Search Results

The YSI 2900C Biochemistry Analyzer is intended for use by trained medical professionals as an in vitro diagnostic device for the quantitative testing of venous whole blood samples, fingerstick capillary whole blood samples, serum and plasma in the laboratory.

For in vitro diagnostic use.

Glucose measurements from the YSI 2900C Biochemistry Analyzer are used in the diagnosis and treatment of carbohydrate metabolism disorders, including diabetes mellitus and idiopathic hypoglycemia, and of pancreatic islet cell tumors. The YSI 2900C Biochemistry Analyzer is not intended for use in the screening or quantitative analysis on neonates. The YSI 2900C Biochemistry Analyzer is not intended for point of care (POC) use.

This test is for prescription use only.

The device under review, the 2900C Biochemistry Analyzer, is a laboratory instrument and In Vitro Diagnostic Device for determining glucose in human whole blood, plasma and serum taken from venous or capillary samples. The 2900C Biochemistry Analyzer is a semi-automated electronic device that incorporates fluidics for sampling, calibrating, and flushing, a membrane-immobilized enzyme-coupled electrochemical detection system with digital electronic control and has graphical user and data interfacing. It is designed for ambient indoor use in a technical laboratory environment.

The system is not intended to be a point of care product and is intended only for professional laboratory use.

The provided text describes the 510(k) premarket notification for the YSI 2900C Biochemistry Analyzer, which is a device for quantitative testing of glucose. It aims to demonstrate substantial equivalence to a predicate device, the YSI 2300 STAT PLUS Analyzer.

Here's a breakdown of the acceptance criteria and study information, based on the provided document:

1. A table of acceptance criteria and the reported device performance

The document does not explicitly present a table of "acceptance criteria" alongside specific numerical "reported device performance" in the format typically seen for algorithm performance (e.g., sensitivity, specificity, AUC thresholds). Instead, the studies aim to demonstrate substantial equivalence to a predicate device. The performance metrics discussed are related to the analytical performance of a biochemistry analyzer.

However, we can extract the precision acceptance criterion and its comparison to the predicate device:

It's important to note that for other tests like Method Comparison, Linearity, Specificity/Interference, etc., the "acceptance criteria" would be met by demonstrating that the new device's performance is comparable to the predicate, often through statistical equivalence testing within defined analytical limits, rather than fixed numerical thresholds for a specific algorithm output. The document states: "The results of the performance testing confirmed that the YSI 2900C Biochemistry Analyzer demonstrates substantial equivalence to the YSI 2300 StatPlus Analyzer."

2. Sample size used for the test set and the data provenance (e.g., country of origin of the data, retrospective or prospective)

The document provides details on the types of studies performed but does not specify the sample sizes used for the test sets in any of the performance studies (Method Comparison, Precision/Reproducibility, Linearity, Specificity/Interference, etc.).

There is no information provided regarding data provenance (e.g., country of origin, retrospective or prospective nature of the samples). The samples are stated to be "human whole blood, plasma and serum taken from venous or capillary samples."

3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts

This section is not applicable in the context of this device. The YSI 2900C Biochemistry Analyzer is an in vitro diagnostic device that quantitatively measures glucose in biological samples. The "ground truth" for its performance is established by comparing its measurements against reference methods or the predicate device, not by expert interpretation of images or clinical data. Therefore, there's no mention of experts establishing a ground truth for a test set in the way one would for an AI-powered diagnostic imaging device.

4. Adjudication method (e.g., 2+1, 3+1, none) for the test set

This section is not applicable for this type of device. Adjudication methods like "2+1" or "3+1" are typically used in studies involving expert readers interpreting medical images or clinical data, where consensus is needed to establish a "ground truth" when individual expert opinions might differ. For a biochemistry analyzer, the validation involves analytical comparisons to reference methods or a predicate device, not human interpretation requiring adjudication.

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 section is not applicable to this device. An MRMC study is designed to evaluate how AI assists human interpretation, typically in diagnostic imaging. The YSI 2900C Biochemistry Analyzer is a laboratory instrument that provides quantitative measurements; it does not involve human "readers" interpreting output in the same way, nor does it provide "AI assistance" for human interpretation. Its performance is evaluated analytically against established methods.

6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done

The device itself is a "standalone" system in its operation for measuring glucose. It performs the analysis and provides quantitative results without direct human "interpretation" of a processed image or complex data set. The performance testing described (Method Comparison, Precision, Linearity, etc.) represents the device's inherent analytical capabilities, which is analogous to "standalone" performance for an IVD. There is no "human-in-the-loop" component for the actual glucose measurement process that would affect its analytical performance.

7. The type of ground truth used (expert consensus, pathology, outcomes data, etc)

The "ground truth" in these performance studies is typically established by:

• Reference Methods: Highly accurate and precise analytical methods (e.g., isotope dilution mass spectrometry) that are considered the gold standard for glucose measurement, or
• Predicate Device: For demonstrating substantial equivalence, the performance of the new device (YSI 2900C) is compared against that of the legally marketed predicate device (YSI 2300 STAT PLUS) across various analytical parameters. The predicate device's established performance serves as the comparative "truth" for demonstrating equivalence.
• Known Concentrations: For linearity and limit studies, samples with precisely known glucose concentrations are used.

The document explicitly states: "A mirror of this testing was also performed concurrently on the predicate device, the YSI 2300 STAT PLUS Biochemistry Analyzer, for demonstration of substantial equivalence." This indicates the predicate device serves as the primary comparative "ground truth" for the equivalence assessment.

8. The sample size for the training set

This section is not applicable in the context of this traditional analytical device. The YSI 2900C Biochemistry Analyzer uses an "enzyme-mediated electrochemical detection" system, not a machine learning or AI algorithm that requires a "training set" of data. Its underlying technology is based on established biochemical and electrochemical principles.

9. How the ground truth for the training set was established

As there is no "training set" for an AI or machine learning algorithm in this device, this question is not applicable. The device's operational parameters and calibration are based on physical chemistry principles and factory calibration/validation, not a data-driven training process.

Ask a specific question about this device

Ask a specific question about this device

The Cholestech LDX™ System is a small, portable analyzer and test cassette system is for in vitro diagnostic use only and should not be used for testing in children under the age of 2 years. The Cholestech LDX™ System is comprised of the Cholestech LDX Analyzer and the following cassettes:

The Lipid Profile GLU cassette is for the quantitative determination of total cholesterol, HDL (high-density Ilpoprotein) cholesterol, triglycerides and glucose in whole blood. The TC/HDL (total cholesterol) ratio and estimated values for LDL (low-density lipoprotein) and non-HDL cholesterol are also reported.

The TC+HDL GLU cassette is for the quantitative determination of total cholesterol, HDL (high-density lipoprotein) cholesterol, and glucose in whole blood.

The TC GLU cassette is for the quantitative determination of total cholesterol and glucose in whole blood.

The Lipid Profile cassette is for the quantitative determination of total cholesterol. HDL (high-density lipoprotein) cholesterol, and triglycerides in whole blood. The TC/HDL (total cholesterol) ratio and estimated values for LDL (low-density lipoprotein) and non-HDL cholesterol are also reported.

The TC+HDL cassette is for the quantitative determination of total cholesterol and HDL (high-density lipoprotein) cholesterol in whole blood.

The TC cassette is for the quantitative determination of total cholesterol in whole blood.

· Cholesterol measurements are used in the diagnosis and treatment of disorders involving excess cholesterol in the blood and lipid and lipoprotein metabolism disorders.

· HDL (lipoprotein) measurements are used in the diagnosis and treatment of lipid disorders (such as diabetes mellitus), atherosclerosis, and various liver and renal diseases.

· Triglyceride measurements are used in the diagnosis and treatment with diabetes mellitus, nephrosis, liver obstruction, other diseases involving lipid metabolism, or various endocrine disorders.

· Glucose measurements are used in the diagnosis and treatment of carbohydrate metabolism disorders including diabetes mellitus, idiopathic hypoglycemia, and of pancreatic islet cell carcinoma.

The Cholestech LDX ™ system combines enzymatic methodology and solid-phase technology to measure total cholesterol, HDL cholesterol, triglycerides and glucose. Samples used for testing can be whole blood from a fingerstick (collected in a lithium heparin-coated capillary tube) or venipuncture. The sample is applied to the Cholestech LDX™ cassette®.

The cassette is then placed into the Cholestech LDX™ Analyzer where a unique system on the cassette separates the plasma from the blood cells. A portion of the plasma flows to the right side of the cassette and is transferred to both the total cholesterol and triglyceride reaction pads. Simultaneously, plasma flows to the left side of the cassette where the low- and very low-density lipoproteins (LDL and VLDL) are precipitated with dextran sulfate (50,000 MW) and magnesium acetate precipitating reagent.The filtrate, containing both glucose and HDL cholesterol, is transferred to both the glucose and HDL cholesterol reaction pads.

The Cholestech LDX ™ Analyzer measures total cholesterol and HDL cholesterol by an enzymatic method based on the method formulation of Allain et al, and Roeschlau. Cholesterol esterase hydrolyzes the cholesterol esters in the filtrate or plasma to free cholesterol and the corresponding fatty acid. Cholesterol oxidase, in the presence of oxygen, oxidizes free cholesterol to cholest-4-ene-3-one and hydrogen peroxide. In a reaction catalyzed by horseradish peroxidase, the peroxide reacts with 4-Aminoantipyrine and N-ethyl-N-sulfohydroxypropyl-m-toluidine, sodium sale (TOOS) to form a purple-colored quinoneimine dye proportional to the total cholesterol and HDL cholesterol concentrations of the sample.

The analyzer measures triglycerides by an enzymatic method based on the hydrolysis of triglycerides by lipase to glycerol and free fatty acids. Glycerol, in a reaction catalyzed by glycerol kinase, is converted to glycerol-3-phosphate. In a third reaction, glycerol-3phosphate is oxidized by glycerol phosphate oxidase to dihydroxyacetone phosphate and hydrogen peroxide. The color reaction utilizing horseradish peroxidase is the same as for the total cholesterol and HDL cholesterol. Estimated LDL cholesterol and non-HDL cholesterol and a TC/HDL ratio are calculated using the measured values for TC, HDL, and Triglycerides.

The analyzer measures glucose by an enzymatic method that uses glucose oxidase to catalyze the oxidation of glucose to gluconolactone and hydrogen peroxide. The color reaction utilizing horseradish peroxidase is the same as that for total cholesterol, HDL cholesterol and triglycerides. The resultant color in all the reactions is measured by reflectance photometry.

A brown (magnetic) stripe on each cassette contains the calibration information required for the Cholestech LDX ™ Analyzer to convert the reflectance reading (% R) to the total cholesterol, HDL cholesterol, triglycerides and glucose concentrations.

The provided text is a 510(k) summary for the Cholestech LDX™ System and primarily discusses device modifications and comparison to a predicate device. It certifies that verification studies were performed as required by risk analysis and all acceptance criteria were met. However, it does not provide the specific details of the acceptance criteria or the reported device performance for these studies. It also does not contain information about the sample size, data provenance, number of experts, adjudication methods, MRMC studies, standalone algorithm performance, or how ground truth was established for test and training sets.

Therefore, based solely on the provided text, I cannot fulfill most of the requested information regarding the study that proves the device meets the acceptance criteria. The document states that such studies were done and met acceptance criteria, but omits the specifics.

Here's what can be inferred or stated from the provided text, and what is missing:

Table of Acceptance Criteria and Reported Device Performance

Information Not Available in the Text: The document explicitly states, "Verification studies were performed as required by risk analysis and all acceptance criteria were met." However, it does not list the specific acceptance criteria (e.g., specific accuracy thresholds, precision ranges, etc.) or the detailed reported device performance (e.g., actual measured accuracy, precision values, etc.) from these studies. The modification pertains to updating the performance claim related to conjugated and unconjugated Bilirubin interference. While it mentions that less than 10% interference was seen at specified levels for various substances, this is a general statement from the predicate device's limitations, not a specific acceptance criterion for the current modification or the exact performance data achieved.

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

Information Not Available in the Text: The document states that "verification studies" were performed, but it does not specify the sample size (e.g., number of patients, number of samples) used for any test set or the provenance of the data (e.g., country of origin, retrospective or prospective nature of the data collection).

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

Information Not Available in the Text: The document details changes to an in vitro diagnostic (IVD) device for measuring cholesterol, triglycerides, and glucose. For IVD devices, ground truth is typically established by reference laboratory methods, not by human experts interpreting images or clinical cases. Therefore, the concept of "experts" as in radiologists or pathologists establishing ground truth is not applicable here. Even if it were (e.g., for method comparison studies requiring expert clinical correlation), the document does not mention any role for experts in establishing ground truth.

4. Adjudication Method for the Test Set

Information Not Available in the Text: Since the ground truth for an IVD device is generally established using reference methods (as opposed to human interpretation needing adjudication), an adjudication method as typically used in AI studies of imaging (e.g., 2+1, 3+1) is not applicable or described in this document.

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

Information Not Applicable/Available in the Text: This is an in vitro diagnostic (IVD) device, not an AI-assisted diagnostic imaging device. Therefore, MRMC studies comparing human readers with and without AI assistance are not relevant to this type of device and are not mentioned in the documentation.

6. If a Standalone (i.e., Algorithm Only Without Human-in-the-Loop Performance) Was Done

Information Not Applicable/Available in the Text: The Cholestech LDX™ System is a chemical analyzer, not an AI algorithm. Its performance is inherent to the device's enzymatic and solid-phase technology. The concept of "standalone algorithm performance" without human-in-the-loop is not directly applicable in the same way it would be for a software-as-a-medical-device (SaMD) that processes and interprets data for human review. The document describes the device's direct measurement capabilities.

7. The Type of Ground Truth Used

Inferred from Text: For an in vitro diagnostic device measuring analytes (cholesterol, HDL, triglycerides, glucose), the ground truth is typically established by reference laboratory methods (e.g., highly accurate and precise methods run on core laboratory instruments). While the document does not explicitly state "reference laboratory comparison" for ground truth, the context of an IVD device submission, especially one measuring these specific analytes, strongly implies this method.

8. The Sample Size for the Training Set

Information Not Applicable/Available in the Text: This is a chemical analyzer, not a machine learning or AI-based device that requires a "training set" in the computational sense. The device's operation is based on established enzymatic and chemical reactions, not on data-driven learning. Therefore, there is no "training set" in the context of AI/ML.

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

Information Not Applicable/Available in the Text: As noted above, there is no "training set" for this type of IVD device in the context of AI/ML. The device's calibration and performance are based on chemical principles and validation studies, not on learning from a training dataset.

Ask a specific question about this device

The i-STAT G cartridge with the i-STAT 1 System is intended for use in the in vitro quantification of glucose in arterial. venous or capillary whole blood in point of care or clinical laboratory settings.

Glucose measurements are used in the diagnosis, monitoring, and treatment of carbohydrate metabolism disorders including, but not limited to, diabetes mellitus, neonatal hypoglycemia, and pancreatic islet cell carcinoma.

The i-STAT G cartridge is used with the i-STAT 1 analyzer as part of the i-STAT 1 System to measure glucose in arterial, venous, or capillary whole blood for the diagnosis, monitoring, and treatment of metabolism disorders including, but not limited to, diabetes mellitus, neonatal hypoglycemia, idiopathic hypoglycemia, and pancreatic islet cell carcinoma.

The i-STAT 1 System is an in vitro diagnostic (IVD) medical device intended for the quantitative determination of various clinical chemistry tests contained within i-STAT cartridges using whole blood. The i-STAT 1 System consists of a portable blood analyzer (i-STAT 1 analyzer), single-use disposable test cartridges (i-STAT cartridges), liquid quality control and calibration verification materials, and accessories (i-STAT 1 Downloader/Recharger, i-STAT Electronic Simulator and i-STAT 1 Printer). The i-STAT 1 Sustem, including the i-STAT G cartridge, is designed for use by trained medical professionals in point of care or clinical laboratory settings and is for prescription use only.

The i-STAT G cartridge contains the required sensors and a fluid pack (calibrant pouch), a sample entry well and closure, fluid channels, waste chamber, and the necessary mechanical features for controlled fluid movement within the cartridge. The test is contained in a single-use, disposable cartridge. All the test steps and fluid movements occur within the i-STAT G cartridge. Cartridges require two to three drops of whole blood applied to the cartridge using a transfer device, by the trained user before the cartridge is placed within the analyzer.

The i-STAT 1 analyzer is a handheld, in vitro diagnostic analytical device designed to run only i-STAT test cartridges. The analyzer interacts with the cartridge to move fluid across the sensors and generate a quantitative result (within approximately 2 minutes).

Here's a breakdown of the acceptance criteria and the study details for the i-STAT G cartridge with the i-STAT 1 System, based on the provided document:

1. Table of Acceptance Criteria and Reported Device Performance

The document does not explicitly state "acceptance criteria" for each performance characteristic in a summarized table within the 510(k) summary. However, it indicates that the studies "met acceptance criteria" or "demonstrated equivalency" to a predicate device or allowable error.
For several tests (e.g., precision, linearity), the results are presented, and the implicit acceptance criterion is that these results fall within acceptable clinical or statistical limits, or demonstrate substantial equivalence to the predicate device.
For the purpose of this response, I will infer the acceptance criteria from the context of how the results are presented and the FDA's ultimate determination of "substantially equivalent."

• Arterial vs. i-STAT CHEM8+: N=173, Slope 1.00, Intercept 1.00, r 1.00
• Venous vs. i-STAT CHEM8+: N=164, Slope 1.00, Intercept 1.50, r 1.00
• Capillary vs. epoc Blood Analysis System: N=234, Slope 1.00, Intercept 2.00, r 1.00 (Table 10) |
  | Matrix Equivalence: Passing-Bablok regression (non-anticoagulated vs. heparinized) with slope ~1, intercept ~0, and high correlation (r ~1). | N=158, r 1.00, Slope 1.00, Intercept 0.00 (Table 11). |
  | EDTA Matrix Equivalence: Passing-Bablok regression (K2EDTA/K3EDTA vs. lithium heparin) with slope ~1, intercept ~0, and high correlation (r ~1). | K2EDTA vs. LiHep: N=43, r 1.00, Slope 1.03, Intercept -1.037 (Table 12)
  K3EDTA vs. LiHep: N=43, r 1.00, Slope 1.03, Intercept 0.015 (Table 13) |

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

• Precision (20-day, aqueous materials): N=80 for each of 5 fluid levels.
• Precision (Multi-site and operator-to-operator, aqueous materials): N=89-90 for each of 5 fluid levels.
• Precision (whole blood):
  • Venous whole blood: N=38 (20-90 mg/dL), N=67 (>90-150 mg/dL), N=32 (>150-250 mg/dL), N=15 (>250-400 mg/dL), N=12 (>400-700 mg/dL).
  • Arterial whole blood: N=9 (20-90 mg/dL), N=94 (>90-150 mg/dL), N=64 (>150-250 mg/dL), N=6 (>250-700 mg/dL).
  • Capillary whole blood: N=33 (20-90 mg/dL), N=53 (>90-150 mg/dL), N=37 (>150-250 mg/dL), N=16 (>250-700 mg/dL).
• Linearity: Whole blood samples of varying glucose levels across the reportable range.
• Limit of Quantitation (LoQ): Whole blood "altered to a low glucose level."
• Limit of Blank/Detection (LoB/LoD): Whole blood "altered to a blank glucose level" and "two (2) low glucose levels."
• Interference: Whole blood samples at low and high glucose levels.
• Oxygen Sensitivity: Whole blood samples "altered to four (4) glucose levels."
• Hematocrit Sensitivity: Whole blood samples at "three (3) hematocrit levels...at four (4) glucose levels."
• Altitude: Whole blood samples "across the reportable range."
• Method Comparison:

  • Arterial/Venous: N=571 (pooled data for i-STAT CHEM8+ comparison)

  • Arterial: N=173 (vs. i-STAT CHEM8+)

  • Venous: N=164 (vs. i-STAT CHEM8+)

  • Capillary: N=234 (vs. epoc Blood Analysis System)

• Matrix Equivalence (non-anticoagulated vs. heparinized): N=158
• EDTA Matrix Equivalence: N=43 for K2EDTA vs LiHep, N=43 for K3EDTA vs LiHep.

Data Provenance:

• Retrospective/Prospective: Primarily involves prospective testing of samples (aqueous materials, whole blood) under controlled conditions, or collection of patient samples for method comparisons.
• Country of Origin: Not specified in the provided text, but implied to be within the scope of where Abbott Point of Care (US company) conducts its clinical evaluations, likely the US or other regions following CLSI guidelines.

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

This information is not detailed in the provided document. The ground truth for analytical performance studies on diagnostic devices like this often relies on:

• Reference methods using highly accurate laboratory analyzers.
• Certified reference materials (like NIST SRM 965 mentioned for traceability).
• Clinically established ranges and allowable errors.
• The "experts" are primarily the laboratory scientists and statisticians who conduct and analyze the studies according to CLSI guidelines.

For method comparison studies, the "ground truth" is typically the result from a legally marketed predicate device (i-STAT CHEM8+ cartridge) or a well-established laboratory reference method (epoc Blood Analysis System). The document does not mention human expert annotation of samples for establishing ground truth, as it's an in-vitro diagnostic device not directly involving image interpretation or clinical decision-making by human experts for the output itself.

4. Adjudication Method for the Test Set

This information is not applicable in the traditional sense for an in-vitro diagnostic glucose test. Adjudication methods (like 2+1 or 3+1) are typically used in clinical studies where human readers or interpreters make subjective assessments that might differ, and a consensus or adjudication process is needed to establish a definitive ground truth. For quantitative measurements like glucose, the "adjudication" is inherent in the analytical process:

• Reference methods provide the comparative "true" value.
• Statistical analyses (e.g., Passing-Bablok regression, precision calculations) determine if the device's results are sufficiently close to this reference or internally consistent.
• Discrepancies would be investigated through quality control and root cause analysis, not expert adjudication of contradictory readings from the device itself.

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

No, an MRMC comparative effectiveness study was not done. This type of study (comparing human readers with and without AI assistance on multiple cases) is relevant for AI-powered diagnostic imaging devices where human interpretation is a key component. The i-STAT G cartridge is an in-vitro diagnostic device that directly measures glucose levels, not an AI assistance tool for human interpretation.

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

Yes, the studies presented are essentially "standalone" performance evaluations. The i-STAT G cartridge with the i-STAT 1 System functions as an automated system for quantifying glucose. The performance data presented (precision, linearity, LoQ, LoB/LoD, interference, sensitivity, altitude, method comparison, and matrix equivalence) are all evaluations of the device's analytical performance on its own, without direct real-time human interpretation of the glucose value itself as the primary output to be compared. Human involvement is for sample collection, device operation, and quality control, but the reported glucose value is generated by the "algorithm" (the device's embedded processes).

7. The Type of Ground Truth Used

The ground truth used for performance validation includes:

• Reference methods: For method comparison studies, the i-STAT CHEM8+ cartridge on the i-STAT 1 System (predicate device) and the epoc Blood Analysis System served as comparative methods to establish ground truth or reference values.
• Certified reference materials/Calibrators: Mention of NIST SRM 965 for traceability and "i-STAT Calibration Verification set" for precision studies indicates the use of highly accurate, traceable materials.
• Spiked samples: For linearity, LoQ, LoB/LoD, and interference studies, samples were "altered" or "spiked" to known concentrations of glucose or interfering substances.
• Statistical models and clinical thresholds: The "allowable error (±Ea)" indicates that performance is judged against pre-defined clinical or statistical acceptance limits.

8. The Sample Size for the Training Set

This information is not provided in the document. The document describes performance validation studies (test sets). For an IVD device like this, the "training set" would refer to data used during the development and optimization phases before the formal validation for regulatory submission. Details on development-stage datasets are generally not included in 510(k) summaries.

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

This information is not provided for the same reasons as point 8.

Ask a specific question about this device

The i-STAT CG8+ cartridge with the i-STAT 1 System is intended for use in the in vitro quantification of glucose in arterial, venous, or capillary whole blood in point of care or clinical laboratory settings.

Glucose measurements are used in the diagnosis, monitoring, and treatment of carbohydrate metabolism disorders including, but not limited to, diabetes mellitus, neonatal hypoglycemia, and pancreatic islet cell carcinoma.

The i-STAT 1 Analyzer is intended for use in the in vitro quantification of various analytes in whole blood or plasma in point of care or clinical laboratory settings. Analyzers and cartridges should be used by healthcare professionals trained and certified to use the system and should be used according to the facility's policies and procedures.

The i-STAT System is for in vitro diagnostics use. Caution: Federal law restricts this device to sale by or on the order of a licensed practitioner.

The i-STAT CG8+ cartridge is used with the i-STAT 1 analyzer as part of the i-STAT 1 System to measure glucose (Glu) in arterial, venous or capillary whole blood.

The i-STAT 1 System is an in vitro diagnostic (IVD) medical device intended for the quantitative determination of various clinical chemistry tests contained within i-STAT cartridges using whole blood. The i-STAT 1 System consists of a portable blood analyzer (i-STAT 1 analyzer), single-use disposable test cartridges (i-STAT cartridges), liquid quality control and calibration verification materials, and accessories (i-STAT 1 Downloader/Recharger, i-STAT Electronic Simulator and i-STAT 1 Printer). The i-STAT 1 System, including the i-STAT CG8+ cartridge, is designed for use by trained medical professionals in point of care or clinical laboratory settings and is for prescription use only.

The i-STAT CG8+ cartridge contains the required sensors, a fluid pack (calibrant pouch), a sample entry well and closure, fluid channels, waste chamber, and the necessary mechanical features for controlled fluid movement within cartridge. The i-STAT cartridge format allows all the tests in the cartridge to be performed simultaneously. All the test steps and fluid movement occur within the i-STAT CG8+ cartridge. Cartridges require two to three drops of whole blood, which are typically applied to the cartridge using a transfer device, by the trained user before the cartridge is placed within the analyzer.

The i-STAT 1 analyzer is a handheld, in vitro diagnostic analytical device designed to run only i-STAT test cartridges. The instrument interacts with the i-STAT CG8+ cartridge to move fluid across the sensors and generate a quantitative result (within approximately 2 minutes).

The acceptance criteria and study proving the device meets those criteria are detailed below, based on the provided FDA 510(k) summary for the i-STAT CG8+ cartridge with the i-STAT 1 System.

Acceptance Criteria and Reported Device Performance

The acceptance criteria are not explicitly listed in a single table with corresponding performance values in the provided document. Instead, performance expectations are implied through the comparison to the predicate device and the presentation of various study results (precision, linearity, detection limits, interference, sensitivity, and method comparison) against established CLSI guidelines or internal thresholds.

However, based on the provided information, a summary of key performance characteristics and their reported values can be presented as follows:

1. Table of Acceptance Criteria and Reported Device Performance (Implied from Study Results)

Note: The acceptance criteria are largely implied by the successful completion and positive results of studies designed according to CLSI (Clinical and Laboratory Standards Institute) guidelines, such as EP05-A3, EP06-Ed2, EP17-A2, EP07-ED3, EP37-ED1, EP09c-ED3, and EP35. Meeting the statistical benchmarks (e.g., specific SD, %CV, R², slope, intercept ranges, or being within "allowable error") for these guidelines indicates that the device's performance is acceptable for its intended use.

Study Details

Here's a breakdown of the study details based on the provided document:

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

• Precision Studies:

  • 20-day Precision (Aqueous): N=80 samples per level.
  • Multi-site/Operator Precision (Aqueous): N=90-96 samples per level (across 3 sites).
  • Whole Blood Precision: Venous (N=29-102 per range), Arterial (N=5-105 per range), Capillary (N=15-107 per range).

• Linearity: Whole blood samples of varying glucose levels. Specific N not provided for this section, but typically multiple points across the range are tested.

• Detection Limits (LoQ, LoB, LoD): Whole blood samples (altered to low/blank glucose levels). Specific N not provided.

• Analytical Specificity (Interference): Whole blood samples. Specific N for each substance not provided, but the study was extensive (Table 8 lists many substances).

• Other Sensitivity Studies:

  • Oxygen Sensitivity: Whole blood samples.
  • Hematocrit Sensitivity: Whole blood samples (at low, mid, high hematocrit levels).
  • Altitude: Whole blood samples.

• Method Comparison (with predicate device):

  • N = 547 (pooled data from arterial, venous, and capillary whole blood specimens).

• Matrix Equivalence:

  • N = 297 (venous and arterial whole blood specimens).

• Data Provenance:

  • The document states "collected across multiple point of care sites" for whole blood precision and method comparison studies. It also mentions "at one site" for the 20-day precision study and "at three (3) sites" for the multi-site precision study.
  • The document does not explicitly state the country of origin of the data or whether the studies were retrospective or prospective. However, clinical studies for 510(k) submissions are typically prospective, especially those involving patient samples collected concurrently with the study.

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

• For this type of in vitro diagnostic device (quantitative glucose measurement), the "ground truth" for the test set is established by comparative reference methods, not by expert consensus or interpretations of images by radiologists.
• The document mentions "a comparative method" (i-STAT CHEM8+ and epoc Blood Analysis System for glucose) for the method comparison study. These are legally marketed, validated laboratory or point-of-care devices that serve as the reference standard for measuring glucose.
• No human experts (like radiologists in an imaging study) are used to establish "ground truth" for quantitative lab tests in this context. Their role might be in collecting samples by healthcare professionals, but not in determining the true value of the analyte.

4. Adjudication Method for the Test Set

• Adjudication methods (e.g., 2+1, 3+1) are primarily relevant for imaging studies where human readers interpret data, and discrepancies need to be resolved.
• For quantitative in vitro diagnostic tests like glucose measurement, adjudication of results in the traditional sense is not applicable. The "ground truth" is determined by the output of a reference instrument or method, and performance is assessed by statistical agreement between the new device and the reference.

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

• No, an MRMC comparative effectiveness study was not done.
• This device is an in vitro diagnostic for quantitative measurement of glucose, not an AI-based imaging or diagnostic device that assists human readers. Therefore, the concept of human readers improving with AI assistance is not relevant to this submission.

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

• Yes, the performance characteristics described (precision, linearity, detection limits, interference, sensitivity) represent the standalone performance of the device (i-STAT CG8+ cartridge with i-STAT 1 System).
• The method comparison study also evaluates the device's performance against a reference method independently. Human involvement is primarily in operating the device and collecting samples, not in interpreting or enhancing the device's quantitative output.

7. The Type of Ground Truth Used

• The ground truth for the device's performance evaluation was established using comparative reference methods (e.g., i-STAT CHEM8+ glucose test, epoc Blood Analysis System) which are considered established and validated methods for glucose quantification.
• This is akin to using a gold standard laboratory test result for comparison.

8. The Sample Size for the Training Set

• The provided document is a 510(k) summary for a point-of-care in vitro diagnostic device, not an AI/Machine Learning device.
• Therefore, there is no "training set" in the context of machine learning model development. The device relies on electrochemical detection mechanisms and pre-calibrated algorithms, not on statistical models trained on large datasets in the way AI applications do.
• The studies described are for validation (test set), not training.

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

• As explained in point 8, there is no "training set" for this device in the machine learning sense.
• The fundamental principles and calibration of the glucose sensor (glucose oxidase-based amperometric peroxide detection) are based on established chemical and electrochemical principles in analytical chemistry, refined and validated during product development.

Ask a specific question about this device

The GEM Premier ChemSTAT is a portable critical care system for use by health care professionals to rapidly analyze lithium heparinized whole blood samples at the point of health care delivery in a clinical setting and in a central laboratory. The instrument provides quantitative measurements of sodium (Na+), Potassium (K+), Ionized Calcium (Ca++), Chloride (Cl-), Glucose (Glu), Lactate (Lac), Hematocrit (Hct), Creatinine (Crea), Blood Urea Nitrogen (BUN), Total Carbon Dioxide (tCO2), pH, and partial pressure of carbon dioxide (pCO2) from arterial and venous heparinized whole blood. These parameters, along with derived parameters, aid in the diagnosis of a patient's acid/base status, electrolyte and metabolite balance.

Electrolytes in the human body have multiple roles. Nearly all metabolic processes depend on or vary with electrolytes:

· Sodium (Na+) measurements are used in the diagnosis and treatment of aldosteronism, diabetes insipidus, adrenal hypertension, Addison's disease, dehydration, inappropriate antidiuretic secretion, or other diseases involving electrolyte imbalance.

· Potassium (K+) measurements are used to monitor electrolyte balance in the diagnosis and treatment of disease conditions characterized by low or high blood potassium levels.

· Ionized calcium (Ca++) measurements are used in the diagnosis and treatment of parathyroid disease, a variety of bone diseases, chronic renal disease and tetany. · Chloride (Cl-) measurements are used in the diagnosis and treatment of electrolyte and metabolic disorders, such as cystic fibrosis and diabetic acidosis.

· Glucose (Glu) measurement is used in the diagnosis, monitoring and treatment of carbohydrate metabolism disturbances including diabetes mellitus, neonatal hypoglycemia, idiopathic hypoglycemia, and of pancreatic islet cell carcinoma.

· Lactate (Lac) measurement is used to evaluate the acid-base status of patients suspected of having lactic acidosis, to monitor tissue hypoxia and strenuous physical exertion, and in the diagnosis of hyperlactatemia.

· Hematocrit (Hct) measurements in whole blood of the packed red cell volume of a blood sample are used to distinguish normal from abnormal states, such as anemia and erythrocytosis (an increase in the number of red cells).

· Creatinine (Crea) measurements are used in the diagnosis and treatment of renal diseases and in monitoring renal dialysis.

· Blood Urea Nitrogen (BUN) or urea measurements are used for the diagnosis, monitoring, and treatment of certain renal and metabolic diseases.

· Total carbon dioxide/tCO2 (also referred to as bicarbonate/HCO3-) is used in the diagnosis, monitoring, and treatment of numerous potentially serious disorders associated with changes in body acid-base balance.

· pH and pCO2 measurements in whole blood are used in the diagnosis and treatment of life-threatening acid-base disturbances.

The GEM Premier ChemSTAT system provides fast, accurate, quantitative measurements of Sodium (Na"), Potassium (K*), Ionized Calcium (Ca*), Chloride (Cl·), Glucose (Glu), Lactate (Lac), Hematocrit (Hct), Creatinine (Crea), Blood Urea Nitrogen (BUN), Total Carbon Dioxide (tCO2), pH, and partial pressure of carbon dioxide (pCO2) from arterial and venous lithium heparinized whole blood.

The provided text describes a Special 510(k) submission for an upgrade to the operating system of the GEM Premier ChemSTAT device. The device itself is an in vitro diagnostic (IVD) system for quantitative measurements of various blood parameters. The submission focuses on the software upgrade rather than a change in the device's fundamental function or performance.

Therefore, the "acceptance criteria" and "reported device performance" in this context refer to the successful verification and validation of the software upgrade and the continued adherence to the established performance of the unmodified device, as the indications for use and performance claims remain unchanged. The study proving this essentially consists of the software verification and validation activities.

Here's the information extracted from the document, tailored to the context of a software upgrade:

1. Table of Acceptance Criteria and Reported Device Performance

Since this is a software upgrade with no changes to the performance claims of the device, the general acceptance criteria are that the upgraded software performs as intended without adversely affecting the device's established performance specifications. The reported device performance is that these criteria were met.

2. Sample size used for the test set and the data provenance (e.g. country of origin of the data, retrospective or prospective)

The document does not specify a "test set sample size" or "data provenance" in the traditional sense for evaluating diagnostic performance. The focus is on software verification and validation. Therefore, the "sample" for testing the software functionality would be the various test cases and scenarios designed to validate the operating system upgrade and its interaction with the GEM Premier ChemSTAT application software.

The document states: "Performance data is limited to Software Verification and Validation as the scope of this Special 510(k) is specific to an operating system upgrade from Fedora 17 Linux to WindRiver LTS 18 Linux."

Further details on the specific number of test cases, the nature of the data (e.g., simulated, actual runs on the device), or its origin are not provided in this summary.

3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts (e.g. radiologist with 10 years of experience)

This information is not applicable to a software operating system upgrade as described. "Ground truth" in the context of expert consensus is typically relevant for diagnostic performance studies where human interpretation or a gold standard reference is needed (e.g., pathology for an imaging device). Here, the "ground truth" is the proper functioning of the software and its integration with the hardware, which is evaluated through engineering and software testing.

4. Adjudication method (e.g. 2+1, 3+1, none) for the test set

This information is not applicable for a software operating system upgrade. Adjudication methods like 2+1 or 3+1 are used in clinical studies to resolve discrepancies in expert interpretation of diagnostic results. Software verification and validation typically rely on predefined test outcomes and engineering assessments.

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 information is not applicable. An MRMC comparative effectiveness study is used to evaluate the impact of an AI algorithm on human reader performance, usually for diagnostic tasks. This submission is for a software operating system upgrade for an existing IVD device, not for a new AI algorithm.

6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done

The concept of "standalone performance" in the context of an algorithm's diagnostic capability (like an AI algorithm) is not directly applicable here. The device itself (GEM Premier ChemSTAT) operates to provide quantitative measurements. The software upgrade ensures the continued, correct operation of the device. The verification and validation activities demonstrate that the upgraded software performs its functions correctly as part of the overall device system.

7. The type of ground truth used (expert consensus, pathology, outcomes data, etc)

For this software upgrade, the "ground truth" is the expected behavior and functionality of the software and the device. This is established through:

• Functional specifications: The software is expected to perform according to its design specifications.
• Risk analysis: The software should not introduce new risks or fail to mitigate existing ones.
• Cybersecurity standards: The software should meet cybersecurity requirements.
• Established device performance: The software upgrade should not negatively impact the established analytical and clinical performance of the GEM Premier ChemSTAT device (which relies on the physical and chemical principles of its measurements).

The document explicitly states that the changes "do not introduce...changes to labeled performance claims." This implies that the performance of the device (e.g., accuracy, precision of Na+, K+, Glu measurements) remains the same as previously cleared, and the software upgrade was validated not to alter these.

8. The sample size for the training set

This information is not applicable. Training sets are used for machine learning models. This submission describes a conventional software operating system upgrade (Fedora 17 Linux to WindRiver LTS 18 Linux) for an existing IVD device, not the development or retraining of a machine learning algorithm.

9. How the ground truth for the training set was established

This information is not applicable, as there is no training set for a machine learning model; it is a software operating system upgrade.

Ask a specific question about this device

The Nova Primary Glucose Analyzer System is indicated for in vitro diagnostic use by healthcare professionals in clinical laboratory setting for the quantitative determination of Glucose in lithium heparinized venous whole blood and plasma.

The measurement of Glucose is used in the diagnosis and treatment of carbohydrate metabolism distuding diabetes mellitus, neonatal hypoglycemia, and idiopathic hypoglycemia, and pancreatic islet cell carcinoma.

The Nova Primary Glucose Analyzer System is a small, portable laboratory glucose analyzer that measures blood glucose levels in lithium heparinized whole blood or plasma utilizing a glucose oxidase based sensor and membrane/cap assembly.

The Nova Primary Glucose Analyzer accepts samples from syringes, blood collection tubes, microcentrifuge tubes, and sample cups. The sample size for analysis is 25 microliters (aspirated volume).

Nova Primary System Components:
The Nova Primary Glucose Analyzer System is comprised of the following components.

• Nova Primary Glucose Analyzer
• Nova Primary Glucose Sensor
• Nova Primary Glucose Membrane
• Nova Primary Calibrator Cartridge
• Optional Barcode Scanner
• IFU/Labeling

Sample Types:
The Nova Primary Glucose Analyzer System accepts lithium heparinized venous whole blood and plasma.

The Nova Primary Glucose Analyzer System by Nova Biomedical Corporation is indicated for in vitro diagnostic use by healthcare professionals in clinical laboratory settings for the quantitative determination of Glucose in lithium heparinized venous whole blood and plasma.

The study presented does not explicitly list acceptance criteria in a tabular format with corresponding performance results. Instead, it describes performance testing conducted to demonstrate substantial equivalence to the predicate device, the YSI 2300 Stat Glucose and L-Lactate Analyzer. The performance tests included:

• Method Comparison Studies
• Precision/Reproducibility Studies
• Linearity Testing
• Specificity / Interference Testing
• Detection Limit
• Shelf Life Stability Testing

The "Summary of Performance Testing" section states, "The results of the performance testing confirmed that the Nova Primary Glucose Analyzer demonstrates substantial equivalence to the YSI 2300 Stat Glucose and L-Lactate Analyzer." This implies that the device met the criteria necessary to be considered substantially equivalent to the predicate device in terms of these performance characteristics. However, specific numerical acceptance criteria and the exact reported device performance metrics are not detailed in the provided text.

Here's an attempt to structure the information based on the provided text, while acknowledging the lack of explicit numerical acceptance criteria:

1. Table of Acceptance Criteria and Reported Device Performance

2. Sample Size for Test Set and Data Provenance

The document does not explicitly state the sample sizes used for the test set or the data provenance (e.g., country of origin, retrospective or prospective nature) for the performance testing. It only mentions that "Testing was completed to show that the Nova Primary Glucose Analyzer demonstrates substantial equivalence to the YSI 2300 Stat Glucose and L-Lactate Analyzer."

3. Number of Experts for Ground Truth and Qualifications

This information is not provided in the document. The device is a laboratory instrument for quantitative determination of glucose, and ground truth would typically be established by recognized reference methods or predicate devices, not by human experts in the diagnostic sense.

4. Adjudication Method for the Test Set

This information is not applicable and not provided. As noted above, the ground truth is based on objective laboratory measurements and reference methods, not subjective assessments requiring adjudication.

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

No, a multi-reader multi-case (MRMC) comparative effectiveness study was not done. This type of study is typically relevant for interpretative diagnostic devices where human reader performance is a key metric, often in image-based diagnostics. The Nova Primary Glucose Analyzer System is a quantitative measurement device.

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

Yes, the performance testing described (Method Comparison, Precision/Reproducibility, Linearity, Specificity/Interference, Detection Limit, Shelf Life Stability) would inherently be a standalone evaluation of the algorithm and device's accuracy and reliability in measuring glucose concentration. The device is designed for automated quantitative measurement.

7. Type of Ground Truth Used

The ground truth for the device's performance is established by comparison to a legally marketed predicate device (YSI 2300 Stat Glucose and L-Lactate Analyzer) and traceability to isotope dilution mass spectrometry (for plasma equivalent glucose results). This suggests established laboratory reference methods and comparison to a 'gold standard' device.

8. Sample Size for the Training Set

The document does not specify a separate "training set" sample size. For an analytical device like this, the "training" (calibration and algorithm development) is typically an internal process, and the performance testing is done on independent samples.

9. How Ground Truth for the Training Set Was Established

The document states that "All glucose measurement algorithms are intended to report plasma equivalent glucose results that are substantially equivalent to the predicate device plasma glucose result, and are traceable to isotope dilution mass spectrometry." This implies that the algorithms were developed or "trained" using data and methods that align with these established analytical principles and reference standards. However, the specific details of how this "ground truth" was established for internal algorithm development (training) are not elaborated.

Ask a specific question about this device

The i-STAT CHEM8+ cartridge with the i-STAT 1 System is in the in vitro quantification of sodium, potassium, chloride, ionized calcium, glucose, blood urea nitrogen, creatinine, hematocrit, and total carbon dioxide in arterial or venous whole blood in point of care or clinical laboratory settings.

Sodium measurements are used for monitoring electrolyte imbalances.

Potassium measurements are used in the diagnosis and clinical conditions that manifest high and low potassium levels.

Chloride measurements are primarily used in the diagnosis, monitoring, and treatment of electrolyte and metabolic disorders including, but not limited to, cystic fibrosis, diabetic acidosis, and hydration disorders.

Ionized calcium measurements are used in the diagnosis and treatment of parathyroid disease, chronic renal disease and tetany.

Glucose measurements are used in the diagnosis, monitoring, and treatment of carbohydrate metabolism disorders including, but not limited to, diabetes mellitus, neonatal hypoglycemia, and pancreatic islet cell carcinoma.

Blood urea nitrogen measurements are used for the diagnosis, and treatment of certain renal and metabolic diseases.

Creatinine measurements are used in the diagnosis and treatment of renal diseases, in monitoring renal dialysis, and as a calculation basis for measuring other urine analytes.

Hematoorit measurements can aid in the determination and monitoring of normal total red cell volume status that can be associated with conditions including anemia and erythrocytosis. The i-STAT Hematocrit test has not been evaluated in neonates.

Carbon dioxide measurements are used in the diagnosis, monitoring, and treatment of numerous potentially serious disorders associated with changes in body acid-base balance.

The i-STAT CHEM8+ test cartridge contains test reagents to analyze whole blood at the point of care or in the clinical laboratory for sodium (Na), potassium (K), chloride (CI), ionized calcium (iCa), glucose (Glu), blood urea nitrogen (BUN), creatinine (Crea), hematocrit (Hct), and total carbon dioxide (TCO2). The test is contained in a single-use, disposable cartridge. Cartridges require two to three drops of whole blood which are typically applied to the cartridge using a transfer device.

The i-STAT 1 Analyzer is a handheld, in vitro diagnostic analytical device designed to run only i-STAT test cartridges. The instrument interacts with the cartridge to move fluid across the sensors and generate a quantitative result (within approximately 2 minutes).

The i-STAT 1 System is comprised of the i-STAT 1 analyzer, the i-STAT test cartridges and accessories (i-STAT 1 Downloader/Recharger, electronic simulator and portable printer). The system is designed for use by trained medical professionals at the patient point of care or in the clinical laboratory and is for prescription use only.

The provided text describes a 510(k) premarket notification for the i-STAT CHEM8+ cartridge with the i-STAT 1 System, specifically addressing the addition of an anticoagulant-free whole blood matrix. The document references several previous 510(k) clearances for various analytical performance characteristics and presents a new "Matrix Equivalence" study for the anticoagulant-free whole blood.

Here's an analysis of the acceptance criteria and study information provided, structured as requested:

1. Table of Acceptance Criteria and Reported Device Performance

The document does not explicitly state pre-defined acceptance criteria for the "Matrix Equivalence" study in a tabulated format. However, it implicitly uses a Passing-Bablok linear regression analysis to demonstrate equivalence. The reported device performance is presented as the results of this regression analysis. We can infer the expected performance from general expectations for method comparisons in analytical chemistry, where a slope close to 1, an intercept close to 0, and a high correlation coefficient (r) indicate good agreement.

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

• Sample Size (Test Set for Matrix Equivalence): The sample sizes vary slightly per analyte:
  • Na, Cl, iCa: 314
  • K, Glu: 313
  • BUN: 310
  • Crea: 312
  • Hct: 311
  • TCO2: 273
• Data Provenance: The study was conducted at "three (3) point of care sites." The document does not specify the country of origin, but given the FDA submission, it is likely the US or a region with equivalent regulatory standards. The data is prospective in nature, as it involved collecting samples (both anticoagulant-free and anticoagulated) for direct comparison.

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

This type of study (matrix equivalence for IVD devices) does not typically involve human experts establishing "ground truth" in the way a diagnostic imaging study would. The ground truth (or reference method) for comparison is the measurement obtained from the previously cleared device using anticoagulated samples. The expertise lies in the calibration of the reference method and the design and execution of the analytical study, not in human interpretation of results.

4. Adjudication Method for the Test Set

Not applicable for this type of analytical method comparison study. Adjudication is relevant for subjective assessments, typically in diagnostic imaging or clinical outcomes, to resolve discrepancies among human readers or between AI and human readers. Here, the comparison is between two quantitative measurement methods.

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

Not applicable. This is an analytical performance study for an in vitro diagnostic device, not a diagnostic imaging or clinical decision support AI.

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

Yes, the analytical performance (precision, linearity, LoQ, LoB/LoD, interference, and method comparison) of the i-STAT CHEM8+ cartridge with the i-STAT 1 System, and its equivalence between different sample matrices, primarily represents standalone performance of the device without human intervention beyond sample collection and device operation. The "Matrix Equivalence" study directly compares the results of the device using two different sample types.

7. The Type of Ground Truth Used (expert consensus, pathology, outcomes data, etc.)

The ground truth for the "Matrix Equivalence" study was established by the mean result from the primary sample, which refers to measurements obtained from whole blood samples collected with balanced heparin or lithium heparin anticoagulant using the previously cleared i-STAT CHEM8+ system. This acts as the "reference method" for comparison to the anticoagulant-free samples.

8. The Sample Size for the Training Set

The document does not explicitly mention a "training set" for the purpose of the Matrix Equivalence study. This study is a validation study demonstrating that a new sample matrix (anticoagulant-free whole blood) yields equivalent results to the established (anticoagulated) sample matrix. The device itself (i-STAT CHEM8+ with i-STAT 1 System) would have undergone extensive development and internal testing (which could be considered a form of "training") prior to its initial clearances (K183678, K183680, K183688, K191298, K191360). The current submission focuses on extending the indications for use.

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

As no explicit "training set" is mentioned for this specific submission's study, this question is not directly answerable from the provided text. The "ground truth" for the reference method within the Matrix Equivalence study, as stated above, derives from the previously cleared performance of the i-STAT CHEM8+ system using anticoagulated samples, which would have been established through robust analytical validation studies (e.g., comparison to laboratory reference methods).

Ask a specific question about this device

The Stat Profile Prime Plus Analyzer System is indicated for use by healthcare professionals in clinical laboratory settings and for point-of-care usage for quantitative determination of Glucose, Lactate, Creatinine, and Blood Urea Nitrogen in heparinized arterial and venous whole blood.

Glucose measurement is used in the diagnosis and treatment of carbohydrate metabolism disturbances including diabetes mellitus, neonatal hypoglycemia, and idiopathic hypoglycemia, and of pancreatic islet cell carcinoma.
Lactate (lactic acid) measurement is used to evaluate the acid-base status of patients suspected of having lactic acidosis.
Creatinine measurement is used in the diagnosis and treatment of certain renal conditions and for monitoring adequacy of dialysis.
Blood Urea Nitrogen measurement is used in the diagnosis and treatment of certain renal and metabolic diseases.

The Stat Profile Prime Plus Analyzer System is a low cost, low maintenance analyzer for hospital laboratory and point-of-care settings. It consists of the analyzer, sensor cartridges, and thermal paper for an onboard printer. Optionally, it provides for reading of barcode labels (such as operator badges and data sheets).

The Stat Profile Prime Plus Analyzer has slots to accommodate two sensor cartridges (Primary and Auxiliary). The analyzer will determine the configuration of the system by detecting which sensor cards are installed.

Primary Sensor Card Port:
There are two options for the primary sensor card:

• Primary Sensor Card 1 shall enable and report the following listed analytes: o Glu. Lactate
• Primary Sensor Card 2 shall enable and report the following listed analytes: o Glu, Lactate

Auxiliary Sensor Card Port:
The reporting of Creatinine and BUN parameters (or not reporting them) shall be determined by the selection of the Auxiliary Sensor Card

• Auxiliary Sensor Card 1 shall enable Creatinine and BUN parameters .
• . Auxiliary Sensor Card 2 shall be a "dummy" sensor card, and will not report any parameters.

As with the predicate, the Stat Profile Prime Plus Analyzer is a blood gas, co-oximetry, electrolyte, chemistry, and hematology analyzer with an enhanced test menu and multiple quality control options. Both traditional internal and external quality control is available, as well as an on-board Quality Management System (QMS), and an electronic monitoring approach that insures the analyzer is working properly at all times.

The Stat Profile Prime Plus Analyzer accepts samples from syringes and open tubes. The minimum sample size for analysis is 135 µL.

Sample collection, preparation and application to the analyzer are the same as for the previously cleared predicate. The end user can select which analytes are to be tested in the panel.

Stat Profile Prime Plus Analyzer System Components:
The Stat Profile Prime Plus Analyzer System is comprised of the following components.

• Stat Profile Prime Plus Analyzer System .
• Primary Sensor Cartridge .
• Auxiliary Sensor Cartridge
• Stat Profile Prime Plus Auto-Cartridge Quality Control Pack ●
• Stat Profile Prime Plus Calibrator Cartridge ●
• Stat Profile Prime Plus External Ampuled Control .
• . IFU/Labeling

This document describes the performance of the Stat Profile Prime Plus Analyzer System for the quantitative determination of Glucose, Lactate, Creatinine, and Blood Urea Nitrogen in heparinized arterial and venous whole blood. The submission specifically focuses on the modification of the device to include Point-of-Care (POC) use.

It is important to note that this document is a 510(k) summary for a medical device. This type of document typically focuses on demonstrating substantial equivalence to a previously cleared device, rather than proving novel clinical effectiveness with large-scale comparative effectiveness studies. Therefore, not all requested sections directly apply, especially those related to AI model development (training set, human-in-the-loop, MRMC studies) common in submissions for AI/ML-based diagnostic software. This device is a point-of-care blood analyzer, not an AI diagnostic software.

Here's the breakdown of the provided information within the context of this device:

1. Table of Acceptance Criteria and Reported Device Performance

The document does not present explicit "acceptance criteria" in a singular table for all parameters as one might find for a pass/fail threshold in an AI/ML study. Instead, performance is demonstrated through various analytical validation studies (precision, linearity, interference, detection limit) and a method comparison study to show equivalence to a predicate device and laboratory methods. The "acceptance criteria" are implicitly met if the performance data supports substantial equivalence for the intended use and accuracy compared to established methods.

For Lactate (specifically detailed in the provided text):

• Within-Run (Quality Control Ampules Levels 4 & 5): CV% ranged from 0.00% to 1.6%.
• Within-Run (Whole Blood from Syringes, 5 samples): CV% ranged from 1.3% to 8.0%.
• Run-to-Run (Auto QC Cartridge Levels 4 & 5): Total Imprecision %CV ranged from 1.4% to 3.5%.
• Run-to-Run (Whole Blood, 5 samples): CV% ranged from 1.8% to 14.7%. Note: Blood #1 showed a higher CV of 13.3-14.7% across devices. |
  | Linearity | Demonstrates linearity across the claimed measurement range, with a high correlation (r-value) to established methods. | - Claimed Measurement Range: 0.3 - 20.0 mmol/L.
• Specimen Range Tested: 0.2 - 23.5 mmol/L.
• Analyzers (PP1, PP2, PP3): r-values were 0.9983, 0.9982, and 0.9988, respectively.
• Comment: Results support the claimed measurement range. |
  | Specificity / Interference | Identifies substances that may interfere and quantifies the highest concentration tested that shows no significant interference. | - Dose response study performed for glycolic acid and hydroxyurea showed interference at all concentrations. (Specific concentrations not given, but the fact of interference is noted).
• Table 7 lists the highest non-interfering concentrations for 23 other common substances (e.g., Acetaminophen, Ascorbic Acid, Bilirubin, Ethanol, Heparin, Ibuprofen). |
  | Detection Limit (LoB, LoD, LoQ) | LoB, LoD, and LoQ should be below the lower limit of the claimed measurement range. | - LoB: 0.0 mmol/L
• LoD: 0.1 mmol/L
• LoQ: 0.1 mmol/L.
• Comment: All are below the claimed lower limit of 0.3 mmol/L. |
  | Method Comparison (POC vs. Lab) | Statistical agreement between the Stat Profile Prime Plus system (operated at POC by various personnel) and laboratory reference methods (implicitly, lab instruments or methodologies used for comparison). | - Analyte: Lactate
• N: 413 samples
• Range: 0.5 - 16.5
• Slope: 1.0181
• Intercept: -0.0796
• r (correlation coefficient): 0.9975
• 95% Confidence Interval Bias (2 mmol/L): 1.9-2.0
• 95% Confidence Interval Bias (6 mmol/L): 6.0-6.1 |
  | Total Imprecision Performance (POC)| Demonstrates acceptable total imprecision when operated by POC personnel. | - Lactate (Level 4): Mean 1.8 mmol/L, Total SD 0.1, Total %CV 3.2%
• Lactate (Level 5): Mean 6.7 mmol/L, Total SD 0.2, Total %CV 3.4%
• Lactate (Linearity Level 4): Mean 16.5 mmol/L, Total SD 0.4, Total %CV 2.2% |
  | Within-Run Whole Blood Precision (POC)| Demonstrates acceptable within-run precision for fresh whole blood samples in POC setting. | - Lactate (7 samples): Mean values varied (e.g., 3.30, 1.43, 2.47, 2.49, 1.24, 4.69, 11.4 mmol/L). Reported %CVs ranged from 4.08% to 6.68%. |

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

The "test set" in this context refers to the samples used in the analytical performance and method comparison studies.

• Sample Sizes:

  • Lactate Precision (Within-Run): 20 replicates for each level/sample type (QC, Ampules, Whole Blood) on 3 analyzers.
  • Lactate Precision (Run-to-Run): QC samples: 240 samples (2 runs/day for 40 runs * 3 analyzers); Whole Blood samples: 30 samples (triplicate analyses over 10 runs * 3 analyzers).
  • Linearity: 9 levels tested on 3 analyzers.
  • Interference: Not explicitly stated for each substance, but dose-response performed for problematic interferents.
  • Detection Limit: LoB: 5 runs on 2 analyzers; LoD: 4 runs on 2 analyzers for 5 low-level samples over 3 days (total 60 replicates per reagent lot); LoQ: 3 runs on 2 analyzers for 4 low-level samples over 3 days (total 36 replicates per reagent lot).
  • Method Comparison (POC vs. Lab): For Lactate: N=413 samples.
  • Total Imprecision (POC): 3 levels of QC/Linearity materials, run in duplicate each day for 20 runs on 3 analyzers.
  • Within-Run Whole Blood Precision (POC): 10 replicate measurements for 7 different whole blood samples at each site, by a minimum of 2 POC operators at 3 sites (total 9 operators).

• Data Provenance:

  • Country of Origin: Not explicitly stated but inferred to be the United States (given the FDA submission and the company address in Waltham, Massachusetts).
  • Retrospective or Prospective: The studies described (e.g., precision, linearity, method comparison) are typically prospective analytical validation studies conducted specifically for regulatory submission, using controlled conditions and fresh/altered samples collected for the purpose of the study. The POC study involved current operations ("discarded blood gas specimens" and "fresh, native and altered whole blood samples") and training of personnel for the study.

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

For this type of in vitro diagnostic device (blood analyzer), "ground truth" is established by comparison to reference methods or calibrated standards rather than expert consensus on images.

• Number of Experts: Not applicable in the context of expert readers for imaging or clinical diagnosis. The "experts" are the laboratory personnel operating the reference instruments and the Point-of-Care personnel who were "trained, qualified staff found in typical POC sites where blood gas analyzers are utilized."
• Qualifications of Experts:
  • Laboratory Personnel: Implied to be qualified professionals operating the reference analyzers (e.g., Stat Profile pHOx Ultra Analyzer System, K110648).
  • Point-of-Care Personnel: A total of 61 Respiratory Care, 12 Nursing, and 1 Exercise Physiology personnel participated from 3 POC settings (Cardiothoracic Intensive Care Unit (CTICU), Emergency Department (ED), Respiratory Therapy Lab (RT)). They are described as "trained, qualified staff."

4. Adjudication Method for the Test Set

Not applicable. This is not an AI/ML diagnostic software involving subjective interpretation or multiple expert reads needing adjudication. Performance is assessed analytically against reference methods or statistical metrics.

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

No, an MRMC comparative effectiveness study was not conducted. This type of study is primarily relevant for AI/ML-based diagnostic software where the AI assists human readers in tasks like image interpretation. This submission is for an in vitro diagnostic device that directly measures analytes in blood. The study compared the device's performance to predicate devices and laboratory methods.

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

While the device's measurement process for each analyte (e.g., impedance, amperometry) can be considered "standalone" in that it performs the measurement algorithmically, the term "standalone performance" in the context of AI/ML usually refers to the accuracy of the algorithm without any human input during the diagnostic process.

For this device, the "measurements" are the direct outputs from the instrument. Its analytical performance (precision, linearity, detection limits, interference) is evaluated intrinsically (standalone from a human operator's interpretation, though a human initiates the test). The "Method Comparison Studies" then compare these outputs to a reference method, which is the closest equivalent to a "standalone" performance assessment for an IVD, demonstrating how well the device matches established lab results.

7. The Type of Ground Truth Used

The ground truth for this device's performance validation is based on reference methods and calibrated standards.

• Reference Methods: The "Method Comparison Studies" compare the Stat Profile Prime Plus Analyzer System's results (both in the lab and at POC) to those obtained from the predicate device (K180340) and/or other validated laboratory methods/instruments. For Lactate, the predicate K110648 (Stat Profile pHOx Ultra Analyzer System) was used for comparison.
• Calibrated Standards: Precision, linearity, and detection limit studies utilize quality control materials, calibrators, and prepared solutions with known concentrations. The linearity study specifically states comparison to "the reference analyzer and/or the product specifications defined in the Stat Profile Prime Marketing Requirements document."

8. The Sample Size for the Training Set

Not applicable. This device is an in vitro diagnostic analyzer (hardware and embedded software for physical measurement), not an AI/ML algorithm that is "trained" on a dataset in the conventional sense. The "development" of its analytical components involves traditional engineering and chemistry, not machine learning training.

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

Not applicable, as there is no "training set" in the context of AI/ML for this device. The physical and chemical principles of measurement (e.g., enzymatic reactions, amperometry, potentiometry) form the basis of the device's function.

Ask a specific question about this device

The i-STAT CHEM8+ cartridge with the i-STAT 1 System is intended for use in the in vitro quantification of glucose, and creatinine in arterial or venous whole blood in point of care or clinical laboratory settings.

Glucose measurements are used in the diagnosis, monitoring, and treatment of carbohydrate metabolism disorders including, but not limited to, diabetes mellitus, neonatal hypoglycemia, and pancreatic islet cell carcinoma.

Creatinine measurements are used in the diagnosis and treatment of renal diseases, in monitoring renal dialysis, and as a calculation basis for measuring other urine analytes.

The i-STAT CHEM8+ test cartridge contains test reagents to analyze whole blood at the point of care or in the clinical laboratory for glucose and creatinine. The test is contained in a single-use, disposable cartridge. Cartridges require two to three drops of whole blood which are typically applied to the cartridge using a transfer device.

The i-STAT 1 Analyzer is a handheld, in vitro diagnostic analytical device designed to run only i-STAT test cartridges. The instrument interacts with the cartridge to move fluid across the sensors and generate a quantitative result (within approximately 2 minutes).

The i-STAT 1 System is comprised of the i-STAT 1 analyzer, the i-STAT test cartridges and accessories (i-STAT 1 Downloader/Recharger, electronic simulator and portable printer). The system is designed for use by trained medical professionals at the patient point of care or in the clinical laboratory and is for prescription use only.

The provided text describes the analytical performance studies for the i-STAT CHEM8+ cartridge with the i-STAT 1 System for measuring glucose and creatinine. While it details numerous performance characteristics, it does not explicitly state "acceptance criteria" for each test. However, the study design and "results met the acceptance criteria" statements imply that certain predefined thresholds were successfully achieved.

Let's break down the information available to address your request:

Acceptance Criteria and Reported Device Performance

The document doesn't provide a consolidated table of explicit acceptance criteria. Instead, it states that results "met the acceptance criteria" for linearity, and it outlines the method for identifying interference (difference between control and test samples outside of the allowed error (Ea)). For precision studies, statistical metrics like Total Standard Deviation (ST) and Coefficient of Variation (CV) are presented, and for method comparison, slope, intercept, and correlation coefficient (r) are given. The implied acceptance is that these values fall within acceptable ranges for a diagnostic device of this type.

Implied Acceptance Criteria and Reported Performance (derived from text):

Study Details:

1. Sample Size used for the Test Set and Data Provenance:

   • Precision (Aqueous): N=80 or 81 for each level of Creatinine and Glucose. Data provenance is not specified (e.g., country of origin) but implied to be laboratory-based ("one site"). It is an analytical performance study, not a clinical study on patient samples.
   • Precision (Whole Blood): Sample sizes vary per site and level (e.g., N=14 to N=21). The study used "venous whole blood (native or altered) samples." The study was conducted at "3 point of care sites."
   • Linearity: Whole blood samples of "varying analyte levels" were prepared. Specific N not provided for this particular section.
   • LoQ, LoB/LoD: Whole blood samples were "altered to low glucose" or "blank" concentrations. Specific N not provided for this section.
   • Hematocrit Sensitivity: Three hematocrit levels evaluated across four glucose levels. Specific N not provided.
   • Oxygen Sensitivity: High and low ranges of oxygen. Specific N not provided.
   • Altitude: Not explicitly stated, but results given for "average measured altitude." Specific N not provided.
   • Interference: Whole blood samples based on CLSI guidelines. Specific N not provided, but multiple substances tested.
   • Method Comparison: N=185 for Glucose, N=180 for Creatinine. Used "Venous and arterial blood specimens" for the i-STAT device and "plasma specimens" for the predicate device. Data provenance is not specified.

   Overall Provenance: The studies are "analytical performance" studies, primarily laboratory-based and conducted on various prepared or native samples. There is no indication of geographic origin or whether samples were retrospective or prospective, though for analytical performance, prospective collection for the purpose of the study is common.

2. 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. The "ground truth" for analytical performance studies is established by reference methods or highly accurate laboratory analyzers (the "comparative method"), not by human experts interpreting images or clinical outcomes. In this case, "ground truth" (or comparative method) for the method comparison study was the Beckman DxC, a laboratory analyzer.

3. Adjudication method (e.g., 2+1, 3+1, none) for the test set:

   • Not applicable. This is an IVD device measuring analytes, not interpretations of images or clinical assessments requiring human adjudication.

4. 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. This is an IVD device for measuring chemical analytes in blood, not an imaging device assisted by AI to be read by human experts.

5. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done:

   • The i-STAT device functions as a standalone analyzer that quantifies glucose and creatinine. The performance studies described (precision, linearity, LoQ, LoD, interference, method comparison) are standalone performance studies of the device's ability to measure these analytes. While a human initiates the test and interprets the quantitative result, the measurement process itself is automated (algorithm only).

6. The type of ground truth used (expert concensus, pathology, outcomes data, etc):

   • For the method comparison study, the ground truth was essentially the measurements from a legally marketed predicate device (Beckman DxC), which serves as the comparative method in analytical validation. For other analytical performance studies (precision, linearity, LoQ/LoD), the "ground truth" is based on the known concentrations of prepared reference materials or controls, or the statistical evaluation of repeated measurements of samples.

7. The sample size for the training set:

   • The documentation does not discuss a "training set" in the context of machine learning or AI models, as this is a traditional IVD device based on electrochemical principles, not an AI/ML device. Therefore, no separate training set is mentioned or applicable in the way it would be for an AI-powered diagnostic.

8. How the ground truth for the training set was established:

   • Not applicable, as there is no explicitly mentioned "training set" in the context of an AI/ML model for this traditional IVD device.

Ask a specific question about this device