Search Results

WARD-CSS is a clinical decision support system that remotely integrates, analyzes and displays continuous vital sign data (via a mobile or web application) from medical devices for nonpediatric hospitalized patients within non-critical care units.

WARD-CSS uses a set of standardized rules based on scientific and clinical evidence to detect and alert on clinically relevant vital sign deviations when used by trained health care professionals in hospitals.

WARD-CSS is not intended to replace current monitoring practices or replace health care professionals' judgment. WARD-CSS is a tool intended to help health care professionals manage monitored patients and make clinical care decisions.

WARD-CSS is a stand-alone software intended for use in continuous monitoring of patients and near real-time analysis of vital signs for the purpose of notifying healthcare professionals in case of clinically relevant vital sign deviations.

WARD-CSS utilizes knowledge-based algorithms to evaluate clinically relevant vital signs deviations to help drive clinical management.

The system is intended to be used as an adjunct to current monitoring practice in the general med/surg floors of the hospitals

The system assists healthcare professionals when monitoring patients on their wards by:

• Providing a real-time monitoring overview of vital signs for all patients. ●
• Alerting the healthcare professionals when a patient deteriorates. ●

The following types of alerts are detected by WARD-CSS in the vital sign data:

• Desaturation ●
• Hypertension
• Hypotension ●
• Bradypnea ●
• Tachypnea ●
• Tachycardia
• Bradycardia
• Hypotension and Bradycardia
• Hypotension and Tachycardia ●
• Bradypnea and Desaturation
• Fever

The WARD-CSS consists of a Mobile App, Web App and Backend Server. The Mobile App is used by healthcare professionals (HCPs) to monitor patients. The HCP will receive notifications of the alerts to their mobile phones. Within this app, the HCP can also document vital signs into an electronic health record system. The Web App is used by administrative users to manage hospitals, wards, users, and monitors. The Backend Server is used to receive and process all incoming data and manage all data used in the apps.

The provided text describes the acceptance criteria and a study to prove the device, WARD-CSS, meets these criteria, primarily focusing on alert reduction.

Here's the breakdown of the information requested:

1. Table of Acceptance Criteria and Reported Device Performance

The acceptance criteria here are implicitly related to the reduction of alert overload, as this is the primary focus of the clinical testing described for WARD-CSS. The performance is measured by the reduction in alert rates compared to a baseline (thresholds only) and an intermediate step (thresholds with time durations).

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

• Sample Size: 794 patients
• Data Provenance: Retrospective analysis of four cohorts from prospective clinical safety studies conducted from 2020-2024. The country of origin is not explicitly stated, but the submission is for the US FDA, implying an interest in data relevant to this regulatory body. The data consists of vital sign data.

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

This information is not provided in the document. The study focuses on quantifying alert rates based on set thresholds and algorithms rather than human expert-established ground truth for specific events that led to the alerts. The 'ground truth' here is the objective vital sign data and the predefined rules/thresholds that trigger alerts.

4. Adjudication Method for the Test Set

This information is not provided as the study's focus is on algorithmic alert reduction based on vital sign data and predefined rules, not on expert adjudication of alert significance.

5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study Was Done, and Effect Size of Human Improvement with AI vs. Without AI Assistance

A MRMC study comparing human readers with and without AI assistance was not done. The study's objective was to quantify the reduction in system-generated alerts due to the WARD-CSS algorithms, not to measure human performance improvement.

6. If a Standalone (Algorithm Only Without Human-in-the-Loop Performance) Was Done

Yes, the study described is a standalone (algorithm only) performance assessment of the WARD-CSS system's alert reduction methodology. It retrospectively analyzes vital sign data against the different alert generation methodologies (thresholds only, thresholds + time durations, and WARD-CSS's full algorithm including alert filters) to demonstrate the reduction in the number of alerts produced by the algorithm itself.

7. The Type of Ground Truth Used

The ground truth used in this study is objective vital sign data combined with predefined, standardized rules and thresholds based on scientific and clinical evidence. The analysis quantifies how often these pre-defined rules would trigger an alert under different algorithmic conditions (basic thresholds, thresholds with time durations, and thresholds with time durations and alert filters). It is not based on expert consensus, pathology, or outcomes data in the traditional sense of clinical event validation.

8. The Sample Size for the Training Set

The document does not explicitly mention a separate training set size. The clinical testing section refers to a "literature review to support software algorithm development and determine the alert thresholds," suggesting that the rules and thresholds were established based on existing clinical knowledge and literature. The 794 patients were used for the retrospective analysis of alert rates, which effectively acts as a test/validation set for the alert reduction logic rather than a dataset for algorithm training in a machine learning sense.

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

Since an explicit training set (for machine learning) is not detailed, the "ground truth" for the algorithm's rules and thresholds was established through a "literature review to support software algorithm development and determine the alert thresholds." This implies that the rules are based on scientific and clinical evidence from medical literature.

Ask a specific question about this device