(150 days)
The Guided Progression Analysis for the Humphrey® Field Analyzer II (HFA II) and Humphrey® Field Analyzer II- i series is a software analysis module that is intended for use as a diagnostic device to aid in the detection and management of ocular diseases including, but not limited to, glaucoma. It is also intended to compare change over time and determine if statistically significant change has occurred.
The Carl Zeiss Meditec, Inc. Guided Progression Analysis is a software analysis module for the Humphrey® Field Analyzer II (HFA II) and Humphrey® Field Analyzer II - i series (HFA II - i) that assists practitioners with the detection, measurement, and management of progression of visual field loss. It aids in assessing change over time, including change from baseline and rate of change. It is intended for use as a diagnostic device to aid in the detection and management of ocular diseases including, but not limited to, glaucoma.
The Carl Zeiss Meditec, Inc. Guided Progression Analysis (GPA) is a software package for the Humphrey Field Analyzer II and II - i series that is designed to help practitioners identify progressive visual field loss in glaucoma patients. GPA compares the visual field test results of up to 14 follow-up tests to an established baseline over time and determines if there is statistically significant change. The GPA printout highlights any changes from baseline that represent larger than expected clinical variability, and it provides simple plain-language messages such as "Possible Progression" or "Likely Progression" whenever changes show consistent and statistically significant loss. The GPA printout also presents the Visual Field Index (VFI), a global index which reports a measure of the patient's remaining useful vision in the form of a percentage, as well as the VFI Rate of Progression plot which provides a trend analysis of the patient's overall visual field history and indicates a 3-5 year projection of the VFI regression line if the current trend continued.
The provided FDA 510(k) summary for the Guided Progression Analysis (GPA) for the Humphrey® Field Analyzer II and II - i series does not detail specific acceptance criteria in a quantitative table format or a standalone study with a predefined set of performance metrics that the device had to meet. Instead, the submission relies on:
- Substantial Equivalence: Demonstrating that the GPA software is functionally equivalent to predicate devices and does not raise new questions regarding safety and effectiveness.
- Clinical Literature Review: Citing published research that discusses GPA's development and its successful use in identifying statistically significant visual field progression, particularly referencing its incorporation of metrics from the Early Manifest Glaucoma Trial (EMGT).
- Sponsored Study on Test-Retest Variability: A study to quantify perimetric test-retest variability in glaucoma subjects, which was used to establish limits for change at different significance levels based on test-retest variability in glaucomatous visual fields. This allows GPA to indicate when change exceeds normal test-retest variability.
Therefore, a table of explicit acceptance criteria and corresponding reported device performance, in the traditional sense of a validation study with pre-defined thresholds, cannot be directly extracted or constructed from the provided text. The "performance" is described in terms of its ability to identify statistically significant change based on established variability data, rather than specific sensitivity/specificity figures against an external gold standard.
However, I can provide a summary of the information available in the document regarding the study that underpins the device's claims.
Acceptance Criteria and Reported Device Performance
As stated, explicit acceptance criteria (e.g., minimum sensitivity, specificity, or accuracy targets) are not provided in this 510(k) summary. The "performance" is intrinsically linked to its ability to identify changes beyond normal test-retest variability.
| Criterion Type | Acceptance Criteria (Not explicitly stated in the document) | Reported Device Performance (as described in the document) |
|---|---|---|
| Efficacy in detecting progression | Implicit: Ability to identify statistically significant visual field progression | GPA incorporates visual field progression metrics successfully used in EMGT. It determines if statistically significant change has occurred by comparing follow-up tests to a baseline and highlighting changes that represent larger than expected clinical variability. Provides plain-language messages like "Possible Progression" or "Likely Progression". |
| Statistical Robustness | Implicit: Ability to distinguish true change from test-retest variability | Results from a sponsored study established limits for change at different significance levels based on test-retest variability in glaucomatous visual fields. GPA indicates when change at a given test location exceeds this test-retest variability. |
| Aid in management | Implicit: Provides actionable information for clinicians | Presents Visual Field Index (VFI), VFI Rate of Progression plot, and trend analysis with 3-5 year projection to aid in estimating future visual status. |
-
Sample size used for the test set and the data provenance:
- Sample Size: 363 qualified glaucoma subjects.
- Data Provenance: Data was collected across a worldwide nine-site study. It is not specified if it was retrospective or prospective, but the description ("Each subject was tested four times within one month") suggests a prospective data collection for the purpose of establishing test-retest variability.
-
Number of experts used to establish the ground truth for the test set and the qualifications of those experts:
- The document does not mention the use of experts to establish a "ground truth" for the test set in the traditional sense of defining disease progression. The study focused on quantifying perimetric test-retest variability in glaucoma subjects. The "ground truth" or reference for this study was the inherent variability of visual field measurements themselves, not an expert-determined clinical diagnosis of progression.
-
Adjudication method for the test set:
- Adjudication methods (like 2+1 or 3+1) are typically used when experts are determining a ground truth for a diagnostic outcome. Since the study focused on quantifying test-retest variability rather than an expert-adjudicated ground truth for progression, no adjudication method is described or implied in the provided text.
-
If a multi-reader multi-case (MRMC) comparative effectiveness study was done, If so, what was the effect size of how much human readers improve with AI vs without AI assistance:
- No MRMC comparative effectiveness study is described in this 510(k) summary. The device, Guided Progression Analysis (GPA), is a software analysis module designed to assist practitioners, but its performance is described in terms of its algorithmic output based on statistical analysis of visual field data, not human reader performance with or without AI (in this case, "AI" refers to the GPA algorithm). The summary indicates that "the results allow HFA GPA to indicate when the change... exceeds the test-retest variability," which implies the software's direct output.
-
If a standalone (i.e., algorithm only without human-in-the-loop performance) was done:
- Yes, the core study mentioned (quantifying perimetric test-retest variability) and the subsequent development of GPA to use this data to identify statistically significant changes represent a standalone algorithmic function. The GPA software, without human intervention in its analysis, compares visual field test results to a baseline and determines statistical significance of changes, providing messages like "Possible Progression" or "Likely Progression." This is an algorithm-only function based on statistical rules derived from the variability study.
-
The type of ground truth used (expert consensus, pathology, outcomes data, etc.):
- The "ground truth" used for the development and validation of the statistical thresholds within GPA was data on perimetric test-retest variability. This means the system's ability to declare "progression" is benchmarked against the statistically expected noise and fluctuations in visual field measurements in glaucoma patients, as determined by the sponsored study. It is not an expert consensus on true progression, pathology, or long-term outcomes data, although the underlying clinical effectiveness of detecting visual field progression is supported by reference to studies like EMGT.
-
The sample size for the training set:
- The document references the Early Manifest Glaucoma Trial (EMGT) literature (citations 3 and 4) as providing the "visual field progression metrics successfully used" in GPA. The EMGT study design document (cited as "Ophthalmology 1999; 106:2144-2153") would contain details about its sample size, which served as a foundational dataset for the conceptual framework of progression analysis incorporated into GPA. However, a specific "training set" sample size for the development of this particular software version (GPA for HFA II) is not explicitly stated in the document beyond the reference to EMGT and the "363 qualified glaucoma subjects" used for the test-retest variability study. The 363 subjects constituted a dataset for establishing variability limits, which are essentially statistical parameters used by the algorithm. It is unclear if these 363 subjects' data was used for training a machine learning model vs. establishing statistical thresholds.
-
How the ground truth for the training set was established:
- Given the reliance on EMGT, the "ground truth" for the principles of progression analysis would have been established within the EMGT, likely through clinical outcomes and expert evaluation of visual field series over time in relation to glaucoma diagnosis and treatment. For the test-retest variability study (the 363 subjects), the "ground truth" was derived from repeated measurements from the same subjects to quantify the inherent variability, rather than a clinical ground truth of progression. The document does not describe a machine learning-style training set with an explicitly adjudicated ground truth for progression; instead, it highlights the incorporation of established clinical knowledge and statistical principles from previous large-scale studies.
§ 886.1605 Perimeter.
(a)
Identification. A perimeter is an AC-powered or manual device intended to determine the extent of the peripheral visual field of a patient. The device projects light on various points of a curved surface, and the patient indicates whether he or she sees the light.(b)
Classification. Class I (general controls). The device is exempt from the premarket notification procedures in subpart E of part 807 of this chapter, subject to the limitations in § 886.9. The device is also exempt from the current good manufacturing practice requirements of the quality system regulation in part 820 of this chapter, with the exception of § 820.180, with respect to general requirements concerning records, and § 820.198, with respect to complaint files.