This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the results posted in the ClinicalTrials.gov record; calculations added by Clinical Biostats are labelled as such.
1. Trial at a Glance
This page covers the registry results for Heidelberg Edge Perimetry (HEP) Detecting Glaucomatous Visual Field Defects (NCT02526654), a randomized na study. The primary outcome is Correlation Coefficient Between HEP and OVF Mean Deviation (MD).
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT02526654 |
| Phase | NA |
| Condition | Glaucoma |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | None (open label) |
| Enrollment | 115 |
| Arms in the primary outcome | 2 |
| Lead sponsor | Wills Eye |
| Results posted | Yes |
2. Clinical Question
Population
The study included glaucoma subjects and healthy controls. The registry describes the glaucoma group as having characteristic glaucomatous disc damage and visual field changes; controls were recruited without glaucoma. The recorded group sizes were 79 glaucoma subjects and 36 healthy controls. The registry does not report further eligibility details in the primary outcome population field.
Intervention
The diagnostic tests of interest were standard automated perimetry using the Heidelberg Edge Perimeter (HEP) and the Octopus Visual Field (OVF) analyzer. Optical Coherence Tomography (OCT) was also used to image the retinal nerve fiber layer. These are measurements and imaging procedures, rather than therapeutic interventions.
Comparator
The primary comparison concerns HEP and OVF mean deviation (MD) measurements. The question is whether their values are correlated within glaucoma subjects and controls. OCT is part of the broader study, but it is not the comparator named in the primary endpoint.
Primary question
How closely do HEP and OVF MD measurements move together in glaucoma subjects and healthy controls? The record reports Pearson correlation coefficients of -0.86 for glaucoma subjects and -0.37 for healthy controls. These coefficients describe association, not direct agreement between the instruments or diagnostic accuracy against an independent reference.
3. Trial Design
The registry describes a randomized, parallel study with no masking. Its title identifies the work as a validation and reproducibility study of the Heidelberg Edge Perimeter in detecting visual field defects in glaucoma. The study was sponsored by Wills Eye, was listed as completed, and recorded enrollment of 115. The recorded study period began in 2014-12, with primary completion in 2015-06.
The study groups were glaucoma subjects and healthy controls. The glaucoma group was described as having characteristic glaucomatous disc damage and visual field changes; controls were described as not having glaucoma. Both groups were to perform visual field testing with HEP and OVF, and OCT was to image the retinal nerve fiber layer. The registry lists these procedures as diagnostic tests.
The primary outcome was the Pearson correlation coefficient between HEP and OVF mean deviation measurements, with a timeframe of the baseline visit, 1 hour. The stated purpose was to assess whether HEP could detect glaucoma as well as OVF, and the description treats closeness of the two parameters as an indication of comparability. The broader study also addressed repeatability of OCT parameters and correlations between an automated structure-function report and clinical impression.
The registry reports randomized allocation, but how randomization related to the glaucoma and healthy-control groups or to the order of diagnostic testing is not reported in the registry record. Details of sequence generation, allocation concealment, masking of assessors, and the test order are also not reported in the registry record. The record does not describe a therapeutic treatment assignment; its listed interventions are diagnostic procedures.
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: Before study tasks, Informed Consent was conducted and obtained according to the International Conference on Harmonization Good Clinical Practices (ICH GCP) guidelines with opportunity for all questions to be answered.
Overall Study
| Stage | Glaucoma Subjects | Healthy Controls |
|---|---|---|
| STARTED | 79 | 36 |
| COMPLETED | 79 | 36 |
| NOT COMPLETED | 0 | 0 |
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.
| Characteristic | Unit | Glaucoma Subjects | Healthy Controls | Total |
|---|---|---|---|---|
| Participants analysed | 79 | 36 | 115 | |
| Age, Categorical: <=18 years | Participants | 0 | 0 | 0 |
| Age, Categorical: Between 18 and 65 years | Participants | 23 | 24 | 47 |
| Age, Categorical: >=65 years | Participants | 56 | 12 | 68 |
| Sex: Female, Male: Female | Participants | 46 | 23 | 69 |
| Sex: Female, Male: Male | Participants | 33 | 13 | 46 |
| Race (NIH/OMB): American Indian or Alaska Native | Participants | 0 | 0 | 0 |
| Race (NIH/OMB): Asian | Participants | 0 | 0 | 0 |
| Race (NIH/OMB): Native Hawaiian or Other Pacific Islander | Participants | 0 | 0 | 0 |
| Race (NIH/OMB): Black or African American | Participants | 21 | 15 | 36 |
| Race (NIH/OMB): White | Participants | 55 | 17 | 72 |
| Race (NIH/OMB): More than one race | Participants | 0 | 0 | 0 |
| Race (NIH/OMB): Unknown or Not Reported | Participants | 3 | 4 | 7 |
| Region of Enrollment: United States | participants | 79 | 36 | 115 |
6. Primary Endpoint
The registry lists the primary outcome as Correlation Coefficient Between HEP and OVF Mean Deviation (MD), measured over Baseline visit, 1 hour.
| Element | Registry entry |
|---|---|
| Result type | Number |
| Measure of spread | not applicable |
| Unit | Pearson's correlation coefficient |
| Endpoint classification | not reduced to a single comparison |
The primary endpoint is Pearson's correlation coefficient between HEP and OVF mean deviation (MD) measurements. MD is the perimetry summary named by the registry. The reported timeframe is the baseline visit, 1 hour. The registry presents a coefficient for each study group: -0.86 for glaucoma subjects and -0.37 for healthy controls.
In general, Pearson's coefficient describes the direction and strength of a linear association between paired measurements. A coefficient with a negative sign indicates that higher values of one measurement tend to occur with lower values of the other, according to the recorded scales. The magnitude describes how closely the paired values follow a linear pattern. The record does not explain the sign convention or provide a plot of the individual paired measurements in the registry record for interpretation.
Correlation is useful for asking whether two instruments produce values that vary together. It is not itself a measure of agreement: instruments can be highly correlated while their values differ systematically. Nor does correlation alone measure sensitivity, specificity, or diagnostic accuracy against an independent reference standard. Those properties require different analyses and definitions, which are not reported for this primary endpoint.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The posted primary results are Pearson correlation coefficients of -0.86 in glaucoma subjects and -0.37 in healthy controls. The negative signs indicate inverse linear associations on the scales as recorded. The glaucoma-group coefficient is farther from zero than the control-group coefficient, indicating a more pronounced linear association in magnitude in that group. This is a descriptive comparison of the posted coefficients, not a reported statistical test of whether the group coefficients differ.
The coefficients do not establish that the two devices agree in their absolute MD values. Pearson correlation captures whether paired measurements vary together, but does not show whether they have the same scale, a shared offset, or sufficiently close values for a particular clinical use. A scatterplot, agreement analysis, and clinically justified limits for acceptable differences would address different aspects of comparability. Those results are not reported in the registry record.
The study description frames the correlation as part of assessing whether HEP can detect glaucoma as well as OVF. A correlation coefficient alone cannot establish equivalent diagnostic performance. It does not directly report how often either procedure identifies glaucoma, nor does it give classification accuracy against a reference standard. The available results therefore speak to association between MD measurements, rather than a complete validation of diagnostic performance.
The record does not provide confidence intervals, standard errors, p-values, or a hypothesis test for the primary coefficients. Consequently, the precision of each coefficient and the uncertainty around their apparent difference cannot be assessed from the posted values alone. The number of paired measurements contributing to each coefficient, and whether all recorded group members contributed complete pairs, are not reported in the registry record.
The primary outcome description includes both glaucoma patients and controls, while the reported results are separated by group. Separate within-group correlations answer whether the instruments' measurements co-vary within each group; they do not answer whether either device distinguishes the groups. The posted results should therefore be read narrowly: they quantify the reported association between the two perimetry MD measures within each group, with uncertainty and broader diagnostic implications left unresolved.
9. Statistical Methods Explained
What does Pearson correlation measure? In general, Pearson's correlation coefficient summarizes the direction and degree of linear association between paired quantitative measurements. It is based on how values vary around their respective averages and is bounded between -1 and 1. Values near either endpoint represent a stronger linear pattern; values near zero represent a weaker linear pattern. The sign indicates direction. These are general properties of the statistic, not additional trial results.
For this study, the paired quantities are HEP MD and OVF MD. The registry reports coefficients of -0.86 for glaucoma subjects and -0.37 for healthy controls. Because both are negative, the reported direction is inverse on the recorded scales. The magnitudes describe the strength of linear association, but do not provide a clinical threshold for acceptable agreement. The registry does not report the raw paired measurements or explain whether any transformation or adjustment was used.
Why is correlation not the same as agreement? In general, correlation is influenced by the range of values in a group. If participants vary widely in underlying status, two methods may show strong correlation even when their measurements differ by a consistent amount. Conversely, restricted variation within a group can make correlation smaller even if individual paired measurements are close. A correlation coefficient does not reveal the average difference between instruments or the range of differences across individuals.
Agreement is usually assessed with methods that examine paired differences directly and define how large a difference can be while remaining acceptable. Diagnostic performance asks another question: whether a test correctly classifies people according to a reference standard. The primary endpoint specified in the registry is correlation, not an agreement statistic or a classification measure. The registry does not report an agreement analysis, diagnostic thresholds, sensitivity, specificity, or a reference-standard comparison for this endpoint.
How should the two group coefficients be compared? The coefficients are descriptive estimates calculated separately for glaucoma subjects and healthy controls. Their numerical difference is not itself a test of a group interaction or a test of equal correlations. In general, formal comparisons require an appropriate statistical procedure and information about the sampling uncertainty and dependence structure. The record does not report a formal comparison, confidence intervals, standard errors, or p-values. It is therefore not possible to determine from the posted coefficients alone whether the apparent difference between groups is more than sampling variation.
What does the reported timeframe mean? The primary outcome lists the baseline visit, 1 hour. This identifies the stated measurement window; it does not establish how much time separated HEP and OVF testing, whether the test order was randomized, or whether the correlation was calculated from a single measurement or an average of repeated measurements. Those procedural details are not reported in the registry record.
What is the role of repeatability? Repeatability concerns how similar measurements are when a measurement procedure is repeated under specified conditions. It is distinct from correlation between two different devices: repeatability evaluates consistency within a method, whereas the primary endpoint evaluates association between methods. The study summary says that test-retest reliability of both perimetry parameters would be investigated, and a secondary outcome names repeatability of OCT parameters. The registry record does not provide a posted repeatability estimate or the statistical method used for it.
The summary also describes investigation of novel optic nerve head imaging parameters and automated structure-function reports. A secondary outcome concerns correlations between the automated report and clinical impression. These broader aims do not change the interpretation of the primary endpoint, and the specific methods and results for them are not reported in the registry record.
What do planning calculations represent? In general, a sample-size planning illustration asks what sample size might be needed under specified assumptions to detect an effect. The planning assumptions associated with this study use a two-sided significance level of 0.05, equal allocation between the two arms, a normal approximation, and fractions of the observed effect. Such illustrations are not observed power and do not establish what assumptions were used to design the trial. The registry does not report the trial's own sample-size justification or statistical plan.
10. Multiplicity, Missing Data and Analysis Populations
The registry names the primary correlation endpoint and lists secondary outcomes concerning repeatability of OCT parameters and correlations between the automated structure-function report and clinical impression. It also records three outcome measures, but the statistical procedures for the secondary outcomes are not reported in the registry record. The record does not describe a prespecified hierarchy, multiplicity adjustment, or how interpretation would account for examining more than one outcome.
Multiplicity matters in general because evaluating several outcomes or several comparisons can increase the chance of obtaining an apparently notable result by chance. Whether adjustment is appropriate depends on the study questions and analysis plan. Without a reported plan, it is not possible to establish whether any adjustment was intended or performed.
The registry does not report missing-data counts, reasons for missingness, or methods for handling incomplete measurements. It does not state whether analyses followed an intention-to-test, per-protocol, or other analysis-population definition. The recorded arm sizes are 79 for glaucoma subjects and 36 for healthy controls, but the number of complete paired HEP and OVF measurements used to calculate each coefficient is not reported in the registry record. Thus, the denominator for the primary correlations cannot be confirmed from the posted information.
There is also no reported account of exclusions after enrollment, protocol deviations, or sensitivity analyses for missing or influential observations. These omissions limit assessment of whether the reported coefficients represent all enrolled participants or a subset with available measurements.
11. Safety Results
The registry reports serious adverse events as affected participants divided by participants at risk.
| Arm | Participants with serious adverse events | Participants at risk | Percentage |
|---|---|---|---|
| Glaucoma Subjects | 0 | 79 | 0.0% |
| Healthy Controls | 0 | 36 | 0.0% |
Deaths
| Arm | Deaths (participants affected of at risk) |
|---|---|
| Glaucoma Subjects | 0 of 79 |
| Healthy Controls | 0 of 36 |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 0% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| Glaucoma Subjects | 0 of 79 |
| Healthy Controls | 0 of 36 |
The registry lists zero serious adverse events among 79 glaucoma subjects at risk and zero among 36 healthy controls at risk. It also lists zero other adverse events in each group and zero deaths in each group. These are the posted counts for the groups and event categories recorded in the safety information.
In general, a zero event count means no events of that category were recorded for the stated group in the registry results. It does not prove that the underlying event risk is zero, and it does not establish that a procedure is free of risk. The precision of an event estimate depends on the observation period, follow-up completeness, event definitions, and how events were collected. Those details are not reported in the registry record.
The listed procedures are diagnostic tests, including perimetry and OCT. The record does not provide event timing, attribution to a specific procedure, definitions for serious or other adverse events, or a narrative of safety monitoring. The counts should therefore be read as registry-reported events, not as a comparative safety analysis or evidence about causation.
12. What the Result Does, and Does Not, Mean
The primary results describe how HEP and OVF mean deviation measurements were associated within the glaucoma and healthy-control groups. The reported coefficients are -0.86 for glaucoma subjects and -0.37 for healthy controls. Their signs indicate inverse associations on the recorded scales, and their magnitudes indicate different strengths of linear association. The registry does not report uncertainty intervals or a test comparing these estimates.
These results do not show whether the devices give interchangeable values. Correlation can be strong even when methods differ systematically, and it does not directly quantify the size of paired measurement differences. The results also do not establish whether HEP identifies glaucoma with the same diagnostic accuracy as OVF. That conclusion would require relevant diagnostic performance analyses and an appropriate reference standard, neither of which is reported for the primary endpoint.
The broader study aimed to examine reproducibility, OCT parameters, and automated structure-function reporting. The presence of those aims does not provide evidence about their findings: the registry record does not report numerical results for the named secondary outcomes. Similarly, the completion status and recorded enrollment describe the study record but do not resolve uncertainty about missing paired measurements or analysis populations.
In general, a useful validation assessment combines association with evidence about agreement, repeatability, and performance for the intended diagnostic task. Each addresses a distinct question. The posted primary coefficients contribute evidence about association only. Without confidence intervals, testing details, the paired-data denominator, and fuller diagnostic analyses, the precision and practical implications of the coefficients remain uncertain. The findings should be interpreted as a limited statistical description of the relationship between the two MD measures, not as proof of equivalence or a clinical recommendation.
13. Why This Trial Matters Statistically
This study raises an important statistical distinction in diagnostic-method research: measurements from two instruments can be associated without agreeing closely enough to be interchangeable. Correlation is a familiar and useful summary, but it addresses co-variation rather than the size of individual differences. Keeping that distinction clear prevents a correlation result from carrying more meaning than its estimand supports.
The separate coefficients for glaucoma subjects and healthy controls also illustrate why group context matters. Correlation depends in part on the distribution and range of measurements within the group being analyzed. As a result, coefficients from groups with different distributions should not automatically be read as a direct comparison of instrument performance. Formal uncertainty and group-comparison methods would be needed to assess whether the estimates differ beyond sampling variation; these are not reported in the registry record.
The study also links functional testing with structural imaging and automated structure-function reporting. Statistically, these are related but distinct measurement questions: association between perimetry devices, repeatability of measurements, and correspondence between automated reports and clinical impression. Naming these aims helps distinguish the different evidence needed for each, while the registry's primary posted results address only the correlation endpoint.
The record therefore offers a useful example of why endpoint definitions, analysis populations, uncertainty estimates, and agreement methods matter in validation studies. The posted coefficients are informative about association, but a fuller interpretation would require additional results and methodological detail that are not reported in the registry record.
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.
Several limitations are specific to the information recorded for this study. The primary outcome is described as a correlation between HEP and OVF MD, but the registry does not report the number of complete paired measurements used for either group. The recorded arm sizes therefore cannot by themselves confirm the endpoint denominators or whether missing measurements affected the estimates.
No confidence intervals, standard errors, p-values, or formal comparison of the group-specific correlations are reported. The coefficients consequently have no displayed measure of precision, and the apparent difference between groups cannot be evaluated inferentially from the registry results. The record also does not provide raw paired measurements, making it impossible to assess outliers, nonlinearity, or the distribution of individual differences.
The endpoint does not directly assess agreement or diagnostic accuracy. The record does not report an independent reference standard, classification thresholds, sensitivity, specificity, or a comparison of diagnostic errors. Details such as the order of HEP and OVF testing, repeat testing procedures, and the relationship of randomization to test administration are not reported in the registry record.
For the broader study aims, numerical results and analysis methods for OCT repeatability and automated structure-function reporting are not reported. The statistical analysis plan, sample-size justification, handling of missing data, and approach to multiple outcomes are also not reported in the registry record. These gaps limit assessment of precision, reproducibility, and the scope of conclusions that can be drawn from the posted results.
15. Related Tutorials
16. Related Statistical Calculators
17. Sources
- ClinicalTrials.gov: Heidelberg Edge Perimetry (HEP) Detecting Glaucomatous Visual Field Defects — NCT02526654.
- Registry record last updated 2019-09-16; status last verified 2019-08.
- Results first posted 2019-09-16.
- 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.
Continue through the Clinical Biostats statistical pathway
Use this trial's endpoint as a starting point for deeper study of effect measures, confidence intervals, and sample-size planning.
18. Record Summary
The registry describes a completed, randomized, parallel study with no masking, sponsored by Wills Eye. It records enrollment of 115 and study dates from 2014-12 to 2015-06. The groups were glaucoma subjects and healthy controls, who were to undergo HEP and OVF visual field testing and OCT imaging.
The primary endpoint was Pearson correlation between HEP and OVF mean deviation measurements at the baseline visit, 1 hour. The posted coefficients were -0.86 for glaucoma subjects and -0.37 for healthy controls. The registry lists secondary outcomes on OCT repeatability and correlations between an automated structure-function report and clinical impression, but does not report numerical results for those outcomes.
Interpretation is limited because confidence intervals, analysis denominators, missing-data handling, and a formal comparison of the group coefficients are not reported in the registry record. The primary correlations describe association, not agreement or diagnostic equivalence.