BARD COMPOSITE PROSTHESIS

• K Number: K971745
• Device Name: BARD COMPOSITE PROSTHESIS
• Manufacturer: DAVOL, INC., SUB. C.R. BARD, INC.
• Date Cleared: 1997-08-06 ( 86 days)
• Product Code: FTL
• Regulation Number: 878.3300
• Type: Traditional
• Panel: SU
• Reference & Predicate Devices: N/A

Intended Use

Reconstruction of soft tissue deficiencies, such as for the repair of hernias and chest wall defects.

Device Description

The proposed Composite Mesh is manufactured from knitted polypropylene monofilament with a diameter of 6 mil. The unique knitting process for the Composite Mesh produces a flat double layer of mesh. This double layer of mesh is knitted and interconnected simultaneously during the knitting process. One side of one layer of mesh is heat bonded to a single layer of expanded polytetrafluoroethylene (e-PTFF).

AI/ML Overview

Here's an analysis of the provided text regarding the acceptance criteria and study for the Bard Composite Prosthesis:

Acceptance Criteria and Reported Device Performance

Acceptance Criteria Category	Specific Criteria/Test	Reported Performance (Bard Composite Prosthesis)	Comparator Device Performance (Predicate)	Equivalence Claim/Outcome
Physical Characteristics	Pore Size	Not explicitly quantified in summary, but assessed.	Assessed against predicate devices.	Comparable to predicate devices.
	Surface Roughness	Not explicitly quantified in summary, but assessed.	Assessed against predicate devices.	Comparable to predicate devices.
	Surface Morphology (SEM)	Not explicitly quantified in summary, but assessed.	Assessed against predicate devices.	Comparable to predicate devices.
Performance	Suture Retention Strength	Tested.	Tested against predicate devices.	Comparable to predicate devices.
	Burst Strength	Tested.	Tested against predicate devices.	Comparable to predicate devices.
Chemical Characteristics	e-PTFE Properties (FTIR, DSC)	Tested.	Tested against Predicate GORE-TEX.	Comparable to Predicate GORE-TEX.
In-Vivo Testing	Organ Adhesion (e-PTFE layer)	Quantitatively measured for e-PTFE layer.	Compared to Predicate GORE-TEX.	Adhesion response comparable to Predicate GORE-TEX.
	Tissue Ingrowth	Qualitatively assessed.	Compared to both predicate devices.	Gross tissue ingrowth comparable to both predicate devices.
	Overall Performance in Simulated Clinical Situation	Assessed.	Not explicitly detailed for comparator.	Not explicitly detailed for comparator.
Biocompatibility	Toxicity	Tested per ISO-10993, Part 1.	N/A	Non-toxic.
	Sensitization	Tested per ISO-10993, Part 1.	N/A	Non-sensitizing.

Study Information

1. Sample Size used for the test set and the data provenance:

   • Test Set Sample Size: Not explicitly stated for any of the individual tests. The document broadly mentions "Laboratory testing" and "in-vivo testing" without providing specific N values for the number of meshes tested, animals used, or human subjects.
   • Data Provenance: The studies were laboratory and in-vivo (animal) tests. The location of these labs or animal facilities is not specified, nor is whether the data is retrospective or prospective, though in-vivo testing inherently suggests prospective data collection in the context of device development.

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

   • This question is not applicable as the studies described are laboratory and in-vivo (animal) tests for device characteristics, not human clinical studies requiring expert ground truth for interpretation of patient data (e.g., imaging, pathology). The ground truth would be the objective measurements from the tests themselves (e.g., burst strength values, FTIR spectra, histological assessment by a veterinary pathologist for in-vivo tests). No "experts" in the clinical sense (like radiologists) were used to establish ground truth for this type of testing.

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

   • None. Adjudication methods are typically employed in clinical studies involving human subject data and expert interpretation (e.g., multiple radiologists reviewing scans and resolving discrepancies). This document describes laboratory and animal studies where direct measurements and observations, likely performed by trained technicians or scientists, serve as the data. Discrepancy resolution for such data would follow standard scientific practice (e.g., re-testing, calibration, statistical analysis), not an adjudication model.

4. 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 is not applicable. This submission is for a surgical mesh, not an AI-powered diagnostic device or a system designed to assist human readers (e.g., radiologists). Therefore, there is no AI component, no human readers in an interpretive capacity, and no effect size regarding AI assistance.

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

   • No. This is not applicable. The device is a physical surgical implant, not an algorithm, so there is no standalone algorithm performance to evaluate.

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

   • Objective Measurements and Observations from Laboratory and In-Vivo Testing:
     • Laboratory Tests: Direct physical measurements (e.g., pore size, surface roughness, suture retention force, burst pressure), and chemical analyses (e.g., FTIR and DSC spectroscopy data).
     • In-Vivo Tests (Animal Model): Quantitative measurements (e.g., organ adhesion scores/measurements) and qualitative assessments (e.g., histological evaluation by a veterinary pathologist for tissue ingrowth characteristics and inflammatory response).
   • The "ground truth" here is the scientific data generated by standardized testing protocols and observed biological responses in the animal models, compared to predicate devices.

7. The sample size for the training set:

   • Not Applicable. The Bard Composite Prosthesis is a physical medical device, not a machine learning model. Therefore, there is no "training set" in the context of AI development.

8. How the ground truth for the training set was established:

   • Not Applicable. As there is no training set (as defined in machine learning), there is no ground truth to establish for it. The development of the device likely involved engineering specifications, materials science, and iterative testing, but not in the framework of a "training set" for an algorithm.

Regulation Number and Section

§ 878.3300 Surgical mesh.

(a)
Identification. Surgical mesh is a metallic or polymeric screen intended to be implanted to reinforce soft tissue or bone where weakness exists. Examples of surgical mesh are metallic and polymeric mesh for hernia repair, and acetabular and cement restrictor mesh used during orthopedic surgery.(b)
Classification. Class II.