Remote Monitoring and Analysis of Gait and Falls Within an Elderly Population (4279)

The investigators aim to do this initial pilot study as an observational prospective cohort study, evaluating elderly patients who have capacity in National Health Service (NHS) rehabilitation and community hospitals. The patients will each be recorded doing simple activities of daily living in two 2 hour sessions using a discrete wireless device. This will generate anonymous data set that can be used to train and refine our machine learning algorithm.

Study Overview

• Status: Unknown
• Conditions: Hip Fractures; Fall; Fall Injury; Accidental Fall
• Intervention / Treatment: Diagnostic test: CUSH

Detailed Description

Phase 1 - Creation The creation of a databank of measurements during stereotypical actions of daily living.

Population recruited (P1 recruits) Patients to complete a demographic questionnaire based on historical falls risk data.

Recruited patients allocated to a mobility/independence subgroup based on the information obtained from the demographic questionnaire completed at the time of recruitment.

Patients may be reclassified to different subgroups to redistribute them according to other factors including frailty, exercise tolerance, balance etc.

Patients given a structured scripted set of simple activities to carry out (such as walking, sitting and picking something off the floor etc.) to cover as many activities of daily of living as possible.

Their actions will be monitored with inertia measurement units (IMU) 1-3 per patient (Each will be off the shelf European Conformity (CE) approved products no bigger than a modern smart phone).

Data is transmitted from the IMUs wirelessly to the investigators' laptop, patients will not be attached or anchored to any secondary devices or cables. The IMUs are attached to the patients with a comfortable lightweight belt design and or non-invasive leg bands.

These scripted sessions will be repeated 5 to 10 times per subject to create an anonymous databank (D1). This will form a reference range per subgroup of "truth" data for the algorithm to search.

Patient identifiable data will be limited to the study questionnaires which will be stored securely at the corresponding study site.

Phase 2 - Validation The validation of the databank while observing previously recorded and not previously recorded subjects in an unscripted session.

Population patients included All patients from phase 1, new patients recruited (P2 recruits)

Newly recruited patients (P2) to complete a demographic questionnaire based on historical falls risk data. These patients will not be immediately allocated to a mobility/independence subgroup, but labelled as unknown (x).

All patients observed individually during daily activities in an unscripted session. Patient to wear single IMU sensor which captures mobility data (D2). The activities the subject undertakes during D2 are manually entered onto timesheet by observer for later cross referencing.

2.1 Patients (P2) who are previously unrecorded and who are unallocated to a mobility/independence subgroup have their unscripted data (D2) analysed by the motion algorithm which stratifies them to a mobility/independence subgroup based on the likeness of their data to that stored in the databank (D1). They are then allocated to a mobility/independence subgroup based on their questionnaire data and the two outcomes are compared to see if they match or differ.

- Can algorithm predict "truth"

2.2 Patients (P1) who have previously been recorded in phase 1 and who are allocated to a mobility/independence subgroup based on their questionnaire data have their unscripted data (D2) from phase 2 analysed by the motion algorithm. This then stratifies them again to a mobility/independence subgroup based on the likeness of their new data (D2) to that stored in the databank (D1). To see if their outcome remains the same or differs.

- Can algorithm confirm "truth"

(See Appendix B. for diagrammatically representation of phase 1 and phase 2.) During the unscripted session (D2) any falls or patient reported loss of balance that occurs will be recorded and time stamped by the observer. These data sets will be labelled as "Falls truth" or "Balance truth" data respectively and added to the anonymous databank for cross-reference.

• Study Type: Observational
• Enrollment (Anticipated): 40

Contacts and Locations

• Study Contact: Name: Kalon Hewage, MBBS BSc; Phone Number: +447776328379; Email: kalon@cushhealth.com
• Study Contact Backup: Name: Sam Fosker, BMBS; Phone Number: +447714091611; Email: Sam@cushhealth.com

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: 65 years and older (OLDER_ADULT)
• Accepts Healthy Volunteers: Yes
• Genders Eligible for Study: All

• Sampling Method: Non-Probability Sample

Study Population

At risk individuals who are able to consent to take part in study.

Description

Inclusion Criteria:

• Aged over 65 years of age
• Able to give informed consent
• Able to mobilise independently or with mobility aid (walking stick, Zimmer frame etc.)

Exclusion Criteria:

• Patients under the age of 65.
• Patients who are bedbound or wheelchair bound.
• Patients with cognitive impairment and are unable to give informed consent.
• Significant medical co-morbidities that make participation in the study unsafe.

Study Plan

How is the study designed?

Number of groups / cohorts

6

Cohorts and Interventions

Group / Cohort	Intervention / Treatment
1; Capable of independent of daily activities and able to mobilise unaided. No previous of falls.	Diagnostic test: CUSH; Machine learning assisted remote monitoring/ telehealth platform to predict and prevent falls
2; Capable of independent of daily activities and able to mobilise unaided. With a previous of atleast one fall.	Diagnostic test: CUSH; Machine learning assisted remote monitoring/ telehealth platform to predict and prevent falls
3; Requires help with most daily activities, mobilises with a single walking stick. No previous falls.	Diagnostic test: CUSH; Machine learning assisted remote monitoring/ telehealth platform to predict and prevent falls
4; Requires help with most daily activities, mobilises with a single walking stick. With a previous of atleast one fall.	Diagnostic test: CUSH; Machine learning assisted remote monitoring/ telehealth platform to predict and prevent falls
5; Requires help with most daily activities. Mobilise with frame or roller frame. No previous falls.	Diagnostic test: CUSH; Machine learning assisted remote monitoring/ telehealth platform to predict and prevent falls
6; Requires help with most daily activities. Mobilise with frame or roller frame. Previous history of at least one fall.	Diagnostic test: CUSH; Machine learning assisted remote monitoring/ telehealth platform to predict and prevent falls

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Validation of gait analysis algorithm by use of Inertial Measurement Unit (IMU)	Validation of a motion analysis algorithm to predict falls risk in an elderly population using IMU data captured from accelerometers and gyroscopes within the IMU. The IMU will allow capture of data from wearers gait and movement.	3 months

Collaborators and Investigators

• Sponsor: CUSH Health Ltd.
• Investigators: Study Director: Sam Fosker, BMBS, CUSH Health; Study Chair: Kalon Hewage, MBBS BSc, CUSH Health

Study record dates

Study Major Dates

• Study Start (ANTICIPATED): September 12, 2018
• Primary Completion (ANTICIPATED): October 31, 2018
• Study Completion (ANTICIPATED): September 30, 2019

Study Registration Dates

• First Submitted: August 31, 2018
• First Submitted That Met QC Criteria: September 19, 2018
• First Posted (ACTUAL): September 21, 2018

Study Record Updates

• Last Update Posted (ACTUAL): September 21, 2018
• Last Update Submitted That Met QC Criteria: September 19, 2018
• Last Verified: September 1, 2018

More Information

Terms related to this study

• Additional Relevant MeSH Terms: Fractures, Bone; Wounds and Injuries; Leg Injuries; Femoral Fractures; Hip Injuries; Hip Fractures
• Other Study ID Numbers: CUSH DHTC 4279

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: NO

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No