(240 days)
The Atellica IM Total hCG (ThCG) assay is for in vitro diagnostic use in the quantitative determination of human chorionic gonadotropin (hCG) in human serum using the Atellica IM Analyzer. The Atellica IM ThCG assay is intended for use as an aid in the early detection of pregnancy.
The Atellica IM Total hCG (ThCG) assay is an in vitro diagnostic assay for the quantitative determination of human chorionic gonadotropin (hCG) in human serum using the Atellica IM Analyzer. The assay utilizes a sandwich immunoassay principle with direct chemiluminescent technology. The reagents include a Lite Reagent containing goat polyclonal anti-hCG antibody labeled with acridinium ester and a Solid Phase containing mouse monoclonal anti-hCG antibody covalently coupled to paramagnetic particles. Ancillary reagents like Atellica IM ThCG DIL are also part of the system.
The Siemens Atellica IM Total hCG (ThCG) assay is an in vitro diagnostic device for the quantitative determination of human chorionic gonadotropin (hCG) in human serum, intended as an aid in the early detection of pregnancy.
Here's an analysis of its acceptance criteria and the study that proves its performance:
1. Table of Acceptance Criteria and Reported Device Performance:
The provided document outlines several performance characteristics, including detection capability, precision, and method comparison, against a predicate device (ADVIA Centaur Total hCG assay).
| Performance Characteristic | Acceptance Criteria (Implicit) | Reported Device Performance |
|---|---|---|
| Detection Capability | ||
| Limit of Blank (LoB) | Not explicitly stated as acceptance criteria, but lowest measurement result for a blank sample expected to be below LoD/LoQ. | 1.5 mIU/mL (IU/L) |
| Limit of Detection (LoD) | Probability of 95% detection for the lowest concentration. | 1.7 mIU/mL (IU/L) (determined using 601 determinations, with 300 blank and 301 low-level replicates, and an LoB of 1.5 mIU/mL (IU/L)). |
| Limit of Quantitation (LoQ) | Lowest amount of hCG with a percent total error of 30%. | 2.6 mIU/mL (IU/L) (determined using multiple samples prepared from WHO reference material (NIBSC code: 99/688) in the interval 1.9–6.7 mIU/mL (IU/L). All samples were assayed in replicates of 5 in each of 2 runs per day using 2 reagent lots, over a period of 5 days). |
| Precision | Not explicitly stated as numerical acceptance criteria, but typical expectations are low %CV for various levels. | Repeatability (Within-run): Serum 1 (2.4 mIU/mL): 8.8% CV; Serum 2 (12.6 mIU/mL): 2.8% CV; Serum 3 (782.0 mIU/mL): 1.8% CV. Control 1 (6.8 mIU/mL): 4.5% CV; Control 2 (23.4 mIU/mL): 2.6% CV; Control 3 (202.1 mIU/mL): 1.8% CV. |
| Within-Lab: Serum 1 (2.4 mIU/mL): 10.5% CV; Serum 2 (12.6 mIU/mL): 3.4% CV; Serum 3 (782.0 mIU/mL): 2.2% CV. Control 1 (6.8 mIU/mL): 5.5% CV; Control 2 (23.4 mIU/mL): 3.5% CV; Control 3 (202.1 mIU/mL): 2.7% CV. | ||
| Method Comparison | Correlation coefficient (r) ≥ 0.95 and a slope of 1.0 ± 0.10 compared to the ADVIA Centaur Total hCG assay. | r = 1.00 (for serum samples, n=115, in the range 2.4-947.7 mIU/mL (IU/L)). |
| Regression Equation: $y = 0.94x + 0.23 mIU/mL (IU/L)$. The slope (0.94) is within the 1.0 ± 0.10 criterion. The intercept (0.23) indicates a small constant bias. | ||
| Interference | Not explicitly stated as acceptance criteria, but generally a bias of ≤ 10% is considered acceptable for common interfering substances. | Most reported bias values for various substances (Acetaminophen, Acetylsalicylic acid, Atropine, Caffeine, EDTA, Ethanol, Gentisic acid, Heparin, Human Serum Albumin, Ibuprofen) are well within the typical ±10% range. For HIL (Hemoglobin, Bilirubin, Lipemia), the assay is designed to have ≤ 10% interference, and reported values are generally within this range (highest reported is 8.9% for Hemoglobin at 5.6 mIU/mL). |
2. Sample Size Used for the Test Set and Data Provenance:
- Detection Capability (LoD): 601 determinations, with 300 blank and 301 low-level replicates. The document doesn't specify data provenance (country of origin or retrospective/prospective).
- Detection Capability (LoQ): Multiple samples prepared from the World Health Organization (WHO) human chorionic gonadotropin reference material (NIBSC code: 99/688). No specific sample count is given for the test set, but it implies a study on prepared samples. Data provenance is implied to be laboratory-based and controlled using a recognized international standard.
- Precision: Not a "test set" from patients, but rather control materials and serum pools. Samples were assayed in duplicate in 2 runs per day for 20 days using two reagent lots on each of two Atellica IM analyzers, leading to
320data points for each serum pool and control (2 duplicates * 2 runs/day * 20 days * 2 analyzers * 1 sample type / 2 for reagent lots = 320 for each condition) - Method Comparison: 115 serum specimens. The data provenance (country of origin, retrospective or prospective) is not specified.
- Interference: "Serum samples" used, concentrations given, but no specific number of unique patient samples mentioned.
- Expected Values: 366 serum samples from 192 apparently healthy non-pregnant females and 174 apparently healthy postmenopausal females. The provenance is not specified. This appears to be a prospective collection for establishing reference ranges.
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications of Those Experts:
This information is not applicable as the Atellica IM Total hCG assay is a quantitative diagnostic test for a biomarker (hCG) in serum. The 'ground truth' for such assays is typically established by:
- Reference Methods: Highly accurate and precise laboratory methods.
- Certified Reference Materials: Materials with an assigned value from an authoritative source (like the WHO reference material mentioned for LoQ).
- Clinical Diagnosis: For outcomes like pregnancy, a clinical diagnosis confirms the state, and the device then quantifies the biomarker associated with that state.
There is no mention of human experts establishing a 'ground truth' in the context of image interpretation or subjective diagnosis for this device.
4. Adjudication Method for the Test Set:
This is not applicable as the Atellica IM Total hCG assay is an automated quantitative assay. There is no subjective interpretation or need for adjudication by experts as would be found in image-based diagnostic systems.
5. 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:
This is not applicable. The Atellica IM Total hCG assay is an automated in vitro diagnostic device, not an AI-assisted diagnostic tool that involves human readers interpreting cases. Therefore, an MRMC study related to human reader improvement with AI assistance is not relevant.
6. If a Standalone (i.e., algorithm only without human-in-the-loop performance) Was Done:
The performance studies described (detection capability, precision, method comparison, interference, expected values) represent the standalone performance of the Atellica IM Total hCG assay. This device is designed to provide quantitative results directly from serum samples through an automated analyzer, without direct human-in-the-loop interpretation that would alter the result. The device performance itself is evaluated.
7. The Type of Ground Truth Used:
- Detection Capability (LoD, LoQ): Determined using prepared samples traceable to the World Health Organization (WHO) 4th International Standard (NIBSC code: 99/688) for human chorionic gonadotropin. This is a form of certified reference material ground truth.
- Precision: Internal serum pools and commercial controls with known concentrations served as the basis for evaluating reproducibility.
- Method Comparison: The ADVIA Centaur Total hCG assay (predicate device) served as the comparative reference, implying that its established measurements were the 'ground truth' for comparison, to show substantial equivalence.
- Interference: Spiked serum samples (serum with known amounts of hCG and interferents added) were used.
- Expected Values: Clinical classification of "non-pregnant females" and "postmenopausal females" served as the basis for establishing reference intervals. This aligns with clinical outcome/classification ground truth.
8. The Sample Size for the Training Set:
The document does not explicitly describe a "training set" in the context of machine learning or AI, as this is an immunoassay device. The studies described are performance validation studies. The calibration of the device (using Atellica IM CAL B) and the determination of its "master curve" can be thought of as akin to a training process for an assay, but no sample size for this is provided.
9. How the Ground Truth for the Training Set Was Established:
As mentioned, there isn't a "training set" in the AI/ML sense. The "ground truth" for the assay's operational parameters (like calibration and standardization) is established through:
- Standardization against the World Health Organization (WHO) 4th IS 75/589 reference material. This is a universally recognized standard, meaning calibrator values are traceable to this international reference.
- Two-point calibration using specific calibrators (Atellica IM CAL B) which themselves are assigned values traceable to the WHO standard.
§ 862.1155 Human chorionic gonadotropin (HCG) test system.
(a)
Human chorionic gonadotropin (HCG) test system intended for the early detection of pregnancy —(1)Identification. A human chorionic gonadotropin (HCG) test system is a device intended for the early detection of pregnancy is intended to measure HCG, a placental hormone, in plasma or urine.(2)
Classification. Class II.(b)
Human chorionic gonadotropin (HCG) test system intended for any uses other than early detection of pregnancy —(1)Identification. A human chorionic goadotropin (HCG) test system is a device intended for any uses other than early detection of pregnancy (such as an aid in the diagnosis, prognosis, and management of treatment of persons with certain tumors or carcinomas) is intended to measure HCG, a placental hormone, in plasma or urine.(2)
Classification. Class III.(3)
Date PMA or notice of completion of a PDP is required. As of the enactment date of the amendments, May 28, 1976, an approval under section 515 of the act is required before the device described in paragraph (b)(1) may be commercially distributed. See § 862.3.