Cobas Influenza A/B Nucleic Acid Test for Use on the Cobas Liat System

• K Number: K191729
• Device Name: Cobas Influenza A/B Nucleic Acid Test for Use on the Cobas Liat System
• Manufacturer: Roche Molecular Systems, Inc.
• Date Cleared: 2019-07-24 ( 27 days)
• Product Code: OCC , OOI
• Regulation Number: 866.3980
• Type: Special
• Panel: MI
• Reference & Predicate Devices: K111386 , K111387

Intended Use

The cobas® Influenza A/B Nucleic acid test for use on the cobas® Liat® System (cobas® Influenza A/B) is an automated multiplex real-time RT-PCR assay for the rapid in vitro qualitative detection and discrimination of Influenza A virus and Influenza B virus RNA in nasopharyngeal swab specimens from patients with signs and symptoms of respiratory infection in conjunction with clinical and evidemiological risk factors. The test is intended for use as an aid in the differential diagnosis of Influenza A and Influenza B in humans and is not intended to detect Influenza C.

Negative results do not preclude Influenza virus infection and should not be used as the sole basis for treatment or other patient management decisions. Conversely, positive results do not rule out bacterial infection or co-infection with other viruses. The agent detected may not be the definite cause of disease.

Performance characteristics for Influenza A were established when Influenza A/H1 and A/H3 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 BSL3+ facility is available to receive and culture specimens.

Device Description

The cobas® Influenza A/B assay is a rapid, automated in vitro diagnostic test for qualitative detection and differentiation of Influenza type A and type B viral RNA. The assay is performed on the cobas® Liat® System. The system automates and integrates sample purification, nucleic acid amplification, and detection of the target sequence in biological samples using real-time RT-PCR assays. The cobas® Liat® Analyzer consists of an instrument and preloaded software for running tests and viewing the results. The cobas® Liat® System consists of the analyzer and a single-use disposable cobas® Influenza A/B assay tube that holds the sample purification and RT-PCR reagents and hosts the sample preparation and RT-PCR processes. Other than adding the sample to the cobas® Influenza A/B assay tube, no reagent preparation or additional steps are required. Because each cobas® Influenza A/B assay tube is self-contained, cross-contamination between samples is minimized. Turnaround time for a test is 20 minutes.

The cobas® Influenza A/B assay includes reagents for the detection and differentiation of Influenza A and B viral RNA in nasopharyngeal swab (NPS) specimens in universal transport media (UTM) from patients suspected of having Influenza. The assay targets a well-conserved region of the matrix gene of Influenza A viral RNA (Inf A target) and non-structural protein (NS) gene of Influenza B (Inf B target). An Internal Process Control (IPC) is also included. The IPC is present to control for adequate processing of the target viruses through all steps of the assay process and to monitor the presence of inhibitors in the RT-PCR reactions.

The cobas® Influenza A/B assay tube uses a flexible tube as a sample processing vessel. It contains all requisite PCR reagents pre-packed in assay tube segments separated by breakable seals. When a cobas® Influenza A/B assay tube containing a raw biological sample is inserted into the cobas Liat® Analyzer. multiple sample processing actuators in the cobas® Liat Analyzer compress the cobas® Influenza A/B assay tube to selectively release the reagents, moving the sample from one segment to the next, and controlling reaction conditions. An embedded microprocessor controls and coordinates these actions to perform all required assay processes, including sample preparation, nucleic acid extraction, target concentration enrichment, inhibitor removal, nucleic acid elution, and real-time PCR. All assay steps are performed within the closed and self-contained cobas® Influenza A/B assay tube, minimizing cross-contamination between samples.

The detection module monitors the reaction in real-time, while an on-board computer analyzes the collected data and outputs an interpreted result. The latter is displayed in the assay report on the integrated LCD touch screen of the cobas® Liat® Analyzer and in an electronic file. The report can be printed directly through a USB or network-connected printer. The results can also be exported to an external server, middleware or data management system, or to a Laboratory Information System (LIS).

AI/ML Overview

The provided document does not contain information related to an AI/ML-based device; instead, it describes a nucleic acid amplification test (NAT) for the detection of Influenza A/B. Therefore, many of the requested details, such as MRMC studies, training set specifics, and expert adjudication, which are pertinent to AI/ML device evaluations, are not applicable to this submission.

However, I can extract the relevant information regarding the device's performance evaluation and acceptance criteria for this specific type of diagnostic test.

Acceptance Criteria and Reported Device Performance

The document describes the cobas® Influenza A/B Nucleic acid test for use on the cobas® Liat® System. This is a real-time RT-PCR assay for the qualitative detection and differentiation of Influenza A and B viral RNA. The evaluation focuses on demonstrating that the performance of the updated software version (FABA v1.35) is equivalent to the previously cleared version (FABA v1.31) and does not negatively impact the assay's existing performance claims.

Since this is a diagnostic test and not an AI/ML device, the performance is primarily evaluated in terms of its analytical and clinical performance as a diagnostic test, rather than typical AI metrics like AUC, sensitivity, or specificity for an AI model. The document states that the overall cobas® Influenza A/B assay performance and claims were not impacted by changes implemented in FABA v1.35.

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

Criteria Type	Specific Acceptance Focus (Implied for Software Update)	Reported Device Performance (with FABA v1.35)
Result Interpretation Logic	Updates to prevent erroneous "Invalid" results and false positives.	The updated Result Interpretation Concept logic, new checks, and cut-offs were implemented to address results erroneously reported as Invalid and prevent False Positives. The evaluation determined that the overall assay performance and claims were not impacted by these changes. Bug fixes were also implemented.
Overall Assay Performance Quality	Maintain equivalent performance to the predicate device (FABA v1.31).	"The result of this evaluation determined that the overall cobas® Influenza A/B assay performance and claims were not impacted by changes implemented in FABA v1.35, when compared to the current commercially available FABA script version."

"Equivalent performance of the modified device and the current commercial device has been demonstrated, and analytical or clinical performance has not changed." |
| Regulatory Equivalence | Substantial equivalence to the predicate device (K111387 and CW150003). | The modified device is substantially equivalent to the predicate device, as originally cleared through K111386 and CLIA waived through CW150003. |

Important Note: The document focuses on demonstrating that the software update (FABA v1.35) does not degrade the already established performance of the cobas® Influenza A/B assay. It does not re-establish the primary performance characteristics (e.g., clinical sensitivity, specificity) from scratch, but rather ensures the new software maintains the established performance. Therefore, detailed performance metrics (like percentages for sensitivity/specificity) for the entire assay are not reiterated in this summary specifically for the software update, as they were presumably part of the original clearance.

Study Details for the Software Update (FABA v1.35)

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

   • The document states that the performance was evaluated using "data from submissions (K111386, CW150003), internal studies, release testing, and the field."
   • Specific sample sizes for these evaluations are not provided in this summary.
   • Data provenance (e.g., country of origin, retrospective/prospective) is not specified for these combined data sources, beyond "internal studies, release testing, and the field." The reference to previous FDA clearances (K111386, CW150003) implies that some data would have been derived from clinical samples used in those original studies, which are typically prospective for diagnostic assay clearances.

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

   • This is a diagnostic assay, and ground truth for such tests is typically established through a combination of reference methods (e.g., viral culture, another cleared PCR assay, or a composite reference standard) rather than human expert consensus on images.
   • Therefore, the concept of "experts" in the context of radiologists or similar interpreters for AI/ML models is not applicable here. The ground truth for influenza detection would be based on the established gold standard for virological diagnosis.

3. Adjudication Method for the Test Set:

   • Not applicable in the context of an AI/ML model's output being adjudicated by human experts. For a diagnostic assay, discrepant results between the investigational device and the reference method would typically undergo further investigation (e.g., retesting, re-extraction), but not "adjudication" by multiple human readers of an output.

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

   • No. This is a diagnostic RT-PCR assay, not an AI/ML system that assists human readers in interpreting medical images or data. Therefore, an MRMC study comparing human readers with and without AI assistance is not relevant or performed for this type of device.

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

   • In a sense, yes, the "algorithm" (the RT-PCR chemistry and instrument's software interpretation) is standalone in that it provides a direct qualitative result (Influenza A detected, Influenza B detected, Negative, Invalid) without requiring human interpretation of raw signals beyond basic quality control. The human "in the loop" is the laboratory technologist performing the test and reviewing the automated result, but not interpreting nuanced images or complex data as in AI applications. The performance evaluation therefore focuses on the accuracy and reliability of these automated results.

6. The Type of Ground Truth Used:

   • The document does not explicitly state the ground truth methodology for the software update evaluation, but for the initial assay clearances (K111386, CW150003) mentioned, the ground truth for an influenza diagnostic test would typically be established by:
     • Composite Reference Standard (CRS): Often involves multiple methods, such as viral culture, another FDA-cleared molecular assay, and/or clinical presentation.
     • Clinical Diagnosis: Corroborated by symptoms and potentially other lab findings.
     • Reference Laboratory PCR: Testing using highly sensitive and specific laboratory-developed or research-use only PCR tests considered to be expert-level.
   • The critical aspect of this submission is demonstrating that the changes in the software (FABA v1.35) do not alter the established performance characteristics, meaning the ground truth from prior studies remained valid against the updated software's performance.

7. The Sample Size for the Training Set:

   • Not applicable in the context of an AI/ML training set. This is a rule-based diagnostic algorithm for PCR analysis, not a machine learning model that requires explicit training data in the same way. The software changes refer to updates in logic and bug fixes, developed based on internal R&D processes and prior performance data, not a "training set" for model parameters.

8. 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 type of device. The logic and cut-offs for the RT-PCR assay are based on established molecular biology principles and analytical validation, refined and verified through various studies to define detection thresholds and interpretation rules.

Regulation Number and Section

§ 866.3980 Respiratory viral panel multiplex nucleic acid assay.

(a)
Identification. A respiratory viral panel multiplex nucleic acid assay is a qualitative in vitro diagnostic device intended to simultaneously detect and identify multiple viral nucleic acids extracted from human respiratory specimens or viral culture. The detection and identification of a specific viral nucleic acid from individuals exhibiting signs and symptoms of respiratory infection aids in the diagnosis of respiratory viral infection when used in conjunction with other clinical and laboratory findings. The device is intended for detection and identification of a combination of the following viruses:(1) Influenza A and Influenza B;
(2) Influenza A subtype H1 and Influenza A subtype H3;
(3) Respiratory Syncytial Virus subtype A and Respiratory Syncytial Virus subtype B;
(4) Parainfluenza 1, Parainfluenza 2, and Parainfluenza 3 virus;
(5) Human Metapneumovirus;
(6) Rhinovirus; and
(7) Adenovirus.
(b)
Classification. Class II (special controls). The special controls are:(1) FDA's guidance document entitled “Class II Special Controls Guidance Document: Respiratory Viral Panel Multiplex Nucleic Acid Assay;”
(2) For a device that detects and identifies Human Metapneumovirus, FDA's guidance document entitled “Class II Special Controls Guidance Document: Testing for Human Metapneumovirus (hMPV) Using Nucleic Acid Assays;” and
(3) For a device that detects and differentiates Influenza A subtype H1 and subtype H3, FDA's guidance document entitled “Class II Special Controls Guidance Document: Testing for Detection and Differentiation of Influenza A Virus Subtypes Using Multiplex Nucleic Acid Assays.” See § 866.1(e) for the availability of these guidance documents.