FDA 510(k) Search - By Innolitics (SaMD and AI Experts)

BacT/ALERT® PF Plus Culture Bottles are used with the BacT/ALERT® Microbial Detection System in qualitative procedures for recovery and detection of aerobic and facultative anaerobic microorganisms (bacteria and yeast) from blood.

Device Description

The new reagent (BacT/ALERT PF Plus Culture Bottle) is an improvement upon the cleared charcoal formulation reagent (BacT/ALERT PF Culture Bottle). The BacT/ALERT PF Culture Bottles are used with the BacT/ALERT Microbial Detection Systems in qualitative procedures for recovery and detection of microorganisms from blood. The predicate BacT/ALERT PF Culture Bottle contains charcoal, for its antimicrobial neutralization properties, in a complex growth medium. Charcoal is eliminated in the proposed BacT/ALERT PF Plus Culture Bottle, and is replaced with two types of adsorbent resins in a complex growth medium. The proposed BacT/ALERT PF Plus Culture Bottle is optimized to increase antimicrobial neutralization properties, and to increase the clarity of Gram stains in comparison to the predicate BacT/ALERT PF Culture Bottle.

The BacT/ALERT Microbial Detection System provides both a microbial detection system and a culture medium bottle with suitable nutritional and environmental conditions for microorganisms commonly encountered in blood (adult and pediatric bottles) or other normally sterile body fluid samples (except urine) (adult bottles only) taken from a patient suspected of having bacteremia/fungemia. An inoculated bottle is placed into the instrument where it is incubated and continuously monitored for the presence of microorganisms that will grow in the BacT/ALERT bottles.

The BacT/ALERT Microbial Detection System utilizes a colorimetric sensor and reflected light to monitor the presence and production of carbon dioxide (CO2) that is dissolved in the culture medium. If microorganisms are present in the test sample, carbon dioxide is produced as the microorganisms mefabolize the substrates in the culture medium. When growth of the microorganisms produces CO2, the color of the gas-permeable sensor installed in the bottom of each culture bottle changes from blue-green to yellow. The lighter color results in an increase of reflectance units monitored by the system. Bottle reflectance is monitored and recorded by the instrument every 10 minutes.

AI/ML Overview

Here's a breakdown of the acceptance criteria and the study information for the BacT/ALERT® PF Plus, as requested:

Acceptance Criteria and Device Performance

Criteria Category	Acceptance Criteria (Implied/Specific)	Reported Device Performance
Analytical Sensitivity (Growth Performance)	The device must demonstrate effective recovery and detection of a representative panel of aerobic and facultative anaerobic microorganisms (bacteria and yeast) from blood. Implicitly, recovery rates for most common pathogens should be high, ideally 100%.	For most tested microorganisms, 100% recovery was observed.
(Table 2):

Staphylococcus aureus: 100.0% (30/30)
Escherichia coli: 100.0% (30/30)
Pseudomonas aeruginosa: 100.0% (12/12)
Klebsiella pneumoniae: 100.0% (12/12)
Candida albicans: 100.0% (30/30)
Streptococcus pneumoniae: 100.0% (30/30)
Staphylococcus epidermidis: 100.0% (12/12)
Enterococcus faecalis: 100.0% (12/12)
Enterococcus faecium: 100.0% (12/12)
Enterobacter cloacae: 100.0% (12/12)
Candida glabrata: 100.0% (12/12)
Haemophilus influenzae: 100.0% (12/12)
Proteus mirabilis: 100.0% (12/12)
Less than 100% detection was observed for Capnocytophaga ochracea, Cardiobacterium hominis, Eikenella corrodens, Haemophilus parainfluenzae, Granulicatella adiacens, and Helicobacter cinaedi. |
| Antimicrobial Neutralization | The device should effectively neutralize a broad range of clinically relevant antimicrobials to allow for microbial growth even in the presence of these substances. | Antimicrobials from numerous categories were neutralized. These include penicillins, glycylcyclines, polyenes, macrolides, triazoles, echinocandins, cefazolin, cefoxitin, ceftaroline, aminoglycosides, fluoroquinolones, lincosamides, glycopeptides, and oxazolidinones.
Neutralization was not achieved for ceftazidime or cefepime.
Less than complete neutralization was observed for ceftotaxime and ceftriaxone (at ranges of 50% PSL to 1-2% PSL depending on microorganism). |
| LoD (Limit of Detection) | The device should reliably detect microorganisms at low concentrations, demonstrating a reproducible Limit of Detection (LoD). Implicitly, at least 95% detection at LoD. | At least 95% detection was achieved at LoD for all tested microorganisms.
(Table 3): LoD ranged from 4 CFU/bottle (E. coli, K. pneumoniae, P. aeruginosa) to 8 CFU/bottle (Enterobacter aerogenes). |
| Within-Laboratory Precision (Repeatability) | The device should demonstrate consistent growth performance (percent recovery and time to detection) when tested repeatedly within the same laboratory under various conditions (e.g., multiple lots, operators, instruments). High percent recovery (ideally 100%) and consistent Time to Detection are expected. | High percent recovery and consistent Time to Detection.
(Table 4):
C. albicans (Fluconazole): 100.0% recovery across 3 lots, Mean TTD 26.0 hours.
E. coli (Amikacin): 100.0% recovery across 3 lots, Mean TTD 12.0 hours.
K. pneumoniae (Levofloxacin): 100.0% recovery across 3 lots, Mean TTD 13.4 hours.
P. aeruginosa (Piperacillin): 99.1% overall recovery, Mean TTD 19.2 hours. (One lot showed 97.2%)
S. pneumoniae (Penicillin G): 100.0% recovery across 3 lots, Mean TTD 13.2 hours.
S. aureus (Vancomycin): 100.0% recovery across 3 lots, Mean TTD 16.9 hours. |
| Reproducibility | The device should demonstrate consistent growth performance across different testing sites, operators, and days. High overall percent recovery is expected. | Overall reproducibility showed high percent recovery (92.0%).
Sites showed varying performance, down to 72.2% for E. aerogenes at Site 2.
(Table 5):
Overall % Recovery: 92.0% (698/759) with 95% CI: 89.8%, 93.8%.
Site 1: 95.4% (206/216)
Site 2: 83.5% (213/255)
Site 3: 96.9% (279/288)
Excluding laboratory errors, E. aerogenes showed 85.0% recovery, while other organisms achieved 100%. |
| Delayed Entry | The device should maintain effective microbial recovery even after delays in loading into the instrument, under various temperature conditions. A cautionary note is expected if performance is significantly degraded under certain conditions. | High recovery across most delayed entry conditions, with one critical caution.
(Table 6):
Control (No delay): 100.0% recovery.
2-8°C, 48 hours hold: 98.6% recovery.
20-25°C, 24 hours hold: 98.0% recovery.
20-25°C, 36 hours hold: 91.9% recovery.
35-37°C, 8 hours hold: 98.9% recovery.
35-37°C, 24 hours hold: 56.6% recovery. CAUTIONARY NOTE issued for this condition.
False positive rate for negative controls was 0.5% (1/221). |
| Clinical Study Performance (Blood Cultures) | The device (PF Plus) should demonstrate comparable or improved performance compared to the predicate device (PF) in a clinical setting for the recovery of microorganisms from blood. | BacT/ALERT PF Plus demonstrated improved or comparable clinical performance over the predicate PF bottle.
(Table 7):
Ratio of True Positives for Significant isolates: 1.182 (PF Plus detected 91 vs PF detected 77).
Ratio of True Positives for Contaminant isolates: 0.828 (PF Plus detected 24 vs PF detected 29).
Ratio of True Positives for Unknown isolates: 1.136 (PF Plus detected 25 vs PF detected 22).
Overall Ratio of True Positives: 1.094 (140/128) with a 95% Cl of (0.954, 1.234).
44 isolates detected only by PF Plus vs 32 isolates detected only by PF.
False positive rate: 0.05% (1/2215) in the clinical study population. |
| Interfering Substances | Potential interfering substances (plasma, blood, blood clots, WBCs) should not interfere with microorganism recovery and detection, nor should they generate false positive results. | No interference or false positives were observed.
Plasma, blood, blood clots, and white blood cells (at concentrations relevant to bacteremia) neither interfered with recovery and detection of organisms nor generated false positive results in the absence of organisms. |

Study Details

2. Sample Size and Data Provenance (Test Set)

Analytical Sensitivity (Growth Performance):
- Sample Size: Varies by microorganism. For most listed, 30 replicates (e.g., 30/30) were tested. For others, 12 replicates (e.g., 12/12) were used.
- Data Provenance: In-house seeded studies with blood obtained from healthy human volunteers. This is a prospective study design using controlled laboratory conditions.
Limit of Detection (LoD):
- Sample Size: A minimum of 30 replicates were tested per species.
- Data Provenance: In-house seeded studies. Bottles at end of shelf life were used, and H. influenzae testing included 4 ml pooled human blood supplementation. This is a prospective study design using controlled laboratory conditions.
Within-Laboratory Precision (Repeatability):
- Sample Size: A minimum of 108 replicates were tested for each organism/antimicrobial combination.
- Data Provenance: In-house seeded studies over 12 days on multiple instruments by multiple operators. Organisms grown in the presence of clinically relevant antimicrobial concentrations. This is a prospective study design using controlled laboratory conditions.
Reproducibility:
- Sample Size: Target of 162 replicates per site for 9 organisms, conducted over 3 days. Total replicates across sites varied per organism, e.g., 72 for S. aureus, 87 for C. albicans, 123 for E. aerogenes.
- Data Provenance: Seeded studies conducted at three sites. This is a prospective multi-site laboratory study.
Delayed Entry:
- Sample Size: For inoculated test bottles, a total of 459 to 458 bottles across various conditions. For negative controls, 221 bottles. Eleven species were tested, with an acceptable range of 30 to 300 CFU per bottle.
- Data Provenance: Seeded studies generated at three sites using human blood from healthy volunteers. This is a prospective multi-site laboratory study.
Clinical Study Results (Blood Cultures):
- Sample Size: 2188 bottle pairs (4376 bottles total) from 1086 pediatric patients. The total reported population for analysis was 2215 results (172 isolates from positive pairs and 2043 negative pairs).
- Data Provenance: Multi-center clinical study conducted at three different geographic sites in the U.S. This is a prospective clinical study with real patient samples.

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

Analytical, LoD, Repeatability, Reproducibility, Delayed Entry Studies: These were seeded studies with known microorganisms and inoculum levels, so the "ground truth" was inherently defined by the experimental setup. No independent experts were needed to establish the ground truth for microbial identity or presence.
Clinical Study: The document states, "Clinical isolates recovered were classified as significant, contaminant, or unknown based on determination by the clinical trial sites." This implies a judgment by clinical laboratory personnel or clinicians at the three study sites. Specific numbers or qualifications (e.g., "radiologist with 10 years of experience") are not provided in the document. It can be inferred that these would be licensed medical professionals (e.g., clinical microbiologists, infectious disease physicians) involved in patient care and laboratory diagnostics.

4. Adjudication Method for the Test Set

Analytical, LoD, Repeatability, Reproducibility, Delayed Entry Studies: Not applicable, as these were seeded studies where the "ground truth" (presence and identity of microorganisms) was controlled by the experimental design. Detection was based on instrument flagging and confirmation by subculture/Gram stain.
Clinical Study: The document specifies: "Subcultures of both bottles were performed when either bottle in the set was determined to be positive by the BacT/ALERT system. A pair of bottles was determined to have a positive status if the subculture of either the PF Plus or PF bottle was positive." This indicates a comparative approach where confirmation by subculture was the primary adjudication for a positive result from the instrument, and then classification (significant, contaminant, unknown) was determined by individuals at the clinical trial sites. No explicit mention of an "adjudication panel" or specific consensus method like 2+1 or 3+1 is provided for the final classification of clinical significance.

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

No, an MRMC comparative effectiveness study was not done in the context of human readers improving with or without AI assistance. This device is a diagnostic instrument (culture bottle and detection system) for microbial growth, not an AI-assisted diagnostic imaging tool for human interpretation. The "comparison" in the clinical study was between two different culture bottle formulations (PF Plus vs. PF) as standalone detection systems.

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

Yes, a standalone performance study was done. The entire analytical and clinical evaluation focuses on the performance of the BacT/ALERT® PF Plus culture bottles when used with the BacT/ALERT® Microbial Detection System as a standalone system. The system automatically monitors for growth, flags positive bottles, and then subculture/Gram stain is performed to identify the organism. The reported "percent recovery" and "time to detection" are purely instrumental/algorithmic outputs, confirmed by microbiological methods. There is no human-in-the-loop interpretative step being evaluated for the device's primary function (detection of microbial growth) itself.

7. The Type of Ground Truth Used (Test Set)

Analytical Sensitivity, LoD, Repeatability, Reproducibility, Delayed Entry:
- Expert Consensus / Microbiological Confirmation: For these in vitro seeded studies, the ground truth was established by known inoculum concentrations of specific microorganisms confirmed by traditional microbiological techniques (e.g., viable plate counts for inoculum, subculture, and Gram stain for confirmation of instrument-flagged bottles).
Clinical Study:
- Microbiological Confirmation & Clinical Judgment:
  - The fact that a sample contained a microorganism was established by subculture of instrument-flagged positive bottles.
  - The significance of the recovered isolates (significant, contaminant, or unknown) was based on clinical determination by the clinical trial sites, which combines microbiological results with patient clinical context.

8. The Sample Size for the Training Set

The document describes performance testing for a device (culture bottle), not an AI/ML algorithm that would typically have a distinct "training set." The "knowledge base specifications utilized by the firmware included changes to the initial value threshold variables used by the firmware algorithm." This suggests that the algorithm itself undergoes adjustments or training, but no specific "training set" sample size is provided in the document. The data presented are for validation/verification (test sets).

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

As mentioned above, the document does not explicitly describe a "training set" or how its ground truth was established. It states that "changes to the initial value threshold variables used by the firmware algorithm" were made. This implies an iterative development process where internal data (likely similar to the analytical studies presented, but not explicitly labeled as "training") would be used to optimize these thresholds. The specific data, methods, or ground truth establishment for this firmware adjustment/training are not detailed in this 510(k) summary.

Summary

Regulation Number and Section

§ 866.2560 Microbial growth monitor.

(a)
Identification. A microbial growth monitor is a device intended for medical purposes that measures the concentration of bacteria suspended in a liquid medium by measuring changes in light scattering properties, optical density, electrical impedance, or by making direct bacterial counts. The device aids in the diagnosis of disease caused by pathogenic microorganisms.(b)
Classification. Class I. With the exception of automated blood culturing system devices that are used in testing for bacteria, fungi, and other microorganisms in blood and other normally sterile body fluids, this device is exempt from the premarket notification procedures in subpart E of part 807 of this chapter.