Search Results

The QuickVue COVID-19 Test is a visually read lateral flow immunoassay device intended for the rapid, qualitative detection of SARS-CoV-2 nucleocapsid protein antigens directly in anterior nasal (nares) swab specimens from individuals with signs and symptoms of COVID-19 within the first 5 days from symptom onset. This test is for nonprescription home use by individuals aged 14 years or older testing themselves, or adults testing individuals aged 2 years or older.

The QuickVue COVID-19 Test does not differentiate between SARS-CoV and SARS-CoV-2.

All negative results are presumptive. Symptomatic individuals with an initial negative test result must be re-tested once between 48 and 72 hours after the first test using either an antigen test for SARS-CoV-2. Negative results do not preclude SARS-CoV-2 infections or other pathogens and should not be used as for treatment. Positive results do not rule out co-infection with other respiratory pathogens.

This test is not a substitute for visits to a healthcare provider or appropriate follow-up and should not be used to determine any treatments without provider supervision. Individuals who test negative and experience continued or worsening COVID-19 like symptoms, such as fever, cough and/or shortness of breath, should seek follow up care from their healthcare provider.

The performance characteristics for SARS-CoV-2 were established from January 2021 to February 2024 when COVID-19 variants Alpha, Delta and Omicron were dominant. Test accuracy may change as new SARS-CoV-2 viruses emerge. Additional testing with a lab-based molecular test (e.g., PCR) should be considered in situations where a new virus or variant is suspected.

The QuickVue COVID-19 Test is a lateral flow immunoassay device intended for the qualitative detection of nucleocapsid protein antigen from SARS-CoV-2. The QuickVue COVID-19 Test does not differentiate between SARS-CoV and SARS-CoV-2.

To begin the test, a self-collected anterior nasal swab sample (in individuals aged 14 and older or individuals between the age of 2 to 14 a swab collected by a parent or guardian), or a healthcare collected anterior nasal swab sample is inserted into the pre-filled reagent tube. The reagent disrupts the virus particles in the specimen, exposing internal viral nucleocapsid antigens. The test strip is then placed in the reagent tube where the viral nucleocapsid antigens in the specimen will react with the reagents in the test strip.

If the extracted specimen contains SARS-CoV-2 viral nucleocapsid antigens, a pink-to-red test line along with a blue procedural control line will appear on the test strip indicating a positive result. If SARS-CoV-2 viral nucleocapsid antigens are not present, or are present at very low levels, only the blue procedural control line will appear.

This document describes the performance of the QuickVue COVID-19 Test, a visually read lateral flow immunoassay device. The acceptance criteria and supporting study details are outlined below.

1. Table of Acceptance Criteria & Reported Device Performance

The acceptance criteria for performance are generally implied by the reported results. For a rapid diagnostic test like this, the key performance metrics are Positive Percent Agreement (PPA) and Negative Percent Agreement (NPA). While explicit numerical acceptance criteria for PPA and NPA are not stated as "acceptance criteria," the study aimed to demonstrate substantial equivalence to a predicate device, which implies similar or better performance. The performance must also be sufficient for its intended use as an over-the-counter home test.

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

• Test Set Sample Size: A total of 780 symptomatic individuals were included in the prospective clinical study that compared the QuickVue COVID-19 Test performance to an EUA extracted SARS-CoV-2 RT-PCR assay.
  • 200 individuals were positive by the comparator assay.
  • 580 individuals were negative by the comparator assay.
• Data Provenance: The study was a prospective clinical study conducted at eight clinical sites. The country of origin is not explicitly stated, but given the FDA review, it is highly likely to be the United States. Samples were collected by lay users (self-collected) or collected for a household member.

3. Number of Experts Used to Establish Ground Truth and Qualifications

The document does not specify the number of experts or their qualifications for establishing the ground truth. The ground truth was established by an EUA extracted SARS-CoV-2 RT-PCR assay, which is considered the gold standard for SARS-CoV-2 detection. Therefore, human expert judgment for result interpretation of the ground truth assay itself is less relevant than with image-based AI studies.

4. Adjudication Method for the Test Set

Not applicable. The ground truth was established by a laboratory-based RT-PCR assay, not through human reader interpretation requiring adjudication.

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

No, an MRMC comparative effectiveness study was not done. This device is a visually-read lateral flow immunoassay intended for direct diagnosis, not an AI-assisted diagnostic tool that would typically involve human readers interpreting AI outputs. The comparison was to a molecular (RT-PCR) test, not a comparative effectiveness study involving human readers with and without AI assistance.

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

The QuickVue COVID-19 Test is a visually read lateral flow immunoassay. It does not involve an algorithm for automated interpretation in its primary intended use. While it has an optional "QVue Mobile Application," the primary interpretation is visual by the user. Therefore, a "standalone algorithm performance" study as seen in AI/software medical devices is not directly applicable in the same way. The performance data presented (PPA, NPA) directly reflects the visually-read results.

7. Type of Ground Truth Used

The type of ground truth used was a lab-based molecular test (RT-PCR). Specifically, an "EUA extracted SARS-CoV-2 RT-PCR assay" was used as the comparator. This is widely considered the gold standard for detecting SARS-CoV-2 viruses.

8. Sample Size for the Training Set

The document describes the performance of a lateral flow immunoassay, not an AI/machine learning model that typically requires a large training set. Therefore, there is no "training set" in the context of an AI model. The development of an immunoassay involves analytical studies and in-house testing to optimize the reagent formulations and visual interpretation, but not a distinct "training set" like in deep learning.

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

As there is no "training set" in the context of an AI model, this question is not applicable. The development of the immunoassay itself relies on established laboratory practices, analytical sensitivity (LoD), inclusivity, and specificity studies, using characterized viral isolates and clinical samples, to ensure the test consistently produces accurate results across various conditions.

Ask a specific question about this device

The Savanna HSV 1+2/VZV Assay is an automated, rapid multianalyte real-time PCR test for the simultaneous qualitative detection and differentiation of herpes simplex virus type 1, herpes simplex virus type 2, and varicella-zoster virus DNA isolated from human cutaneous or mucocutaneous lesion samples obtained from symptomatic patients suspected of active herpes simplex virus 1, herpes simplex virus 2 and/or varicella-zoster infection. This in vitro diagnostic test is intended to aid in the diagnosis of patients with signs or symptoms of herpes simplex virus type 1, herpes simplex virus type 2, and varicella-zoster virus infection.

The Savanna HSV 1+2/VZV Assay is intended to aid in the diagnosis of herpes simplex virus 1, herpes simplex virus 2 and varicella-zoster virus active infections. The results of this test should not be used as the sole basis for diagnosis, treatment or other management decisions and must be combined with clinical observations, patient history and/or epidemiological information. Negative results do not preclude herpes simplex virus type 1, herpes simplex virus type 2, or varicella-zoster virus infection that is not detected by a cutaneous or mucocutaneous lesion swab specimen. Positive results do not rule out co-infection with other organisms. Additional laboratory testing (e.g., viral culture, immunoassay, serology) may be necessary for patient evaluation. Savanna HSV 1+2/VZV Assay is for professional use. The Savanna HSV 1+2/VZV Assay is intended for use only with the Savanna instrument.

Warning: The Savanna HSV 1+2/VZV Assay is not intended for use with the cerebrospinal fluid (CSF) or to aid in the diagnosis of HSV or VZV infections of the central nervous system (CNS). The Savanna HSV 1+2/VZV Assay is not intended for use in prenatal screening.

The Savanna HSV 1+2/VZV Assay consists of a single, self-contained assay cartridge employing real-time PCR technology for use with the Savanna instrument to detect and differentiate DNA from herpes simplex virus type 1, herpes simplex virus type 2 and varicella-zoster virus. In approximately 24 minutes, this platform extracts, amplifies and detects DNA present in cutaneous or mucocutaneous lesion swab specimens obtained from symptomatic patients and placed in transport media.

To initiate the assay, a patient cutaneous or mucocutaneous lesion swab specimen in transport medium is transferred via the supplied transfer pipette to the Liguid Sample Port of the Test Cartridge. The user closes the Sample Port and inserts the Test Cartridge into the Savanna instrument, initiating sample processing. The sample containing human DNA is pushed out of the Sample Port by lysis buffer, which then rehydrates the Process Internal Control (IC), and together with the paramagnetic nucleic acid binding particles, are pumped into the extraction chamber. The solution is mixed, and virus and/ or bacteria are further lysed by sonication within the extraction chamber. Specimen and IC DNA are bound to, washed and then eluted off the paramagnetic particles. The purified specimen DNA and IC solution is used to rehydrate four individual lyophilized master mixes. Each master mix is pumped into a PCR chamber and Taq-man® multiplex real-time PCR reactions are carried out under optimized conditions, generating amplicons for the targeted pathogen (if present) and the Process Internal Control (IC).

Each master mix contains primers and dual-labeled probes unique for the pathogen targets and the IC. The probes are labeled with a fluorophore on one end and a quencher on the other end. Target DNA sequences are amplified by pathogen-specific primers and detected by correspondingly specific fluorescence probes. The IC targets are also amplified by specific primers and detected by an IC-specific fluorescence probe. A polymerase included in the master mix cleaves the probes bound to complementary DNA sequences, separating the fluorophore from the quencher. This step generates a signal and if it surpasses multiple defined thresholds, the sample is reported as positive for the detected target sequence.

The Savanna instrument will display the test results (Positive or Invalid) on the main bay screen.

The provided text describes the performance characteristics and clinical studies of the Savanna HSV 1+2/VZV Assay, which is an in vitro diagnostic (IVD) device for the detection and differentiation of herpes simplex virus type 1, herpes simplex virus type 2, and varicella-zoster virus DNA. It is not an AI/ML device in the context of typical FDA AI/ML medical device submissions, as it is a molecular diagnostic test. Therefore, many of the requested points regarding AI/ML device evaluation (like human-in-the-loop, expert consensus for ground truth, MRMC studies, number of experts for ground truth, adjudication methods, and sample size for training set) are not applicable to this type of device.

However, I can extract the acceptance criteria and performance data for this diagnostic assay as presented. The study aims to demonstrate substantial equivalence to a predicate device, the Solana HSV 1+2/VZV Assay (K162451), not to prove performance of an AI model.

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

The document does not explicitly state "acceptance criteria" in a tabulated format for the clinical performance. Instead, it presents the results of the clinical studies with Positive Percent Agreement (PPA) and Negative Percent Agreement (NPA) values with their confidence intervals, comparing the Savanna Assay to a commercially available RT-PCR comparator method. For laboratory studies, the acceptance criteria are implied by the results (e.g., >95% positivity for LoD, 100% detection for inclusivity, no interference/cross-reactivity).

I will present the clinical performance results as the "reported device performance" against an implied high standard of agreement with the comparator method.

Table: Reported Device Performance (Clinical Study 1 & 2)

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

• Clinical Study 1 (Fresh Samples):

  • Sample Size: 590 residual specimens (207 cutaneous, 383 mucocutaneous).
  • Data Provenance: United States, from symptomatic patients suspected of HSV-1, HSV-2, or VZV infection. Retrospective (residual specimens).

• Clinical Study 2 (Frozen Residual Samples):

  • Sample Size: 154 evaluable samples.
  • Data Provenance: Not explicitly stated but implies U.S. as it's a supplement to Clinical Study #1. Retrospective (residual frozen samples).

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

This is not applicable as this is an IVD device, not an AI/ML device relying on human expert labels for ground truth. The ground truth is established by a comparator FDA-cleared nucleic acid amplification test.

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

Not applicable. The ground truth is determined by the results of an FDA-cleared comparator nucleic acid amplification test, not by human adjudication of images.

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 IVD device, not an AI/ML device intended to assist human readers.

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

The device is a standalone diagnostic assay (real-time PCR test on the Savanna instrument). Its performance is evaluated independently against a comparator method. The results presented (PPA, NPA) represent this standalone performance.

7. The type of ground truth used

The ground truth for the clinical studies was established by comparison to FDA-cleared nucleic acid amplification tests (comparator method).

8. The sample size for the training set

Not applicable. This is not an AI/ML device that requires a "training set" in the context of machine learning model development. The robust performance of an IVD often relies on extensive analytical studies (LoD, inclusivity, cross-reactivity, precision, etc.) and clinical validation.

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

Not applicable, as there is no "training set" in the AI/ML sense. The analytical performance was established through various laboratory studies (e.g., Limit of Detection, Inclusivity, Cross-Reactivity, Precision) using well-characterized viral strains and matrices.

Ask a specific question about this device

The Sofia 2 SARS Antigen+ FlA is a lateral flow immunofluorescent sandwich assay that is used with the Sofia 2 instrument for the rapid, qualitative detection of SARS-CoV-2 nucleocapsid protein antigens directly in anterior nasal swab specimens from individuals with signs and symptoms of upper respiratory infection (i.e., symptomatic) when testing is started within 6 days of symptom onset. The test is intended for use as an aid in the diagnosis of SARS-CoV-2 infections (COVID-19) in symptomatic individuals when either: tested at least twice over three days with at least 48 hours between tests; or when tested once, and negative by the Sofia 2 SARS Antigen+ FIA and followed up with a molecular test.

The test does not differentiate between SARS-CoV and SARS-CoV-2.

A neqative test result is presumptive, and it is recommended these results be confirmed by a molecular SARS-CoV-2 assay.

Positive results do not rule out co-infection with bacteria or other viruses and should not be used as the sole basis for treatment or other patient management decisions.

Performanc characteristics for SARS-CoV-2 were established during the 2021-2022 SARS-CoV-2 pandemic when SARS CoV 2 Omicron was the predominant SARS-CoV-2 variant in circulation. When other SARS-CoV-2 virus variant are emerging, performance characteristics may vary.

This test is intended for prescription use only and can be used in Point-of-Care settings.

Sofia 2 SARS Antigen+ FIA Control Swab Set:

The Sofia 2 SARS Antigen+FIA Control Swabs are intended to be used as quality control samples with the Sofia 2 SARS Antigen+ FIA and are representative of positive and negative test samples. These Controls may be used to demonstrate that the reagents and assay procedure perform properly.

The Sofia 2 SARS Antigen+FIA is based upon a lateral flow technology that employs immunofluorescence technology in a sandwich design that is used with Sofia 2 to detect nucleocapsid protein from the SARS-CoV-2 virus in human anterior nasal swab specimens.

The patient sample is placed in the Reagent Tube, during which time the virus particles in the sample are disrupted, exposing internal viral nucleoproteins. After disruption, the sample is dispensed into the Test Cassette sample well. The Test Strip is composed of the following biochemical components dried and immobilized onto the nitrocellulose membrane: 1) sample pad that receives the specimen: 2) a label pad that contains detection fluorescent micro-particles, coated with monoclonal antibodies that are specific for SARS-CoV-2 nucleocapsid antigen; 3) embedded monoclonal antibodies specific for SARS-CoV-2 nucleocapsid antigen to capture the antigen-microparticle complex at the tocation. The sample pad facilitates migration of the sample fluid across the nitrocellulose strip into the absorbent pad (See Figure 4-1 in attachment). The test strip also contains a desiccant that does not participate in the assay but serves as a stabilizing agent during storage.

Sample is applied to in the sample well and migrates through a test strip, then passes through the test and control lines. If SARS-CoV-2 viral antigen is present, they will be bound by the fluorescent microparticles io the label pad reqion, forming an antigen-microparticle complex.

The test line is coated with monoclonal antibodies that are specific to SARS-CoV-2 nucleocapsid antigen and is intended to capture the antigen-microparticle complex. If SARS-CoV-2 viral antigen is not present, the fluorescent microparticles will not be trapped by the capture antibodies nor detected by Sofia 2.

The Sofia 2 SARS Antigen+ FIA employs antibody tagged microparticles dyed with a fluorescent compound, to be detected and read by the Sofia 2 reader instrument. The Sofia 2 analyzers automatically scan/image the test strip, collect and analyze the fluorescence data, and report the result as either positive, negative, or invalid.

Additionally, the Sofia 2 Antigen+ FIA utilizes a reference line for the Sofia 2 reader (to locate the test line and negative control line) and a procedural control (to assess for sample presence and adequate sample flow). No colored lines will be visible in the test window of the fluorescent assay cassette, thereby preventing visual interpretation of the test results. The operator must use the Sofia 2 analyzer to obtain a test result.

The Sofia 2 SARS Antigen+ FIA Control Swabs are intended to be used as quality control samples representative of positive and neqative test samples, to demonstrate that the reagents are that the assay procedure is correctly perform.

This document is a 510(k) Premarket Notification from the FDA regarding the Sofia 2 SARS Antigen+ FIA. It primarily details the device's indications for use, its description, and a comparison to a legally marketed predicate device.

Here's an analysis of the provided information to address your request:

1. Table of Acceptance Criteria and Reported Device Performance

The document does not explicitly state "acceptance criteria" in terms of performance metrics (e.g., sensitivity, specificity thresholds) for the currently reviewed device (K233688). This submission (K233688) is for a modification to the labeling of an already cleared device (DEN220039). The core performance characteristics were established during the clearance of the predicate device (DEN220039).

The "510(k) Summary" section explicitly states under "Non-Clinical and/or Clinical Tests Summary & Conclusions": "Not applicable." This indicates that no new performance studies were conducted or provided with this specific submission because the design, material, chemical composition, and principle of operation are unchanged from the predicate device. The changes are limited to a minor revision of the intended use statement on labeling.

Therefore, performance data to directly prove acceptance criteria for this specific submission is not provided in this document. The document refers to the established performance characteristics from the predicate device, stating: "Performanc characteristics for SARS-CoV-2 were established during the 2021-2022 SARS-CoV-2 pandemic when SARS CoV 2 Omicron was the predominant SARS-CoV-2 variant in circulation."

If we were to infer the implied acceptance criteria from the predicate device's clearance, it would be that the device performs "as safe, as effective, and performs as well as the predicate device." However, no specific numerical performance metrics are listed in this document.

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

As stated above, no new performance studies were conducted for this specific submission (K233688). The document explicitly states "Not applicable" for clinical tests. Therefore, there is no sample size for an additional test set, nor new data provenance provided here. Performance characteristics were established during the 2021-2022 SARS-CoV-2 pandemic for the predicate device.

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

Since no new studies were conducted, this information is not provided in the document.

4. Adjudication Method for the Test Set

Since no new studies were conducted, this information is not provided in the document.

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

No MRMC study is mentioned or implied for this submission. The device is an in vitro diagnostic (IVD) rapid antigen test read by an instrument (Sofia 2 analyzer), not typically involving multiple human readers for interpretation in the same way an imaging AI would. The Sofia 2 analyzer automatically scans, collects, analyzes fluorescence data, and reports results as positive, negative, or invalid, preventing visual interpretation.

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

Yes, this device inherently operates in a "standalone" fashion as an algorithm-only device in terms of result interpretation. The Sofia 2 analyzer processes the test strip by detecting and reading fluorescent microparticles, analyzing the fluorescence data, and automatically reporting the result without human interpretation of the signal on the test strip. The human interaction is limited to collecting the sample, performing the assay steps, and reading the final result from the instrument.

7. The Type of Ground Truth Used

The document indicates the test is intended as an aid in diagnosis, and negative test results are presumptive and "recommended these results be confirmed by a molecular SARS-CoV-2 assay." This strongly implies that the ground truth for establishing the device's performance characteristics (for the predicate device) would have been molecular SARS-CoV-2 assays (e.g., RT-PCR), which are considered the gold standard for SARS-CoV-2 detection.

8. The Sample Size for the Training Set

No information about a training set or its sample size is provided in this document, as no new performance studies were conducted for this submission. The initial development and training (if applicable, for the Sofia 2 analyzer's signal processing algorithm) would have occurred prior to the predicate device's clearance.

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

Since no information on a training set is provided, the method for establishing its ground truth is also not mentioned. It is likely that for the predicate device, if machine learning was involved in interpreting the fluorescent signal, the training data's ground truth would have been established against molecular SARS-CoV-2 assays, similar to the test set ground truth.

Ask a specific question about this device

The Lyra RSV + hMPV Assay is a multiplex Real-Time PCR (RT-PCR) assay for the qualitative detection and identification of respiratory syncytial virus (RSV) and human metapneumovirus (hMPV) ribonucleic acid (RNA) extracted from nasal and nasopharyngeal swab specimens from patients with signs and symptoms of respiratory infection. This in vitro diagnostic test is intended to aid in the differential diagnosis of RSV and hMPV infections in humans in conjunction with clinical and epidemiological risk factors. This test is not intended to differentiate the two subtypes of RSV or the four genetic sub-lineages of hMPV.

Negative results do not preclude RSV infection and/or hMPV infection and should not be used as the sole basis for diagnosis, treatment or other patient management decisions.

Conversely, positive results do not rule-out bacterial infection with other viruses. The agent detected may not be the definite cause of disease. The use of additional laboratory testing and clinical presentation must be considered in order to obtain the final diagnosis of respiratory viral infection.

The Lyra RSV + hMPV Assay can be performed using ether the Life Technologies QuantStudio Dx RT-PCR Instrument, the Applied Biosystems 7500 Fast Dx RT-PCR Instrument, or the Cepheid SmartCycler II System.

The Lyra RSV + hMPV Assay detects viral nucleic acids that have been extracted from a patient sample using the NucliSENS® easyMAG® or NucliSENS® EMAG® automated extraction platform. A multiplex RT-PCR reaction is then performed in a single tube generating amplicons for each of the target viruses present in the sample. This reaction is performed utilizing either the Cepheid SmartCycler® II, the Applied Biosystems 7500 Fast DX, or the Life Technologies QuantStudio" Dx. Identification of RSV and hMPV and the PRC occurs by the use of target-specific primers and fluorescent-labeled probes that hybridize to conserved regions in the genomes of RSV and hMPV and the PRC.

The acceptance criteria and the study proving the device meets these criteria for the Lyra RSV + hMPV Assay are detailed below. It's important to note that the provided documents primarily describe a change to an existing device (the addition of a new extraction platform, BioMerieux NucliSENS EMAG) and compare its performance to the previously cleared predicate device. Therefore, the "acceptance criteria" and "reported device performance" are framed around this equivalency.

1. Table of Acceptance Criteria and Reported Device Performance

Based on the information provided, the overall acceptance criterion is equivalent performance of the modified Lyra RSV + hMPV Assay (with the new extraction platform) to the predicate device (Quidel RSV+hMPV Assay using the original extraction platform). The performance is assessed through "non-clinical and clinical verification and validation activities."

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

The exact sample sizes for the "Limit of Detection Equivalency Study" and "Clinical Equivalence Study" are not explicitly stated in the provided 510(k) summary.

• Test Set Sample Size: Not specified.
• Data Provenance: Not specified (e.g., country of origin, retrospective or prospective). The nature of "clinical equivalence study" typically implies prospective or retrospectively collected clinical samples, but details are lacking.

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

This information is not provided in the given text. For a PCR-based assay like this, ground truth is typically established using a highly sensitive and specific reference method (e.g., another validated PCR assay or sequencing), rather than expert clinical consensus in the traditional sense of image interpretation.

4. Adjudication Method for the Test Set

This information is not provided in the given text. Given that this is a molecular diagnostic assay, adjudication methods like N+1 for expert review (common in imaging studies) are generally not applicable. Instead, the "ground truth" would be determined by the reference method itself.

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

No, an MRMC comparative effectiveness study was not done. This type of study is primarily relevant for medical imaging interpretation where different human readers interpret cases, often with and without AI assistance. This device is a molecular diagnostic assay (RT-PCR) and does not involve human interpretation of complex images in the same way.

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

Yes, implicitly. The device itself is an in vitro diagnostic test, specifically a Real-Time PCR assay. Its performance (qualitative detection of RNA) is inherently "standalone" as it produces a result without direct human interpretive input during the assay itself. The studies ("Limit of Detection Equivalency Study" and "Clinical Equivalence Study") would evaluate the accuracy of the assay's output against a reference method, which is a standalone performance assessment.

7. The Type of Ground Truth Used

The type of ground truth used is not explicitly stated but can be inferred to be a highly reliable reference method, likely another validated molecular diagnostic test (e.g., gold standard PCR assay or culture/sequencing) for detecting RSV and hMPV. For Limit of Detection studies, ground truth would be established by precisely quantifying viral RNA in samples.

8. The Sample Size for the Training Set

This information is not applicable in the context of this device. The Lyra RSV + hMPV Assay is a Real-Time PCR assay, which is a biochemical reaction-based test, not an AI/Machine Learning algorithm that requires a "training set" in the traditional sense. The "training" for such an assay involves optimization of reagents, primers, probes, and reaction conditions.

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

This information is not applicable as there is no "training set" in the context of an AI/ML algorithm for this device. The development of the assay involves establishing analytical ground truth through various laboratory experiments to ensure the primers and probes are specific and sensitive to the target viruses, and that the internal control functions correctly.

Ask a specific question about this device

The Sofia 2 SARS Antigen+ FIA is a lateral flow immunofluorescent sandwich assay that is used with the Sofia 2 instrument for the rapid, qualitative detection of SARS-CoV-2 nucleocapsid protein antigens directly in anterior nasal swab specimens from individuals with signs and symptoms of upper respiratory infection (i.e., symptomatic) when testing is started within 6 days of symptom onset. The test is intended for use as an aid in the diagnosis of SARS-CoV-2 infections (COVID-19) in symptomatic individuals when tested at least twice over three days with at least 48 hours between tests.

The test does not differentiate between SARS-CoV and SARS-CoV-2.

A negative test result is presumptive, and it is recommended these results be confirmed by a molecular SARS-CoV-2 assay. Negative results do not preclude SARS-CoV-2 infections and should not be used as the sole basis for treatment or other patient management decisions.

Positive results do not rule out co-infection with other respiratory pathogens.

Performance characteristics for SARS-CoV-2 were established during the 2021-2022 SARS-CoV-2 pandemic when SARS-CoV-2 Omicron was the predominant SARS-CoV-2 variant in circulation. When other SARS-CoV-2 virus variant are emerging, performance characteristics mav vary.

This test is intended for prescription use only and can be used in Point-of-Care settings.

The Sofia 2 SARS Antigen+FIA is based upon a lateral flow technology that employs immunofluorescence technology in a sandwich design that is used with Sofia 2 to detect nucleocapsid protein from the SARS-CoV-2 virus in human anterior nasal swab specimens.

The patient sample is placed in the Reagent Tube, during which time the virus particles in the sample are disrupted, exposing internal viral nucleoproteins. After disruption, the sample is dispensed into the Test Cassette sample well. The Test strip is composed of the following biochemical components dried and immobilized onto the nitrocellulose membrane: 1) sample pad that receives the specimen; 2) a label pad that contains detection fluorescent micro-particles, coated with monoclonal antibodies that are specific for SARS-CoV-2 nucleocapsid antigen: 3) embedded monoclonal antibodies specific for SARS-CoV-2 nucleocapsid antigen to capture the antigen-microparticle complex at the test line location. The sample pad facilitates migration of the sample fluid across the nitrocellulose strip into the absorbent pad. The test strip also contains a desiccant that does not participate in the assay but serves as a stabilizing agent during storage.

Sample is applied to in the sample well and migrates through a test strip, then passes through the test and control lines. If SARS-CoV-2 viral antigen is present, they will be bound by the fluorescent microparticles in the label pad region, forming an antigen-microparticle complex. The test line is coated with monoclonal antibodies that are specific to SARS-CoV-2 nucleocapsid antigen and is intended to capture the antigen-microparticle complex. If SARS-CoV-2 viral antigen is not present, the fluorescent microparticles will not be trapped by the capture antibodies nor detected by Sofia 2.

The Sofia 2 SARS Antigen+FIA employs antibody-tagged microparticles dyed with a fluorescent compound, to be detected and read by the Sofia 2 reader instrument. The Sofia 2 analyzers automatically scan/image the test strip, collect and analyze the fluorescence data, and then calculate and report the result as either positive, negative, or invalid.

Additionally, the Sofia 2 Antigen+ FIA utilizes a reference line for the Sofia 2 reader (to locate the test line and negative control line) and a procedural control (to assess for sample presence and adequate sample flow). No colored lines will be visible in the test window of the fluorescent assay cassette, thereby preventing visual interpretation of the test results. The operator must use the Sofia 2 analyzer to obtain a test result.

The Sofia SARS Antigen FIA Control Swabs are intended to be used as quality control samples representative of positive and negative test samples, to demonstrate that the reagents are functional and that the assay procedure is correctly perform.

Acceptance Criteria and Device Performance for Sofia 2 SARS Antigen+ FIA

The Sofia 2 SARS Antigen+ FIA is a qualitative lateral flow immunoassay designed for rapid detection of SARS-CoV-2 nucleocapsid protein antigens in anterior nasal swab specimens. The following details outline the acceptance criteria and the studies conducted to prove the device meets these criteria.

1. Acceptance Criteria and Reported Device Performance

• Low Positive samples (1 x LoD) should demonstrate an expected positive agreement (e.g., >95%).
• Moderate Positive samples (3 x LoD) should demonstrate high expected positive agreement (e.g., >95%).
• Zero invalid test results throughout the study (or very low %). | - Negative samples: 99.4% expected negative agreement (159/160)
• High Negative samples (0.04 x LoD): 95.0% expected negative agreement (152/160)
• Low Positive samples (1 x LoD): 98.1% expected positive agreement (157/160)
• Moderate Positive samples (3 x LoD): 99.4% expected positive agreement (159/160)
• 0 invalid test results (out of 640 replicates) |
  | Reproducibility (Inter-site)| - High Negative samples (0.04 x LoD) should demonstrate reasonable negative agreement across sites and operators.
• Low Positive samples (1 x LoD) should demonstrate high positive agreement across sites and operators.
• Moderate Positive samples (3 x LoD) should demonstrate high positive agreement across sites and operators.
• Zero invalid test results throughout the study (or very low %). | - Negative samples: 100.0% expected negative agreement (120/120)
• High Negative samples (0.04 x LoD): 55.0% expected negative agreement (66/120) - Note: This is lower than typical ideal scenarios, but likely deemed acceptable given the nature of a "high negative" near the detection limit.
• Low Positive samples (1 x LoD): 99.2% expected positive agreement (119/120)
• Moderate Positive samples (3 x LoD): 99.2% expected positive agreement (119/120)
• 0 invalid test results (out of 480 samples) |
  | Analytical Specificity (Cross-reactivity & Interference) | - No cross-reactivity with a defined panel of common respiratory pathogens (bacteria, viruses, fungus) at specified concentrations.
• No interference from common endogenous and exogenous substances found in nasal specimens at specified concentrations.
• 100% negative agreement in the absence of SARS-CoV-2.
• 100% positive agreement in the presence of SARS-CoV-2. | - All 28 organisms/viruses tested showed 100.0% negative agreement (5/5 replicates) for cross-reactivity and 100.0% positive agreement (5/5 replicates) for interference.
• All 13 endogenous/exogenous substances tested showed 100.0% positive agreement (5/5 replicates) and 100.0% negative agreement (5/5 replicates). |
  | Limit of Detection (LoD) | - The device should consistently detect SARS-CoV-2 at a specific low concentration (LoD) with high positivity (e.g., 95% or 100%) in confirmatory studies.
• Negative clinical matrix should consistently result in negative readings. | - Confirmed LoD: 1.44 x 10^4 TCID50/mL.
• At confirmed LoD: 100% positivity (20/20 replicates) across both tested lots.
• NCM (Negative Clinical Matrix): 0% positivity (0/5 replicates). |
  | High-dose Hook Effect | - No false negative results should be observed at very high concentrations of SARS-CoV-2. | - All spiked samples from 10X LoD up to a maximum virus concentration (unspecified highest concentration) were 100% positive (5/5 replicates each). No hook effect observed. |
  | Inclusivity | - The device should detect various clinically relevant SARS-CoV-2 strains/variants (e.g., Delta, Omicron BA.1, BA.2) with high positivity. | - Heat-inactivated SARS-CoV-2 (isolate Italy-INMI1): 100.0% positivity (5/5 replicates) at 2.43E+05 TCID50/mL.
• Heat-inactivated SARS-CoV-2 (Delta B.1.617.2): 100.0% positivity (5/5 replicates) at 1.00E+04 TCID50/mL.
• Heat-inactivated SARS-CoV-2 (Omicron BA.1): 100.0% positivity (5/5 replicates) at 2.36E+04 TCID50/mL.
• Heat-inactivated SARS-CoV-2 (Omicron BA.2): 100.0% positivity (5/5 replicates) at 8.22E+03 TCID50/mL. |
  | Clinical Performance | - Demonstrate acceptable Positive Percent Agreement (PPA) and Negative Percent Agreement (NPA) compared to a highly sensitive RT-PCR comparator, within defined confidence intervals. The specific thresholds would be pre-defined by regulatory guidelines (e.g., FDA's EUA templates or De Novo requirements for diagnostics). While not explicitly stated as a numerical criterion, the presented results suggest approval, implying the performance met the agency's expectations for a Class II device. | - PPA (Clinical Sensitivity): 89.0% (97/109; 95% CI: 81.7% - 93.6%)
• NPA (Clinical Specificity): 99.6% (470/472; 95% CI: 98.5% - 99.9%) |

2. Sample Size and Data Provenance

Test Set Sample Sizes:

• Precision/Repeatability: 640 replicates (160 replicates per analyte level/kit lot for 4 levels x 2 lots).
• Reproducibility: 480 replicates (120 replicates per analyte level x 4 levels).
• Analytical Specificity (Cross-reactivity & Interference):
  • Cross-reactivity: 5 replicates per organism/virus (28 tested) = 140 replicates for negative agreement. 5 replicates per organism/virus + 2xLoD SARS-CoV-2 (25 tested, some "Not Tested" for interference) = 125 positive replicates for interference.
  • Interfering Substances: 5 replicates per substance (13 tested) for both positive and negative agreement = 130 replicates.
• Limit of Detection (LoD):
  • Preliminary LoD: 5 replicates per dilution (6 dilutions) x 2 lots = 60 replicates.
  • Confirmatory LoD: 20 replicates at preliminary LoD concentration for each of 2 lots = 40 replicates.
• High-dose Hook Effect: 5 replicates per concentration (4 concentrations) = 20 replicates.
• Inclusivity: 5 replicates per strain/variant (4 tested) = 20 replicates.
• Clinical Study (Accuracy): 581 evaluable subjects.

Data Provenance:

• Country of Origin: Not explicitly stated for analytical studies, but given the FDA review, it is implicitly expected to be a regulated environment. For the clinical study, it was a multi-center study in a "CLIA-waived" setting, which refers to US clinical laboratories.
• Retrospective or Prospective:
  • Analytical Performance Studies (Precision, Reproducibility, Analytical Specificity, LoD, Hook Effect, Inclusivity): These are typically prospective, lab-controlled experiments with contrived samples designed specifically for the study.
  • Clinical Studies: "multi-center, prospective study conducted from August 2021 to November 2022."

3. Number of Experts and Qualifications for Ground Truth

• Analytical Studies: For analytical studies (Precision, Reproducibility, LoD, Cross-reactivity, Interference, Hook Effect, Inclusivity), the ground truth is established by the known concentrations of spiked analytes (e.g., heat-inactivated SARS-CoV-2, other microbes, interfering substances). This does not involve human experts establishing ground truth beyond standard laboratory technician expertise in preparing and measuring these concentrations.
• Clinical Study: The ground truth for the clinical study was established by an "highly sensitive Emergency Use Authorization (EUA) authorized RT-PCR comparator assay." This is a laboratory-based molecular test, often considered the gold standard for SARS-CoV-2 detection, and does not involve human experts in the conventional sense of image adjudication or clinical consensus. The RT-PCR results themselves serve as the ground truth.

4. Adjudication Method for the Test Set

• Analytical Studies: Not applicable. Ground truth is determined by precise laboratory methods and known concentrations.
• Clinical Study: Not applicable. The comparator RT-PCR assay is a definitive laboratory test; there is no mention of a human-based adjudication process for the RT-PCR results.

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

• No, a MRMC comparative effectiveness study was not done.
• This device is an automated, instrument-read diagnostic assay (Sofia 2 SARS Antigen+ FIA). The results are generated by the instrument (Sofia 2 Analyzer) based on fluorescent signals and programmed algorithms, not by human interpretation of images or signals. Therefore, a human-in-the-loop study to assess how human readers improve with AI assistance is not relevant to this type of device. The study evaluates the standalone performance of the device compared to a reference method (RT-PCR).

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

• Yes, the performance data presented for the Sofia 2 SARS Antigen+ FIA are standalone performance, i.e., algorithm only without human-in-the-loop performance.
• The "Sofia 2 analyzers automatically scan/image the test strip, collect and analyze the fluorescence data, and then calculate and report the result as either positive, negative, or invalid." The operator only uses the analyzer to obtain the result; they do not visually interpret the test or make diagnostic decisions based on human observation of the test strip.

7. Type of Ground Truth Used

• Analytical Studies: The ground truth was based on known concentrations of heat-inactivated SARS-CoV-2 or other microorganisms/substances in controlled laboratory settings. This is a form of "spiked sample" ground truth.
• Clinical Study: The ground truth was established using an "highly sensitive Emergency Use Authorization (EUA) authorized RT-PCR comparator assay." This is a laboratory-based molecular diagnostic method, considered the gold standard for detecting SARS-CoV-2 RNA.

8. Sample Size for the Training Set

• No information is provided about a specific "training set" related to an AI/algorithm development in the context of machine learning. The algorithms referenced ("method-specific algorithms," "software specific cutoff," "specific algorithms") for the Sofia 2 analyzer relate to processing fluorescent signals and determining thresholds for positive/negative results, which are typically derived from extensive analytical characterization and optimization using laboratory-controlled samples, rather than a distinct "training set" in the machine learning sense often seen with image-based AI. The document implies these algorithms are developed and refined through the analytical performance studies (such as LoD, linearity, etc.) using various known concentrations.

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

• As mentioned above, a traditional "training set" as understood in a machine learning context is not explicitly described. The "ground truth" for establishing the device's operational algorithms and cut-offs would have been determined through laboratory experiments with precisely known concentrations of analytes. This process involves:
  • Identifying the Limit of Detection (LoD).
  • Evaluating signal response across a range of concentrations.
  • Using reference lots and known values during development to establish parameters for calculations and cut-offs.
  • "Final cut-off values were further validated as part of the analytical and clinical studies." This suggests an iterative process of development, testing, and validation against a known ground truth (spiked samples for analytical studies, RT-PCR for clinical).

Ask a specific question about this device

The Lyra Influenza A+B Assay is a multiplex Real Time RT-PCR assay for the in vitro qualitative detection and differentiation of influenza A and influenza B viral RNA in nasal and nasopharyngeal swabs from patients with signs and symptoms of respiratory infection. This test is intended for use as an aid in the differential diagnosis of influenza A and influenza B viral infections in humans in conjunction with clinical and endemiological risk factors. The assay does not detect the presence of influenza C virus.

Negative results do not preclude influenza virus infection and should not be used as the sole basis for diagnosis, treatment or other patient management decisions.

Performance characteristics for influenza A were established during the 2013 influenza seasons when influenza A/H3 and 2009 H1N1 influenza were the predominant influenza A viruses in circulation. When other influenza A viruses are emerging, performance characteristics may vary.

If infection with a novel influenza A virus is suspected based on current clinical and epidemiological screening criteria recommended by public health authorities, specimens should be collected with appropriate infection control precautions for novel virulent Influenza viruses and sent to state or local health department for testing. Viral culture should not be attempted in these cases unless a BSL 3+ facility is available to receive and culture specimens.

The assay can be performed using either the Life Technologies QuantStudio Dx, the Applied Biosystems 7500 Fast Dx or the Cepheid SmartCycler II.

The Lyra Influenza A+B Assay detects viral RNA that have been extracted from a patient sample using the NucliSENS easyMAG or EMAG automated extraction platform. A multiplex RT-PCR is carried out under optimized conditions in a single tube generating amplicons for each of the target viruses present in the sample. This reaction is performed utilizing either the Life Technologies QuantStudio™ Dx, the Applied Biosystems® 7500 Fast Dx, or the Cepheid® SmartCycler® II. Identification of influenza A occurs by the use of target specific primers and a fluorescent-labeled probe that hybridizes to a conserved influenza A sequence within the matrix protein gene. Identification of influenza B occurs by the use of target specific primers and fluorescent-labeled probes that will hybridize to a conserved influenza B sequence within the neuraminidase gene.

The provided text describes a 510(k) premarket notification for the Lyra Influenza A+B Assay. The notification primarily focuses on a modification to the device, specifically the inclusion of a new nucleic acid extraction platform (BioMerieux NucliSENS EMAG) while maintaining the original intended use and other core functionalities.

Here's an analysis of the acceptance criteria and study information based on the provided text:

1. Table of Acceptance Criteria and Reported Device Performance

The document explicitly states that "All verification and validation activities were performed in accordance with relevant standards, established plans, protocols, and Design Control procedures. Testing verified all acceptance criteria were met." However, the specific quantitative acceptance criteria (e.g., sensitivity, specificity thresholds) are not detailed in the provided text. Similarly, the specific quantitative reported device performance metrics (e.g., exact sensitivity and specificity values) from the studies are not presented in the provided summary.

The summary only states: "Non-clinical and clinical verification activities conducted with the Lyra Influenza A+B Assay demonstrate that the modified device met predetermined acceptance criteria, supporting equivalency of the modified device to the cleared device."

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

The text mentions a "Clinical Equivalency Study" but does not provide details on the sample size used for its test set or the data provenance (e.g., country of origin, retrospective/prospective nature).

3. Number of Experts and Qualifications for Ground Truth

The document does not provide information on the number of experts used to establish ground truth for the test set or their qualifications.

4. Adjudication Method for the Test Set

The document does not provide information on the adjudication method used for the test set.

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

The text does not indicate that a multi-reader multi-case (MRMC) comparative effectiveness study was done. The focus is on the performance of the assay itself, not human reader improvement with/without AI assistance.

6. Standalone Performance Study

The studies mentioned ("Limit of Detection Equivalency Study" and "Clinical Equivalency Study") describe the performance of the device (algorithm/assay only), implying a standalone performance evaluation. However, the text does not explicitly use the term "standalone" or specify that it was "algorithm only without human-in-the-loop performance." Given that it's an RT-PCR assay, its performance by definition is standalone.

7. Type of Ground Truth Used

The text does not explicitly state the type of ground truth used for the clinical equivalency study. For diagnostic assays like this, ground truth is typically established through a combination of:

• Confirmatory laboratory methods (e.g., viral culture, another highly sensitive and specific PCR method, or sequencing as a gold standard).
• Clinical diagnosis by a physician.

However, this information is not provided.

8. Sample Size for the Training Set

The text does not provide information on a training set sample size. This is a modification to an existing assay, and the studies mentioned are verification/validation studies for the modification, not development studies for a new algorithm that would typically involve a separate training set.

9. How Ground Truth for the Training Set was Established

Since no training set information is provided, there is no information on how its ground truth was established.

Summary of Missing Information:

A significant amount of detail regarding the studies, particularly the quantitative acceptance criteria, reported performance, sample sizes, and ground truth methodologies, is not present in the provided FDA 510(k) summary letter. The letter serves as an approval notification and summary of the device's substantial equivalence, focusing on the change (new extraction platform) rather than a comprehensive, detailed clinical study report.

Ask a specific question about this device

Sofia 2 Campylobacter FIA employs immunofluorescence for the rapid qualitative detection of a Campylobacter-specific antigen in human fecal specimens. Sofia 2 Campylobacter FIA is designed to detect C. jejuni, C. coli, C. lari and C. upsalients with signs and symptoms of gastroenteritis. The test is intended for use with preserved fecal specimens in transport media and unpreserved fecal specimens. Test results should be considered in conjunction with clinical findings and patient history.

The Sofia 2 Campylobacter FIA employs immunofluorescence technology that is used with Sofia 2 for the rapid qualitative detection of Campylobacter jejuni, Campylobacter lari, and Campylobacter upsaliensis specific antigens in fecal samples.

The patient's sample is placed in the Specimen Tube containing the Specimen Solution to dilute, making the antigenic components more accessible to the specific antibodies. An aliquot of the diluted sample is dispensed through a filter to remove particulates, making them more compatible for testing, into the Test Cassette sample well. From the sample well, the sample migrates through a test strip containing various unique chemical environments. If Campylobacter jejuni, Campylobacter coli, Campylobacter lari, or Campylobacter upsaliensis specific antigens are present, they will be bound by antibodies coupled to fluorescent microparticles that migrate through the test strip. The fluorescent microparticles containing bound proteins will be captured by antibodies at a defined location on the test strip where they are detected by Sofia 2. If Campylobacter jejuni, Campylobacter coli, Campylobacter lari, or Campylobacter upsaliensis specific antigens are not present, the fluorescent microparticles will not be trapped by the capture antibodies nor detected by Sofia 2.

The Test Cassette is placed inside of Sofia 2 for automatically timed development (WALK AWAY Mode), or pre-incubated on the bench top prior to loading into Sofia 2 (READ NOW Mode), where Sofia 2 will scan, measure, and interpret the immunofluorescent signal using method-specific algorithms. Sofia 2 will display the test results (Positive, Negative, or Invalid) on the screen.

The fluorescence signal obtained with this assay is invisible to the unaided eye. The test results can only be obtained with the proper use of Sofia 2.

The Sofia 2 Campylobacter FIA is a device for the rapid qualitative detection of Campylobacter-specific antigens. It is designed to detect C. jejuni, C. coli, C. lari, and C. upsaliensis from patients with signs and symptoms of gastroenteritis in human fecal specimens (preserved or unpreserved).

Here's an analysis of the acceptance criteria and the study data provided:

1. Table of Acceptance Criteria and Reported Device Performance

Based on the provided information, the key performance metrics are Sensitivity and Specificity from the Prospective Clinical Study.

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

• Test Set Sample Size for Clinical Study: 811 total specimens.
  • 191 fresh, neat specimens
  • 620 fresh specimens in transport media
• Data Provenance: Prospective Clinical Study (likely multi-center, though specific countries are not mentioned). All specimens were "fresh".
• Test Set Sample Size for Archived Study: 70 frozen, characterized specimens.
  • 35 culture-negative specimens preserved in transport media.
  • 35 positive specimens (11 in transport media, 24 neat fecal specimens).
• Data Provenance for Archived Study: Archived, frozen specimens, characterized at a central laboratory.

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 of experts or their qualifications for establishing the ground truth.

4. Adjudication method for the test set

• For the Prospective Clinical Study: The primary comparison was against culture and identification. For discordant specimens (Sofia Positive/Culture Negative), species-specific RT-PCR and bi-directional sequence analysis were used for confirmation. In this case, 6 discordant specimens were identified as positive for Campylobacter (3 C. jejuni, 2 C. upsaliensis, and 1 C. coli) by this additional molecular testing. This suggests a form of discordant analysis/adjudication.
• For the Archived Clinical Study: Ground truth was established by culture and further characterized by RT-PCR and bi-directional sequencing. All 35 positive specimens were confirmed by all methods, and all 35 negative specimens yielded 100% correlation with all test methods.

5. If a multi reader multi case (MRMC) comparative effectiveness study was done, If so, what was the effect size of how much human readers improve with AI vs without AI assistance

This is not applicable as the Sofia 2 Campylobacter FIA is an automated IVD device. It does not involve human readers interpreting results with or without AI assistance in the context of diagnostic imaging, for example. The Sofia 2 instrument scans, measures, and interprets the immunofluorescent signal using method-specific algorithms and displays the result.

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

Yes, the device operates in a standalone manner. The Sofia 2 instrument itself provides the final "Positive, Negative, or Invalid" result after processing the test cassette. This performance directly reflects the algorithm's output. The performance data (Sensitivity, Specificity) from the clinical studies represents this standalone performance.

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

• Prospective Clinical Study: Culture and identification was the primary reference method. For discordant results (Sofia Positive/Culture Negative), species-specific RT-PCR and bi-directional sequence analysis were used as a confirmatory ground truth.
• Archived Clinical Study: Culture for initial characterization, and further confirmed by RT-PCR and bi-directional sequencing.

8. The sample size for the training set

The document does not explicitly state the sample size used for the training set of the device's algorithms. The provided data focuses on the validation of the finalized device.

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

The document does not provide details on how the ground truth for the training set was established. It focuses on the analytical and clinical validation of the device.

Ask a specific question about this device

The Sofia 2 Lyme FIA employs immunofluorescence for the rapid differential detection of human IgM and IgG antibodies to Borrelia burgdorferi from finger-stick whole blood specimens from patients suspected of B. burgdorferi infection of at least 2 weeks' duration. The Sofia 2 Lyme FIA is intended for use as an aid in diagnosis of Lyme disease. A negative result does not preclude infection with B. burgdorferi. Positive results must be confirmed by testing with a corresponding second-tier B. burgdorferi Western blot assay. Test results are to be used in conjunction with information obtained from the patient's clinical evaluation and other diagnostic procedures. The assay is to be performed on the Sofia 2 instrument. Professional guidelines should be consulted regarding testing and treatment for Lyme disease when acute B. burgdorferi infection is suspected.

The Sofia Lyme Control Set is intended for use as assayed quality control materials to verify the performance of the Sofia Lyme FIA and Sofia 2 Lyme FIA test system.

The Sofia 2 Lyme FIA is an immunofluorescence-based, lateral flow assay for detection of IgM and/or IgG antibodies to Borrelia burgdorferi in patient specimens. Reagents for the assay are ready-to-use and provided in the kit. The assay uses a bidirectional test strip format to detect both IgM and IgG antibodies to B. burgdorferi. One side of the test strip detects IgM antibodies to B. burgdorferi and the other side of the test strip detects IgG antibodies to B. burgdorferi. To perform the test, the patient finger-stick whole blood specimen is obtained with the provided Capillary Tube (a.k.a. whole blood separator device). The Capillary Tube stands or is held vertically to allow the blood to drain. This device separates the red blood cells from the whole blood specimen using a gravimetric flow through a sample pad coated with rabbit anti-human red blood cell (RBC) antibodies. The user dispenses all of the Reagent Solution into the Reagent Tube and inserts the Capillary Tube into the Reagent Tube and shakes the tube vigorously. Two drops of diluted sample are dispensed into the round sample well located near the center of the Test Cassette. The Test Cassette is loaded into Sofia 2 in either the READ NOW Mode or WALK AWAY Mode. In READ NOW Mode, the user allows the cassette to develop on the countertop for 15 minutes. In WALK AWAY Mode, the user immediately after adding the specimen to the cassette, the cassette is inserted into Sofia 2. Sofia 2 will analyze the test strip at 3, 5, 8, 10, and 15 minutes until both IgM and IgG positive results are received. This feature allows for earlier read times. Each Sofia 2 Lyme FIA kit will contain one Positive and one Negative Control—each provided in separate dropper bottles. The external controls will be provided separately as well in a Sofia Lyme Control Set. The Positive and Negative QC controls are formulated with patient Lyme IgM and IgG positive serum that are diluted into 1X PBS and 0.3% Microcide is added to the solution as an antimicrobial. The External Controls will be tested by adding 2 drops to the test cassette.

Here's a breakdown of the acceptance criteria and study information for the Sofia 2 Lyme FIA device, based on the provided text:

1. Table of Acceptance Criteria and Reported Device Performance

The document doesn't explicitly state "acceptance criteria" in a single table, but rather presents performance data that would be used to demonstrate these criteria. Based on the "Substantial Equivalence Information" and "Performance Data" sections, the implied acceptance criteria are comparable performance to the predicate devices (Vidas Lyme IgM and Vidas Lyme IgG) in terms of precision, reproducibility, analytical specificity, clinical sensitivity, and method comparison (PPA and NPA).

Since specific numerical acceptance thresholds are not provided, the table below summarizes the reported performance values that are implicitly demonstrating compliance with "good" or "comparable" performance.

Notes on Inconsistencies: There is a minor inconsistency in the "Precision – Within Run" table for IgG Low Positive (C95) results. The "Run 2" column shows 46/48 (95.8%), while the "Total (n=96)" shows 100%. This might be a typo in the table, or the "Total" is derived differently.

2. Sample Size and Data Provenance

• Precision and Reproducibility:

  • Sample Size: Contrived samples. For precision, 96 tests per sample level (48 in Run 1, 48 in Run 2). For reproducibility, 30 tests per sample level per site (total 90 tests per sample level across 3 sites).
  • Data Provenance: Not explicitly stated, but "Contrived samples were prepared" suggests laboratory-generated, possibly in the US (Quidel site).

• Matrix Equivalency:

  • Sample Size: 321 patients
  • Data Provenance: "A field study was conducted." Location not specified, but likely within the US.

• CDC Reference Panel:

  • Sample Size: IgM: 190 Negative Controls, 60 Early Lyme EM Positive, 30 Late Lyme. IgG: 190 Negative Controls, 60 Early Lyme EM Positive, 30 Late Lyme.
  • Data Provenance: CDC Reference Panel (implies US government source).

• Specificity Study:

  • Sample Size: 200 samples (100 endemic, 100 non-endemic).
  • Data Provenance: Samples "obtained from asymptomatic (healthy, normal) populations in both endemic and non-endemic regions." Regions are not specified.

• Sensitivity Study:

  • Sample Size: 95 "well-characterized clinically or culture confirmed Lyme disease samples."
  • Data Provenance: Not explicitly stated, but described as "well-characterized clinically or culture confirmed Lyme disease samples."

• Method Comparison (Clinical Study):

  • Sample Size: Total 326 samples for IgM (62 Positive, 29 Equivocal, 233 Negative by Predicate). Total 324 samples for IgG (54 Positive, 270 Negative by Predicate).
  • Data Provenance: "prospectively collected specimens from subjects suspected of having Lyme disease." Country of origin not specified, but typically US for FDA 510(k) submissions unless stated otherwise.

3. Number of Experts and Qualifications for Ground Truth

• The document does not specify the number of experts used or their qualifications for establishing ground truth for the test set samples in any of the clinical or specificity/sensitivity studies.
• For the CDC Reference Panel, the panel itself implies expert consensus or well-established diagnostic criteria as its ground truth, but the individual experts involved are not mentioned.
• For the Sensitivity Study, "well-characterized clinically or culture confirmed Lyme disease samples" implies diagnosis by medical professionals following established guidelines, but specific expert details are absent.

4. Adjudication Method for the Test Set

• The document does not explicitly mention any adjudication method (e.g., 2+1, 3+1) for resolving discrepancies in the test set's ground truth determination. Ground truth is inferred primarily from clinical diagnosis, culture confirmation, or established reference panels/predicate assays.

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

• No, a MRMC comparative effectiveness study was not explicitly described. The studies focus on the analytical and clinical performance of the device itself, and its comparison to predicate devices, rather than a study evaluating improvement in human reader performance with or without AI assistance. The device is an automated immunofluorescence assay (FIA), not an AI-assisted diagnostic imaging or interpretation tool.

6. Standalone Performance

• Yes, standalone performance was done. The entire "Performance Data" section (sections a-j) describes the performance of the Sofia 2 Lyme FIA algorithm/device in a standalone capacity, without human interpretation of the results to establish the initial diagnosis. The device's output (positive/negative) is directly compared to various ground truth standards.

7. Type of Ground Truth Used

The types of ground truth used vary by study:

• Precision/Reproducibility: Contrived samples (known concentrations/status).
• Matrix Equivalency: Clinical diagnosis based on patient outcome or other diagnostic procedures (implied by comparing finger-stick whole blood to serum/plasma, which are standard clinical samples).
• CDC Reference Panel: "Clinical Status" (established by CDC, likely involving clinical diagnosis and confirmed lab results) and "Western Blot" (a gold standard serological confirmation for Lyme disease).
• Specificity Study: Asymptomatic/healthy status from endemic and non-endemic populations.
• Sensitivity Study: "Well-characterized clinically or culture confirmed Lyme disease samples."
• Method Comparison: Predicate assays (Vidas Lyme IgM and IgG) as a benchmark, and second-tier Western Blot for confirmation of first-tier positive or equivocal results.

8. Sample Size for the Training Set

• The document does not explicitly state the sample size used for a "training set." This device is an immunoassay, not a machine learning or AI model in the typical sense that requires a training set of images or data for algorithm development. The "Assay Cutoff" study mentions establishing cutoffs, which is an optimization process, but not a "training set" in the context of deep learning.

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

• Given that a specific "training set" for an AI/ML model is not described, the method for establishing its ground truth is not applicable/not provided. The "Assay Cutoff" study involved testing "matched finger-stick whole blood, serum and plasma samples with the aim of setting the whole blood assay cutoff so that the clinical performance... is statistically similar to the FDA cleared Sofia Lyme FIA serum and plasma assay." This suggests that the ground truth for establishing cutoffs was based on a combination of clinical performance (likely against a known diagnosis or reference method) and comparison to an already cleared device.

Ask a specific question about this device

Ask a specific question about this device

The Sofia Lyme FIA employs immunofluorescence for the rapid differential detection of human IgM and IgG antibodies to Borrelia burgdorferi from serum and plasma specimens from patients suspected of B. burgdorferi infection. This qualitative test is intended for use as an aid in the diagnosis of Lyme disease. A negative result does not preclude infection with B. burgdorferi. All positive results for IgM and/or IgG should be further tested by a corresponding second-tier western blot assay. Test results are to be used in conjunction with information obtained from the patient's clinical evaluation and other diagnostic procedures.

The Sofia Lyme FIA may be used with Sofia or Sofia 2.

The Sofia Lyme FIA is an immunofluorescence-based, lateral flow assay for detection of IgM and/or IgG antibodies to Borrelia burgdorferi in patient specimens. Reagents for the assay are ready-to-use and provided in the kit.

The assay uses a bidirectional test strip format to detect both IgM and IgG antibodies to B. burgdorferi. One side of the test strip detects IgM antibodies to B. burgdorferi and the other side of the test strip detects IgG antibodies to B. burgdorferi.

To perform the test, a patient serum or plasma specimen is obtained and added to a pre-filled vial containing the lyme running buffer. The diluted sample is then pipetted into the round sample port in the center of the Test Cassette.

The Test Cassette is loaded into Sofia 2 in either the READ NOW Mode or WALK AWAY Mode. In READ NOW Mode, the user allows the cassette to develop on the countertop for 10 minutes. In WALK AWAY Mode, the user immediately after adding the specimen to the cassette is inserted into Sofia 2. Sofia 2 will analyze the test strip at 3, 5, 8, and 10 minutes until both IgM and IgG positive results are received. This feature allows for earlier read times.

Each Sofia Lyme FIA kit will contain one Positive and one Negative Control-each provided in separate dropper bottles. The Positive QC control is formulated with patient Lyme IgM and IgG positive plasma diluted into 1X PBS, and 0.3% Microcide is added to the solution as an antimicrobial. The Negative QC control is formulated with patient negative serum diluted into 1X PBS and 0.3% Microcide is added to the solution as an antimicrobial. External Controls will be tested by adding 2 drops to the test cassette.

Here's an analysis of the provided text to extract the acceptance criteria and study details for the Sofia Lyme FIA device.

Note: The provided document is a 510(k) summary for a medical device. It focuses on demonstrating substantial equivalence to a predicate device, rather than proving that the device meets specific acceptance criteria in the way one might see for a novel AI device with clearly defined performance targets post-development. The performance data section describes the types of studies performed and their general positive outcomes ("good performance," "very good," "comparable performance"), but it does not specify quantitative acceptance criteria or report specific numbers for sensitivity, specificity, or predictive values from a test set against a pre-defined ground truth. The acceptance criteria referred to here are implicitly those required to demonstrate substantial equivalence for regulatory approval, which are less stringent than the "acceptance criteria" usually found in the context of AI/ML model validation against specific quantitative metrics.

Therefore, the answers below are based on the available information, highlighting what is present and noting what is not explicitly stated in this type of regulatory document.

Acceptance Criteria and Device Performance

Based on the nature of a 510(k) submission, the "acceptance criteria" are broad and relate to demonstrating substantial equivalence to a legally marketed predicate device. The performance data presented focuses on analytical and clinical performance attributes that support this claim. Specific quantitative acceptance values are not explicitly listed in the summary, but rather implied by the successful completion of the studies and the conclusion of substantial equivalence.

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

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

• Test Set Sample Size: Not explicitly stated in the summary. The "Method Comparison" mentions "prospectively collected specimens from subjects suspected of having Lyme disease," but does not provide a number. "Specificity Study" used "samples obtained from asymptomatic (healthy, normal) populations," also without a specified sample size.
• Data Provenance:
  • Country of Origin: Not specified.
  • Retrospective or Prospective: The "Method Comparison" study explicitly states "prospectively collected specimens." The "Specificity Study" refers to "samples obtained from asymptomatic (healthy, normal) populations," likely indicating prospective collection for that purpose.

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

• This device is an in vitro diagnostic (IVD) test, which directly measures biomarkers (antibodies) in patient samples. The concept of "experts" establishing ground truth for individual cases, as in interpreting medical images, does not directly apply here. The performance is evaluated against established methods (the predicate device) or by spiking samples with known concentrations.
• The "ground truth" for evaluating an IVD is typically the true antibody status (positive/negative) determined by reference methods (like Western Blot as mentioned for confirmatory testing) or by known sample characteristics (e.g., healthy controls, defined positive controls). The summary alludes to results being used "in conjunction with information obtained from the patient's clinical evaluation and other diagnostic procedures," implying clinical diagnosis and confirmatory tests serve as the ultimate clinical ground truth, but this wasn't directly adjudicated by "experts" for the test set in the same way a radiologist would read an X-ray.

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

• Not applicable in the context of this IVD device. Adjudication methods like 2+1 or 3+1 are typically used for subjective assessments (e.g., image interpretation) where human readers establish ground truth. For an IVD, the "ground truth" is derived from a reference laboratory test or clinical definition, not subjective expert consensus on the diagnostic output itself.

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 study was not done. This device is not an AI-assisted diagnostic tool that supports human readers in interpretation. It's an automated immunoassay.

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

• This device is inherently "standalone" in its measurement. It's an automated fluorescent immunoassay (FIA) system where the Sofia 2 analyzer performs the reading and provides a result. There isn't a separate "algorithm" being validated outside of the full Sofia Lyme FIA-Sofia 2 system. The performance studies described (analytical, precision, specificity, method comparison, reproducibility) are all evaluating the performance of this system as a whole.

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

• For the Method Comparison study, the ground truth was likely established by the comparative performance against the predicate devices (Vidas Lyme IgG and Vidas Lyme IgM tests), which are themselves established diagnostic tests.
• For the Specificity Study, the ground truth for these samples was "asymptomatic (healthy, normal) populations," implying that the absence of the disease (and thus, antibodies) served as the ground truth.
• For the overall clinical utility, the "Indications for Use" states that the test results are to be used "in conjunction with information obtained from the patient's clinical evaluation and other diagnostic procedures," and "All positive results for IgM and/or IgG should be further tested by a corresponding second-tier western blot assay." This implies that confirmatory testing (like Western Blot), along with clinical evaluation, contribute to the definitive diagnosis or "ground truth" in clinical practice. The studies themselves don't explicitly detail how definitive ground truth for each case was established beyond comparison with predicate or known healthy status.

8. The sample size for the training set

• This information is not provided because the Sofia Lyme FIA is not an AI/ML device that requires a separate "training set" for model development. It's a chemical immunoassay, and its "development" involves assay optimization and validation rather than model training.

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

• Not applicable, as there is no "training set" in the AI/ML sense for this device.

Ask a specific question about this device