FDA 510(k) Search - By Innolitics (SaMD and AI Experts)

The Tina-quant Albumin Gen. 2 assay is an immunoturbidimetric assay intended for the quantitative determination of albumin in serum, plasma, urine, and CSF on Roche/Hitachi cobas c systems. Measurement of albumin aids in the diagnosis of kidney and intestinal diseases.

Device Description

The Tina-quant Albumin Gen. 2 assay employs an immunoturbidimetric test in which anti-albumin antibodies react with the antigen in the sample to form antigen/antibody complexes which, following agglutination are determined turbidimetrically.

AI/ML Overview

The provided document describes the Tina-quant Albumin Gen. 2 assay, an immunoturbidimetric test for quantitative determination of albumin in human urine, serum, plasma and CSF on Roche/Hitachi cobas c systems.

Here's an analysis of the acceptance criteria and the studies performed, based on the provided text:

1. Table of Acceptance Criteria and Reported Device Performance

The document presents separate substantial equivalence comparisons for the urine, serum/plasma, and CSF applications of the Tina-quant Albumin Gen. 2 assay against predicate devices. The "acceptance criteria" are implied by the measured performance parameters, as they are being compared to the predicate device and found to be substantially equivalent.

Table 1: Urine Application Performance Comparison

Feature	Acceptance Criteria (Predicate: Hitachi Microalbumin Urine Assay - K932950)	Reported Device Performance (Tina-quant Albumin Gen. 2 - Urine Application)
Precision
Within-run (mg/L)	Mean = 9.0, SD = 0.29, CV = 3.2%	Mean = 30.7, SD = 0.2, CV = 0.8%
	Mean = 22.1, SD = 0.30, CV = 1.4%	Mean = 108, SD = 1, CV = 0.7%
	Mean = 81.1, SD = 0.67, CV = 0.8%	Mean = 14.3, SD = 0.2, CV = 1.6%
		Mean = 252 mg/L, SD = 4, CV = 1.6%
Total (mg/L)	Mean = 9.0, SD = 0.92, CV = 10.1%	Mean = 31.2, SD = 0.5, CV = 1.7%
	Mean = 22.1, SD = 1.15, CV = 5.2%	Mean = 105, SD = 1, CV = 1.2%
	Mean = 81.1, SD = 0.78, CV = 1.0%	Mean = 13.6, SD = 0.4, CV = 2.8%
		Mean = 60.6, SD = 1.4, CV = 2.3%
Analytical Sensitivity	Lower Detection Limit = 3 mg/L	LoB = 2 mg/L, LoD = 3 mg/L, LoQ = 12 mg/L
Analytical Specificity	No interference from 18 common drugs	No interference at common therapeutic concentrations
		No unflagged high-dose hook effect up to 40000 mg/L
Interferences		Criterion: Recovery within ± 10%
Icterus	No significant interference up to 25 mg/dL	No significant interference up to I index of 50 (approx. 50 mg/dL conjugated bilirubin)
Hemolysis	Hemoglobin levels >300 mg/dL cause significant positive interference	No significant interference up to H index of 400 (approx. 400 mg/dL hemoglobin)
Lipemia	No significant interference up to an L index of 200	No interference by acetone, ammonia chloride, calcium, creatinine, y-globulin, glucose, urea, uric acid, urobilinogen (specific concentrations given)
Method Comparison	Linear Regression: y = 1.028x - 4.13 mg/L, r = 0.999	Passing Bablock: y = 1.023x + 3.64 mg/L, $\tau$ = 0.984
	(Predicate on Hitachi 917 analyzer)	(Tina-quant on c501 analyzer vs. Hitachi Microalbumin on Hitachi 917)

Table 2: Serum/Plasma Application Performance Comparison

Feature	Acceptance Criteria (Predicate: Behring N Antiserum to Human Albumin Assay - K972929)	Reported Device Performance (Tina-quant Albumin Gen. 2 - Serum/Plasma Application)
Precision
Intra-assay (g/L)	Mean: 46.5; CV: 4.3%
Inter-assay (g/L)	Mean: 44.7; CV: 4.4%
Repeatability (Within-run) (mg/L)	Not directly comparable, but predicate shows ~4% CV	Mean = 39.9, SD = 0.5, CV = 1.2%
		Mean = 66.6, SD = 1.4, CV = 1.2%
		Mean = 27.6, SD = 0.4, CV = 1.3%
		Mean = 62.5 mg/L, SD = 0.9, CV = 1.5%
Intermediate Precision (Total) (mg/L)	Not directly comparable, but predicate shows ~4% CV	Mean = 42.3, SD = 0.9, CV = 2.0%
		Mean = 70.5, SD = 1.6, CV = 2.2%
		Mean = 7.78, SD = 0.74, CV = 9.5%
		Mean = 36.2, SD = 0.7, CV = 2.1%
Analytical Sensitivity	Established by lower limit of reference curve, depends on protein concentrations	LoB = 1 g/dL, LoD = 2 g/dL, LoQ = 3 g/dL
Analytical Specificity	No interference from commonly used drugs is known.	No interference at common therapeutic concentrations
Interferences	Turbidity and particles may interfere. Bovine serum albumin may disturb.	Criterion: Recovery within ± 10%
Icterus	Not specified in detail, but general interference noted.	No significant interference up to I index of 60 (approx. 60 mg/dL conjugated bilirubin)
Hemolysis	Not specified in detail, but general interference noted.	No significant interference up to H index of 1000 (approx. 1000 mg/dL hemoglobin)
Lipemia	Not specified in detail, but general interference noted.	No significant interference up to L index of 1500 (approx. 1500 mg/dL intralipid)
Rheumatoid factors	Not specified.	≤ 1200 IU/mL do not interfere.
Method Comparison	Predicate calibrates by serial dilutions	Passing Bablock: y = -0.1320 + 0.9600x, $\tau$ = 0.919
		Linear Regression (provided): y = -0.0095 + 0.9572x, r = 0.993
	(Tina-quant on c501 analyzer vs. nephelometric Albumin test)	(Tina-quant on c501 analyzer vs. nephelometric Albumin test)

Table 3: CSF Application Performance Comparison

Feature	Acceptance Criteria (Predicate: Behring N Antiserum to Human Albumin Assay - K972929)	Reported Device Performance (Tina-quant Albumin Gen. 2 - CSF Application)
Precision	Not specified for CSF in predicate.
Repeatability (Within-run) (mg/L)		Mean = 99.2, SD = 1.4, CV = 1.4%
		Mean = 174, SD = 3, CV = 1.7%
		Mean = 383, SD = 4, CV = 1.0%
		Mean = 454 mg/L, SD = 4, CV = 0.8%
Intermediate Precision (Total) (mg/L)		Mean = 91.0, SD = 2.9, CV = 3.2%
		Mean = 389, SD = 7, CV = 1.7%
		Mean = 166, SD = 4, CV = 2.3%
		Mean = 366, SD = 5, CV = 1.3%
Analytical Sensitivity	Established by lower limit of reference curve, depends on protein concentrations	LoB = 20 mg/L, LoD = 36 mg/L, LoQ = 95 mg/L
Analytical Specificity	No interference from commonly used drugs is known.	No interference at common therapeutic concentrations
Interferences	Not specified for CSF.	Criterion: Recovery within ± 10%
Hemolysis		No significant interference up to H index of 1000 (approx. 1000 mg/dL hemoglobin)
Icterus		No significant interference up to I index 60 (approx. 600 mg/L conjugated and unconjugated bilirubin)
High dose hook-effect		No false result without a flag up to 30000 mg/L albumin concentration
Method Comparison	Predicate calibrates by serial dilutions.	Passing Bablock: y = 1.000x - 8.75 mg/L, $\tau$ = 0.936
		Linear Regression (provided): y = 0.991x + 0.301 mg/L, r = 0.992
	(Tina-quant on c501 analyzer vs. nephelometric Albumin test)	(Tina-quant on c501 analyzer vs. nephelometric Albumin test)

2. Sample Size Used for the Test Set and Data Provenance

Urine Application Method Comparison:
- Sample Size: n = 125
- Provenance: Not explicitly stated (e.g., country of origin, prospective/retrospective). The comparison is between the Tina-quant Albumin Gen. 2 assay on the c501 analyzer and the Hitachi Microalbumin assay on the Hitachi 917 analyzer (K953239). The sample concentrations were between 12.3 and 386 mg/L.
Serum/Plasma Application Method Comparison:
- Sample Size: n = 77
- Provenance: Not explicitly stated. The comparison is between the Tina-quant Albumin Gen. 2 assay on the cobas c501 analyzer and a nephelometric Albumin test. The sample concentrations were between 5.70 and 100 g/L.
CSF Application Method Comparison:
- Sample Size: n = 85
- Provenance: Not explicitly stated. The comparison is between the Tina-quant Albumin Gen. 2 assay on the cobas c501 analyzer and a nephelometric Albumin test. The sample concentrations were between 115 and 2640 mg/L.
Precision, Sensitivity, Specificity, Interferences:
- The sample sizes for these internal studies are not explicitly mentioned in the provided text for each specific test, but they are implied to be sufficient for analytical validation. Provenance is also not explicitly stated.

3. Number of Experts Used to Establish Ground Truth and Their Qualifications

The document describes an in-vitro diagnostic device for quantitative determination of albumin. The "ground truth" here is the actual albumin concentration in the samples, established by reference methods or predicate devices, rather than expert interpretation of images or other subjective data.
Therefore, the concept of "experts" to establish a subjective ground truth (like radiologists for imaging) is not directly applicable to this type of analytical device. The references are to other cleared devices (predicate devices) or established analytical methods.

4. Adjudication Method for the Test Set

As the device provides a quantitative measurement, and the "ground truth" is established by a reference method or predicate device performance, there is no mention of an "adjudication method" in the sense of reconciling differences between multiple human interpreters. This concept is typically relevant for subjective assessments (e.g., medical image interpretation).

5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study

No MRMC comparative effectiveness study was done. This type of study is relevant for devices that assist human readers (e.g., AI in radiology). The Tina-quant Albumin Gen. 2 assay is a standalone diagnostic test performed by an automated analyzer, not an assistive technology for human interpreters.

6. Standalone (Algorithm Only Without Human-in-the-Loop Performance)

Yes, the performance data presented is for the standalone device (Tina-quant Albumin Gen. 2 assay on the Roche/Hitachi cobas c systems) without human intervention for the measurement itself. The results are quantitative values generated by the automated system. The accuracy is assessed by correlation with predicate devices/methods.

7. The Type of Ground Truth Used

The ground truth for the performance studies (specifically method comparisons) was established by comparison to existing, legally marketed predicate devices or established nephelometric albumin tests. This implies that the predicate devices/methods themselves served as the reference standard for the "true" albumin values in the samples.
- For urine, the predicate was the Hitachi Microalbumin urine assay (K932950).
- For serum/plasma and CSF, the predicate was the Behring Nephelometric method (K972929).

8. The Sample Size for the Training Set

The document describes a 510(k) submission for substantial equivalence. This is for a conventional in-vitro diagnostic (IVD) assay, not a machine learning or AI-based device that typically involves "training sets." Therefore, the concept of a "training set" in the context of machine learning is not applicable here. The assay relies on established immunoturbidimetric principles, and performance is validated through analytical studies on test samples.

9. How the Ground Truth for the Training Set Was Established

As explained in point 8, the concept of a "training set" is not applicable to this device submission.

Summary

Regulation Number and Section

§ 866.5040 Albumin immunological test system.

(a)
Identification. An albumin immunological test system is a device that consists of the reagents used to measure by immunochemical techniques the albumin (a plasma protein) in serum and other body fluids. Measurement of albumin aids in the diagnosis of kidney and intestinal diseases.(b)
Classification. Class II (special controls). The device is exempt from the premarket notification procedures in subpart E of part 807 of this chapter subject to § 866.9.