(30 days)
The Access Estradiol assay is a paramagnetic particle, chemiluminescent immunoassay for the quantitative determination of estradiol levels in human serum, using the Access Immunoassay System.
The Access® Estradiol reagents and the Access® Immunoassay Analyzer comprise the Access® Immunoassay System for the quantitative determination of Estradiol levels in human serum.
This document describes the validation of the ACCESS® Estradiol assay, an immunoassay for determining estradiol levels in human serum.
Acceptance Criteria and Reported Device Performance
The acceptance criteria for the device are implied by the "Summary of Studies" section, where specific performance metrics are reported. There are no explicit "acceptance criteria" stated as thresholds; rather, the study presents the observed performance and concludes that the device is substantially equivalent to a predicate device.
| Performance Metric | Acceptance Criteria (Implied) | Reported Device Performance |
|---|---|---|
| Precision (Within Run) | CV within acceptable clinical ranges for estradiol measurement | Ranges from 19.7% CV (at 39 pg/ml) to 3.1% CV (at 2167 pg/ml) |
| Precision (Total) | CV within acceptable clinical ranges for estradiol measurement | Ranges from 20.0% CV (at 39 pg/ml) to 5.0% CV (at 2918 pg/ml) |
| Accuracy (Spiking Recovery) | Recovery between 80% and 120% | Ranging from 80% to 115% |
| Accuracy (Dilution Recovery) | Mean recovery close to 100% | Mean recoveries of 108% to 112%. Average dilution recovery at 1:2 was 106%. For samples >500 pg/ml, average % recovery was 107%. |
| Correlation with Predicate Device | Strong positive correlation (e.g., r > 0.95) | r = 0.98 |
| Analytical Sensitivity (Lowest Detectable Level) | Detectable level at or below clinically relevant thresholds | 20 pg/ml |
| Analytical Specificity (Cross-Reactivity with Estrone Sulfate) | Negligible or no detectable cross-reactivity | Not detectable when spiked at 3600 pg/ml |
| Analytical Specificity (Cross-Reactivity with Estriol Sulfate) | Low percentage apparent cross-reactivity | Apparent cross-reactivity of 0.004% when spiked at 10,000,000 pg/ml |
Study Details
-
Sample sizes used for the test set and the data provenance:
- Precision studies: The number of samples for precision studies is not explicitly stated, but typically involves multiple replicates of several control levels or patient samples. The data provenance is not specified (e.g., country of origin, retrospective/prospective).
- Accuracy (Spiking recovery): Ten serum samples were used. Data provenance not specified.
- Accuracy (Dilution recovery): 2 patient samples and 2 patient pools were used, along with 10 patient samples with estradiol >500 pg/ml. Data provenance not specified.
- Correlation: 103 samples were run. Data provenance not specified.
- Analytical Sensitivity: Implies samples were used to determine the lowest detectable level. Number of samples not specified.
- Analytical Specificity: S0 calibrator was spiked with estriol sulfate and estrone sulfate. Number of samples/spiked conditions not explicitly detailed beyond the two substances.
-
Number of experts used to establish the ground truth for the test set and the qualifications of those experts:
This is an in vitro diagnostic (IVD) device (immunoassay). The "accuracy" and "sensitivity" of such devices are typically established against reference methods or through spiked samples with known concentrations, rather than through expert consensus on images or clinical conditions. Therefore, the concept of "experts establishing ground truth for the test set" in the context of interpretation of results (like in image analysis) does not directly apply here. The "ground truth" for quantitative measurements like estradiol is the actual concentration determined by highly accurate methods (e.g., mass spectrometry, or certified reference materials) or by the known concentration of spiked analytes. The document does not specify details about these reference methods or expert panel for ground truth. -
Adjudication method for the test set:
Not applicable (N/A) for this type of quantitative immunoassay study. Adjudication methods like 2+1 or 3+1 are typically used for qualitative or interpretive assessments, such as in image reading where multiple experts might disagree on findings. Here, results are quantitative values compared to known or reference values. -
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:
Not applicable (N/A). This is an immunoassay, not an AI-assisted diagnostic tool involving human readers interpreting cases. -
If a standalone (i.e. algorithm only without human-in-the-loop performance) was done:
Yes, this is implicitly a standalone study. The performance characteristics (precision, accuracy, sensitivity, specificity, correlation) are reported for the ACCESS® Estradiol assay system itself (reagents + analyzer) as a device, independent of human interpretation or intervention in the measurement process beyond operating the instrument and following procedures. -
The type of ground truth used (expert consensus, pathology, outcomes data, etc.):
The ground truth used here consists of:- Known concentrations: For spiking recovery studies (estradiol spiked into serum samples) and analytical specificity studies (estrone sulfate, estriol sulfate spiked into calibrator).
- Reference measurements: The Abbott IMx® Estradiol Kit served as a comparator for the correlation study, implying it was considered a widely accepted and validated method for measuring estradiol. For precision and dilution recovery, the ground truth is the inherent concentration of estradiol in the samples, measured multiple times or diluted to known proportions.
-
The sample size for the training set:
Not applicable (N/A) as this is a traditional immunoassay validation report, not one describing a machine learning or AI-driven device that requires a distinct "training set." The development of the assay itself involves optimization and characterization over many samples, but these are not formally called "training sets" in the context of this type of document. -
How the ground truth for the training set was established:
Not applicable (N/A). See point 7.
§ 862.1260 Estradiol test system.
(a)
Identification. An estradiol test system is a device intended to measure estradiol, an estrogenic steroid, in plasma. Estradiol measurements are used in the diagnosis and treatment of various hormonal sexual disorders and in assessing placental function in complicated pregnancy.(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 § 862.9.