Optimized Novel Wireless Assessment of Respiratory and Circulatory Dirstress - Alarm Validation

Validation of the Accuracy and Relevance of Continuous Vital Sign Monitoring Using an AI-algorithm in Hospitalized Patients

The ONWARD study evaluates user experiences from continuous vital sign monitoring alarms sent to the clinical staff in case of deviating vital signs at the general hospital ward. User experiences are evaluated though a 5-point likert questionnaire.

The ONWARD study evaluates user experiences from continuous vital sign monitoring alarms sent to the clinical staff in case of deviating vital signs at the general hospital ward. User experiences are evaluated though a 5-point likert questionnaire.

The study will focus on adult, hospitalized patients who are monitored at the general hospital ward with continuous, real-time vital sign monitoring combined with the WARD-CSS (WARD clinical support system) that relays alerts to the clinical staff in case of significant deviations. The evaluation focus on the user perceived relevance and usefulness of the alerts generated by the system which includes a proprietary novel alerts logic. Specifically, investigators aim to assess how users evaluate the clinical relevance of triggered change-based alerts and the helpfulness of the accompanying alert descriptions.

During monitoring, the app will be used according to the intended use and additionally two questions will be prompted if alerts are sent.

Patients monitored will have an expected length of stay at the hospital, which supports at least one day of continuous monitoring. Readings from the WARD-CSS is not blinded for treating staff to get feedback on the use of the application. This study will include up to 100 patients.

An alerting feedback questionnaire will be constructed as simple feedback buttons/sliders, prompted after an alert is 'accepted' in the app. The feedback buttons will only be prompted for a representative proportion of the alerts. This will be done at random based on the alert frequencies determined in the retrospective analysis. This way, a close to equal amount of feedback responses will be prompted and obtained for different alert types with different occurrences. The feedback will be linked to the alert in the app, such that feedback can be submitted at any point in time, thus not disturbing the nurse while medical intervention is provided.

The feedback will be constructed with the following two interactions for the clinical staff.

• Alert relevance in relation to deterioration of the patient (5-point Likert scale).
• Relevance of the description following alerts in relation to the helpfulness in the patient assessment/care (5-point Likert scale)

Inclusion criteria:

Patients can be included if they meet all the following criteria:

• Age ≥ 18 years
• Patients assessed by the clinical staff as having an acute condition with risk for clinical deterioration.
• Expected admission for more than 24 hours to a general ward.

Exclusion criteria:

Patients are excluded if they meet any of the following criteria:

• Having implanted electronic devices (such as pacemaker or Implantable Cardioverter Defibrillator (ICD) device).
• Known pregnancy or breastfeeding
• Inability to give informed consent.
• The participant expected not to cooperate with study procedures.