LogoFDA 510(k) Search
K Number
K133728
Device Name
DIMENSION VISTA CREATININE (CRE2) FLEX REAGENT CARTRIDGE
Manufacturer
Siemens Healthcare Diagnostics Inc.
Date Cleared
2014-01-24

(49 days)

Product Code
CGX,PAN
Regulation Number
862.1225
Type
Traditional
Panel
CH
Reference & Predicate Devices
K061238
AI/MLSaMDIVD (In Vitro Diagnostic)TherapeuticDiagnosticis PCCP AuthorizedThirdpartyExpeditedreview
Intended Use

The CRE2 method is an in vitro diagnostic test for the quantitative measurement of creatinine in human serum, plasma, and urine on the Dimension Vista® System. Creatinine measurements are used in the diagnosis and treatment of certain renal disease, in monitoring renal dialysis, and as a calculation basis for measuring other urine analytes.

Device Description

The Dimension Vista® Creatinine (CRE2) Flex® reagent cartridge uses a modified kinetic Jaffe technique. In the presence of a strong base such as sodium hydroxide, picrate reacts with creatinine to form a red chromophore. The rate of increasing absorbance at 510 nm due to the formation of this chromophore is directly proportional to the creatinine concentration in the sample and is measured using a bichromatic (510, 577 nm) rate technique.

AI/ML Overview

Here's a breakdown of the acceptance criteria and study information for the Dimension Vista® System Creatinine (CRE2) Flex® reagent cartridge, based on the provided text:

Acceptance Criteria and Device Performance

Acceptance Criteria CategorySpecific CriteriaReported Device Performance
Method Comparison (Serum)Slope close to 1, Intercept close to 0, Correlation Coefficient close to 1 when compared to predicate device (Vista CREA Assay)Vista CREA Assay (Serum): Range 0.38 – 18.93 mg/dL, Slope = 1.02, Intercept = -0.11 mg/dL, Correlation Coefficient = 1.000 (n=140)
Method Comparison (Urine)Slope close to 1, Intercept close to 0, Correlation Coefficient close to 1 when compared to predicate device (Vista CREA Assay)Vista CREA Assay (Urine): Range 8.39 – 299.60 mg/dL, Slope = 1.05, Intercept = -4.29 mg/dL, Correlation Coefficient = 0.999 (n=112)
Method Comparison (IDMS)Slope close to 1, Intercept close to 0, Correlation Coefficient close to 1 when compared to IDMS Reference MethodIDMS Reference Method (Serum): Range 0.18–6.32 mg/dL, Slope = 1.06, Intercept = -0.03 mg/dL, Correlation Coefficient = 0.997 (n=48)
Serum Plasma EquivalencySlope close to 1, Intercept close to 0, Correlation Coefficient close to 1 for matched serum vs. lithium heparin plasma samplesSlope = 1.03, Intercept = -0.002, Correlation Coefficient (r) = 0.998 (Range 0.418 - 17.9, n=56)
PrecisionWithin-Lab %CV values for various serum and urine pools and control materials. (Implicitly, these should be within acceptable limits for a diagnostic test, though specific numerical criteria for acceptance are not explicitly stated as a single value)Reported %CVs: Serum Pool 1 (2.7%), Serum Pool 2 (0.7%), BioRad Multiqual Level 1 (4.0%), Level 2 (2.5%), Level 3 (1.1%), Urine Pool 1 (3.6%), Urine Pool 2 (1.0%), BioRad Liquicheck Level 1 (2.2%), Level 2 (2.0%). (All values are low, indicating good precision.)
Limit of Blank (LoB)Consistent with claim of 0.05 mg/dL (serum), 1.0 mg/dL (urine)LoB (Serum): 0.030 mg/dL (consistent with 0.05 mg/dL claim)
LoB (Urine): 0.438 mg/dL (consistent with 1.0 mg/dL claim)
Limit of Detection (LoD)Consistent with claim of 0.1 mg/dL (serum), 2.0 mg/dL (urine)LoD (Serum): 0.056 mg/dL (consistent with 0.1 mg/dL claim)
LoD (Urine): 0.828 mg/dL (consistent with 2.0 mg/dL claim)
Limit of Quantitation (LoQ) Serum/PlasmaAllowable total error for creatinine published by CLIA, CAP, AAB, NYS, and WSLH is +/- 0.3 mg/dL or +/- 15%. (LoQ should meet these recommendations).LoQ (Serum/Plasma): 0.15 mg/dL based on allowable total error of 0.15 mg/dL. This meets the recommendations.
Limit of Quantitation (LoQ) UrineNo recognized allowable total error limit for urine creatinine measurements. (Implicit acceptance is that the determined LoQ is appropriate and justified).LoQ (Urine): 5.00 mg/dL based on allowable total error of 3.00 mg/dL.
Linearity (Serum)Slope close to 1, Intercept close to 0, Correlation Coefficient close to 1 across the assay rangeLinearity (Serum): Range 0.00 – 21.9 mg/dL, Slope = 0.996, Intercept = -0.05, Correlation Coefficient = 1.0 (N=12)
Linearity (Urine)Slope close to 1, Intercept close to 0, Correlation Coefficient close to 1 across the assay rangeLinearity (Urine): Range 1.71 – 338 mg/dL, Slope = 0.995, Intercept = 0.47, Correlation Coefficient = 1.0 (N=12)
Interfering SubstancesBias exceeding 10% is considered interference. Dilution studies determine the level at which spiked substances no longer display significant interference.Demonstrated concentrations where various substances cause 10% bias, the concentration at which interference is no longer significant through dilution. For example, Acetone interferes at 150 mg/dL (18% bias) but not at 75 mg/dL (9% bias). Bilirubin (unconjugated) at 40mg/dL shows -29.6% bias at low creatinine, but 20mg/dL shows -0.2% bias.
HIL InterferenceBias exceeding 10% is considered interference.Intralipid 20% interferes at 2000 mg/dL and 1500 mg/dL (at low creatinine concentration) (15.5% and 15.1% bias, respectively), but not at 1500 mg/dL (at high creatinine concentration) (9.5% bias) or 1000 mg/dL (7.0% bias). Hemoglobin shows no significant interference at 1000 mg/dL (~7% and 3% bias). Bilirubin (unconjugated) at 40 mg/dL shows -29.6% interference at low creatinine levels. Bilirubin (conjugated) shows 16.1% interference at 40 mg/dL at low creatinine levels.
Expected Values (Serum & Plasma)≤10% of samples within the study should fall outside of established range.Males: 5 (9.4%) out of 53 samples outside range (meets ≤10%).
Females: 2 (5.0%) out of 40 samples outside range (meets ≤10%). (Supports published ranges).
Expected Values (Urine)≤10% of samples within the study should fall outside of established range.Males: 2 (9.1%) out of 22 samples outside range (meets ≤10%).
Females: 1 (5.0%) out of 20 samples outside range (meets ≤10%). (Supports published ranges).
Reportable Range Dilution (Serum)Mean percent recovery of 90-110% for autodiluted samples.Mean of N=5 reps: 16.5 µL (normal volume) = 16.5; 6.6 µL (auto-dilute volume) = 16.0; % Bias = -3.0%. (Implicitly meets recovery criteria, as bias is very low). Supports extended range to 40.0 mg/dL.
Reportable Range Dilution (Urine)Mean percent recovery of 90-110% for autodiluted samples.Mean of N=5 reps: undiluted (normal) = 268; auto-dilute 3x with water = 245; % Bias = -8.6%. (Implicitly meets recovery criteria, as bias is very low). Supports extended range to 900 mg/dL.

Study Details

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

    • Method Comparison (Predicate Device):
      • Serum: 140 patient samples. Remnant de-identified serum samples. No patient history. Inclusion/exclusion criteria not applicable. Both native and spiked samples.
      • Urine: 112 patient samples. Remnant de-identified urine samples (implied). No patient history. Inclusion/exclusion criteria not applicable.
    • Method Comparison (IDMS Reference Method):
      • Serum: 48 patient samples. Remnant de-identified serum samples. No patient history. Inclusion/exclusion criteria not applicable. All native samples.
    • Serum Plasma Equivalency: 56 matched serum and lithium heparin plasma samples. All samples fresh and never frozen. 8 spiked sample sets.
    • Precision: Not a direct "test set" in the sense of patient samples, but testing involved: 2 serum pools, 3 levels of BioRad Multiqual material, 2 levels of BioRad Liquicheck material, and 2 urine pools. Each tested multiple times (over 20 days, 2 runs, 2 samples per run).
    • Limit of Blank/Detection (Serum): 4 samples with no analyte (N=5 reps for 3 days), 4 low patient serum samples (N=5 reps for 3 days).
    • Limit of Blank/Detection (Urine): 4 samples with no analyte (N=5 reps for 3 days), 4 low patient urine samples (N=5 reps for 3 days).
    • Linearity: 12 equally spaced samples for serum, 12 for urine.
    • Analytical Specificity (Interference): Low (~1.5 mg/dL) and High (~5.0 mg/dL) creatinine serum pools; Low (~40 mg/dL) and High (~175 mg/dL) creatinine urine pools. Number of individual samples/tests per interferent is not explicitly stated but implied to be sufficient for creating means and controls.
    • Expected Values:
      • Serum and Plasma: 53 male donors, 40 female donors (all normal healthy adults).
      • Urine: 22 male donors, 20 female donors (all normal healthy adults).
    • Reportable Range Dilution: N=5 replicates for each dilution study (serum and urine).

    Data Provenance: The studies were conducted internally by Siemens Healthcare Diagnostic Inc. R&D organization personnel. Remnant de-identified samples were used for method comparison studies, implying retrospective or archived samples. Patient history was not obtained. The origin of the patient samples (e.g., country) is not specified, but the internal R&D location would presumably be Newark, DE, USA based on the applicant's address.

  2. Number of experts used to establish the ground truth for the test set and the qualifications of those experts:
    The document does not mention the use of experts to establish ground truth for the test set in the context of clinical interpretation. For method comparison, the "ground truth" for the new device was established by comparison to a legally marketed predicate device (Dimension Vista® Creatinine (CREA) Flex® reagent cartridge) and the IDMS reference method. These are analytical "gold standards" rather than expert interpretation.

  3. Adjudication method (e.g., 2+1, 3+1, none) for the test set:
    Not applicable. The reported studies are analytical performance evaluations of a quantitative laboratory test, not diagnostic interpretations requiring 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 a quantitative in vitro diagnostic test for creatinine, not an imaging or interpretive device that would typically involve human readers or AI assistance in the way an MRMC study would assess.

  5. If a standalone (i.e., algorithm only without human-in-the-loop performance) was done:
    Yes, these are all standalone (algorithm only) performance studies. The device itself performs the quantitative measurement of creatinine from samples without human interpretation in the loop of the measurement.

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

    • For Method Comparison studies: The ground truth was established by two primary methods:
      • Comparison to a predicate device (Dimension Vista® Creatinine (CREA) Flex® reagent cartridge).
      • Comparison to the IDMS reference method (Isotope Dilution Mass Spectrometry), which is a highly accurate and precise analytical method often considered a "gold standard" for creatinine measurements.
    • For Expected Values studies: Published reference ranges in scientific literature (e.g., Tietz 1999, Clin Chem 54:3 (2008), Clin Chem Acta 344 (2004) 137-148) served as the basis for validating the established ranges for the new device.
    • For LoQ validation: Allowable total error limits published by regulatory and accreditation bodies (CLIA, CAP, AAB, NYS, WSLH) were used. The reference values are traceable to IDMS.

  7. The sample size for the training set:
    Not applicable in the context of typical AI/machine learning studies. This device is a reagent cartridge for an automated clinical chemistry system, not an AI or machine learning algorithm that undergoes 'training' in the conventional sense with labeled data. The development and internal refinement of the reagent formulation and measurement parameters would be analogous to "training" but is not quantified in terms of a discrete sample set size like this.

  8. How the ground truth for the training set was established:
    Not applicable for the same reasons as above. The "ground truth" for developing such a system would involve rigorous analytical chemistry principles, extensive experimentation, and calibration using traceable standards (e.g., NIST SRM 914a (IDMS assigned crystalline creatinine standard)).

§ 862.1225 Creatinine test system.

(a)
Identification. A creatinine test system is a device intended to measure creatinine levels in plasma and urine. 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.(b)
Classification. Class II.