
• Registry record last updated 2023-11-07; status last verified 2023-01.
• Results first posted 2023-11-07.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

18. Record Summary

NCT03035292 is a completed proof-of-concept diagnostic study titled “Infrared Choroidal Reflectance Camera for the Detection of Childhood Cataract.” It compared red-reflex and infrared-reflex assessment in children attending a paediatric ophthalmology clinic. The registry states that 110 children aged 1 month to 5 years were eligible and that children with previous intra-ocular surgery were excluded.

Each child’s right eye was assessed by a medical student using both methods, with findings compared against an ophthalmologist’s examination. The primary outcome counted true positives, true negatives, false positives, and false negatives over a 1 day timeframe. Ethnicity-stratified outcomes were also listed.

The primary outcome counts and accuracy estimates are not reported in the registry record. The record lists 0 serious adverse events, 0 other adverse events, and 0 deaths among 110 at risk; further safety follow-up details are not reported.