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

The LabonaCheck™ A1c is intended for the quantitative measurement of glycated hemoglobin in venous whole blood and capillary fingerstick samples. This device is intended for multiple patient, professional use. Measurement of percent glycated hemoglobin (HbA 1c) is effective in monitoring long-term glucose control in individuals with diabetes mellitus by using the LabonaCheck™ A1c. Only auto-disabling, single use lancing devices should be used with this system.

Device Description

The system consists of HbA1c Analyzer (MH 200) and HbA1c Test kit. The LabonaCheck™ AIc HbAic Analyzer measures the coloured responses of the LabonaCheck™ A1c HbA1c Test Kit by spectral reflectance.

AI/ML Overview

The LabonaCheck™ A1c device is intended for the quantitative measurement of glycated hemoglobin (HbA1c) in venous whole blood and capillary fingerstick samples for professional use. The provided document details several performance characteristic studies.

Here's an analysis of the acceptance criteria and study data:

1. Table of Acceptance Criteria and Reported Device Performance

The document does not explicitly state formal acceptance criteria in a dedicated section with pass/fail thresholds for all studies. However, the manufacturer's conclusions and noted percentage differences are used to infer an implied acceptance criterion where applicable.

Performance Characteristic	Implied Acceptance Criteria	Reported Device Performance
Precision/Reproducibility	Not explicitly stated, but typical expectations for HbA1c CV% are generally below 3-5% for clinical utility.	Internal Study (Within-run, Day-to-day, and Total Precision):

Reference A1c 4.3% (Mean 4.4%): Total Precision CV 3.0%
Reference A1c 6.2% (Mean 6.1%): Total Precision CV 2.6%
Reference A1c 8.9% (Mean 9.1%): Total Precision CV 1.9%
Reference A1c 11.7% (Mean 11.6%): Total Precision CV 2.2%
Reference A1c 14.3% (Mean 14.2%): Total Precision CV 1.6%
External Study (Point-of-Care Sites):
Interval 1 (approx 4.8% HbA1c): Overall CV% between 2.9% - 3.5% (across operators and sites)
Interval 2 (approx 8.8% HbA1c): Overall CV% between 2.7% - 3.6% (across operators and sites)
Interval 3 (approx 13.2% HbA1c): Overall CV% between 2.7% - 3.5% (across operators and sites) |
| Linearity/Assay Measuring Range | The study aims to support linearity from 4% to 15%. % Recovery between real and expected values should be close to 100%. (Implied successful if sponsor claims support for the range). | Supported linearity from 4% to 15%.
Expected 3.2% (Real 3.5%): 108% Recovery
Expected 5.6% (Real 5.8%): 103% Recovery
Expected 8.0% (Real 8.1%): 101% Recovery
Expected 10.4% (Real 10.5%): 101% Recovery
Expected 12.7% (Real 12.9%): 102% Recovery
Expected 15.1% (Real 15.4%): 102% Recovery
Expected 17.5% (Real 17.5%): 100% Recovery |
| Detection Limit | A reportable range of 4% to 15% A1c. | LoB Mean: 3.10%, LoD Mean: 4.12%. This suggests the lower end of the reportable range (4%) is supported. |
| Analytical Specificity (Interference) | Recovery within ±10% of control results for interfering substances. (For Hemoglobin F: average bias within ±10% up to 22.7% HbF. For variants C, D, E, S, Carbamylated Hb, and Rheumatoid factor: within ±10% of reference difference). For Hemoglobin concentration: within 10% difference. | Known Substances: All tested substances (Ascorbic acid, Bilirubin, Glucose, Hemoglobin (up to 20g/dL), Lipid, Albumin, K3EDTA, Heparin, Sodium fluoride, Sodium citrate, Acetaminophen, Metformin, Acetylsalicylic acid, Glybenclamide, Ibuprofen) showed no significant interference (implied within ±10%).
Hemoglobin Variants:
HbF: Average bias within ±10% up to 22.7%. The device is affected by HbF > 20%.
HbC, D, E, S: Satisfied acceptance criteria (within ±10% of reference difference).
Other interfering factors:
Carbamylated Hemoglobin & Rheumatoid Factor: Satisfied acceptance criteria (within ±10% of reference difference).
Hemoglobin Concentration: Satisfied acceptance criteria (within 10% difference) for concentrations ranging from 8.7g/dL to 21.0g/dL. The claimed functional range is 10.0g/dL - 20.0g/dL. |
| Incubation Time Study | Reference difference within 5% when R1 reagent and whole blood sample incubated for 2 to 3 minutes. (Implied acceptance met if results within 5%). | When R1 reagent and whole blood sample incubated for 2 to 3 minutes, reference difference was within 5%. (Acceptance criteria satisfied). |
| Method Comparison (Accuracy) w/ Predicate Device | Not explicitly stated but typical metrics for point-of-care devices include a certain percentage of results within ±X% or correlation coefficient/slope/intercept criteria. For this submission, comparison to predicate and reference devices is key. For the difference % to predicate, 90% or more within 10% is often an acceptance. | LabonaCheck™A1c vs. Tosoh G7: 100% (50/50) within ±10% difference. Slope 1.0552, Intercept -0.4360, Correlation 0.9708.
NycoCard vs. Tosoh G7: 96% (48/50) within ±10% difference. Slope 1.0613, Intercept -0.3304, Correlation 0.9695.
LabonaCheck™A1c vs. NycoCard: 96% (48/50) within ±10% difference. Slope 0.9873, Intercept -0.0516, Correlation 0.9873. |
| Point-of-Care Study (Accuracy) | Slope (0.9-1.1), Intercept (less than 1%), Correlation coefficient results of three POC sites satisfied the acceptance criteria. | Venous Blood Samples (All sites combined): Slope 1.0005 (95% CI 0.9796~~1.0215), Intercept -0.0292 (95% CI -0.2036~~-0.1451), Correlation 0.9870. (All within acceptance criteria).
Capillary Blood Samples (All sites combined): Slope 0.9692 (95% CI 0.9481~~0.9903), Intercept 0.4230 (95% CI 0.2523~~0.5937), Correlation 0.9859. (All within acceptance criteria). |
| Matrix Comparison | Not explicitly stated, but the sponsor concluded that the stated anticoagulants can be used, implying satisfactory performance based on the linear regression analysis. | K3EDTA vs. Sodium Heparin (combined results): Slope 0.9682 (95% CI 0.9279~~1.0085), Intercept 0.3190 (95% CI -0.0570~~0.6951), Correlation 0.9670.
K3EDTA vs. NaF (combined results): Slope 0.9487 (95% CI 0.9033~~0.9941), Intercept 0.4364 (95% CI 0.0127~~0.8602), Correlation 0.9569. (Concluded that these anticoagulants can be used). |

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

Precision Studies (Internal): The internal study details whole blood samples, tested per day using two meters and two lots of reagents. The table shows 5 reference A1c levels. The specific number of distinct patient samples is not explicitly stated, but rather "whole blood samples" were used, with replicates for statistical analysis, implying a controlled, prospective study design. Provenance is internal (CERAGEM Medisys Inc.).
Precision Studies (External): 3 intervals of HbA1c EDTA blood were analyzed. The specific number of distinct patient samples is not stated directly, but it specifies "three intervals of HbA1c EDTA blood." This was a multi-site prospective study conducted at three Point-of-Care (POC) sites. Provenance is external (three POC sites).
Linearity/Assay Measuring Range: 7 mixed control solutions were used. This is likely a prospective, controlled study using prepared samples. The provenance is internal to the manufacturer.
Detection Limit: A zero sample (blank) and five low HbA1c samples (n=60, Ns=60 for each, assayed twice a day in replicates of ten for three days). This is prospective, controlled study. Provenance is internal to the manufacturer.
Analytical Specificity (Interference):
- Common Substances: Whole blood sample pools (3 intervals) were spiked with interfering substances. This is a controlled, prospective study. Provenance is internal to the manufacturer.
- Hemoglobin Variants (F, C, D, E, S): Samples provided by the National Glycohemoglobin Standardization Program (NGSP). This implies retrospective use of well-characterized samples from the NGSP program. Provenance: NGSP (likely international, given their role in standardization).
- Carbamylated Hemoglobin & Rheumatoid Factor: 10 samples (whole blood and spiked samples). This is a controlled, prospective study. Provenance is internal to the manufacturer.
- Hemoglobin Study: 50 whole blood samples. This is a prospective study using patient samples. Provenance is internal to the manufacturer.
Incubation Time Study: Whole blood samples of 3 HbA1c concentrations (Low, Medium, High). This is a controlled, prospective study. Provenance is internal to the manufacturer.
Method Comparison w/ Predicate Device: 50 EDTA venous whole blood patient samples (25 at each of two POC sites). Samples spanned 4.2-14.5% A1c. This is a prospective study using patient samples. Provenance is two Point-of-Care (POC) sites.
Point-of-Care Study (Accuracy):
- Venous Blood: 120 EDTA venous whole blood samples (40 at each of three POC sites). Samples spanned 4.5-14.6% A1c. This is a prospective study using patient samples. Provenance is three Point-of-Care (POC) sites.
- Capillary Blood: 120 capillary blood samples (40 at each of three POC sites). Samples spanned 4.2-14.8% A1c. This is a prospective study using patient samples. Provenance is three Point-of-Care (POC) sites.
Matrix Comparison: Venous whole blood samples (K3EDTA, Sodium heparin, NaF) collected from 40 donors. Samples spanned 4.5-14.1% A1c. This is a prospective study using patient samples. Provenance is internal to the manufacturer.

3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications

The studies primarily involve quantitative measurements for HbA1c.

For accuracy studies (Method Comparison, POC Study, Matrix Comparison), the ground truth was established by a reference method, Tosoh G7, and the predicate device, NycoCard® HbA1c, not by human experts interpreting results. The Tosoh G7 is likely an NGSP-certified laboratory method, considered a gold standard for HbA1c. The NycoCard® HbA1c is a legally marketed predicate device.
For Precision studies, there isn't a "ground truth" in the sense of expert interpretation; rather, the "reference A1c (%)" values are used as targets for reproducibility.
While operators (6 intended users across 3 sites for external precision and POC studies) performed the tests, they were not establishing ground truth, but rather generating test results to be compared against established reference values. The specific qualifications of these operators are not detailed beyond "intended users."

4. Adjudication Method for the Test Set

Not applicable. This device provides a quantitative measurement of a biomarker (HbA1c). The "ground truth" is established by laboratory reference methods (Tosoh G7, NGSP certified methods) or comparison to a predicate device, not by expert consensus or adjudication of qualitative interpretations.

5. If a Multi Reader Multi Case (MRMC) Comparative Effectiveness Study was Done

No, an MRMC comparative effectiveness study was not explicitly done. The studies mentioned involve multiple operators or sites (e.g., 6 operators at 3 POC sites), but this is primarily to assess reproducibility and accuracy in a real-world setting, not to measure the human diagnostic improvement with vs. without AI assistance. The device is a standalone quantitative analyzer, not an AI-assisted diagnostic aid for human readers.

6. If a Standalone (i.e., algorithm only without human-in-the-loop performance) was Done

Yes, the core of the submission provides standalone performance of the LabonaCheck™ A1c device. The studies evaluate the device's accuracy, precision, linearity, detection limits, and interference characteristics directly, comparing its results against reference methods or a predicate device. Human involvement is as operators performing the test, not as interpreters whose performance is being augmented by an algorithm.

7. The Type of Ground Truth Used

The ground truth used for these studies is primarily reference method data and, in some cases, predicate device data.

For accuracy and method comparison studies, the Tosoh G7 (a laboratory reference method, likely NGSP-certified) served as the primary reference for establishing "true" HbA1c values.
For interference studies of hemoglobin variants, NGSP-provided samples were used, with comparisons made against NGSP data (results of Primus - the predicate manufacturer).
For precision and linearity, controlled samples with known or targeted HbA1c concentrations were used.

8. The Sample Size for the Training Set

This submission describes a medical device (HbA1c analyzer and test kit) that performs a chemical assay for quantitative measurement. It is not an AI/ML algorithm that requires a "training set" in the conventional sense of machine learning. Therefore, a training set sample size is not applicable to this type of device.

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

As stated above, this is not applicable as the device is not an AI/ML algorithm requiring a training set. The device operates based on a boronate affinity assay, a chemical method.

Summary

Regulation Number and Section

§ 864.7470 Glycosylated hemoglobin assay.

(a)
Identification. A glycosylated hemoglobin assay is a device used to measure the glycosylated hemoglobins (A1a , A1b , and A1c ) in a patient's blood by a column chromatographic procedure. Measurement of glycosylated hemoglobin is used to assess the level of control of a patient's diabetes and to determine the proper insulin dosage for a patient. Elevated levels of glycosylated hemoglobin indicate uncontrolled diabetes in a patient.(b)
Classification. Class II (performance standards).