Search Results

The Tosoh Automated Glycohemoglobin Analyzer HLC-723GR01 is intended for in vitro diagnostic use for the quantitative measurement of % hemoglobin A1c (HbA1c) (DCCT/NGSP) and mmol/mol hemoglobin A1c (IFCC) in human venous whole blood and hemolysate specimens using ion-exchange high performance liquid chromatography (HPLC).

This test is to be used as an aid in diagnosis of diabetes and identifying patients who may be at risk for developing diabetes, and for monitoring of long-term blood glucose control in individuals with diabetes mellitus.

The Tosoh Automated Glycohemoglobin Analyzer HLC-723GR01 is intended for in vitro diagnostic use for the quantitative measurement of % hemoglobin A1c (HbA1c) (DCCT/NGSP) and mmol/mol hemoglobin A1c (IFCC) in human venous whole blood and hemolysate specimens using ion-exchange high performance liquid chromatography (HPLC). The Tosoh Automated Glycohemoglobin Analyzer HLC-723GR01 analyzer is to be used in the clinical laboratory setting. The analyzer hemolyzes and dilutes whole blood specimens and subsequently injects the diluted and hemolyzed sample into the injection valve and onto the TSKgel GR01 HbA1c Column. Manually hemolyzed and prediluted samples can be loaded in sample cups directly for analysis on the analyzer. Separation of hemoglobin fractions is achieved by using differences in ionic interactions between the cation exchange group on the column resin surface and the hemoglobin components by a step gradient elution with three elution buffers of different ionic strengths. Changes in light absorbance at 415nm caused by hemoglobin fractions eluting from the column are monitored. The GR01 software has chromatographic peak retention time windows for the six expected hemoglobin fractions, A1a, A1b, HbF, LA1c+, SA1c, and A0. In addition, the software has retention time windows for the presumptive identification of the common hemoglobin variants, D, S, C, and E. An analysis requires 50 seconds per sample to complete. A printout of the results includes the sample ID, date, percentage, and retention time of each hemoglobin fraction, the SA1c percentage along with a chromatogram of the elution pattern of the hemoglobin fractions. The analyzer is equipped with external USB ports. These can be used to store assay results, to update and backup program versions, to attach an external device, for example an external printer or an optional handheld barcode scanner, and for software installation or backup of data. The last 150,000 assay results are automatically saved to the analyzer's internal memory and can be retrieved as list data in .CSV format and chromatograms in .PDF format.

The Tosoh Automated Glycohemoglobin Analyzer HLC-723GR01 system consists of the following components.

• GR01 HbA1c Elution Buffer No. 1, No. 2, No. 3
• TSKgel® GR01 HbA1c Column
• Hemoglobin A1c Controls Levels 1 and 2
  • 21 CFR 862.1660, Product Code JJX
• Hemoglobin A1c Calibrator Set
  • 21 CFR 862.1150, Product Code JIT
• Hemolysis and Wash Solution (L) and (LL)
  • 21 CFR 864.8540, Product Code GGK

N/A

Ask a specific question about this device

Immunoassays for the in vitro quantitative determination of the soluble fms like tyrosine kinase-1/placental growth factor (sFlt-1/PlGF) ratio in human serum.

The sFlt-1/PlGF ratio is indicated as an aid in the risk assessment of pregnant women, with a singleton pregnancy (23+0 to 34+6/7 weeks' gestation) hospitalized for hypertensive disorders of pregnancy (preeclampsia, chronic hypertension with or without superimposed preeclampsia, or gestational hypertension), to develop preeclampsia with severe features within two weeks from testing. The sFit-1/PlGF ratio should be used in conjunction with clinical assessment and routine laboratory testing.

The electrochemiluminescence immunoassay "ECLIA" is intended for use on cobas e immunoassay analyzers.

The Elecsys sFlt-1 and Elecsys PlGF assays employ a sandwich principle using electrochemiluminescence immunoassay "ECLIA" technology. The total duration of each assay is 18 minutes. Samples are incubated with biotinylated and ruthenium-labeled monoclonal antibodies specific to sFlt-1 or PlGF, forming a sandwich complex. Streptavidin-coated microparticles are added, binding the complex to the solid phase. The microparticles are magnetically captured, unbound substances are removed, and a voltage is applied to induce chemiluminescent emission, which is measured by a photomultiplier. Results are determined via a calibration curve generated by 2-point calibration and a master curve provided via the reagent barcode. The reagents for each assay are combined in a "rackpack".

Here's a breakdown of the acceptance criteria and study information for the Elecsys sFlt-1 and Elecsys PlGF assays, based on the provided document.

Acceptance Criteria and Device Performance

Study Details

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

   • Clinical Performance Study (PRAECIS): N=556 pregnant women.
   • Data Provenance: The document does not explicitly state the country of origin for the PRAECIS study's clinical data, but it refers to the study as a "validation of the Roche sFlt-1/PlGF ratio cut-off." The reference range study (mentioned as a separate activity) collected samples from "$US" (United States) to establish normal reference ranges. Given the context, it's highly probable the PRAECIS clinical validation study also involved US patient data, though not explicitly stated for that specific study section.
   • Retrospective or Prospective: The document describes the study as "Praecis 'Preeclampsia Risk Assessment: Evaluation of Cut-offs to Improve Stratification'" and states it included women "admitted to the hospital with (or develop while hospitalized) a hypertensive disorder." This strongly suggests a prospective design within the hospital setting, though "retrospective" is not explicitly excluded for patient inclusion/data collection aspects. However, the nature of "risk assessment...to develop preeclampsia with severe features within two weeks from testing" implies a prospective observation period.

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

   • The document does not specify the number of experts or their qualifications used to establish the ground truth for the clinical test set. The ground truth (preeclampsia with severe features within two weeks) would typically be established by clinicians based on diagnostic criteria, but further details are not provided.

3. Adjudication Method for the Test Set:

   • The document does not specify an adjudication method (e.g., 2+1, 3+1, none) for establishing the clinical ground truth.

4. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study:

   • No, a Multi-Reader Multi-Case (MRMC) comparative effectiveness study was not mentioned or conducted in this submission. This device is an in-vitro diagnostic (IVD) immunoassay, not an imaging AI device that relies on human reader interpretation. Therefore, the concept of "human readers improve with AI vs without AI assistance" is not directly applicable here.

5. Standalone (Algorithm Only Without Human-in-the-Loop Performance) Study:

   • Yes, the clinical performance evaluation (PRAECIS study) presents the standalone performance of the sFlt-1/PlGF ratio at a specified cutoff (38). It assesses the test's ability to predict clinical outcome (preeclampsia with severe features) directly. The "human-in-the-loop" aspect is that the results are "used in conjunction with clinical assessment and routine laboratory testing" by a physician, but the performance statistics presented (Sensitivity, Specificity, NPV, PPV) are for the assay's output itself.

6. Type of Ground Truth Used:

   • Clinical Outcome Data: The ground truth used was the "development of preeclampsia with severe features within two weeks from testing," as defined by established medical guidelines (likely ACOG, given its mention for the predicate device). This is a clinical outcome.

7. Sample Size for the Training Set:

   • The document does not explicitly describe a separate "training set" for an algorithm in the traditional sense, as these are immunoassays. The clinical validation study (PRAECIS) evaluates the performance of the established cutoff (38) for the sFlt-1/PlGF ratio. The cutoff itself might have been derived from earlier studies or internal research, which would implicitly act as a form of "training" or "development" data, but it's not detailed here with sample sizes. The reference range study for healthy pregnant women used 380 evaluable subjects to establish expected values.

8. How the Ground Truth for the Training Set Was Established:

   • As there isn't an explicitly described "training set" for an algorithm to learn from, this question largely pertains to the development of the assay and its interpretive cutoffs. The cutoff of 38 for the sFlt-1/PlGF ratio is stated to be "validated" in the PRAECIS study. The process for initial establishment or derivation of this cutoff (38) is not detailed in this document. It's common for such cutoffs to be determined through prior multicenter clinical trials and statistical analysis to achieve optimal predictive performance targets.

Ask a specific question about this device

B·R·A·H·M·S™ CgA II KRYPTOR™ is an automated immunofluorescent assay using Time-Resolved Amplified Cryptate Emission (TRACE™) technology for quantitative determination of Chromogranin A concentration in human serum.

B·R·A·H·M·S™ CgA II KRYPTOR™ is to be used in conjunction with other clinical methods as an aid in monitoring of disease progression during the course of disease and treatment in patients with gastroentero-pancreatic neuroendocrine tumors (GEP-NETs, grade 1 and grade 2).

The B-R-A-H-M-S CgA II KRYPTOR assay is based on the formation of a complex comprised of a Chromogranin A (CgA) analyte "sandwiched" between two monoclonal mouse anti-CgA antibodies. One of the antibodies (537/H2) is directed at the epitope AA124–144 and labelled with DiSMP cryptate, the other antibody (541/E2) binds to AA280-301 and is labelled with Alexa Fluor®647.

The measurement principle is based on a non-radiative energy transfer from a donor (cryptate) to an acceptor (Alexa Fluor™647) when they are part of an immunocomplex (TRACE technology (Time-Resolved Amplified Cryptate Emission)).

The fluorescent signal is proportional to the concentration of the analyte to be measured.

With this principle B-R-A-H-M-S CgA II KRYPTOR is a homogenous one-step immunoassay for the quantification of CgA II in human serum. The linear direct measuring range of the assay is from 20-3,000 ng/mL, going up to 1,000,000 ng/mL with automated dilution. Results can be retrieved after a 29 min incubation time.

Here's an analysis of the acceptance criteria and study findings for the B.R.A.H.M.S CgA II KRYPTOR device, based on the provided FDA 510(k) summary:

Acceptance Criteria and Reported Device Performance

Note: The provided document primarily describes analytical performance criteria and clinical performance measures (sensitivity, specificity) rather than explicit "acceptance criteria" in a pass/fail format for clinical decision-making. However, the sensitivity and specificity values obtained from the clinical study serve as the reported device performance against which implicit clinical acceptance would be judged. The analytical performance metrics are generally presented as numerical results meeting industry standards (CLSI guidelines).

Study Details:

1. Sample size used for the test set and the data provenance:

   • Clinical Study (for Sensitivity and Specificity): 153 adult GEP-NET patients (grade 1 and 2), with 459 total observations (likely reflecting multiple monitoring visits per patient). The study was described as a prospective study.
   • Clinical Cut-off Derivation: 102 patients with diagnosed well-differentiated G1 and G2 GEP-NETs. This was a retrospective, bicentric observational pilot study.
   • Reference Range Determination: 206 samples from self-declared healthy individuals. Data provenance is USA.
   • Analytical studies: Various sample sizes were used, often involving replicates of pooled or individual human serum samples. For example, LoQ used 420 total replicates from 7 different pools of human serum samples.
   • Provenance for analytical samples: Not explicitly stated but generally implied to be from human subjects, for instance, "human serum samples".

2. 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):

   • For the clinical study, tumor progression was classified by RECIST 1.1 criteria. This implies that experts (typically radiologists or oncologists) were involved in interpreting imaging (CT/MRI) according to these established criteria to determine the ground truth for tumor progression.
   • The document does not specify the direct number of experts or their specific qualifications (e.g., "radiologist with 10 years of experience"). However, RECIST 1.1 is an internationally recognized standard for evaluating cancer treatment response based on imaging, implying adjudication by qualified personnel.

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

   • The ground truth for tumor progression in the clinical studies was established using RECIST 1.1 criteria based on standard imaging (CT/MRI).
   • The document does not explicitly state an adjudication method like "2+1" or "3+1" for discordant interpretations if multiple readers were involved in RECIST assessment. However, RECIST guidelines themselves are designed to standardize interpretation, and clinical trials often employ independent central review or consensus panels for definitive RECIST ratings, though this specific detail is not provided here.

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:

   • No, an MRMC comparative effectiveness study was not done. This device is an in vitro diagnostic (IVD) for quantitative determination of Chromogranin A concentration in human serum, intended to be used in conjunction with other clinical methods as an aid in monitoring. It is not an AI-assisted imaging device or a device that directly aids human readers in interpreting images.

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

   • This is an IVD assay, which functions as a "standalone" measurement of a biomarker in serum. The results are generated by the automated instrument (B.R.A.H.M.S KRYPTOR compact PLUS analyzer) without direct human interpretation of the measurement itself. However, the device's output (CgA concentration) is explicitly stated to not be used for standalone diagnosis or monitoring but "in conjunction with other clinical methods." So while the analytical measurement is standalone, the clinical interpretation for decision-making is not.

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

   • For the clinical performance evaluation (sensitivity and specificity for tumor progression), the ground truth was imaging-based tumor assessment using RECIST 1.1 criteria. This is a form of expert assessment based on a standardized methodology, often relying on radiologists and oncologists to interpret imaging studies.

7. The sample size for the training set:

   • This document describes an IVD device submission, not a machine learning/AI device. Therefore, the concept of a "training set" for an algorithm in the typical AI sense does not directly apply. The development and validation of the assay itself would have involved numerous samples for optimization and establishment of analytical performance characteristics, but these are not referred to as a "training set" here.

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

   • As addressed above, the concept of a "training set" in the context of machine learning/AI is largely inapplicable here. The development of the assay's analytical characteristics (e.g., linearity, precision, detection limits) would be established through standard laboratory practices and reference materials, for which "ground truth" is defined by known concentrations or established analytical methods.

Ask a specific question about this device

Immunoassay for the in vitro quantitative determination of antibodies to thyroglobulin in human serum and plasma. The anti-Tg autoantibodies determination is used as an aid in the detection of autoimmune thyroid diseases in conjunction with other laboratory and clinical findings.

The electrochemiluminescence immunoassay "ECLIA" is intended for use on the cobas e 411 immunoassay analyzer.

The Elecsys Anti-Tg immunoassay makes use of a competitive test principle using biotinylated human antigen and monoclonal human anti-Tg antibodies labeled with a ruthenium complex. The Elecsys Anti-Tg immunoassay is intended for the quantitative determination of antibodies to thyroglobulin in human serum and plasma. It is intended for use on the cobas e immunoassay analyzers.

Results are determined via a calibration curve which is instrument-specifically generated by 2 point calibration and a master curve provided via the reagent barcode or e barcode.

The provided text describes the performance evaluation of the Elecsys Anti-Tg immunoassay, a diagnostic device, and its acceptance criteria. Here's a breakdown of the requested information based on the text:

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

The document does not explicitly present a single table labeled "Acceptance Criteria" with corresponding "Reported Device Performance" in a direct side-by-side format. Instead, it describes various performance evaluations and states whether "All predefined acceptance criteria was met" for each. However, we can reconstruct a table based on the provided details for the non-clinical performance evaluation.

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

• Test Set Sample Sizes:
  • Precision: Not explicitly stated, but includes "Human serum 1-5" and "PC THYRO1-2" (presumably replicates for each, as per CLSI EP05-A3 which requires sufficient replicates).
  • Lot-to-lot Reproducibility: "three reagent lots" (number of samples per lot not specified).
  • Analytical Sensitivity (LoB, LoD, LoQ): Not explicitly stated, determined according to CLSI EP17-A2 which has sample size recommendations.
  • Linearity: "Six dilution series" using "native human serum samples and sample pools" (number of samples/pools not specified).
  • Endogenous Interferences: Not explicitly stated per substance, but mentions "Six endogenous substances."
  • Analytical Specificity/Cross-Reactivity: Not explicitly stated (for anti-TPO).
  • Exogenous Interferences: "17 commonly and 14 specially used pharmaceutical compounds" (number of samples not stated).
  • Sample Matrix Comparison: "blood from 13 donors" (tested across serum, K2-EDTA, K3-EDTA plasma, and serum separation tubes from 3 manufacturers).
  • Method Comparison to Predicate: "total of 129 human serum samples."
• Data Provenance: The document does not specify the country of origin for the data or whether the studies were retrospective or prospective. It is a "510(k) Summary" for an FDA submission, reporting on laboratory performance studies. Given they are "non-clinical performance evaluation," these are typically controlled laboratory studies.

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 device is an in-vitro diagnostic (IVD) immunoassay. The ground truth for such devices is established through analytical testing against reference materials, established methods, and clinical samples with known characteristics, not typically by expert consensus in the same way as an imaging AI. The "ground truth" here is the precise concentration or presence/absence of the analyte (thyroglobulin antibodies) as determined by the study's reference method or expected values for standards/controls.

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

Not applicable for this type of IVD analytical performance study. Adjudication methods like 2+1 or 3+1 are common in clinical trials or imaging studies where expert readers interpret results, but not for direct quantitative measurements from an immunoassay. The acceptance criteria are based on statistical analysis of quantitative results (e.g., precision, linearity, recovery, regression).

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 immunoassay for determining antibody levels, not an imaging device or an AI intended to assist human readers. Hence, no MRMC study was performed, and human reader improvement with AI assistance is not relevant to this device's evaluation.

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

This entire non-clinical performance evaluation section (4. NON-CLINICAL PERFORMANCE EVALUATION) describes the standalone performance of the Elecsys Anti-Tg immunoassay (a device, not an AI algorithm). The measurements are performed by the "cobas e 411 immunoassay analyzer," which acts as the "algorithm" or automated system. There's no human "in the loop" for the direct measurement results themselves, though human operators are involved in running the tests and interpreting the results in a clinical setting.

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

The ground truth for the analytical performance studies (precision, linearity, sensitivity, interferences, stability) is based on:

• CLSI guidelines: The studies adhere to specific Clinical and Laboratory Standards Institute (CLSI) guidelines (e.g., EP05-A3, EP17-A2, EP06-Ed2), which define how such analytical characteristics are determined using reference materials, spiked samples, and statistical methods.
• Reference Standards/Materials: Implied in sections like "Traceability/Standardization" against the NIBSC 65/93 Standard, and the use of calibrators (Anti-Tg CalSet) and controls (PreciControl ThyroAB).
• Known Sample Characteristics: For linearity, samples with varying known concentrations are typically used. For interference studies, samples spiked with known interferents are used.
• Predicate Device Comparison: For method comparison, the predicate device's results serve as a comparative reference.

8. The sample size for the training set

Not applicable in the context of an immunoassay. This device is an in-vitro diagnostic test kit (reagents) used on an existing analyzer. It does not involve a "training set" in the machine learning sense. The "development" or "training" of such a diagnostic involves optimizing the chemical and biological components of the assay (reagents, antibodies, detection method) and calibrating the system across a range of known concentrations.

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

As there is no "training set" in the AI/machine learning sense, this question is not applicable. The development process for an immunoassay involves extensive research and development to create reagents that accurately quantify the target analyte. Calibration is done using reference materials with assigned values, and the assay's performance characteristics (as detailed in section 4) are then rigorously validated.

Ask a specific question about this device

The HAVAb IgG II assay is a chemiluminescent microparticle immunoassay (CMA) used for the qualitative detection of IgG antibody to hepatitis A virus (IgG anti-HAV) in human adult and pediatric (4 through 21 years) serum (collected in serum and serum separator tubes) and plasma (collected in sodium heparin, lithium heparin separator, dipotassium EDTA, and tripotassium EDTA tubes) from patients with signs and symptoms or at risk for hepatitis A on the Alinity i system.

The HAVAb IgG II assay is used to determine the immune status of individuals to hepatitis A virus (HAV) infection. Warning: This assay has not been cleared for use in screening blood, plasma, or tissue donors. This assay camot be used for the diagnosis of acute HAV infection.

Assay performance characteristics have not been established when the HAVAb IgG II assay is used in conjunction with other hepatitis assays.

The HAVAb IgG II assay is a chemiluminescent microparticle immunoassay (CMIA) used for the qualitative detection of IgG antibody to hepatitis A virus (IgG anti-HAV) in human adult and pediatric (4 through 21 years) serum (collected in serum and serum separator tubes) and plasma (collected in sodium heparin, lithium heparin separator, dipotassium EDTA, and tripotassium EDTA tubes) from patients with signs and symptoms or at risk for hepatitis A on the Alinity i system. The kit includes reagents (Microparticles, Conjugate, Assay Diluent), Calibrator, and Controls. The assay is an automated, two-step immunoassay.

Considering the provided document, the device described is an in vitro diagnostic (IVD) assay (HAVAb IgG II) for the qualitative detection of IgG antibody to hepatitis A virus (IgG anti-HAV). The FDA 510(k) summary focuses on establishing substantial equivalence to a predicate device, not on meeting specific acceptance criteria in the context of an AI/ML medical device's performance evaluation against clinical ground truth.

Therefore, many of the requested criteria (like ground truth establishment by experts, adjudication methods, multi-reader multi-case studies, and separate training/test sets with their ground truth establishment) are generally not applicable to the performance claims made for this in vitro diagnostic device in this FDA submission. The document describes analytical and clinical performance studies typical for an IVD, focusing on agreement with a predicate device and reproducibility/precision, rather than predictive performance of an AI model against a complex clinical endpoint established by human experts.

Based on the provided text, here's a breakdown of the information that is applicable and a clear indication where the requested information is not present or relevant to this type of device submission:

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

The document doesn't explicitly state "acceptance criteria" in a table format that would typically be found for an AI/ML model for diagnostic imaging (e.g., target sensitivity/specificity values). Instead, it presents performance data for comparison to a predicate device and for reproducibility. The implicit "acceptance criterion" for a 510(k) is demonstrating "substantial equivalence" to a legally marketed predicate device.

However, we can infer some performance metrics presented as evidence of equivalence:

2. Sample sizes used for the test set and the data provenance

• Clinical Performance Test Set (Agreement with Predicate):

  • Individuals at Increased Risk of HAV Infection: n=250
  • Individuals with Signs and Symptoms of Hepatitis Infection: n=499
  • Pediatric Population: n=105
  • Total for Agreement Study: 250 + 499 + 105 = 854 specimens.
  • Data Provenance: Prospective multi-center study.
    • Increased Risk of HAV: collected in California, Colorado, Florida, Illinois, Massachusetts, North Carolina, and Texas.
    • Signs and Symptoms of Hepatitis: collected in California, Colorado, Florida, Illinois, Massachusetts, and Texas.
    • Pediatric Population: collected in the US (California and Massachusetts) and Belgium.

• Precision/Reproducibility Test Sets:

  • Within-Laboratory Precision: n=80 per sample/control for a representative combination (tested over 20 days, 2 replicates/day). Total tested for this study was 4 reagent/calibrator/instrument combinations.
  • System Reproducibility: n=360 per sample/control (tested at 3 sites, with 4 replicates/run, 2 runs/day, 5 days).

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

• Not applicable / Not stated. For this IVD device, the "ground truth" for the clinical performance study was the result produced by the legally marketed predicate device (ARCHITECT HAVAB-G). This is a common practice for 510(k) submissions for IVDs. There were no human experts adjudicating results for the purpose of establishing a "ground truth" beyond what the predicate device determined.

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

• Not applicable. As the ground truth was the predicate device's result, no human adjudication method (like 2+1, 3+1) was used or described.

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 in vitro diagnostic assay, not an imaging AI device intended to assist human readers. Therefore, an MRMC study and evaluation of human reader improvement with AI assistance were not performed.

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

• Yes, in spirit. The device (HAVAb IgG II assay) functions as a standalone test; it's an automated immunoassay that generates a qualitative result (Reactive or Nonreactive) without human interpretation in the loop influencing the output beyond sample collection and instrument operation. Its performance was assessed directly against the predicate device's results.

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

• The "ground truth" for the clinical performance comparison was the results from a legally marketed predicate device (ARCHITECT HAVAB-G assay). In essence, the performance of the new device was compared to the established performance of the predicate. This is a common form of "truth" in demonstrating substantial equivalence for IVDs.

8. The sample size for the training set

• Not explicitly stated in terms of a "training set" for model development. This is an immunoassay, not an AI/ML model where a distinct training dataset size is typically reported. The document describes analytical verification and clinical performance studies, not model training.

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

• Not applicable. As this is not an AI/ML device in the sense of a machine learning model requiring a training set with a ground truth established by experts for supervised learning, this information is not provided. The development process for an immunoassay does not involve "training data" in the same way an AI/ML model does.

Ask a specific question about this device

The ARCHITECT CMV IgG assay is a chemiluminescent microparticle immunoassay (CMIA) used for the qualitative detection of IgG antibodies to cytomegalovirus in human serum, serum separator, and plasma tubes (lithium heparin, lithium heparin separator, and tripotassium EDTA) on the ARCHITECT i System.

The ARCHITECT CMV IgG assay is to be used as an aid in the diagnosis of infection with cytomegalovirus and as an aid in the determination of serological status to cytomegalovirus in individuals including women of child-bearing age.

The ARCHITECT CMV IgG assay has not been cleared for use in screening blood, plasma, or tissue donors.

The ARCHITECT CMV IgG assay is a chemiluminescent microparticle immunoassay (CMIA) for the qualitative detection of IgG antibodies to cytomegalovirus. The reagent kit contains microparticles coated with CMV virus lysate, murine anti-human IgG acridinium-labeled conjugate, and assay diluent. The assay is performed on the ARCHITECT i System. The presence or absence of anti-CMV IgG is determined by comparing the chemiluminescent relative light unit (RLU) in the reaction to a cutoff RLU from an active calibration.

Here's an analysis of the acceptance criteria and study information for the ARCHITECT CMV IgG device, based on the provided text:

1. Table of Acceptance Criteria and Reported Device Performance

The acceptance criteria are not explicitly stated as numerical targets in the document. However, based on the non-clinical and clinical performance summaries, we can infer performance goals for the device's accuracy and precision. The "Predicate Device" (ADVIA Centaur CMV IgG Assay, K181213) serves as a benchmark for substantial equivalence.

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

• Within-Laboratory Precision (20-Day): 117-120 replicates per panel/control across 3 reagent lots/calibrator lot combinations.
• Analytical Specificity (Interference): Not explicitly stated, but each substance was tested at 3 analyte levels.
• Analytical Specificity (Other Conditions): 187 specimens.
• CDC Panel Agreement: 80 samples (either CMV IgG negative or CMV IgG positive). Data provenance is the Centers for Disease Control and Prevention (CDC). The document doesn't specify if this was retrospective or prospective, but the description "masked, characterized serum panel" suggests it was a curated, retrospective panel.
• Clinical Percent Agreement:
  • Routine Order: 591 specimens collected in the US and 198 specimens collected outside of the US.
  • Pregnant Females: 200 specimens collected in the US.
  • Further Evaluation: 4 specimens (3 routine order, 1 pregnant female) with equivocal/grayzone results by the comparator assay.
  • Data provenance for the clinical study is multi-site (Indianapolis Indiana, Lewisville Texas, and Palo Alto California) for the US samples, and unspecified international locations for the "outside of the US" routine order specimens. This was a clinical study (method comparison), suggesting prospective collection relative to the comparison against the investigational assay.

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

There is no information provided about experts being used to establish a ground truth for the test set in the conventional sense (e.g., for image interpretation).

• For the CDC Panel Agreement, the CDC provided "their result interpretation for each sample," implying the CDC's own characterization (likely by reference methods/consensus) served as the ground truth. No specific number or qualifications of experts are given.
• For the Clinical Percent Agreement, the ground truth was established by a "current, FDA-cleared, commercially available anti-CMV IgG assay" (the comparator assay), essentially establishing a reference standard from an already approved diagnostic device. For 4 equivocal/grayzone samples, "consensus testing" with "2 additional current, FDA-cleared, commercially available anti-CMV IgG assays" was used. The document does not mention human experts establishing ground truth for these studies.

4. Adjudication Method for the Test Set

• CDC Panel Agreement: The results were compared directly to CDC's result interpretation. No explicit adjudication method is described beyond this direct comparison.
• Clinical Percent Agreement: The primary comparison was against a single comparator assay. For specific equivocal/grayzone cases (4 specimens), a form of adjudication was used where the result was "negative by consensus testing" or "negative based on the consensus result from the comparator assay and 2 additional current, FDA-cleared, commercially available anti-CMV IgG assays." This implies a form of multi-comparator consensus for ambiguous cases, but not necessarily human expert adjudication in the traditional sense.

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 MRMC comparative effectiveness study was done. This device is an automated in vitro diagnostic (IVD) assay (chemiluminescent microparticle immunoassay), not an AI-powered diagnostic imaging device or tool that assists human readers. Therefore, the concepts of human readers, AI assistance, and effect size in improving human reader performance are not applicable.

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

The device itself, the ARCHITECT CMV IgG assay on the ARCHITECT i System, is a standalone algorithm/device for qualitative detection. Its performance is evaluated entirely as an automated system without human interpretation in the loop influencing the direct output of IgG antibody detection. The "human-in-the-loop" would be the clinician interpreting the result provided by the device (e.g., Reactive, Nonreactive, Grayzone/Equivocal) in the context of a patient's overall clinical picture, but not in generating the result itself.

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

• CDC Panel Agreement: Ground truth was based on CDC's own "result interpretation" of their characterized serum panel. This likely represents a highly robust reference method or internal consensus.
• Clinical Percent Agreement: Ground truth was primarily established by a "current, FDA-cleared, commercially available anti-CMV IgG assay" (comparator assay). For a few ambiguous cases, consensus results from multiple FDA-cleared comparator assays were used. This is a form of reference method comparison.

8. The Sample Size for the Training Set

The document does not explicitly mention a "training set" in the context of machine learning or AI. As this is a traditional immunoassay, it undergoes development and validation using various biological samples. However, no specific number is provided for samples used during the development or optimization phases that could be analogized to a training set. The clinical study samples described are specifically for performance evaluation (test set).

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

As no explicit training set is defined (because it's not an AI/ML device), this question is not applicable. The assay's parameters (e.g., cutoff values) would have been established during product development using a large set of characterized samples and optimized for performance, but this is a different process than establishing ground truth for an AI training set.

Ask a specific question about this device

The Tosoh Automated Glycohemoglobin Analyzer HLC-723G8 is intended for in vitro diagnostic use for the measurement of % hemoglobin A1c (HbA1c) (DCCT/NGSP) and mmol/mol hemoglobin A1c (IFCC) in venous whole blood specimens using ion-exchange high-performance liquid chromatography (HPLC). This test is an aid in diagnosis of diabetes and identifying patients who may be at risk for developing diabetes, and for monitoring of long-term blood glucose control in individuals with diabetes mellitus.

The Tosoh Automated Glycohemoglobin Analyzer HLC-723G8 is an automated High-Performance Liguid Chromatography (HPLC) system that separates and reports stable hemoglobin A1c (sA1c) percentage in venous whole blood. The operational portion of the G8 is composed of a sampling unit, liquid pump, degasser, column, detector, microprocessors, sample loader, smart media card, operation panel, and a printer. The Tosoh Automated Glycohemoglobin Analyzer HLC-723G8 uses ion-exchange HPLC for rapid, accurate, and precise separation of the stable form of HbA1c (sA1c) from other hemoglobin fractions. The G8 uses a non-porous cation exchange column and separates the hemoglobin components in the blood. Separation is achieved by utilizing differences in ionic interactions between the cation and exchange group on the column resin surface and the hemoglobin components in a step gradient elution. The hemoglobin fractions (designated as A1a. A1b. F. LA1c+, SA1c, A0, and, if present, H-V0, H-V2, H-V2 and H-V3) are subsequently removed from the column by performing a step-wise elution gradient using the varied salt concentrations in the Variant Elution Buffers HSi 1, 2 and 3. The peaks, H-V0, H-V1, H-V2 and H-V3 are typically presumptive HbAD, HbAS, HbAC and HbAE respectively. The software compares the retention times of hemoglobin fractions in a sample to the expected "windows of retention" and labels each fraction that correctly elutes within a defined expected window of retention. The software designates a hemoglobin fraction as POX (where X is the order of the peak as it elutes from the column) if it does not match a defined window of retention. All automated processes in the G8 are controlled by internal microprocessors, using software downloaded via a smart media card. The result report is printed and can be stored on the instrument. The data can be transmitted to a host computer through a bi-directional interface. The result report includes the sample ID, date, percentage and retention time of each fraction of hemoglobin, sA1c percentage and total A1 percentage, along with a chromatogram of the elution pattern of the hemoglobin fractions. If a sample contains a hemoglobin variant, the column elutes the fraction depending upon its charge.

The provided text describes the non-clinical performance testing of the Tosoh Automated Glycohemoglobin Analyzer HLC-723G8 (subject device) to support its substantial equivalence to a predicate device. This document focuses on the analytical performance of a diagnostic device rather than an AI/ML powered device, so some of the specific questions regarding AI/ML study design (e.g., number of experts, adjudication methods, MRMC studies) are not applicable.

Here's the information extracted from the document:

1. Acceptance Criteria and Reported Device Performance

The acceptance criteria are generally implied by the statement "All performance testing results met their pre-determined acceptance criteria." While explicit numerical acceptance criteria for each test are not listed in a consolidated table, the discussion throughout the "Summary of Non-Clinical Performance Testing" implicitly defines them through the methodology and results. For example, for precision/repeatability, the claim of "imprecision at ≤ 2%" was a pre-established criterion. Similarly, for hemoglobin variant interference, "Non-clinically significant interference was defined as <= 6% relative difference in the results from a comparative method."

It's important to note that the provided text is a 510(k) summary, which often summarizes detailed study reports. The full reports would contain explicit acceptance criteria thresholds.

Table of Acceptance Criteria (Implied) and Reported Device Performance:

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

• Precision/Repeatability Study:
  • Sample Size: Four concentrations of HbA1c% in K2EDTA venous whole blood. A total of 720 measurements per concentration were performed (three analyzers, over 20 non-consecutive days, with three reagent lots, samples run in duplicate, two times per day).
  • Data Provenance: The study was implicitly conducted retrospectively on collected samples. The country of origin is not explicitly stated, but the testing was performed at "Tosoh Bioscience, Inc." laboratories, likely in the US (given the South San Francisco address).
• Method Comparison Study:
  • Sample Size: 220 K₂EDTA venous whole blood specimens with HbA1c concentrations spanning the measuring range (4.0-16.9%). Each specimen was tested in duplicate.
  • Data Provenance: The study was conducted at "two separate sites," one a "moderate complexity clinical laboratory" (for the candidate device) and the other an "NGSP SRL" (for the comparator device). The precise country of origin is not explicitly stated. The specimens were collected and likely used retrospectively for the comparison.
• Endogenous Interfering Substances / Hemoglobin Variant Interference Studies:
  • Sample Size:
    • Endogenous interference: Not explicitly stated per substance, but specimens were "spiked with increasing amounts of the interfering substance."
    • Hemoglobin Variant Interference:
      • HbC: 26 samples
      • HbD: 24 samples
      • HbE: 26 samples
      • HbS: 29 samples
      • HbA2: 20 samples
      • HbF: 21 samples
  • Data Provenance: Not explicitly stated, but implies lab-tested, potentially spiked, and clinical samples from unknown origin (likely retrospective).

3. Number of Experts and Qualifications for Ground Truth:

This device is an automated in vitro diagnostic analyzer for quantitative measurement of HbA1c. The "ground truth" for such devices is typically established by:

• Reference methods (e.g., NGSP SRL certified methods).
• Traceability to international standards (DCCT, IFCC).
• Direct analytical measurements and statistical comparisons, rather than expert consensus on image interpretation.

Therefore, the concept of "number of experts used to establish the ground truth" similar to that in an AI imaging study is not applicable here. The ground truth is the chemical measurement itself, validated against recognized reference standards.

4. Adjudication Method for the Test Set:

Not applicable, as this is an analytical device for quantitative measurement, not an interpretative AI/ML system requiring human adjudication of results. The "adjudication" is through statistical agreement with reference methods and assessment against predetermined analytical performance criteria.

5. If a multi-reader multi-case (MRMC) comparative effectiveness study was done:

No. An MRMC study is relevant for AI/ML systems that assist human readers in tasks like image interpretation. This device performs automated quantitative measurements, and its performance is evaluated against reference methods and analytical standards, not human readers.

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

Yes, the studies described are standalone performance studies of the device (Tosoh Automated Glycohemoglobin Analyzer HLC-723G8) without human intervention in the measurement process itself. The device is an automated HPLC system.

7. The Type of Ground Truth Used:

The ground truth for evaluating the subject device's performance is based on:

• Reference Methods: Specifically, the NGSP SRL (National Glycohemoglobin Standardization Program Secondary Reference Laboratory) using the Trinity Biotech Premier Hb9210™ HbA1c Analyzer (for method comparison) and Primus Model CLC 330 (for some hemoglobin variant interference testing), and the Bio-Rad VARIANT II TURBO HbA1c Kit - 2.0 (for some hemoglobinopathy interference testing).
• Traceability and Standardization: The device's results are traceable to the Diabetes Control and Complications Trial (DCCT) reference method and IFCC, and it is certified by the NGSP.
• Analytical Standards: For precision, linearity, and interference, the "ground truth" reflects the known concentrations of spiked samples or validated concentrations in patient samples, against which the device's measurements are compared.

8. The Sample Size for the Training Set:

The provided document describes non-clinical performance testing to support substantial equivalence after a software modification (v5.24). It does not provide information on the training set used for the development of the device's algorithms or software. This 510(k) summary is focused on verification and validation of the modified device.

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

This information is not provided in the 510(k) summary. Given that the device is an automated HPLC system, its "training" per se would involve optimization of its operational parameters, calibration curves, and potentially peak detection/integration algorithms, likely using well-characterized samples with known HbA1c concentrations established by reference methods. However, the details of such a "training set" and its ground truth establishment are not discussed here.

Ask a specific question about this device

The Atellica® IM BRAHMS Procalcitonin (PCT) assay is for in vitro diagnostic use in the quantitative determination of procalcitonin in human serum and plasma (EDTA, lithium heparin, and sodium heparin) using the Atellica® IM Analyzer.

The Atellica IM BRAHMS PCT assay is intended for use, in conjunction with other laboratory findings and clinical assessments, as an aid in:

· The risk assessment of critically ill patients on their first day of ICU admission for progression to severe sepsis and septic shock.

• Assessing the cumulative 28-day risk of all-cause mortality for patients diagnosed with severe sepsis or septic shock in the ICU or when obtained in the emergency department or other medical wards prior to ICU admission, using percent change in PCT level over time.

· Decision-making on antibiotic therapy for patients with suspected or confirmed lower respiratory tract infections (LRTI) – defined as community-acquired pneumonia (CAP), acute bronchitis, and acute exacerbation of chronic obstructive pulmonary disease (AECOPD) - in an inpatient setting or an emergency department.

• · Decision-making on antibiotic discontinuation for patients with suspected or confirmed sepsis.

The Atellica IM BRAHMS PCT assay is a 2-site sandwich immunoassay using direct chemiluminescent technology that uses 3 mouse monoclonal antibodies specific for PCT. The first antibody, in the Lite Reagent, is a mouse monoclonal anti-PCT antibody labeled with acridinium ester. The second and third antibodies, in the ancillary reagent, are mouse monoclonal anti-PCT antibodies labeled with fluorescein. The immunocomplex formed with PCT is captured with mouse monoclonal anti-fluorescein antibody coupled to paramagnetic particles in the Solid Phase.

A direct relationship exists between the amount of PCT present in the patient sample and the amount of relative light units (RLUs) detected by the system.

The provided text describes the Siemens Healthcare Diagnostics Atellica IM B.R.A.H.M.S Procalcitonin (PCT) assay, an in vitro diagnostic device. The study presented focuses on the analytical and clinical performance of this assay, comparing it to a predicate device and evaluating its utility in specific clinical scenarios related to sepsis and respiratory tract infections.

Here's an analysis of the acceptance criteria and the study proving the device meets them:

Acceptance Criteria and Reported Device Performance

The acceptance criteria for this device are implicitly derived from the performance goals demonstrated in the analytical and clinical studies. Since this is a 510(k) submission, the primary goal is to show substantial equivalence to a legally marketed predicate device (B.R.A.H.M.S PCT sensitive KRYPTOR). The performance characteristics are presented as meeting industry standards (CLSI guidelines) and showing comparable results to the predicate.

Here's a table summarizing the analytical performance with implicit acceptance criteria (typically, these would be defined before the study in a protocol, often as a deviation tolerance from the predicate or a specific measure):

Detailed Study Information:

1. Sample sizes used for the test set and the data provenance:

   • Analytical Performance:
     • Precision: 5 contrived human serum samples, 80 replicates per sample.
     • Linearity: 9 samples (prepared from high and low human serum pools).
     • Dilution Recovery: 5 human serum samples.
     • Detection Limits (LoD): 360 determinations (160 blank, 200 low-level replicates).
     • Endogenous/Therapeutic Drug/Cross-Reactivity Interference: Not specified beyond "serum pools" and "serum and plasma."
     • Heterophile Interference: HAMA and RF positive patient samples + control samples.
     • Method Comparison: 623 native human serum samples with assigned values from the predicate device.
     • Matrix Comparison: 51 matched specimen sets in 4 tube types (Serum, EDTA plasma, Lithium Heparin Plasma, Sodium Heparin Plasma).
     • Expected Values/Reference Interval: 144 serum samples from normal subjects.
   • Clinical Studies:
     • 28-day Mortality: 858 adult patients (>18 years) diagnosed with severe sepsis or septic shock admitted to ICU. The analysis population comprised 598 subjects.
     • Antibiotic Therapy/Discontinuation Support: Systematic literature reviews and meta-analyses:
       • Study-level meta-analysis (Initiation/Discontinuation): 11 RCTs (4090 patients) published 2004-2016.
       • Patient-level meta-analysis (Initiation): 13 RCTs (3142 patients) published 2004-2011.
       • Patient-level meta-analysis (Discontinuation): 5 RCTs (598 patients) published 2008-2010.
   • Data Provenance:
     • Analytical Data: In-house or contracted lab studies (implied to be prospective given the detailed methodology descriptions).
     • Clinical Data: "Banked specimens that were collected from subjects... and included as part of the intention-to-diagnose population from the BRAHMS MOSES study (DEN150009)." This indicates retrospective use of previously collected clinical trial data. The original MOSES study subjects were recruited across 13 investigational sites in the US. The meta-analyses for antibiotic guidance consolidated data from previously published RCTs.

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

   • For the analytical test sets, ground truth is established by standard laboratory methods and reference materials (e.g., recombinant PCT, serum pools), not human experts.
   • For the clinical study, the ground truth for patient outcomes (e.g., severe sepsis/septic shock diagnosis, 28-day all-cause mortality, need for ICU care) was presumably established by the clinical assessments and medical records from the original BRAHMS MOSES study that provided the banked specimens. The document does not specify the number or qualifications of clinicians/experts who established these initial diagnoses and outcomes for the MOSES study, but it's implied to be standard clinical practice within a multi-site clinical trial.
   • For the meta-analyses, the "ground truth" for antibiotic management decisions and patient outcomes (mortality, LOS, etc.) derived from the various RCTs where those decisions and outcomes were prospectively recorded by the clinical teams involved in those trials.

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

   • For the analytical studies, the "ground truth" values are quantitative measurements or characteristics of samples, typically not subject to human adjudication in the same way as, for example, image interpretation. Internal laboratory quality control and statistical methods (e.g., regression analysis, CV calculations) serve as "adjudication."
   • For the clinical study, the data were obtained from previously collected "banked specimens" and "clinical assessments" from the BRAHMS MOSES study. The document does not describe a new adjudication process for this particular 510(k) submission's analysis of this retrospective data. The rigor of the original MOSES study's data collection and endpoint ascertainment would be relevant but is not detailed here.
   • The meta-analyses relied on data from published RCTs; any adjudication would have occurred within the original trials.

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:

   • No. This device is an in vitro diagnostic (IVD) assay that measures a biomarker (Procalcitonin) in human specimens. It is not an AI-assisted diagnostic imaging or pathology device that requires human "readers." Therefore, an MRMC study is not applicable and was not performed.

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

   • Yes, implicitly. The performance characteristics (precision, linearity, detection limits, interference, method comparison) are "standalone" results of the assay itself, demonstrating its analytical performance when run on the Atellica IM Analyzer. The clinical studies demonstrate the performance of the assay's results when used in conjunction with "other laboratory findings and clinical assessments," but the assay itself generates the PCT value independently of human interpretation of that specific run.

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

   • Analytical Studies: "Ground truth" for analytical studies is typically established by meticulously prepared reference materials, spike-in methods, and comparison to a well-validated predicate device using a large sample size of native patient samples.
   • Clinical Studies:
     • For the 28-day mortality prediction, the ground truth was outcomes data (survival status at 28 days) derived from a pre-existing clinical study (BRAHMS MOSES study).
     • For antibiotic guidance, the insights were derived from meta-analyses of published Randomized Controlled Trials (RCTs). The "ground truth" here is the aggregated, prospectively collected patient outcomes (antibiotic use, duration, mortality, complications, LOS) from these trials where patients were managed either by standard care or PCT-guided protocols.

7. The sample size for the training set:

   • This document describes a 510(k) premarket notification for an IVD assay, not an AI/ML device that typically requires a distinct "training set." The performance studies for an IVD device like this are primarily for validation and demonstrate clinical and analytical performance. There is no explicit "training set" in the context of machine learning.
   • However, if one were to consider the development of the assay's methodologies, the calibration and control materials (e.g., PCT MCM) would be used for standardization and quality control, which could be seen as analogous to internal "training" or optimization data during product development, but no specific sample size for this is detailed as "training set."

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

   • As noted above, the concept of a "training set" with ground truth establishment in the ML sense is not directly applicable to this type of IVD device submission. The assay is a chemical measurement system. Its "truth" is established via analytical validation showing its accuracy, precision, and comparability to reference methods/predicate devices.

Ask a specific question about this device

QUANTA Flash Calprotectin is a chemiluminescent immunoassay for the quantitative determination of fecal calprotectin in extracted human stool samples. Elevated levels of fecal calprotectin, in conjunction with clinical findings and other laboratory tests, can aid in the diagnosis of inflammatory bowel disease (IBD) (ulcerative colitis and Crohn's disease), and in the differentiation of IBD from irritable bowel syndrome (IBS).

QUANTA Flash Calprotectin Calibrators are intended for use with the QUANTA Flash Calprotectin Reagents for the determination of fecal calprotectin levels in extracted stool samples. Each calibrator establishes a point of reference for the working curve that is used to calculate unit values.

QUANTA Flash Calprotectin Controls are intended for use with the QUANTA Flash Calprotectin Reagents for quality control in the determination of fecal calprotectin levels in extracted stool samples.

QUANTA Flash Calprotectin Extraction Buffer is intended for use with the QUANTA Flash Calprotectin Reagents as sample extraction solution.

The principle of the assay is chemiluminescent microparticle immunoassay, a variation of solid phase immunoassay. The QUANTA Flash® Calprotectin assay is designed to run on the BIO-FLASH® instrument. This platform is a fully automated closed system with continuous load and random access capabilities that automatically processes the samples, runs the assay and reports the results. It includes liquid handling hardware, luminometer and computer with software-user interface. The QUANTA Flash® Calprotectin assay utilizes a reagent cartridge format, which is compatible with the BIO-FLASH® instrument.

Calprotectin-specific capture antibodies are coated on to paramagnetic beads, which are stored in the reagent cartridge under conditions that preserve the antibody in its reactive state. Prior to use in the BIO-FLASH® system, the reagent pack containing all the necessary assay reagents is mixed thoroughly by being inverted several times. The sealed reagent tubes are pierced with the reagent cartridge lid, and the reagent cartridge is loaded onto the instrument. Reagents are calibrated when the lot is first used. A patient extracted stool sample is prediluted by the BIO-FLASH® with sample buffer in a disposable plastic cuvette. Small amounts of the diluted patient extracted stool, the beads, and the assay buffer are all combined into a second cuvette, and mixed. This cuvette is then incubated at 37°C. The beads are magnetized and washed several times. Isoluminol conjugated monoclonal antibodies are then added to the cuvette, and again incubated at 37°C. The beads are magnetized and washed repeatedly. The isoluminol conjugate is oxidized when Trigger 1 (Fe(II)coproporphyrin in sodium hydroxide solution) and Trigger 2 (urea-hydrogen peroxide in sodium chloride solution) are added to the cuvette, and the flash of light produced from this reaction is measured as Relative Light Units (RLU) by the BIO-FLASH® optical system. The measured RLU is proportional to the amount of bound isoluminol conjugate, which is in turn proportional to the amount of calprotectin antigen captured by the antibodies (anti-calprotectin polyclonal antibodies in this case) on the beads. For quantitation, the QUANTA Flash® Calprotectin will utilize a predefined lot specific Master Curve that is uploaded onto the instrument through the reagent cartridge barcode. The Master Curve is generated by Inova Diagnostics for each reagent pack lot with in-house Standards with assigned unit values (ng/mL). The RLU and assigned ng/mL values of the Standards are used to create a 4 parameter logistic curve. These four parameters are embedded in the reagent pack barcode. When the lot is used the first time, the Calibrators are run, and based on the results obtained on the Calibrators, an instrument specific Working Curve is created; The Working Curve is used to calculate units (ng/mL) based on RLU values obtained on each sample. The obtained ng/mL values will be converted to mg/kg by a calculation that takes into account the dilution of the samples. This unit conversion is calculated automatically by the software.

Here's an analysis of the provided text, extracting the requested information about acceptance criteria and the study proving device performance:

1. Table of Acceptance Criteria and Reported Device Performance

The device is an in vitro diagnostic (IVD) test system, so performance metrics like sensitivity and specificity are evaluated, alongside analytical performance criteria common for laboratory assays.

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

• Clinical Performance Test Set:

  • Sample Size: A total of 165 characterized samples were included in the validation set.
  • Data Provenance: Samples came from studies performed at two different sites (Site A and Site B) and a commercial source.
    • Site A: Samples 1 to 107. Inclusion criteria for IBD patients included suspicion of IBD, a calprotectin test request, and underwent ileocolonoscopy. Patients diagnosed with IBD or without ileocolonoscopy were excluded.
    • Site B: Samples 108 to 175. Sixty-eight consecutive samples requested for a calprotectin determination test by a gastroenterologist. Patients with excessive mucous, unclear diagnosis, or without colonoscopy were excluded.
  • Retrospective/Prospective: The description implies a retrospective collection of "characterized samples" from previously conducted studies or existing cohorts. For Site A, samples were "recruited over a period of 6 months," suggesting prospective collection during that period, but then used retrospectively for this validation. For Site B, "consecutive samples requested" also suggests a retrospective collection from a past clinical workflow.

• Method Comparison Test Set:

  • Sample Size: 137 samples out of the 165 clinical validation study samples. Of these, 77 samples fell within the AMR of both assays.
  • Data Provenance: Same as the clinical performance test set (two sites and commercial source).

• Precision and Reproducibility Studies: Sample numbers are specified per study (e.g., 8 samples for within-laboratory precision, 8 samples for between-sites, 8 samples for between-lots, 5 samples for extraction reproducibility). These are generally manufactured or spiked samples, not from patients.

• LoD/LoQ: 2 low-level samples for LoQ, 60 blank samples and 60 low-level samples (per lot) for LoD.

• Linearity: Three extracted stool samples and three recombinant antigen samples.

• Recovery: Seven extracted stool samples.

• Interference: Six human stool specimens (one high positive, one moderately positive, one low positive, one near cut-off, one indeterminate, one negative).

• Sample Stability: Eight extracted human stool samples (negative, indeterminate, around cut-off, positive) and three extracted stool samples (indeterminate, around cut-off, high positive) for frozen stability.

• Reference Range Establishment: 61 subjects (presumably healthy donors and some with specific benign conditions).

• Expected Values (Normal Population): 164 apparently healthy stool donors.

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

• Clinical Performance Ground Truth:

  • For IBD diagnosis at Site A, "Senior gastroenterologists performed all endoscopies and findings were documented in a computer-based database. The final diagnosis of IBD (i.e. CD and UC) was independently made by a pathologist or gastroenterologist who was blinded for calprotectin results." This indicates multiple Senior Gastroenterologists and Pathologists/Gastroenterologists were involved. Specific years of experience are not mentioned, but "Senior" implies significant experience.
  • For IBD diagnosis at Site B, "Patients were diagnosed after exclusion of organic pathology on the basis of routine blood tests, thyroid function tests, serological screening for coeliac disease, stool examination for bacteria and parasites, ultrasound examination, and eventually colonoscopy using the ROME III criteria. The diagnosis of IBD was made upon clinical, endoscopic and histological findings as described in J Crohn Colitis 2012:6: 965-990 (ulcerative colitis) and J Crohn Colitis 2010:4:7-27 (Crohn's disease)." This implies multiple Gastroenterologists and Pathologists following established diagnostic guidelines.

• Reference Range Establishment: For the 61 subjects used to establish the cut-off, the description indicates they were "apparently healthy donors" or had specific benign conditions (e.g., Squamous Cell Carcinoma, Glandular Polyp, Hyperplastic Polyp, Adenoma). The diagnosis here would be based on standard clinical and pathological evaluations, though specific experts for this ground truth are not detailed.

4. Adjudication Method for the Test Set

• For the IBD diagnosis at Site A, the final diagnosis was "independently made by a pathologist or gastroenterologist who was blinded for calprotectin results." This suggests a consensus-based approach or review by independent experts, but a specific (e.g., 2+1, 3+1) method is not explicitly stated. The term "independently made" implies that the diagnosis of IBD (considered the ground truth) was confirmed by one or more experts who were shielded from the calprotectin results.

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

• No, an MRMC comparative effectiveness study involving human readers with and without AI assistance was not done. This device is an automated in vitro diagnostic test (chemiluminescent immunoassay) for measuring fecal calprotectin, not an AI-powered image analysis or decision support system that directly assists human readers in real-time interpretation. The studies focus on the analytical and clinical performance of the assay itself compared to a predicate device or clinical diagnosis.

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

• Yes, performance metrics like precision, reproducibility, LoD, LoQ, linearity, recovery, interference, and stability are all "standalone" performance measures of the device itself (its reagents and instrument) without human-in-the-loop directly influencing the measurement result.
• The "Clinical Performance Characteristics" (sensitivity, specificity) are also standalone performance data for the device's ability to differentiate IBD from controls, given a clinical ground truth.
• The "Comparison with predicate device" also serves as a standalone comparison of the new device against an existing, legally marketed device.

7. Type of Ground Truth Used

• Clinical Performance / Clinical Validation: The ground truth for IBD diagnosis was established through comprehensive medical work-up, including "clinical findings and other laboratory tests," "endoscopic and histologic analysis, radiologic work-up and laboratory tests including ileocolonoscopy" (Site A), and "exclusion of organic pathology on the basis of routine blood tests, thyroid function tests, serological screening for coeliac disease, stool examination for bacteria and parasites, ultrasound examination, and eventually colonoscopy using the ROME III criteria" (Site B). This is robust clinical diagnosis/outcomes data, supported by pathology and endoscopy.
• Analytical Performance (e.g., LoD, LoQ, Linearity, Recovery, Interference, Stability): The ground truth for these studies involves pre-defined concentrations of calprotectin (e.g., recombinant calprotectin antigen, spiked samples) or known conditions (e.g., specific concentrations of interferents, different storage temperatures/durations). This is essentially known concentrations/states as ground truth.
• Reference Range / Cut-off: The cut-off was established based on a reference population (61 subjects) and further informed by 31 diagnosed IBD patients. The definition of "negative," "indeterminate," and "positive" ranges are based on statistical analysis of these populations and clinical considerations.

8. Sample Size for the Training Set

The document does not explicitly describe a "training set" in the context of an AI/machine learning model. The device is a chemiluminescent immunoassay run on an automated instrument.

However, the "Master Curve" for the assay is established during manufacturing using "in-house Standards with assigned unit values". This master curve can be considered analogous to a foundational calibration or "training" for the assay's quantitative functionality.

• Master Curve Standards: Seven calprotectin standards (0.0 ng/mL to 3478.3 ng/mL) are used to create the lot-specific Master Curve. The number of samples for generating this master curve is not explicitly stated beyond "7 Standards", but it is a manufacturing process.
• Calibrators and Controls Value Assignment: Calibrators and controls are assigned values by testing on at least two instruments, on at least two lots of reagent cartridge, in replicates of 5 to obtain a minimum of 10 data points. This process of value assignment, while not for an AI training set, is how the reference values are established for the functional units of the assay.

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

As noted above, there isn't a traditional "training set" for an AI model. For the functional calibration of the assay:

• Master Curve Ground Truth: The "Master Curve Standards" are "in-house Standards with assigned unit values (ng/mL)." This effectively means the ground truth for the Master Curve is based on traceable, pre-assigned concentrations of calprotectin.
• Calibrators and Controls Ground Truth: The "Calibrator and Control values are directly traceable to the in-house Standards that are used to create the Master Curves for the QUANTA Flash Calprotectin assay." This means their ground truth is rooted in the same pre-assigned values of the in-house standards.

Ask a specific question about this device

N Latex FLC kappa and lambda assays: In-vitro diagnostic reagents for the quantitative determination of free light chains (FLC), type kappa or type lambda, in human serum and EDTA plasma by means of particle-enhanced immunonephelometry using the BN Systems. FLC measurements are used as an aid in the diagnosis of multiple myeloma (MM) and amyloidosis (AL).
N FLC Supplementary Reagent: Supplementary reagent for the immunonephelometric determination of free light chains (FLC), type kappa and type lambda on BN Systems. A mixture of both supplementary reagents is used to suppress interference by rheumatoid factors and human anti-mouse antibodies (HAMA).
N FLC Standard SL: Establishment of reference curves for the determination of free light chains (FLC), type kappa and type lambda on the BN Systems.
N FLC Controls SL1 and SL2: The N FLC Controls SL1 and SL2 are for use as assayed accuracy controls in the determination of free light chains (FLC), type kappa and type lambda by immunonephelometry with the BN Systems.

The N Latex FLC (free light chain) assays are in vitro diagnostic reagents for the quantitative determination of free light chains, type kappa or type lambda, in human serum and EDTA plasma by means of particle-enhanced immunonephelometry using the BN™ II and BN ProSpec® Systems. Used in conjunction with other clinical and laboratory findings, FLC measurements are used as an aid in the diagnosis of multiple myeloma (MM) and amyloidosis (AL). Used in conjunction with the assay reagents, N FLC Standard SL is for use in the establishment of reference curves for the determination of free light chains, type kappa and type lambda on the BN™ II and BN ProSpec® Systems. The N FLC Control SL 1 and 2 products are for use as assayed accuracy controls and precision controls in the determination of free light chains, type kappa and type lambda by immunonephelometry with the BN™ II and BN ProSpec® Systems. The FLC test systems are based upon the principles of particle-enhanced immunonephelometry. Polystyrene particles coated with monoclonal antibodies to human free light chains, type kappa or lambda, respectively, are agglutinated when mixed with samples containing FLC. These aggregates scatter a beam of light passed through the sample. The intensity of the scattered light is proportional to the concentration of the respective protein in the sample. The result is evaluated by comparison with a standard of known concentration.

The provided text describes the Siemens N Latex FLC kappa and N Latex FLC lambda assays, along with their associated calibrators and controls. These devices are intended for the quantitative determination of free light chains (FLC) in human serum and EDTA plasma, used as an aid in diagnosing multiple myeloma (MM) and amyloidosis (AL).

Here's an analysis of the acceptance criteria and the study that proves the device meets them:

1. Table of Acceptance Criteria and Reported Device Performance

The acceptance criteria are generally implied by the performance characteristics presented in the study. For analytical performance, typical acceptance limits for precision (CV%), linearity, and interference are industry standards for IVD devices. For clinical performance, the reported sensitivity and specificity values against clinical diagnosis are the acceptance metrics.

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

• Precision and Reproducibility:
  • Sample Size: Serum samples were obtained from commercial sources. Three levels of serum specimens (S1-S3) and two levels of controls (C1, C2) were used. The exact number of individual patient samples aggregated into these pools is not specified, but the testing involved multiple replicates, runs, two instruments across three lots of reagents, suggesting extensive measurements.
  • Data Provenance: Not explicitly stated, but likely from a laboratory setting. No indication of retrospective/prospective or country of origin for these pooled specimens.
• Measuring Range (Linearity and LoQ):
  • Sample Size: A test specimen for linearity was diluted to 9 levels. Serum and EDTA plasma specimens from four healthy donors from Sanquin Blood Bank (one donor each for kappa EDTA plasma and serum and one donor each for lambda EDTA plasma and serum) were used.
  • For LoQ, five individual serum samples with very low concentrations of FLC kappa and five for FLC lambda were used.
  • Data Provenance: Fresh human serum and EDTA plasma. Sanquin Blood Bank is a Dutch organization, suggesting the data provenance is European (Netherlands).
• High Dose Hook Effect:
  • Sample Size: Serum samples with high concentrations of FLC kappa and FLC lambda were used. The number of samples is not specified.
  • Data Provenance: Not specified.
• Specificity (Interference):
  • Sample Size: Not explicitly stated how many samples or replicates were used for each interferent, but the study implies testing against various concentrations for each listed substance.
  • Data Provenance: Not specified.
• Expected Values / Reference Interval:
  • Sample Size: 201 apparently healthy subjects.
  • Data Provenance: US-population. This was a prospective study to establish reference intervals.
• Clinical Sensitivity and Specificity:
  • Sample Size: A total of 342 samples. This included:
    • 96 samples from Multiple Myeloma patients.
    • 83 samples from AL Amyloidosis patients.
    • 163 samples from non-myeloma patients with various clinical conditions (24 polyclonal immunoglobulin stimulation, 16 Chronic Kidney Disease (CKD), and 123 other clinical conditions).
  • Data Provenance: Not explicitly stated, but these are patient samples. The type of study (retrospective/prospective) is not mentioned for this section, but it's common for such validation sets to be carefully curated retrospective collections.
• Method Comparison with Predicate Device:
  • Sample Size: 152 serum samples from patients with monoclonal gammopathy.
  • Data Provenance: Patients with monoclonal gammopathy. Not specified if retrospective or prospective or country of origin.

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

The document does not specify the number or qualifications of experts used to establish the "ground truth" for the test set in the conventional sense of a diagnostic imaging or pathology study.

• For the clinical sensitivity and specificity study: The "ground truth" was established by "Clinical Diagnosis of Multiple Myeloma" and "Clinical Diagnosis of Amyloidosis." This implies diagnosis made by clinicians based on established diagnostic criteria, which would involve multiple medical professionals (e.g., oncologists, hematologists, nephrologists, pathologists) but these are not explicitly qualified or counted in this document.
• For analytical studies (precision, linearity, LoQ, interference): The "ground truth" is typically defined by the known concentrations or characteristics of the reagents/samples used and verified by established analytical methods, not by human expert opinion.

4. Adjudication Method for the Test Set

Not applicable for this type of in-vitro diagnostic device study. Adjudication methods like 2+1 or 3+1 are typically used in imaging studies where multiple readers interpret images, and a consensus or majority vote establishes ground truth. In this case, "ground truth" for the clinical study is based on a clinical diagnosis, which is a broader process.

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 is relevant for comparing human reader performance with and without AI assistance, typically in medical imaging. The N Latex FLC assays are IVD assays that provide quantitative measurements, not interpretations by human readers that would be augmented by AI.

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

Yes, the studies described are for the standalone performance of the N Latex FLC assays. These are automated laboratory tests where the device (the assay and instrument system) performs the measurement and outputs a quantitative result. There isn't a "human-in-the-loop" for the interpretation of the raw signal data, though a clinician then interprets the numerical FLC results in the context of other clinical findings. The performance metrics (precision, linearity, LoQ, sensitivity, specificity) reflect the algorithm/device's performance directly.

7. The Type of Ground Truth Used

• Clinical Sensitivity and Specificity: "Clinical Diagnosis of Multiple Myeloma" and "Clinical Diagnosis of Amyloidosis." This would typically be based on a combination of clinical signs, symptoms, other laboratory tests, bone marrow biopsy results, and imaging studies, as per established medical guidelines.
• Analytical Studies (Precision, Linearity, LoQ, Hook Effect, Interference): Ground truth was established by known concentrations of analytes in prepared samples or by the absence/presence of interfering substances at specified levels. For example, linearity samples were prepared by serial dilution from a high concentration, and LoQ samples were prepared to known very low concentrations. For interference, substances were added at specific concentrations.
• Reference Intervals: Based on measurements from an "apparently healthy subject" population (201 subjects from a US-population).

8. The Sample Size for the Training Set

The document does not explicitly mention a "training set" in the context of an AI/ML algorithm. This device is an immunoassay using a well-established technology (particle-enhanced immunonephelometry). Such systems are typically developed and validated using calibration and verification samples, but not "training sets" in the machine learning sense. The N FLC Standard SL is used for establishing reference curves (calibration), which is a form of "training" for quantitative measurement but not for a high-level diagnostic algorithm that learns from data.

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

As noted above, a "training set" in the AI/ML context is not directly applicable here.

• For the calibration (N FLC Standard SL): "Calibration of the assay is traceable to an internal master calibrator." This master calibrator's "ground truth" (assigned value) would have been established through a rigorous internal development and characterization process, likely involving primary reference materials or highly characterized analytical methods, though the specific details are not provided in this summary.
• For the controls (N FLC Control SL1 and SL2): "The concentration of the free light chains (FLC), type kappa and type lambda is calibrated against standard preparations and is lot-dependent." This means their "ground truth" values are assigned relative to the calibration curve established using the N FLC Standard SL, which in turn refers to the internal master calibrator.

Ask a specific question about this device