Search Results

The LifeShield™ Infusion Safety Software Suite is a collection of software products that facilitates networked communication between compatible systems. The Infusion Safety Software Suite provides trained healthcare professionals the ability to manage data for compatible infusion pumps. All data entry and validation of infusion parameters on compatible infusion pumps is performed by a trained healthcare professional. LifeShield™ Infusion Safety Software Suite is indicated for use in patients including adult, pediatric and neonate undergoing infusion therapy with connected compatible infusion pumps (as per the indications for use specified for the compatible infusion pump).

The LifeShield™ Drug Library Management (DLM) software product is intended to be used by pharmacists to create, configure, edit, and manage drug library data, including infusion pump settings, for use with compatible infusion pumps. Drug library contents are constructed based on the healthcare provider's defined best practices.

The LifeShield™ Clinical Dashboards & Reports (CDR) software product provides trained healthcare professionals with the capability to view and manage infusion information collected from compatible infusion pumps. Healthcare professionals may choose to use the collected infusion information to support continuous quality improvement programs, or to analyze and trend various aspects of the infusion pumps and therapies used. It is not intended to be a substitute for good clinical management practices, nor does its operation create decisions or treatment pathways.

The LifeShield™ Data Flow Management (DFM) software product is intended to facilitate bi-directional communication with compatible infusion pumps, information technology systems, and other LifeShield™ Infusion Safety Software Suite products. LifeShield™ DFM provides a way to automate the programming of infusion parameters, thereby decreasing the number of manual steps necessary to enter infusion data. LifeShield™ DFM forwards infusion-related information received from the infusion pump to compatible information technology systems.

Device Description

The LifeShield™ Infusion Safety Software Suite is a cloud-based platform provided as a software-as-a-service (SaaS) designed to be compatible with the Plum Duo™ Precision IV infusion pump and Plum Solo™ Precision IV infusion pump. The LifeShield Infusion Safety Software Suite is hosted by Amazon Web Services (AWS) as its cloud provider.

LifeShield™ Infusion Safety Software Suite consists of a collection of software services which, when used together, provide a comprehensive set of data management capabilities for trained healthcare professionals when working with infusion pumps. LifeShield™ Infusion Safety Software Suite does not remotely control or program the infusion pump or provide the ability to remotely manage pump alarms such as real-time monitoring, clearing and silencing alarms.

The LifeShield™ Drug Library Management (DLM) software is used by pharmacists to create approved drug libraries that can be downloaded by the infusion pumps. The latest software version introduces enhancements to its user interface and additional drug library settings for support of Plum Duo™ and Plum Solo™ pumps.

The LifeShield™ Clinical Dashboards & Reports (CDR) software is used by clinical administrators to view infusion or device-related information received from the infusion pumps via the LifeShield™ DFM. The information presented by the software does not create decisions or treatment pathways for patients. The latest version of the software improves the data presented for ongoing infusions and dashboards.

LifeShield™ Data Flow Management (DFM) software facilitates bidirectional communications between the infusion pump and hospital information systems (HIS); it routes pharmacy-validated orders to the connected pumps and infusion-related information to the HIS. The latest software version adds the ability to forward alarms to the HIS.

AI/ML Overview

The provided FDA 510(k) clearance letter for the LifeShield™ Infusion Safety Software Suite (K242117) does not contain the specific details required to construct a table of acceptance criteria with reported device performance, nor does it detail a study that proves the device meets specific performance criteria in terms of accuracy or clinical outcomes.

This document primarily focuses on demonstrating substantial equivalence to a predicate device (K223606) based on non-clinical testing (design verification, design validation, cybersecurity, risk management, and human factors) and software modifications, rather than performance metrics related to clinical accuracy or diagnostic capabilities (which would be more common for AI/ML devices in image analysis, for example).

The submission explicitly states:

"Summary of Clinical Testing: Not applicable. A clinical study is not required for this submission to support substantial equivalence."

Therefore, I cannot fulfill all parts of your request as the information is not present in the provided text. The device is a software suite for managing infusion pump data, not an AI/ML diagnostic or predictive tool that would have performance metrics like sensitivity, specificity, or AUC against a ground truth of disease.

However, I can extract information related to the types of testing performed and what they aimed to prove, which indirectly serve as acceptance criteria for this type of software device.

Analysis of Acceptance Criteria and Study Details based on the Provided Document

Given the nature of the device (infusion safety software suite) and the information provided in the 510(k) summary, the "acceptance criteria" here are aligned with software quality, safety, and functionality, rather than clinical performance metrics typically associated with AI/ML diagnostic tools.

The study proving the device meets acceptance criteria is a comprehensive set of non-clinical tests described below, rather than a single clinical trial.

1. Table of Acceptance Criteria and the Reported Device Performance

Since this is not an AI/ML diagnostic device with performance metrics like accuracy, sensitivity, or specificity against disease presence, the "acceptance criteria" relate to software design, functionality, safety, and usability. The document reports that all these tests "pass established acceptance criteria."

Acceptance Criteria Category	Specific Criteria (Inferred from testing)	Reported Device Performance
Design Requirements	Adherence to design specifications; proper function of new features (e.g., UI enhancements, new drug library settings, alarm forwarding).	Passed: "Design verification tests pass established acceptance criteria, confirming the subject device meets design requirements."
Software Quality	Conformance to IEC 62304:2015 software development process; reliability.	Passed: "Software verification follows the software development process outlined in IEC 62304:2015." "Verification activities also include software verification, performance, reliability, compatibility, and interoperability tests."
Intended User Needs/Use Environment	Device meets all intended users' needs for its intended use and environment.	Passed: "Design validation tests pass established acceptance criteria, confirming the subject device meets all intended users' needs for the device's intended use and intended use environment."
Cybersecurity	Reasonable security against threats; protection of data.	Passed: "Cybersecurity evaluation and testing demonstrate that the software is reasonably secure."
Risk Management	Identification and mitigation of risks (ISO 14971:2019); effectiveness of mitigations.	Passed: "Risk management activities...concludes that the subject device is reasonably safe." "Mitigations...tested for correct implementation and effectiveness."
Human Factors/Usability	Effectiveness of user interface design for new features and critical tasks, per FDA guidance and standards (IEC 62366-1:2020, ANSI/AAMI HE75:2009/(R)2018).	Passed: "Human Factors study demonstrates the effectiveness of the user interface design for additional features and their associated critical tasks."

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

Test Set Sample Size: The document does not specify a "sample size" in terms of patient data or clinical cases. For software verification and validation, the "test set" would consist of numerous test cases, simulated use scenarios, and functional tests. These numbers are not detailed in the summary.
Data Provenance: Not applicable in the traditional sense of patient data. The testing is focused on software functionality, safety, and usability. There is no mention of retrospective or prospective data collection from patients or clinical settings for performance evaluation.

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

Ground Truth Establishment: For this type of software, "ground truth" isn't established by clinical experts in the same way it would be for a diagnostic image. Instead, it's defined by design specifications, regulatory requirements, industry standards (e.g., IEC 62304, ISO 14971, IEC 62366-1), and best practices in software engineering and cybersecurity.
Experts: The development team, regulatory affairs, quality assurance, and potentially third-party cybersecurity and human factors experts would define and verify these "ground truths" (i.e., whether the software behaves as intended and safely). The specific number or qualifications of these internal or external experts is not detailed in the provided K summary. The human factors study implies the use of representative users (trained healthcare professionals), but not necessarily "experts" defining a ground truth about a medical condition.

4. Adjudication Method for the Test Set

Not applicable in the context of clinical expert adjudication of medical cases. Adjudication in software testing would involve bug reporting, resolution, and re-testing processes managed by the development and QA teams, not described here.

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

No, an MRMC study was not done. The document explicitly states: "Summary of Clinical Testing: Not applicable. A clinical study is not required for this submission to support substantial equivalence." An MRMC study is a type of clinical study often done for diagnostic AI to compare human performance with and without AI assistance. This device is not a diagnostic AI.

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

This concept is not directly applicable. The "LifeShield™ Infusion Safety Software Suite" is inherently a human-in-the-loop system designed to assist trained healthcare professionals. It manages data, facilitates communication, and supports drug library management but does not "decide" or "treat" on its own. Its "standalone" performance would pertain to its functional correctness in processing data and facilitating communication, which is covered by the general "design verification" and "performance" testing.

7. The Type of Ground Truth Used

The "ground truth" for this software pertains to its functional correctness, adherence to specifications, safety requirements, cybersecurity posture, and usability standards. It is not based on clinical "outcomes data" or "pathology" in the medical sense, but rather on:
- Design Specifications: Whether the software performs its programmed functions as intended.
- Regulatory Standards: Compliance with relevant medical device software (IEC 62304), risk management (ISO 14971), and human factors (IEC 62366-1) standards.
- User Needs/Requirements: Whether the software meets the needs of its intended users in its intended environment.

8. The Sample Size for the Training Set

Not applicable. This device is a software suite, not an AI/ML model that undergoes a "training phase" with a distinct "training set" of data to learn patterns. The software's capabilities are based on explicit programming and configuration, not machine learning from a dataset.

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

Not applicable, as there is no "training set" for this type of software.

In summary, the FDA 510(k) clearance for the LifeShield™ Infusion Safety Software Suite is based on a demonstration of substantial equivalence to a predicate device through extensive non-clinical software verification and validation, risk management, cybersecurity testing, and human factors analysis. The "acceptance criteria" are related to software quality, safety, and functionality, rather than clinical performance metrics against a medical ground truth or AI model training data.

Ask a Question

Ask a specific question about this device

K Number

K223606

Device Name

LifeShield Infusion Safety Software Suite

Manufacturer

ICU Medical, Inc.

Date Cleared

2023-08-24

(265 days)

Product Code

Regulation Number

Type

Panel

Reference & Predicate Devices

K141102

Intended Use

The LifeShield™ Clinical Dashboards & Reports (CDR) software product provides trained healthcare professionals with the capability to view and manage infusion collected from compatible infusion pumps. Healthcare professionals may choose to use the collected information to support continuous quality improvement programs, or to analyze and trend various aspects of the infusion pumps and therapies used. It is not intended to be a substitute for good clinical management practices, nor does its operation create decisions or treatment pathways.

Device Description

The LifeShield™ Infusion Safety Software Suite is a cloud-based platform provided as a software-as-a-service (SaaS) designed to be compatible with the Plum Duo™ infusion pump. The LifeShield Infusion Safety Software Suite is hosted by Amazon Web Services (AWS) as its cloud provider. LifeShield™ Infusion Safety Software Suite consists of a collection of software services which, when used together, provide a comprehensive set of data management capabilities for trained healthcare professionals when working with infusion pumps. LifeShield™ Infusion Safety Software Suite does not remotely control or program the infusion pump or provide the ability to remotely manage pump alarms such as real-time monitoring, clearing and silencing alarms.

The LifeShield™ Drug Library Management (DLM) software product is used by pharmacists to create approved drug libraries that can be downloaded by the infusion pumps. Drug libraries contain information on medications along with rulesets and associated clinical care areas (CCA) defined by pharmacists in accordance to their facility's best practices. Certain infusion pump parameters are also defined in the drug library. The LifeShield™ Device Manager (DM) and LifeShield™ Data Flow Management (DFM) are used to make the drug libraries available for the infusion pump to download and install. Download and installation of a drug library to the infusion pump establishes alert parameters for a medication that is being programmed for infusion. Additionally, the infusion pump applies the userdefined drug library settings for the configurable features of the pump.

The LifeShield™ Clinical Dashboards & Reports (CDR) software is used by clinical administrators to view infusion or device-related information received from the Plum Duo™ infusion pump via the LifeShield™ DFM. It provides a near real-time view of ongoing infusions and their status; a view of all infusion pumps with their asset information and operational status; dashboards that provides easy navigation of key infusion or asset metrics; and an analytics viewer that users can use to view historical infusion and/or asset information. Healthcare professionals may choose to use the collected infusion information to support continuous quality improvement programs, or to analyze and trend various aspects of the infusion pumps and therapies used. The information presented by the software does not create decisions or treatment pathways for patients. LifeShield™ CDR is able to display infusion and infusion pump information from a single or multiple facilities within the customer account.

LifeShield™ Infusion Safety Software Suite can be configured to interface with a facility's Hospital Information System (HIS) / EHR system to support auto-programming and infusion documentation. When the autoprogramming feature license is enabled, the LifeShield™ Infusion Safety Software Suite can receive a pharmacy-validated order (also referred as auto-program order) from the HIS/EHR and route it to the infusion pump where the therapy program is pre-populated with physician-prescribed medication and infusion parameters, helping to reduce manual entry by the clinician when programming the pump. LifeShield™ Infusion Safety Software Suite does not modify the contents of the auto-program order received from the HIS/EHR.

When the infusion documentation feature license is enabled, the LifeShield™ Infusion Safety Software Suite forwards the infusion data it receives from the infusion pump to the HIS/EHR system to support the facility's documentation of infusion information and HIS/EHR dashboards. Infusion data includes infusion status (e.g. volume change) and events (e.g. infusion start, stop, complete). LifeShield™ Infusion Safety Software Suite does not modify the contents of the infusion data sent to the HIS/EHR.

AI/ML Overview

Here's an analysis of the acceptance criteria and study information for the LifeShield™ Infusion Safety Software Suite based on the provided document:

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

The document primarily focuses on demonstrating substantial equivalence to a predicate device rather than presenting specific quantitative acceptance criteria and device performance metrics in a direct table format. However, it implicitly states that the device met acceptance criteria through various tests.

Here's an interpreted table based on the non-clinical testing summary:

Acceptance Criteria Category	Reported Device Performance (as stated in the document)
Design Requirements Met	Design verification tests passed established acceptance criteria, confirming the subject device meets design requirements.
Software Verification	Software verification followed the software development process outlined in IEC 62304:2015.
Performance, Reliability, Compatibility	Verification activities also included performance, reliability, compatibility tests, and systems integration tests with the compatible infusion pump, which were passed.
System Integration	Systems integration tests with the compatible infusion pump were passed.
Cybersecurity	Cybersecurity evaluation and testing demonstrate that the software is reasonably secure.
User Needs Met (Design Validation)	Design validation tests passed established acceptance criteria, confirming the subject device meets all intended users' needs.
Risk Management	Risk mitigations have been incorporated into the design and have been tested for correct implementation and effectiveness as part of design verification and validation.
User Interface Effectiveness (Human Factors)	Human Factors studies demonstrate the effectiveness of the user interface design for key features and their associated critical tasks.

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

The document does not specify the sample size used for any of the test sets (design verification, design validation, cybersecurity, human factors). It also does not mention the country of origin of the data or whether the data was retrospective or prospective.

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

The document does not provide information on the number of experts used or their specific qualifications for establishing ground truth for any of the tests. It refers to "trained healthcare professionals" for the intended use and "pharmacists" for drug library creation but does not detail their involvement in formal ground truth establishment for testing.

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

The document does not describe any adjudication method used for establishing ground truth or evaluating test results.

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:

The document states, "A Clinical Study is not required for this submission to support substantial equivalence." Therefore, no MRMC comparative effectiveness study was done, and there is no information on human reader improvement with or without AI assistance. This device is a software suite for managing infusion pump data, not an AI-powered diagnostic or assistive tool for human interpretation.

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

The device is described as a "software-as-a-service (SaaS)" that facilitates data management and communication for infusion pumps. It explicitly states that "All data entry and validation of infusion parameters on compatible infusion pumps is performed by a trained healthcare professional." While portions of the software may operate automatically (e.g., routing data), the core functionality involves human interaction and oversight. Therefore, a purely standalone algorithm-only performance study as typically understood for AI algorithms would not be applicable or detailed here. The "non-clinical testing" and "design verification" cover its functional performance.

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

The document does not explicitly state the type of "ground truth" used for its internal verification and validation studies. Given the nature of the device (infusion safety software), the "ground truth" would likely involve:

Design requirements and specifications: For design verification, the software's output and functionality are compared against its documented requirements.
User needs: For design validation, the software's ability to meet the needs of trained healthcare professionals (pharmacists, clinical administrators) for tasks like drug library management, data viewing, and auto-programming would be assessed.
Industry standards: Adherence to standards like IEC 62304 for software and ISO 14971 for risk management implies these standards served as a form of "ground truth" for compliance.

8. The sample size for the training set:

As this is not an AI/ML device that requires a training set in the conventional sense, the document does not mention a training set or its sample size. The software development process aligns with IEC 62304, focusing on verification and validation of designed functionality.

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

Not applicable, as there is no mention of a training set.

Ask a Question

Ask a specific question about this device

Page 1 of 1