Decommissioned due to not enough staff across the study to adequately support it.

Monitoring of Biologic Signals of Elderly Patients

The Body Information Management Systems (BIMS) are advanced biological sensors which monitor human behaviour, activity and basic vital signs. There is a combination of four separate ultra-sensitive weighing scales attached to each of the individual legs of a hospital bed measuring:

• Micro-changes in weight distribution
• Activity/sleep pattern

In addition to the bed-based system, there will be a wrist worn wearable (Samsung fit gear) that will monitor participants when they are both in and out of bed. These are about the size of a small watch, light weight, water resistant, and will be worn 24/7 by participants. These measure:

• Movement and sleep
• Heart rate The set up will remain flexible throughout the study allowing re-design to accommodate the needs of participants and staff.

Device testing

Devices (BIMS & wearables) have been tested at the UoS (University of Surrey) 'living lab', a mock 'real world' environment created to test reliability and communication interfacing between the device, gateway, servers and user interface. The BIMS device itself has been tested to ensure reliability and sensitivity of measurement and communication of data. As part of this pilot, further testing of the sensitivity and specificity of the BIMS system will be undertaken.

The bed sensor system does not have CE certification. MinebeaMitsumi acknowledge and confirm they retain responsibility for the functionality and safety of the BIMS systems and associated equipment.

For the purposes of the present project, the devices are not intended to investigate, diagnose or treat any disease, injury or condition, thus are not defined as a medical device. They collect data for informational purposes only.

This study is interested in the quality of the data produced in an inpatient environment, user experience and workforce impact. The study shall assess the utility of the data in an inpatient environment.

The Samsung fit gear wearable is a mass produced activity tracker product that has been CE marked. Twenty of these wearables have been purchased by SABP to be deployed in technology based projects and are covered by the project's insurance.

4.1.2 The User Interface

Data will be collated via the Internet of Things to a dashboard view (the user interface) which will be monitored by staff on the ward and the research workers on the ward involved in this specific project, including the study specific Research Assistant (RA), Research nurse (RN) and Health Care Assistant (HCA)

As the study progresses, clinical and technical knowledge gained will facilitate the development of algorithms that may in the future have the potential to inform clinical decision-making. Actual changes to any care pathway would be beyond the current scope of this project.

The User Interface

Part of the research aim is to develop a user interface with which to present the data. This is likely to take the form of a user interface that shows the relevant bed being monitored and the relevant biometrics based on changes in weight distribution. Alongside this representation will be additional data derived from any wearable worn by the patient occupying that particular bed.

4.1.3 Deployment

Once the testing of the devices and data flow has been completed within the UoS living lab, the devices will be deployed onto the ward. Members of UoS Technical Team will deploy the devices on the ward and provide any maintenance or trouble-shooting as required.

Mock ward set-ups have been conducted to assess any requirements for changes to the infrastructure and to ensure risk and safety issues have been considered. As a part of this process of continual development, electric leads between the devices are being housed within the structure of the beds, and the device control unit is being secured to the underside or headboard of the bed so it cannot be removed by patients.

Once the investigators are in a position to deploy, the Consultant Psychiatrist and Ward Manager will advise on which patients will be appropriate for the project. Prospective participants will be informed about the project, provided accessible information, and consent will be sought. Deployment will occur within 2 weeks of consent.

4.1.4 Reporting Issues

The BIMS system will be configured in such a way so as to send an automatic communication to the UoS Technical Team when there is a device or data sending issue. User interface issues or suggested developments will be recorded by the Clinical Project Coordinator, Researcher Assistant and Ward staff involved in the project, and transferred to the UoS Technical Team for discussion. There will also be a regular Project Board and Steering Board to update on progress and address any ongoing issues.

4.1.5 BIMS Development

Part of the present project is to develop the BIMS systems alongside feedback provided during the project.

Machine learning algorithms:

Requirements and insights will be provided throughout the project. The clinical team will review received data and advise the UoS Technical Team on development of algorithms and Machine Learning to advance the monitoring and alert functionality. Data from initial stages of the project will facilitate the design of clinically-led and patient/psychiatric condition-driven algorithms. The main work will focus on developing adaptive Machine Learning solutions to assist in extracting and identifying changes in health status, associated patterns and anomalies from the raw data (or pre-processed signals). Over the course of the project it is hoped that the system will indicate data patterns relevant to specific conditions, and detect trends and changes using personalised patient specific algorithms. MinebeaMitsurni will provide the initial signal processing work to UoS, based on prior testing, and SABP will provide clinical expertise to develop how the data is presented.

New algorithms will be tested, deployed and refined based on a review of their impact on understanding of participant presentation and treatment decision-making. Evaluation feedback will be provided to UoS by clinicians and service users. The research nurse will gather the data which will be fed back to UoS.

Where algorithms monitoring certain data fields highlight a change in health status or activity levels, an alert may present on the user interface to indicate to clinicians a change has occurred. These alerts, and their associated algorithms, will be continually developed and refined as the project progresses.

4.1.7 Naturalistic data

Naturalistic data will be gathered with which to make comparisons with those using the BIMS technology. This shall comprise of information regarding the average time of ward stays over the past year.

4.2 Recruitment

Service Users will be approached for consent to the study if they are allocated one of the five beds where devices are installed. They will be given the participant information sheet and given verbal information about the study and given the opportunity to ask questions. If they wish to proceed, they will be asked to sign an informed consent form and baseline information will be obtained.

If the service user allocated to the device bed does not wish to participate in the study, other service users on the ward will be approached to participate. If another individual wishes to provide consent for the study, the devices will be moved onto their bed.

Withdrawal

A participant may withdraw from the study at any point without giving a reason. They may also be withdrawn from the study for reasons including, but not limited to:

1. Withdrawal of consent
2. Death
3. Serious Adverse events affecting participation in the study. This decision will be made between the PI, ward staff, monitors and research worker.
4. Participant becomes ineligible (see Inclusion/Exclusion Criteria)
5. Inability to continue participation due to loss of physical capacity
6. Participants fail to comply with protocol requirements

4.3 Measurements

Service users (BIMS)

• Demographics (including diagnosis)
• HADS - Hospital Anxiety and Depression Scale (weekly) (Zigmond & Snaith, 1983)
• PSQI - Pittsburgh Sleep Quality Index (Weekly) (Buysse, Reynolds, Monk, Berman & Kupfer, 1989)

The HADS is a 14-item scale with 7 questions on anxiety and 7 on depression. A meta- analysis found it to have good internal consistency for the depression scale and anxiety scale with mean Cronbach's alphas of .82 and .83 respectively (Bjelland, Dahl, Haug & Neckelmann, 2002). This scale was chosen as includes both anxiety and depression so negates the need for separate, longer scales.

The PSQI is a 24-item scale measuring sleep disturbance. It has been shown to have good test-retest reliability (.87) and good validity (Backhaus, Junghanns, Broocks, Riemann & Hohagen, 2002).

Clinical information will also be gathered to assess the potential impact of the BIMS system on treatment pathways. This may include medication changes, and severity and progression of illness. Clinical information will also be collected to help explore the future utility of the systems in an inpatient environment.

The total length of stay of each participant will be recorded at the point of discharge.

Due to the developmental nature of the study, the information gathered at each interview may vary due to the present requirements of the study.

Staff

Staff will have two main semi-structured interviews, one at the beginning (their expectations of the devices) and one at the end of the study (their actual experience of the devices).

The interview schedule at the end of the project is subject to change depending on the development of the system during the project. Interviews will be recorded and transcribed for analysis by a member of the sponsors research team. Participants will be informed of this. Interviews have been chosen as the method of data collection to gain as much depth as possible in the data which is necessary for the small sample size of the pilot

In addition to this, staff will be consulted on a fortnightly basis regarding the impact of the BIMS system. If staff leave their role during the study the final interview will happen before they leave and their length of time in the study will be noted.

The chief investigator or an appointed individual will be interviewed on a fortnightly basis regarding the use of the BIMS on the ward including its use by staff any issues encountered and any developments that are required.

Other

In addition to the impact on service users and staff the present project also aims to evaluate the cost of care, resource management and quality of care. This will be derived from qualitative and quantitative analysis on the semi-structured data.

Inclusion Criteria-

For service users:

• Female
• In-patient on Spencer Ward, Abraham Cowley Unit, SABP.
• Presenting with any psychiatric illness
• Willing to provide informed consent
• For participants assessed to lack capacity, a consultee must advise if they believe the individual would provide consent or not

For staff:

• Male or Female
• Staff working on Spencer Ward, Abraham Cowley Unit, SABP.
• Willing to provide informed consent
• 18 years of age or over

For Carers:

• Male or female
• Carer, family member or friend of service user on Spencer Ward
• Willing to provide informed consent
• 18 years of age or over

Exclusion Criteria-

Participants not meeting inclusion criteria