Search Results

NG-Test® CTX-M MULTI is an in vitro rapid and visual immunochromatographic assay for the qualitative detection of CTX-M enzymes (groups 1, 2, 8, 9, and 25) from pure colonies of Enterobacterales suspected of ESBL production when grown on the following media:

• 5% sheep blood agar or MacConkey agar (16-24 hours)
• HardyCHROM™ ESBL agar (18-24 hours)

The NG-Test® CTX-M MULTI is intended as an aid for infection control in the detection of CTX-M enzymes-producing organisms (Enterobacterales) in healthcare settings. NG-Test® CTX-M MULTI is not intended to guide or monitor treatment. A positive or negative NG-Test® CTX-M MULTI test result does not rule out the presence of other mechanisms of antibiotic resistance. NG-Test® CTX-M MULTI should be used in conjunction with other laboratory tests including phenotypic antimicrobial susceptibility testing.

NG-Test® CTX-M MULTI is an in vitro rapid and visual immunochromatographic assay for the qualitative detection of CTX-M enzymes (groups 1, 2, 8, 9, and 25) from pure colonies of Enterobacterales suspected of ESBL production after culturing on agar and processed in an extraction buffer. The device consists of a sample port, sample and conjugate pad, and nitrocellulose test strip, which are contained within a plastic cassette, in addition to reagents for liquid extraction. The result can be read 15 minutes after adding the sample to the sample well. A positive result on the NG-Test® CTX-M MULTI occurs when two red lines appear, one on the control (C) region and one on the test (T) region. A negative result occurs when only the control line is observed and indicates the sample does not contain any target CTX-M enzymes, or the CTX-M enzymes are present at a non-detectable level. If the control line does not appear, the test result is invalid.

Monoclonal antibodies that recognize the five major CTX-M groups are immobilized on a nitrocellulose membrane. Free monoclonal antibodies are present in the conjugate pad and labeled with colloidal gold. Upon addition of the processed sample to the sample pad, the capillary action of the nitrocellulose draws the sample through the mobile antibodies in the conjugate pad and the immobile antibodies on the test strip. The immobilized control antibodies capture any mobile antibodies that do not bind to the test line.

This document describes the performance of the NG-Test® CTX-M MULTI device, an in vitro rapid immunochromatographic assay for detecting CTX-M enzymes.

Here is a breakdown of the requested information:

1. Acceptance Criteria and Reported Device Performance

The acceptance criteria for the device are implied by the reported sensitivity and specificity values. While explicit "acceptance criteria" are not listed as target percentages, the study results demonstrate high performance that led to clearance.

Table of Performance Data (Clinical Study - Blood and MacConkey Agar):

Table of Performance Data (Seeded Study - HC ESBL Agar):

Note: The lower bound of the specificity for HC ESBL agar (87.9%) is acknowledged as being below 90% due to a lower quantity of CTX-M negative isolates in this specific part of the study, and additional CTX-M negative isolates were evaluated in the cross-reactivity study to compensate.

2. Sample Size and Data Provenance

• Test Set Sample Size:
  • Clinical Study (Blood and MacConkey Agar): 309 Enterobacterales isolates.
  • Seeded Study (HardyCHROM™ ESBL Agar): 193 clinical isolates.
• Data Provenance:
  • Country of Origin: Not explicitly stated, but the study was conducted at "three geographically diverse hospitals," implying the data is from a clinical setting.
  • Retrospective or Prospective: A mix was used:
    • "Prospectively-collected" bacterial isolates were used for the performance evaluation.
    • "Stock bacterial isolates" were also used. This suggests a combination of prospective collection and retrospective use of archived isolates.

3. Number of Experts and Qualifications for Ground Truth

• The document does not specify the number of experts used to establish ground truth or their qualifications. The ground truth relies on laboratory methods (PCR and AST).

4. Adjudication Method for the Test Set

• The document does not describe an adjudication method for the test set results. The comparison is between the device's reading and the reference method (PCR and AST).

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

• A MRMC comparative effectiveness study involving human readers with vs. without AI assistance was not conducted. This device is an immunochromatographic assay, not an AI-based diagnostic with human-in-the-loop performance.
• However, a reproducibility study involved multiple readers: "The testing was done with one operator and two readers, blinded to each other's results, per site." This was to demonstrate reproducibility of the device's results, not to compare human performance with and without the device.

6. Standalone (Algorithm Only) Performance

• This device is a standalone diagnostic test (an immunochromatographic assay). Its performance is evaluated directly without a human-in-the-loop component for result interpretation beyond visual reading. The reported sensitivity and specificity refer to the device's performance as a standalone test.

7. Type of Ground Truth Used

• Reference standard:
  • PCR (Polymerase Chain Reaction): Used to detect blaCTX-M genes, considered the definitive molecular method. Isolates with a positive PCR result were also sequenced to determine the CTX-M variant.
  • Antimicrobial Susceptibility Testing (AST): Performed according to CLSI M100, 34th edition breakpoints. A positive comparator method result was defined as any Enterobacterales that produced a positive PCR result and was not susceptible to at least one of the antimicrobial agents.
  • Identification (ID): Performed using FDA-cleared ID systems.

8. Sample Size for the Training Set

• The document does not specify a separate "training set" size. As this is an immunochromatographic assay, it doesn't involve machine learning model training in the typical sense. The analytical reactivity study and analytical specificity study essentially serve as extensive performance characterization using known strains, which could be conceptually similar to a "development" or "characterization" set for more traditional devices.
  • Analytical Reactivity: 57 strains of Enterobacterales characterized to harbor CTX-M enzymes.
  • Analytical Specificity: 55 organisms with other resistance mechanisms.

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

• For the analytical reactivity and analytical specificity studies (which characterize the device's performance with known strains, akin to a training/development set in ML contexts):
  • Strains were "characterized to harbor CTX-M enzymes" or "exhibit antimicrobial resistance mechanisms other than CTX-M." This characterization would typically involve established molecular methods (e.g., PCR, sequencing) and phenotypic susceptibility testing performed by a reference lab or research institution. The document states "Each organism was incubated..." and tested, implying these were well-characterized isolates.

Ask a specific question about this device

NG-Test CARBA 5 is an in vitro rapid and visual multiplex immunochromatographic assay for the qualitative detection and differentiation of five common carbapenemases (KPC, OXA-48-like, VIM, IMP and NDM) from carbapenem non-susceptible pure bacterial colonies when grown on the following media:

• 5% sheep blood agar or MacConkey agar (16-24 hours) for testing Enterobacteriaceae and Pseudomonas aeruginosa
• HardyCHROM™ CRE agar (18-24 hours) for testing E. coli and KES (Klebsiella aerogenes, . Klebsiella oxytoca, Klebsiella pneumoniae, Enterobacter cloacae complex, and Serratia marcescens).
  The NG-Test CARBA 5 is intended as an aid for infection control in the detection of carbapenemase-producing Enterobacteriaceae and Pseudomonas aeruginosa in healthcare settings. NG-Test CARBA 5 is not intended to guide or monitor treatment for carbapenem non-susceptible bacterial infections. A positive or negative NG-Test CARBA 5 test result does not rule out the presence of other mechanisms of antibiotic resistance. NG-Test CARBA 5 should be used in conjunction with other laboratory tests including phenotypic antimicrobial susceptibility testing.

NG-Test CARBA 5 is an in vitro rapid and visual multiplex immunochromatographic assay that detects one or more of the five common types of carbapenemase enzymes (KPC (K), OXA-48-like (O), IMP (I), VIM (V), NDM (N)) in bacterial colonies. Liquid extraction buffer is used as a cell lysing solution when mixed with colonies. Monoclonal antibodies that individually recognize each of the five carbapenemases are immobilized on a nitrocellulose membrane. Free monoclonal antibodies are present in the sample pad and labelled with colloidal gold. Upon addition of colonies mixed with extraction buffer to the sample pad, the capillary action of the nitrocellulose draws the sample the mobile antibodies and immobile antibodies on the test strip. The immobilized control antibodies capture any mobile antibodies that run through the sample pad and nitrocellulose without binding to other test lines. A positive result occurs when a red line appears on the control region and one or more lines appear in the test regions (K. O. V. I. or N) and indicates that the sample contains one or more carbapenemases. A negative result occurs when only the control line is observed and indicates that the sample does not contain any of the five carbapenemases. If the control line does not appear, the test result is invalid. The sample pad, and nitrocellulose strip are contained within a plastic cassette. After addition of colonies in extraction buffer to the sample port, a result can be read after 15 minutes.

The NG-Test CARBA 5 is an in vitro rapid and visual multiplex immunochromatographic assay designed to detect five common carbapenemase enzymes (KPC, OXA-48-like, IMP, VIM, NDM) in bacterial colonies. The device's performance was evaluated through a multi-center study to establish its acceptance criteria.

1. Table of Acceptance Criteria and Reported Device Performance

The acceptance criteria for the NG-Test CARBA 5 are based on Positive Percent Agreement (PPA) and Negative Percent Agreement (NPA) compared to a composite reference method. The reported device performance met or exceeded these criteria for both Enterobacteriaceae and P. aeruginosa across different culture media.

Table 1: Acceptance Criteria and Reported Device Performance

(Note: "Implicit" acceptance criteria are derived from the fact that the device was found substantially equivalent, meaning it met FDA's expectations for performance for this class of device. Exact pre-defined numerical thresholds for acceptance were not explicitly stated in the provided text, but the reported performance generally reflects high agreement to the reference methods.)

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

• Agar: 5% Sheep Blood or MacConkey agar:
  • Total organisms initially tested: 310
  • Organisms included in analysis: 309 (1 Pseudomonas species other than P. aeruginosa excluded)
  • Enterobacteriaceae: 240 isolates (yielding 244 results due to co-production of two carbapenemases in four isolates).
  • P. aeruginosa: 69 isolates.
  • Provenance: Prospectively-collected and stock bacterial isolates from three geographically diverse hospitals.
• Agar: HardyCHROM™ CRE agar (from stool specimens):
  • Total organisms initially enrolled: 186
  • Organisms included in analysis: 185 (1 organism unavailable)
  • Raw Stool: 180 organisms (184 target results) for E. coli and KES (Klebsiella aerogenes, K. oxytoca, K. pneumoniae, Enterobacter cloacae complex, and Serratia marcescens).
  • C&S Cary Blair Stool: 178 organisms (182 target results) for E. coli and KES.
  • Provenance: This data was collected internally using the same bacterial isolates (from blood and MacConkey agar study) cultured on HardyCHROM™ CRE after being seeded in raw and C&S Cary Blair stool.

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

The document does not explicitly state the number or qualifications of experts (e.g., radiologists, microbiologists) who established the ground truth. However, the ground truth was established through a "composite reference method" combining:

• An FDA-cleared device: Xpert Carba-R by Cepheid (PCR for KPC, OXA-48 or 181, IMP, VIM, NDM). This is a molecular method, so human expertise here would primarily be in running the test and interpreting its output according to established protocols.
• Modified carbapenem inactivation method (mCIM) and EDTA carbapenemase inactivation method (eCIM) as described by CLSI M100, S29. These are phenotypic tests requiring skilled microbiologists for execution and interpretation.
• Antibiotic susceptibility testing results to ertapenem, imipenem, and meropenem. This also requires expertise in microbiology and adherence to CLSI guidelines.
• Identity and susceptibility of organisms were confirmed using FDA-cleared ID and AST systems. This implies expert oversight or use of automated systems with established performance.

For discrepant analysis, alternative PCR assays and bidirectional sequencing were used, which would involve highly specialized molecular microbiologists.

4. Adjudication Method for the Test Set

The study employed discrepant analysis to resolve inconsistencies between the NG-Test CARBA 5 results and the initial composite reference method results. This means that when the device result did not agree with the initial ground truth, further, more definitive tests (like alternative PCR assays and bidirectional sequencing) were performed. The results from these additional, often more gold-standard methods, were then used to adjudicate the true status of the isolate, effectively updating the ground truth for those specific cases. This "2+1" or "3+1" type of adjudication (NG-Test vs. Composite Reference, then a tie-breaker/definitive test) was used to ensure the most accurate ground truth for performance calculation.

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

The provided text does not describe a multi-reader multi-case (MRMC) comparative effectiveness study involving human readers with and without AI assistance (as the NG-Test CARBA 5 is an immunochromatographic assay, not an AI-driven image analysis tool requiring human interpretation assistance). The device provides a visual, rapid result, and its performance is evaluated against laboratory gold standards, not through human reader performance improvement.

6. Standalone Performance Study

Yes, the performance study directly evaluates the standalone performance of the NG-Test CARBA 5 device. The provided tables and metrics (PPA, NPA) reflect the algorithm's performance (in this case, the immunochromatographic assay's ability to detect carbapenemases) without human intervention in the result determination process beyond the initial preparation and visual reading of the test strip. The operators were blinded to the expected result during setup and interpretation, indicating an objective assessment of the device's inherent performance.

7. Type of Ground Truth Used

The ground truth used was a composite reference method consisting of:

• Molecular (PCR): Xpert Carba-R (FDA-cleared device).
• Phenotypic (Enzymatic Activity): Modified Carbapenem Inactivation Method (mCIM) and EDTA Carbapenemase Inactivation Method (eCIM) as per CLSI guidelines.
• Phenotypic (Susceptibility Testing): Antibiotic susceptibility testing results (ertapenem, imipenem, meropenem).
• Discrepant Analysis: For conflicting results, more definitive molecular techniques (alternative PCR and bidirectional sequencing) were employed to establish the definitive truth.

This multifaceted approach to ground truth ensures robust and accurate classification of carbapenemase production.

8. Sample Size for the Training Set

The provided document does not mention a training set or its sample size. This is typical for in vitro diagnostic (IVD) devices like the NG-Test CARBA 5 which are based on biochemical or immunological principles, rather than machine learning algorithms that require explicit training data. The "analytical reactivity" section describes testing against a panel of 92 characterized strains, which could be considered an internal validation or analytical performance assessment, but not a "training set" in the machine learning sense.

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

As no training set (in the context of machine learning) is described, there is no information on how its ground truth was established. The "Analytical Reactivity" study used 92 strains "characterized to have a target carbapenemase," implying that their carbapenemase status was already established by other reference methods prior to this specific evaluation. This characterization would have been done using methods similar to the composite reference described for the main performance study (e.g., PCR, sequencing, phenotypic tests).

Ask a specific question about this device

RAPIDEC® CARBA NP is a phenotypic (colorimetric) in vitro diagnostic test for the qualitative detection of carbapenemase enzymes in Enterobacteriaceae and Pseudomonas aeruginosa colonies that have elevated MIC values to any carbapenem. RAPIDEC® CARBA NP is performed on pure colonies grown on non-selective sheep blood agar culture media.

RAPIDEC® CARBA NP is intended as an aid in the prevention and control of infection caused by carbapenemase-producing Enterobacteriaceae and Pseudomonas aeruginosa.

RAPIDEC® CARBA NP is not intended to guide or monitor the treatment for these bacterial infections. A negative result does not preclude the presence of carbapenemase enzymes. The ability of RAPIDEC® CARBA NP to detect carbapenemase enzymes encoded by genetic markers other than KPC, NDM, OXA-48, VIM, and IMP has not been established. RAPIDEC® CARBA NP testing should be used in conjunction with other laboratory tests including antimicrobial susceptibility testing.

The RAPIDEC® CARBA NP strip is composed of 5 wells prepared with premeasured portions of the necessary substrates for the reactions. In addition, the kit contains the necessary accessories for performing the test.

In order to rehydrate the dry reagents and initiate the reactions, wells a, b and c are filled with 100 uL of API Suspension Medium (purified water). The strip is left at room temperature for 4-10 minutes to allow the dry reagents to reconstitute in the wells. The bacterial inoculum suspension is prepared in well c until the turbidity equals well b. Well c contains the lysis buffer. In order to achieve lysis of the inoculum suspension, which enables the extraction of the enzyme, the strip is left at room temperature for additionally 30 minutes.

As the next step, 25 uL of the lysed inoculum suspension is transferred to wells d and e and 25 µL from well a (phenol red solution) is also transferred to wells d and e. The strip is incubated for 30 minutes at 33-38°C to allow for the hydrolysis to occur and change in color of the phenol red solution in case of presence of a carbapenemase enzyme. The initial reading is performed after 30 minutes of incubation. In case of a negative or doubtful reaction, the strip is re-incubated for an additional 1 hour and 30 minutes before performing the final reading.

The hydrolysis acidifies the medium which results in the change in color of the pH indicator. Reading is performed by comparing the colors in wells d and e. The test is positive when a significant variation in color is observed between the two wells. For example, the control well is red and the test well has changed to yellow.

The document provided describes the RAPIDEC® CARBA NP device, an in vitro diagnostic test for the qualitative detection of carbapenemase enzymes. This is a 510(k) submission, meaning the manufacturer is demonstrating that their new device is substantially equivalent to a legally marketed predicate device, rather than proving its safety and effectiveness de novo. Therefore, the study focuses on establishing performance characteristics compared to a reference method and demonstrating substantial equivalence.

Here's an analysis of the acceptance criteria and the study that proves the device meets them, based on the provided text:

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

The document implicitly defines acceptance criteria through the reported analytical and clinical performance. While explicit numerical acceptance criteria (e.g., "sensitivity must be >X%") are not laid out in a dedicated table, the "Results of QC testing...met the pre-defined acceptance criteria (≥95% agreement)" suggests this was a criterion for analytical performance. For clinical performance, the reported agreement percentages themselves serve as the outcome against which an implicit set of criteria would have been measured.

Here’s a summary of the performance characteristics:

Acceptance Criteria (Implied) and Reported Device Performance for RAPIDEC® CARBA NP

• 97.9% (91/93) at 30 minutes (2 false positives).
• 93.6% (87/93) at 2 hours (6 false positives).
  For Intended Organisms (Enterobacteriaceae & P. aeruginosa):
• 100% (67/67) at 30 minutes.
• 98.5% (66/67) at 2 hours (1 false positive: AmpC producing Morganella morganii). |
  | Agar Compatibility | No significant differences in performance when using different culture media (Trypticase Soy agar + 5% sheep blood (TSS) vs. Columbia agar + 5% sheep blood (COS)). | Positive rate for carbapenemase producers: 94.6% (87/92) for TSS and 92.4% (85/92) for COS at 30 min. Positive rate increased to 98.9% (91/92) for both media at 2 hours. Negative rate for non-carbapenemase producers: 100% for both media at both time points. "The results indicates no significant differences between TSS and COS." |
  | Clinical Performance | | |
  | Overall Agreement (Clinical) | High agreement with composite reference method. No explicit numerical criterion. | Overall (All samples):
• Routine Subculture: 98.7% (451/457)
• Short Subculture: 98.0% (440/449)
  Enterobacteriaceae:
• Routine Subculture: 98.5% (388/394)
• Short Subculture: 97.7% (383/392)
  Pseudomonas aeruginosa:
• Routine Subculture: 100% (63/63)
• Short Subculture: 100% (57/57) |
  | Sensitivity (Clinical) | High sensitivity (agreement for positive samples). No explicit numerical criterion. | Overall (Positive samples):
• Routine Subculture: 99.6% (264/265) (1 false negative, KPC-producing Enterobacteriaceae)
• Short Subculture: 98.5% (257/261) (4 false negatives: 1 KPC, 1 NDM, 1 VIM, 1 OXA-48 from Enterobacteriaceae) |
  | Specificity (Clinical) | High specificity (agreement for negative samples). No explicit numerical criterion. | Overall (Negative samples):
• Routine Subculture: 97.4% (187/192) (5 false positives, all Enterobacteriaceae)
• Short Subculture: 97.3% (183/188) (5 false positives, all Enterobacteriaceae). |
  | Reproducibility | High reproducibility across sites and operators. No explicit numerical criterion. | Routine Subculture: 98.2% (884/900)
  Short Subculture: 99.1% (892/900) |

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

• Test Set Sample Size:

  • Analytical Reactivity (Challenge testing): 43 carbapenemase producers.
  • Cross-reactivity: 93 non-carbapenemase producing strains (67 Enterobacteriaceae/P. aeruginosa, 26 other organisms).
  • Agar Culture Media Compatibility Studies: 106 strains (92 carbapenemase producers, 14 non-carbapenemase producers).
  • Clinical Studies: 457 strains (394 Enterobacteriaceae, 63 P. aeruginosa).
  • Reproducibility: A panel of 10 organisms (9 Enterobacteriaceae + 1 P. aeruginosa), tested in triplicates for 5 days at 3 sites (total 900 measurements per subculture procedure).

• Data Provenance: The document does not explicitly state the country of origin for the clinical or analytical data. It mentions "5 sites" for the clinical study and "each study site" for QC, implying a multi-site study, but locations are not specified. The study appears to be prospective as it involves the evaluation of the device on collected strains and controlled experiments (e.g., QC, challenge testing, clinical evaluation).

3. 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 number or qualifications of experts used to establish the ground truth. The ground truth for the clinical study was established by a composite reference method consisting of three laboratory tests: carbapenem MIC, CLSI® Carba NP, and PCR for carbapenemase genetic markers. The "final composite reference result based on agreement of at least two of the three tests" does not explicitly involve human expert review for establishing "ground truth," but rather relies on the combined results of established laboratory methods.

For the reproducibility study, results were interpreted by "2 operators" but their qualifications are not specified beyond being "operators."

4. Adjudication method for the test set

For the clinical study, the adjudication method for the composite reference ground truth was: agreement of at least two of the three tests (carbapenem MIC, CLSI® Carba NP, and PCR).

For the reproducibility study, the results were interpreted by "2 operators" and their interpretations were then compared ("blinded to each other"), suggesting a form of comparative adjudication for reproducibility but not for establishing primary ground truth.

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, a multi-reader multi-case (MRMC) comparative effectiveness study was not done for this device. This is an in vitro diagnostic device (a laboratory test kit), not an AI-assisted imaging device or a device that directly supports human reader interpretation in a complex diagnostic task like radiology. Therefore, questions regarding "human readers improve with AI vs without AI assistance" are not applicable to this device type. The device provides a direct result (positive/negative for carbapenemase) which then informs downstream clinical decisions.

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

This refers to an in vitro diagnostic device, not typically an "algorithm only" device in the context of imaging or clinical decision support AI. The device itself is a test kit where reactions occur and a visual color change is observed. The "performance" described is of the device as a whole. While human operators perform the test and interpret the "visual based on color change," the core "algorithm" is the biochemical reaction itself. The analytical and clinical studies assess the performance of this system (device + human interpretation). No separate "algorithm only" study is described because the "algorithm" here is the chemical reaction and visual readout, which inherently involves human observation.

7. The type of ground truth used

The ground truth used was a composite reference method comprised of:

• Carbapenem MIC (Imipenem, Meropenem, Ertapenem and/or Doripenem).
• CLSI® Carba NP (a standardized carbapenemase test).
• Carbapenemase genetic markers by Polymerase Chain Reaction (PCR).

This is a form of expert consensus based on the agreement of established laboratory tests, rather than, for example, pathology or outcomes data.

8. The sample size for the training set

The document does not explicitly describe a separate "training set". In the context of a 510(k) submission for an in vitro diagnostic device, often the manufacturer develops and refines the device, and then tests its performance on a "test set" (clinical and analytical studies) to demonstrate substantial equivalence. If any internal development/optimization data was used, it's not detailed or referred to as a "training set" in this regulatory submission. The reported studies are validation studies on a specific set of samples.

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

Since a distinct "training set" is not described, the method for establishing its ground truth is also not detailed. Any internal development ground truth would likely have been established using similar reference methods to the ones described for the validation set.

Ask a specific question about this device