Search Results

Light Diagnostics™ Human Metapneumovirus DFA Kit is intended for the detection and identification of human metapneumovirus (hMPV) in direct respiratory specimen cell preparations from nasopharyngeal swabs from patients with febrile respiratory illness. This assay detects but is not intended to differentiate the four recognized genetic sub-lineages of hMPV.

Negative results do not preclude hMPV infection and should not be used as the sole basis for diagnosis, treatment or other management decisions. It is recommended that specimens found to be negative after examination of the direct specimen results be confirmed by FDA cleared hMPV molecular assay.

For In Vitro Diagnostic Use.

Light Diagnostics Human Metapneumovirus DFA Kit utilizes a single reagent for the detection and identification of human metapneumovirus. The fluorescein labeled monoclonal antibodies, specific for human metapneumovirus will bind to viral antigen in human metapneumovirus infected cells. Unbound reagent is removed by rinsing with phosphate-buffered saline (PBS/Tween 20). Illumination allows visualization of the antigenantibody complex by fluorescence microscopy. When a FITC filter set is used, the human metapneumovirus antibody complex will exhibit an apple-green fluorescence. Uninfected cells stain a dull red due to the presence of Evans blue in the reagent.

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

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

• Sample Size for Test Set: 411 nasopharyngeal swab specimens.
• Data Provenance: Retrospective clinical specimens were collected from two sites:
  • Site One: A regional medical center in southeastern Canada (208 specimens).
  • Site Two: A hospital laboratory in the northeastern United States (199 specimens).
  • Site Three: Additional 200 specimens were submitted for testing, of which four were nasopharyngeal swabs.

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

4. Adjudication method for the test set

The ground truth for the test set was established using a composite algorithm based on cell culture results and a validated RT-PCR method followed by bidirectional sequencing for confirmation and identification of human metapneumovirus. This implies a hierarchical or multi-modal adjudication process rather than an expert-based one.

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

A multi-reader multi-case (MRMC) comparative effectiveness study was not done. This study focuses on the performance of a direct immunofluorescence assay (DFA) device against a composite reference method, not human reader performance with or without AI assistance.

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

This study evaluates a diagnostic kit which is inherently intended for use by a human operator (reading fluorescence microscopy slides). Therefore, a purely standalone "algorithm only" performance (without human-in-the-loop) was not conducted in the context of an automated AI system. The performance presented for the Light Diagnostics™ hMPV DFA Kit represents its performance when used as intended by an operator.

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

The type of ground truth used was a composite algorithm based on cell culture results and a validated RT-PCR method followed by bidirectional sequencing for confirmation and identification of human metapneumovirus.

8. The sample size for the training set

The document does not explicitly mention a separate "training set" sample size for an AI algorithm. This submission describes the validation of a diagnostic kit (DFA) which relies on antibody-antigen binding observed via microscopy by a human, not a machine-learning based algorithm that would typically require a training set. If the "training set" refers to the data used to develop the assay's reagents, it is not specified.

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

As there is no explicit mention of a training set for an AI algorithm, the method for establishing its ground truth is not applicable/not provided in this document. The assay's development (e.g., selection of monoclonal antibodies) would have been guided by traditional laboratory methods for identifying hMPV, but this is not akin to a machine learning training set ground truth establishment.

Ask a specific question about this device

The Diagnostic Hybrids, Inc. device, D3 FastPoint L-DFA RSV/MPV Identification Kit is intended for the qualitative identification of respiratory syncytial virus and human metapneumovirus in nasal and nasopharyngeal swabs and aspirates/washes specimens from patients with signs and symptoms of respiratory infection by direct detection of immunofluorescence using monoclonal antibodies (MAbs).

It is recommended that specimens found to be negative for respiratory syncytial virus after examination of the direct specimen result be confirmed by cell culture. Specimens found to be negative for human metapneumovirus after examination of the direct specimen results should be confirmed by an FDA-cleared human metapneumovirus molecular assay. Negative results do not preclude respiratory syncytial virus and human metapneumovirus infection and should not be used as the sole basis for diagnosis, treatment or other management decisions.

The D3 FastPoint L-DFA RSV/MPV Identification Kit uses a blend (called a "L-DFA Reagent'") of viral antigen-specific murine monoclonal antibodies that are directly labeled with either R-phycoerythin (PE) (respiratory syncytial virus) or fluorescein isothiocyanate (FITC) (human metapneumovirus) for the rapid identification of respiratory syncytial virus and human metapneumovirus in nasal and nasopharyngeal swabs and aspirates from patients with signs and symptoms of respiratory infection.

The cells to be tested are derived from respiratory specimens from patients with signs and symptoms of respiratory infection. The cells are permeabilized and stained concurrently in a liquid suspension format with the L-DFA Reagent. After incubating at 35℃ to 37℃ for 5-minutes, the stained cell suspensions are rinsed with 1X PBS. The rinsed cells are pelleted by centrifugation and then re-suspended with the Resuspension Buffer and loaded onto a specimen slide well. The cells are examined using a fluorescence microscope. Cells infected with RSV will exhibit golden-yellow fluorescence due to the PE. Cells infected with hMPV will exhibit apple-green fluorescence due to the FITC. Non-infected cells will exhibit red fluorescence due to the Evans Blue counter-stain. Nuclei of intact cells will exhibit orange-red fluorescence due to the propidium iodide.

Acceptance Criteria and Study for D3 FastPoint L-DFA RSV/MPV Identification Kit

1. Table of Acceptance Criteria and Reported Device Performance

The provided document doesn't explicitly state numerical acceptance criteria for all performance metrics. However, for a 510(k) submission, implied acceptance is often "substantially equivalent" to predicate devices, and for clinical performance, high sensitivity and specificity are expected. The reproducibility study explicitly aims for 100% agreement.

Note: The "acceptance criteria" are inferred based on standard expectations for diagnostic device performance and the detailed reporting of study outcomes, particularly the 100% agreement for reproducibility and the very high sensitivity/specificity for RSV. The lower sensitivity for hMPV in NP swab samples might be within acceptable limits given the context of medical device approval for challenging targets.

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

• Reproducibility Test Set:
  • Sample Size: A reproducibility panel consisting of 5 members (low RSV, low hMPV, mixed RSV/hMPV, mixed hMPV/RSV, negative). Each panel member was tested daily in two separate runs for 5 days by 4 different laboratories, resulting in 40 total runs. This yielded 280 total tests (across all panel members and runs) with individual results reported for expected positive and negative wells.
  • Data Provenance: The study was conducted at four different laboratories. The document does not specify the country of origin but implies U.S. clinical laboratories (referencing "U.S. clinical laboratories" for clinical performance). It's a prospective study in the sense that the testing itself was performed to assess reproducibility.
• Limit of Detection (LoD) Test Set:
  • Sample Size: Dilution series of infected model cells were used. For each virus (RSV and hMPV A1), 10 replicate microscope slides were prepared for each dilution level. The specific number of dilutions isn't explicitly stated as a single number but spanned from 1000 infected cells/mL down to 0.8 or 1.5 infected cells/mL, with 10 replicates for each dilution.
  • Data Provenance: Laboratory study, likely internal to the manufacturer or a contracted lab.
• Analytical Reactivity (Inclusivity) Test Set:
  • Sample Size: 3 RSV virus strains and 4 hMPV virus strains were evaluated. For each strain, "low concentration infected cell suspensions (approximately 4% cells infected, 25 to 50 infected cells)" were prepared.
  • Data Provenance: Laboratory study.
• Clinical Performance Test Set:
  • Sample Size: 1519 total respiratory specimens (nasal and nasopharyngeal swabs and aspirates/washes).
  • Data Provenance: Prospective studies at 4 geographically diverse U.S. clinical laboratories during the 2009 respiratory virus seasons (January 2009 - March 2009). The specimens were "excess, remnants of respiratory specimens that were prospectively collected from symptomatic individuals suspected of respiratory infection, and were submitted for routine care or analysis by each site, and that otherwise would have been discarded." Individual specimens were delinked from patient identifiers.

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 for establishing ground truth as a separate role. Instead, the ground truth for clinical performance was established using a composite comparator method:

• RSV: Direct Specimen Fluorescent Antibody (DSFA) test with an FDA-cleared predicate device, followed by viral culture confirmation of all negatives from the comparator DSFA test.
• hMPV: DSFA with an FDA-cleared predicate device, followed by confirmation of all negative specimens from the comparator DSFA test using a validated hMPV real-time RT-PCR, which was then followed by bi-directional sequencing analysis.

This implies that the "ground truth" was determined by the results of these established and confirmed laboratory methods, rather than by human expert consensus or adjudication of raw images/output from the test device solely.

4. Adjudication Method for the Test Set

For the clinical studies, an explicit "adjudication method" involving human experts reviewing conflicting results is not detailed. Instead, a composite comparator algorithm was used to define "true positive" and "true negative":

• "True positive" for RSV was defined as any sample that tested positive by the comparator DSFA test or viral culture.
• "True positive" for hMPV was defined as any sample that tested positive by the comparator DSFA test OR had bi-directional sequencing data meeting pre-defined quality acceptance criteria that matched hMPV sequences in GenBank.
• "True negative" was defined as any sample that tested negative by both the comparator DSFA test and either viral culture (for RSV) or the hMPV real-time RT-PCR comparator assay (for hMPV).

This approach essentially pre-defines how discordant results between screening and confirmatory tests contribute to the final ground truth, replacing a separate human adjudication step.

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

No, a Multi-Reader Multi-Case (MRMC) comparative effectiveness study was not conducted. This study is an evaluation of an in-vitro diagnostic device (IVD), specifically a direct fluorescent antibody (DFA) test, which is read by trained laboratory personnel, but the study focuses on the device's performance against comparator methods, not on comparing reader performance with and without AI assistance (as it is not an AI device).

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

Yes, the studies presented are essentially standalone performance evaluations of the device. The D3 FastPoint L-DFA RSV/MPV Identification Kit is an immunofluorescent assay where a human technician observes fluorescent staining patterns under a microscope. However, the performance metrics (sensitivity, specificity) are for the device's ability to detect the viral antigens in specimens, without involving a study design where human readers using the device are compared to human readers using another method, or AI assistance. The results in the tables reflect the diagnostic performance of the kit itself when used according to its instructions.

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

The ground truth used for the clinical performance studies was based on a composite comparator method combining:

• FDA-cleared predicate DFA devices
• Viral culture (for RSV)
• Validated real-time RT-PCR with bi-directional sequencing analysis (for hMPV)

This is a form of reference standard composite, aiming for a highly accurate and confirmed diagnosis of the presence or absence of the target viruses.

8. The Sample Size for the Training Set

The document does not explicitly describe a separate "training set" or "validation set" in the context of machine learning. This is a traditional in vitro diagnostic device, not an AI/ML-based device.

All the described studies (reproducibility, LoD, analytical reactivity, clinical performance) contribute to the overall evidence for the device. The 1519 clinical specimens (fresh nasal/nasopharyngeal wash/aspirate and swab specimens) can be considered the test set for evaluating clinical performance.

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

As noted above, there is no explicit "training set" mentioned in the context of machine learning. The studies assess the performance of the pre-developed D3 FastPoint L-DFA RSV/MPV Identification Kit using the specified ground truth methods.

Ask a specific question about this device

The Diagnostic Hybrids, Inc. device, D3 FastPoint L-DFA Respiratory Virus Identification Kit is intended for the qualitative identification of influenza A virus, influenza B virus, respiratory syncytial virus, human metapneumovirus, adenovirus and to screen for the presence of parainfluenza virus types 1, 2, and 3 in nasal and nasopharyngeal swabs and aspirates/washes specimens from patients with signs and symptoms of respiratory infection by direct detection of immunofluorescence using monoclonal antibodies (MAbs).

It is recommended that specimens found to be negative for influenza A virus, influenza B virus, respiratory syncytial virus, adenovirus or parainfluenza viruses after examination of the direct specimen result be confirmed by cell culture. Specimens found to be negative for human metapneumovirus after examination of the direct specimen results should be confirmed by an FDA cleared human metapneumovirus molecular assay. Negative results do not preclude respiratory virus infection and should not be used as the sole basis for diagnosis, treatment or other management decisions.

Performance characteristics for influenza A virus detection and identification were established when influenza A (H3N2) and influenza A (H1N1) were the predominant influenza A strains circulating in the United States. Since influenza strains display antigenic drift and shift from year to year, performance characteristics may vary. If infection with a novel influenza A virus is suspected, based on clinical and epidemiological screening criteria communicated by public health authorities, collect specimens following appropriate infection control precautions and submit to state or local health departments, for testing. Viral culture should not be attempted in these cases unless a BSL 3+ facility is available to receive and culture specimens.

The D3 FastPoint L-DFA Respiratory Virus Identification Kit uses three blends (each called a "L-DFA Reagent") of viral antigen-specific murine monoclonal antibodies that are directly labeled with either R-PE (influenza A virus, respiratory syncytial virus, and parainfluenza virus) or fluorescein (influenza B virus, metapneumovirus, and adenovirus) for the rapid identification of respiratory viruses in nasal and nasopharyngeal swabs and aspirates from patients with signs and symptoms of respiratory infection.

Kit Components:

1. D3 FastPoint L-DFA Influenza A/Influenza B Reagent, 4.0-mL. One dropper bottle containing a mixture of PE-labeled murine monoclonal antibodies directed against influenza A virus antigens and FITC-labeled murine monoclonal antibodies directed against influenza B virus antigens. The buffered, stabilized, aqueous solution contains Evans Blue and propidium iodide as counter-stains and 0.1% sodium azide as preservative.
2. D3 FastPoint L-DFA RSV/MPV Reagent, 4.0-mL. One dropper bottle containing a mixture of PE-labeled murine monoclonal antibodies directed against respiratory syncytial virus antigens and FITC-labeled murine monoclonal antibodies directed against metapneumovirus antigens. The buffered, stabilized, aqueous solution contains Evans Blue and propidium iodide as counter-stains and 0.1% sodium azide as preservative.
3. D3 FastPoint L-DFA PIV/Adenovirus Reagent, 4.0-mL. One dropper bottle containing a mixture of PE-labeled murine monoclonal antibodies directed against parainfluenza virus types 1, 2, or 3 antigens and FITClabeled murine monoclonal antibodies directed against adenovirus antigens. The buffered, stabilized, aqueous solution contains Evans Blue and propidium iodide as counter-stains and 0.1% sodium azide as preservative.
4. 40X PBS Concentrate, 25-mL. One bottle of 40X PBS concentrate containing 4% sodium azide (0.1% sodium azide after dilution to 1X using de-mineralized water).
5. Re-suspension Buffer, 6.0-mL. One bottle of a buffered glycerol solution and 0.1% sodium azide.
6. D3 FastPoint L-DFA Respiratory Virus Antigen Control Slides, 5-slides. Five individually packaged control slides containing 6 wells with cell culture-derived positive and negative control cells. Each positive well is identified as to the virus infected cells present, i.e., influenza A virus, influenza B virus, respiratory syncytial virus, metapneumovirus, parainfluenza virus, and adenovirus. The negative wells contain noninfected cells. Each slide is intended to be stained only one time.

The cells to be tested are derived from respiratory specimens from patients with signs and symptoms of respiratory infection. The cells are permeabilized and stained concurrently in a liquid suspension format in 3 separate vials, each containing one of the 3 above reagents. After incubating at 35℃ to 37℃ for 5 minutes, the stained cell suspensions are rinsed with 1X PBS. The rinsed cells are pelleted by centrifugation and then re-suspended with the resuspension buffer and loaded onto a specimen slide well. The cells are examined using a fluorescence microscope. Cells infected with influenza A virus, respiratory syncytial virus, or parainfluenza virus types 1, 2 and 3 will exhibit goldenyellow fluorescence due to the PE. Cells infected with influenza B virus, metapnemovirus or adenovirus will exhibit apple-green fluorescence due to the FITC. Non-infected cells will exhibit red fluorescence due to the Evans Blue counter-stain. Nuclei of intact cells will exhibit orange-red fluorescence due to the propidium iodide.

Here's a summary of the acceptance criteria and study details for the D3 FastPoint L-DFA Respiratory Virus Identification Kit, based on the provided text:

1. Table of Acceptance Criteria and Reported Device Performance

The document does not explicitly state pre-defined acceptance criteria (e.g., "Sensitivity must be >X%"). Instead, it presents the results of reproducibility and clinical performance studies, implying that these results met the necessary standards for clearance. The "Reported Device Performance" column reflects the results from the clinical sensitivity and specificity studies presented in Tables 5.10-5.15 (for NP wash/aspirate) and 5.16-5.21 (for NP swab). For analytical performance, the reproducibility and LOD results are provided.

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

• Clinical Test Set:

  • Total Specimens Evaluated: 1519
  • Provenance: Prospectively collected excess remnants of respiratory specimens from symptomatic individuals suspected of respiratory infection.
  • Country of Origin: 4 geographically diverse U.S. clinical laboratories.
  • Retrospective or Prospective: Prospective studies (January 2009 - March 2009).
  • Specific Breakdown for Clinical Performance Tables:
    • NP Wash/Aspirate (Sites 1, 2, and 3 combined): Number of specimens varies per virus (e.g., 637 for Flu A, 694 for hMPV).
    • NP Swab (Sites 3 and 4 combined): Number of specimens varies per virus (e.g., 689 for Flu A, 675 for hMPV).

• Analytical Test Set (Reproducibility):

  • Sample Size: 5 randomized panel members for each of the 3 panels (Influenza A/B, RSV/hMPV, HPIV/Adenovirus). Each panel was tested daily in two separate runs for 5 days by four different laboratories (40 total runs per virus group). This means 40 replicates for each panel member for a given virus.
  • Provenance: Proficiency-level antigen control slides with infected cells.

• Analytical Test Set (Limit of Detection):

  • Sample Size: 10 replicate microscope slides for each dilution level of 8 characterized respiratory virus isolates.
  • Provenance: Dilution series of infected model cells.

• Analytical Test Set (Analytical Reactivity/Inclusivity):

  • Sample Size: Low concentration infected cell suspensions (approximately 4% cells infected, 25-50 infected cells) for various viral strains for each of the 3 reagents.
  • Provenance: Culture isolates of various influenza A (13 strains), influenza B (7 strains), RSV (3 strains), hMPV (4 strains), HPIV (3 strains), and Adenovirus (10 strains).

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 clinical test set. Instead, it defines "True" positive and "True" negative based on a composite comparator method:

• For Influenza A, Influenza B, RSV, Parainfluenza, and Adenovirus: Direct Specimen Fluorescent Antibody (DSFA) test with an FDA cleared device, and viral culture confirmation of all negatives (as determined by the comparator DSFA test).
• For Human Metapneumovirus (hMPV): DSFA with an FDA cleared device, and confirmation of all negative specimens (as determined by the comparator DSFA test) using a validated hMPV real-time RT-PCR followed by bi-directional sequencing analysis.

4. Adjudication Method for the Test Set

The adjudication method used for establishing the ground truth for the clinical test set was a composite comparator method. This means results from multiple established methods (DSFA, viral culture, and for hMPV, RT-PCR with sequencing) were combined to determine the "true" status of a specimen. There is no mention of a specific expert panel adjudication method like "2+1" or "3+1" for interpreting these comparator results in case of discrepancies; rather, the definition of "true" positive/negative indicates the hierarchy of methods (e.g., positive by DSFA or viral culture is "true" positive).

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 a diagnostic kit read by a human using a fluorescence microscope, but the study focuses on the kit's performance against comparator methods, not on human reader improvement with or without AI assistance.

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

This is not applicable as the D3 FastPoint L-DFA Respiratory Virus Identification Kit is a diagnostic kit that relies on a human reading the results under a fluorescence microscope. It is not an AI algorithm. The performance presented is of the kit as used by a human.

7. The Type of Ground Truth Used

The ground truth used for the clinical test set was a composite comparator method, which included:

• FDA cleared Direct Specimen Fluorescent Antibody (DSFA) tests
• Viral Culture
• Validated hMPV real-time RT-PCR followed by bi-directional sequencing analysis (for hMPV only)
• NCBI GenBank database matching with acceptable E-values for bi-directional sequencing data.

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 development. This device is a diagnostic reagent kit, not an AI algorithm. The studies conducted are for analytical and clinical validation of the kit itself.

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

As there is no "training set" in the context of this diagnostic kit, this question is not applicable. The ground truth for the comparator methods was established using established laboratory techniques and FDA-cleared devices, as described in point 7.

Ask a specific question about this device

The Diagnostic Hybrids, Inc. device, D3 DFA Metapneumovirus Identification Kit, is intended for the qualitative detection and identification of human metapneumovirus (hMPV) in nasal and nasopharyngeal swabs and aspirates/washes or cell culture. The assay detects hMPV antigens by immunofluorescence using a blend of three monoclonal antibodies (MAbs), from patients with signs and symptoms of acute respiratory infection. This assay detects but is not intended to differentiate the four recognized genetic sub-lineages of hMPV.

Negative results do not preclude hMPV infection and should not be used as the sole basis for diagnosis, treatment or other management decisions. It is recommended that specimens found to be negative after examination of the direct specimen results be confirmed by an FDA cleared hMPV molecular assay.

The D DFA Metapneumovirus Identification Kit uses a blend of three hMPV antigen-specific murine MAbs that are directly labeled with fluorescein for detection of hMPV. The reagent detects but does not differentiate between the four recognized subtypes of hMPV (subtypes A1, A2, B1, and B2).

Kit Components:

• 1. Metapneumovirus DFA Reagent, 5-mL. One dropper bottle containing a blend (see below for MAb discussion) of fluorescein-labeled murine monoclonal antibodies directed against MPV. The buffered, stabilized, aqueous solution contains Evans Blue as a counter-stain and 0.1% sodium azide as a preservative.
• 2. hMPV Antigen Control Slides, 5 slides. Five individually packaged control slides, each with a well containing cell culture-derived MPV positive cells and a well containing cell culture-derived negative cells. Each slide is intended to be stained only one time. Control material has been treated to be non-infectious; however normal laboratory precautions are required when the material is handled.
• 3. 40X PBS Concentrate, 25-mL. One bottle containing a 40X concentrate consisting of 4% sodium azide (0.1% sodium azide after dilution to 1X using de-mineralized water) in PBS.
• 4. Mounting Fluid, 7-mL. One dropper bottle containing an aqueous, buffer-stabilized solution of glycerol and 0.1% sodium azide.

The cells to be tested, derived from a clinical specimen or cell culture, are placed onto a glass slide, allowed to air dry and are fixed in acetone. The Metapneumovirus DFA Reagent is added to the cells which are then incubated for 15 to 30 minutes at 35℃ to 37℃ in a humidified chamber or humidified incubator. The stained cclls are then washed with the diluted phosphate buffered saline (PBS), a drop of the supplied Mounting Fluid is added and a coverslip is placed on the prepared cells. The cells are examined using a fluorescence microscope. The hMPV infected cells will fluoresce apple-green. Uninfected cells will contain no fluorescence but will be stained red by the Evans Blue counter-stain.

It is recommended that specimens found to contain no fluorescent cells after examination of the direct specimen be confirmed by an FDA cleared hMPV molecular assay.

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

Device Name: D3 DFA Metapneumovirus Identification Kit

1. Table of Acceptance Criteria and Reported Device Performance

The acceptance criteria are implied by the performance results required for clearance. As this is not a modern AI/ML device, the performance metrics are clinical sensitivity and specificity, or Positive Percent Agreement (PPA) and Negative Percent Agreement (NPA). Acceptance usually implies that these metrics meet certain thresholds, particularly with lower bounds of the 95% Confidence Interval (CI) demonstrating adequate performance.

For Direct Specimen Testing (Clinical Studies Sites 1-3):

*Note: For Study Site 3, Positive Percent Agreement (PPA) and Negative Percent Agreement (NPA) were reported instead of sensitivity and specificity due to the comparator method used.

For Cultured Cells Testing (Clinical Study Site 4):

For Reproducibility (Analytical Performance):

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

The clinical performance studies used the following sample sizes for the test set:

• Study Site 1: 1482 fresh nasal wash/nasopharyngeal aspirate specimens.
• Study Site 2: 368 fresh nasal/nasopharyngeal swab specimens.
• Study Site 3: 32 fresh nasal wash/nasopharyngeal aspirate specimens and 66 fresh nasal/nasopharyngeal swab specimens.
• Study Site 4 (Cultured Cells): 74 freeze-thawed nasopharyngeal swab specimens that were cultured.

Data Provenance: The data was collected during prospective studies at 3 geographically diverse U.S. clinical laboratories (Study Sites 1-3) during the 2005-2006 and 2006-2007 respiratory virus seasons (December 2005 - April 2006 and December 2006 - March 2007). Study Site 4, for cultured cells, was performed at DHI during the 2007-2008 respiratory virus season (January - April 2008). Specimens were "excess, remnants of respiratory specimens that were prospectively collected from symptomatic individuals suspected of respiratory infection, and were submitted for routine care or analysis by each site, and that otherwise would have been discarded." Individual specimens were de-linked from all patient identifiers. All clinical sites were granted waivers of informed consent by their IRBs.

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

The document does not explicitly state the number of experts or their qualifications for establishing the ground truth. It describes the ground truth as composite comparator methods or a single comparator assay in different sites:

• Sites 1 and 2: Ground truth was a composite comparator method consisting of viral culture and a validated real-time RT-PCR comparator assay with bi-directional sequencing analysis. "True" hMPV positive was defined as positive by viral culture OR positive by RT-PCR with sequencing matching hMPV. "True" hMPV negative was defined as negative by both viral culture and RT-PCR.
• Site 3 and 4: Ground truth was based solely on the validated hMPV real-time RT-PCR followed by bi-directional sequencing analysis comparator assay. Positive was defined by sequencing data matching hMPV, and negative by negative RT-PCR.

While these methods are considered reference standards in microbiology, the document does not specify human expert involvement in interpreting these results or adjudicating discrepancies, beyond the inherent expertise in running and interpreting these laboratory assays.

4. Adjudication Method for the Test Set

The document does not explicitly detail an "adjudication method" in the sense of multiple human readers resolving disagreements, as would be typical for image-based AI studies. Instead, the ground truth itself is a carefully defined reference standard.

• For Sites 1 and 2, the ground truth was a composite definition:
  • Positive: Viral culture positive OR Real-time RT-PCR positive with bi-directional sequencing matching hMPV.
  • Negative: Viral culture negative AND Real-time RT-PCR negative.
• For Sites 3 and 4, the ground truth was based on the hMPV real-time RT-PCR followed by bi-directional sequencing analysis comparator assay. Positive was defined by acceptable sequencing data matching hMPV, and negative by negative RT-PCR.

This structure inherently handles potential discrepancies between methods by prioritizing certain outcomes (e.g., a positive by either method for the composite 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, an MRMC comparative effectiveness study involving human readers and AI assistance was not done. This device is a diagnostic kit (an immunofluorescence assay), not an AI-powered system designed to assist human readers in interpreting complex data. The clinical studies evaluated the standalone performance of the kit itself against a reference standard.

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

Yes, the studies described are essentially standalone performance studies for the D3 DFA Metapneumovirus Identification Kit. The kit's results (fluorescence detection by microscopy) were compared directly against the established ground truth without any involvement of a human-in-the-loop for interpreting the kit's results in the context of an AI system. The "algorithm" here is the biochemical and optical detection mechanism of the DFA test combined with human interpretation of fluorescence under a microscope, as per standard laboratory practice.

7. The Type of Ground Truth Used

The type of ground truth used varied slightly across study sites but primarily involved molecular and classical microbiological methods:

• Clinical Study Sites 1 and 2: Composite Ground Truth combining:
  • Viral Culture: A classical microbiological method for isolating and identifying viruses.
  • Validated Real-time RT-PCR followed by bi-directional sequencing analysis: A molecular method to detect hMPV nucleic acid, with sequencing confirming the identity.
• Clinical Study Sites 3 and 4: Molecular Ground Truth based solely on a validated hMPV real-time RT-PCR followed by bi-directional sequencing analysis. This method is considered a highly specific and sensitive reference standard.

8. The Sample Size for the Training Set

No information about a "training set" for an algorithm is provided. This device is a diagnostic kit (DFA assay), not an AI/ML model that undergoes a training phase.

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

As there is no "training set" for an AI/ML algorithm in this context, this question is not applicable. The device relies on direct antigen detection via immunofluorescence, not on learned patterns from a training dataset.

Ask a specific question about this device