Search Results

Intended Use:
The ASTar System is intended to be used for the automated quantitative susceptibility testing for most clinically significant microorganisms. The ASTar System does not provide organism identification.

Indications for Use:
The ASTar System, comprised of the ASTar Instrument with the ASTar BC G- Kit (ASTar BC G- Consumable kit, ASTar BC G- Frozen insert, and ASTar BC G-Kit software), utilizes high-speed, time-lapse microscopy imaging of bacteria for the in vitro, quantitative determination of antimicrobial susceptibility of on-panel gram-negative bacteria. The test is performed directly on positive blood culture samples signaled as positive by a continuous monitoring blood culture system and confirmed to contain gram-negative bacilli by Gram stain. Organism identification is required for AST result interpretation and reporting.

Test results from the ASTar BC G- Kit should be interpreted in conjunction with other clinical and laboratory findings. Standard laboratory protocols for processing positive blood cultures should be followed to ensure availability of isolates for supplemental testing. Sub-culturing is necessary to support further testing for: bacteria and antimicrobials not on the ASTar BC G- panel, where inconclusive results are obtained, epidemiologic testing, recovery of organisms present in microbial samples, and susceptibility testing of bacteria in polymicrobial samples.

The ASTar BC G- Kit tests the following antimicrobial agents with the following bacterial species:

Amikacin: Citrobacter freundii, Enterobacter cloacae complex, Klebsiella aerogenes, Klebsiella oxytoca, Klebsiella pneumoniae, Proteus mirabilis, Pseudomonas aeruginosa, Serratia marcescens
Ampicillin: Escherichia coli, Proteus mirabilis
Ampicillin-sulbactam: Escherichia coli, Klebsiella oxytoca, Klebsiella pneumoniae, Proteus mirabilis, Proteus vulgaris
Aztreonam: Citrobacter koseri, Enterobacter cloacae complex, Escherichia coli, Klebsiella aerogenes, Klebsiella oxytoca, Klebsiella pneumoniae, Proteus mirabilis, Proteus vulgaris, Serratia marcescens
Cefazolin: Klebsiella pneumoniae
Cefepime: Citrobacter freundii, Escherichia coli, Klebsiella aerogenes, Klebsiella oxytoca, Klebsiella pneumoniae, Proteus mirabilis, Proteus vulgaris, Pseudomonas aeruginosa, Serratia marcescens
Ceftazidime: Enterobacter cloacae complex, Escherichia oxytoca, Klebsiella pneumoniae, Proteus mirabilis, Proteus vulgaris, Serratia marcescens
Ceftazidime-avibactam: Citrobacter freundii, Citrobacter koseri, Enterobacter cloacae complex, Klebsiella oxytoca, Proteus mirabilis, Pseudomonas aeruginosa, Serratia marcescens
Cefuroxime: Escherichia coli, Klebsiella oxytoca, Klebsiella pneumoniae, Proteus mirabilis
Ciprofloxacin: Citrobacter koseri, Enterobacter cloacae complex, Escherichia coli, Klebsiella aerogenes, Klebsiella oxytoca, Klebsiella pneumoniae, Proteus mirabilis, Proteus vulgaris, Pseudomonas aeruginosa, Serratia marcescens
Gentamicin: Citrobacter freundii, Citrobacter koseri, Klebsiella oxytoca, Klebsiella pneumoniae, Proteus mirabilis, Proteus vulgaris, Pseudomonas aeruginosa, Serratia marcescens
Levofloxacin: Citrobacter freundii, Citrobacter koseri, Enterobacter cloacae complex, Escherichia coli, Klebsiella aerogenes, Klebsiella oxytoca, Klebsiella pneumoniae, Proteus mirabilis, Proteus vulgaris, Pseudomonas aeruginosa, Serratia marcescens
Meropenem: Acinetobacter baumannii, Citrobacter freundii, Citrobacter koseri, Escherichia coli, Proteus mirabilis, Proteus vulgaris, Pseudomonas aeruginosa, Serratia marcescens
Meropenem-vaborbactam: Citrobacter freundii, Citrobacter koseri, Enterobacter cloacae complex, Escherichia coli, Klebsiella aerogenes, Klebsiella pneumoniae, Proteus mirabilis, Serratia marcescens
Piperacillin-tazobactam: Citrobacter koseri, Escherichia coli, Klebsiella pneumoniae, Proteus mirabilis, Proteus vulgaris, Serratia marcescens
Tigecycline: Citrobacter freundii, Citrobacter koseri, Enterobacter cloacae complex, Escherichia coli, Klebsiella aerogenes, Klebsiella oxytoca, Klebsiella pneumoniae, Serratia marcescens
Tobramycin: Citrobacter freundii, Citrobacter koseri, Enterobacter cloacae complex, Escherichia coli, Klebsiella pneumoniae, Proteus mirabilis, Serratia marcescens
Trimethoprim-sulfamethoxazole: Enterobacter cloacae complex, Escherichia coli, Klebsiella aerogenes, Klebsiella oxytoca, Klebsiella pneumoniae, Proteus vulgaris

ASTar System is a fully automated system for antimicrobial susceptibility testing (AST). It consists of the ASTar Instrument which is used in combination with dedicated application kits. The ASTar BC G- Kit consists of the ASTar BC G- Consumable kit, ASTar BC G- Frozen insert, and ASTar BC G-Kit software which must be installed on the instrument to process the kit.

The system provides robust and consistent inoculum preparation for AST and utilizes high-speed, time-lapse microscopy imaging of pathogens in broth microdilution to determine minimum inhibitory concentration (MIC) and qualitative susceptibility results. Organism identification using an approved method is required to be entered into the ASTar Instrument for results to be reported.

The instrument is designed to carry out sample preparation of up to six samples in parallel, using a dedicated ASTar Cartridge consumable for each sample. In the subsequent AST culturing step, the instrument transfers the prepared sample into a second dedicated consumable, referred to as the ASTar Disc. Up to 12 Discs can be incubated simultaneously in the system. The processed samples can be in different stages of the processing protocol. New samples can be loaded in a random-access manner when there are available slots. Processing of loaded samples will, in most cases, start shortly after loading. If six samples are started at the same time limitations given by the sample scheduler will result in a queue. The operator interacts with the instrument via the touchscreen display by which the operator controls the instrument.

ASTar BC G- Kit is used for in vitro determination of antimicrobial susceptibility testing of commonly isolated bacteria derived from positive blood culture samples confirmed positive for Gram-negative bacteria by Gram stain. The antimicrobial and organism combinations are listed in Table 1. Reportable ranges for each antimicrobial are listed in Table 2.

To start an analysis approximately 1 mL of a positive blood culture, confirmed Gram-negative by Gram stain is pipetted into the ASTar Cartridge by the operator and loaded into the system, from which the system purifies and quantifies the bacterial concentration is adjusted to the appropriate inoculum concentration and produces an inoculum for analysis of non-fastidious organisms. The bacterial suspensions are transferred automatically to the ASTar Disc and antimicrobial susceptibility testing is performed based on a defined short-term protocol. Results are available within approximately six hours. Bacterial growth and response to relevant concentrations of different antimicrobial drugs are measured throughout the incubation period, using a high-performance optical detection system in combination with image analysis algorithms. The system generates an MIC and further qualitative susceptibility results (i.e., S, I, R) for the tested antimicrobials when applicable. The qualitative results are determined based on established breakpoints stipulated by applicable authorities, i.e., FDA, CLSI or EUCAST. FDA Susceptibility Testing Interpretive Criteria (STIC), aka "breakpoints" are found in Table 3.

The provided text describes the performance characteristics of the ASTar BC G- Kit and ASTar Instrument, primarily focusing on its antimicrobial susceptibility testing (AST) capabilities. While it details various studies, it does not describe an AI/ML device that utilizes a test set with ground truth experts. Instead, it describes a medical device for in vitro quantitative determination of antimicrobial susceptibility based on time-lapse microscopy imaging.

Therefore, many of the requested points, such as "number of experts used to establish ground truth," "adjudication method," "MRMC comparative effectiveness study," "standalone (algorithm only) performance," and "sample size for the training set" (for an AI model), are not applicable to this document as it does not describe an AI/ML-driven diagnostic device in the traditional sense.

However, I will extract relevant information about the device's acceptance criteria and studies to the best of my ability, interpreting "acceptance criteria" as performance metrics for this type of medical device.

Key Information from the Document:

The ASTar System is an automated system for antimicrobial susceptibility testing (AST) that uses high-speed, time-lapse microscopy imaging of bacteria to determine Minimum Inhibitory Concentration (MIC) and qualitative susceptibility results (S, I, R).

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

The document defines acceptance criteria primarily through performance metrics like Essential Agreement (EA) and Category Agreement (CA) compared to a reference method (frozen Broth Micro-Dilution, BMD), along with rates for Very Major (VMJ) discordant results, Major (MAJ) discordant results, and Minor (MIN) discordant results.

While a single explicit "acceptance criteria table" is not provided with specific pass/fail percentages before results, the overall performance table (Table 16) implicitly represents the success or failure against internal performance goals. The FDA's Special Controls guidance (referenced in 8.5.8) would typically outline such criteria. Based on the "Conclusions" section, the device was deemed "substantially equivalent," implying these metrics were acceptable.

Here's a summary of the reported device performance from Table 16, which reflects the met acceptance criteria for the clinical study:

Table: Reported Device Performance (Summary from Table 16)

Note: Some "poor performance" combinations (EA <90% or higher VMJ/MAJ rates) are noted as having limitation statements in the device labeling, indicating that these specific combinations might have had acceptance criteria adjusted or were deemed acceptable if the overall performance portfolio met the standard for substantial equivalence.

Reproducibility Study Acceptance Criteria:

• Overall reproducibility of ≥ 95% based on the number of results that fall within ±1 doubling dilution between the test MIC result and test MIC mode.

Reproducibility Study Performance (Aggregated, Table 5):

• Best case scenario: All antimicrobials showed ≥95% reproducibility (range 95.1% to 100%).
• Worst case scenario: Most antimicrobials showed ≥95% (range 88% to 100%). Two were slightly below: Ceftazidime (88%) and Ceftazidime-avibactam (90.7%).

Blood Culture Bottle Compatibility Acceptance Criteria:

• Overall essential agreement (EA) as compared to reference MIC obtained by frozen broth microdilution according to CLSI M07 shall be ≥90% for each antimicrobial, stratified by bacteria.
• Percentage of MIC values within ±1 doubling dilution of the mode MIC for each antimicrobial/bottle were determined, with the expectation of high agreement.

Blood Culture Bottle Compatibility Performance (Table 7):

• EA with BMD: Ranged from 97.2% to 99.8%. Some individual antimicrobial/bottle/isolate combinations had EA <90% (e.g., Tobramycin / BACTEC Standard Anaerobic: K. pneumoniae QM2403 (0/3)), but overall, the performance met the criterion.
• MIC values ±1 from mode: Ranged from 98.9% to 100% across different bottle types, indicating similar performance.

Sample Stability Acceptance Criteria:

• 95% of MIC values within ±1 doubling dilution of the mode MIC of initial samples (loaded <1 hour).

Sample Stability Performance (Table 8):

• 16-24 hours at room temperature: 100% pass rate.
• 16-24 hours at 35°C: 99.6% pass rate.

Interfering Substances Acceptance Criteria: (Implicitly, high pass rate expected)

• The comparison was to control samples; a MIC value within ±1 doubling dilution from the control value was considered a "pass."

Interfering Substances Performance (Table 11):

• Pass rates were 100% for most substances, except RBCs (99.1%). All were deemed acceptable.

Interfering Antibiotics Acceptance Criteria:

• 95% pass rate as compared to control samples without interfering antibiotics.

Interfering Antibiotics Performance (Table 13):

• All six evaluated antibiotic/BCB-combinations had overall pass rates of 96.2% to 100%. Some individual combinations fell below 90% (e.g., Cefotaxime / BACTEC: Trimethoprim-sulfamethoxazole 77.8%), but the overall criterion (per combination type) was met.

Carry Over and Cross Contamination Acceptance Criteria: (Implicitly, close to 100% pass rate expected)

• MIC for the susceptible isolate for each antimicrobial must be within ±1 doubling dilution of the control mode MIC to pass.

Carry Over and Cross Contamination Performance:

• 99.7% pass rate (307/308) for susceptible isolate MIC value. No carry over or cross contamination observed.

Set Inoculum for AST Acceptance Criteria:

• For starting bacterial concentration >5 x 10^7 CFU/mL, assess and adjust successfully at high rate, producing an inoculum within acceptance ranges.
• For concentrations <5 x 10^6 CFU/mL, acceptable to be aborted.

Set Inoculum for AST Performance (Table 14):

• For samples with starting bacterial concentration >5 x 10^7 CFU/mL: 95.8% (23/24) completed concentration adjustment, and 100% (23/23) of those produced an inoculum within acceptance ranges.
• For samples with starting bacterial concentration 5 x 10^6 to < 4.99 x 10^7 CFU/mL: 75% (9/12) completed concentration adjustment, 88.9% (8/9) within range.
• For samples with starting bacterial concentration < 5 x 10^6 CFU/mL: 0% completed concentration adjustment and 36/36 were aborted, as expected.

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

• Clinical Study Test Set Sample Size: A total of 1,068 samples were enrolled. After exclusions (off-panel organisms, contamination, non-viable isolates, protocol deviations), 880 samples were included in the performance analysis. This included:

  • 256 fresh, positive blood cultures (Fresh PBC)
  • 223 contrived with clinical stock isolates
  • 401 contrived blood cultures with challenge isolates.
  • 933 valid samples were analyzed (907 produced at least a partial AST result).

• Data Provenance: The study involved both prospective collection (fresh, left-over samples from patients with suspected bacteremia, implicitly "clinical samples") and retrospective/contrived samples (clinical stock isolates and challenge isolates).

• Country of Origin: The initial study was conducted at four sites: three external clinical sites in the United States (US) and one internal site in Sweden. The supplemental testing phase was conducted at three sites: two of the original external clinical sites in the United States (US) and one internal site in Sweden.

3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts (e.g. radiologist with 10 years of experience)

This is not applicable in the traditional sense of human experts annotating images for ground truth in an AI/ML context. The ground truth for this device's performance is established by a gold standard laboratory method: frozen Broth Micro-Dilution (BMD) performed according to CLSI M07 11th Edition. This is a well-established, standardized microbiological method, not reliant on human expert interpretation of a visual output.

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

Not applicable in the human-in-the-loop expert adjudication sense for an AI/ML device. For the reference BMD, if a Mode MIC could not be established with the first set of three replicates, a second set of three frozen replicates was tested. If a Mode MIC still could not be established, the Median from all six plates was used. This is a form of scientific replication/adjudication.

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 device is an automated system for AST, not an AI assistance tool for human interpretation of medical images or other diagnostics. There are no "human readers" directly assisted by AI in its primary function as described.

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

This is an automated device that provides quantitative MIC and qualitative S/I/R results. Its "performance" (EA, CA, etc.) is inherently the standalone performance of the system against the gold standard BMD. The "imaging of bacteria" and "image analysis algorithms" mentioned in the device description imply automated processing, which is essentially "algorithm only" performance for generating the MIC. While "Organism identification is required for AST result interpretation and reporting" (and likely entered by a human user), the core AST algorithm itself acts in a standalone manner once the species ID is provided.

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

The ground truth used for performance evaluation was frozen Broth Micro-Dilution (BMD) results run in triplicate according to CLSI M07 11th Edition. This is a laboratory gold standard method for antimicrobial susceptibility testing.

8. The sample size for the training set

The document does not describe an AI/ML model in the sense of a deep learning model requiring a discrete "training set" of labeled data for model development. The "image analysis algorithms" mentioned are part of the core technology of the ASTar Instrument. If these algorithms involved machine learning, no details about a specific "training set" size are provided in this document. Given it's a 510(k) submission for a non-AI/ML device, such details are typically not required or relevant in the same way as for AI/ML device submissions.

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

As there's no explicitly defined AI "training set" described, this point is not applicable. The device's performance is inherently compared to the established laboratory gold standard (BMD).

Ask a specific question about this device