ANZCTR - Registration

Registration number

ACTRN12616001003460

Ethics application status

Approved

Date submitted

23/07/2016

Date registered

29/07/2016

Date last updated

29/07/2016

Type of registration

Retrospectively registered

Titles & IDs

Public title

Efficacy, feasibility and usability of a facial recognition technology integrated mobile app for identifying the presence and severity of pain in institutionalised patients with dementia.

Scientific title

FAcial recognition-based and Computerised Evaluation of PAIN in INstitutionalised DEMENTIA residents

Secondary ID [1]

Nil Known

Universal Trial Number (UTN)

U1111-1185-6973

Trial acronym

FAcial recognition-based and Computerised Evaluation of PAIN in INstitutionalised DEMENTIA residents (FACE of PAIN IN DEMENTIA) Project

Linked study record

Health condition

Health condition(s) or problem(s) studied:

Dementia

Pain

Behavioural problems

Condition category

Condition code

Neurological

Dementias

Intervention/exposure

Study type

Interventional

Description of intervention(s) / exposure

The study consists of 3 phases
I. Pre-implementation phase
1) Face to face training of “Pain Champions” (half day)
2) Retrospective clinical audit (8 weeks prior to the commencement of the clinical trial undertaken by a research assistant employed by the project team)
3) Care staff familiarisation period with the use and operation of the ePAT (2 weeks)
4) Baseline data collection (Week 0)
II.Clinical implementation (8 weeks)
1) Commencement of the clinical trial and introduction of the tool (Week 1-Week 8)
III.Post-implementation (8 weeks)
1) Cessation of the clinical trial and return to standard (normal) practice.
2) Data collection points at Week 9

Intervention: Electronic Pain Assessment Tool (ePAT) is a point of care app which uses automated facial recognition technology in real-time to identify facial micro-expressions indicative of the presence of pain. These data are then used in combination with other non-facial pain cues such as vocalisation, movements and behaviours to automatecally calculate a pain severity score.

The ePAT consists of 6 domains, which are derived from American Geriatric Society's Indicators of Persistent Pain in Older Adults. For each pain assessment encounter, staff member videos the resident's face for 10 seconds which is analysed by the app in real-time to identify facial micro-expressions. Then, staff member navigates through other non-facial domains to answer Yes/No questions for a checklist of indicators, using the app, to arrive at a pain intensity score.

Pain Champions
As a strategy of implementation, nurse from each care home will be nominated as “Pain Champion”. Nomination will occur when an expression of interest of an employee nurse is initially raised with the nurse manager of the care home to participate in the project. According to Knowledge for Health (K4Health) initiative, “champions are charismatic opinion leaders who advocate for a particular program, policy or technology”. Champions should be characterised by “passion, persistence and persuasiveness”. The role of “Pain Champions” is to administer the study where they work and train and encourage the rest of the staff within the care home to perform electronic pain assessments. Within each care home, each champion will be finally selected by the general manager based on certain criteria agreed on by both parties.

Face-to-face Training of Pain Champions will be conducted by the researcher. It is comprised of 2 parts:
1) Theoretical session
This session includes lectures, educational videos and case studies and has the following four modules:

Module 1: Background about pain
Definition of pain, subjective nature of pain, pathophysiology of pain, types of pain, causes of pain, consequences of pain (including behavioural change),

Module 2: Pain in dementia and residential care
Prevalence of pain in dementia, pain in mild dementia, pain in moderate-to-severe dementia, pain assessment tools (self-rating instruments e.g. VDS or FPS), (behavioural or observational (proxy) pain assessment tools e.g. Abbey, ePAT), challenges of currently available pain assessment methods

Module 3: Pain management in residential care
Discussion about current recommendations from clinical guidelines, and different therapeutic options available for pain management in residential care such as pharmacotherapy (e.g. opioids or paracetamol) and non-drug therapy (e.g. message or essential oil therapy)

Module 4: Overview of study protocol for ePAT implementation
The study protocol will be worked through step by step and staff members will have the opportunity to complete data collection forms.

2) Practical session
During this session the following will be conducted:

Module 1: Conditions that might affect the operation of the ePAT

Module 2: Live demo of the ePAT

Module 3: Hand-on practice using the ePAT by staff in pairs (i.e. completing facial assessments and questionnaire checklist domains) and on residents (if possible)

Participants will be sampled using non-probability convenience sampling strategy. Participants will be served as own control and will be subjected to the ePAT as an intervention of measuring pain. The ePAT uses the combination of a digital AGS-based checklist and automated facial recognition technology in a smart device to capture micro-facial expressions, behaviours and physical features indicative of pain. Residents’ pain will be assessed by aged care staff for 8 weeks whenever is suspected during rest and on movement (upon receiving routine care activities known to be associated with pain e.g. assisted transfer, repositioning, bathing and toileting or when a resident performing his/her own daily activities e.g. self-transfer or walking within the care home) as per usual practice. There is no independent control group in this study.

Communication with GPs
GPs will be formally consented to be part of the study as they will be making decisions regarding ePAT scores. Each GP will be briefed about a previous validation study and the objectives and implementation of current trial. Participating GPs will be regularly consulted about ePAT pain scores as per standard practice, with particular focus on those residents with behavioural issues. The outcome of consultation together with respective reasons for selecting therapeutic actions (if any) will also be documented on the Pain Management Record Form. Treating GPs will be under no pressure to recommend anything under the umbrella of the trial but rather will do so when necessary as part of standard care. Feedback about the clinical utility of the tool will be gathered using focus group interview.

Responding to Pain Rating in Patients with Cognitive Impairment during the Clinical Trial
During implementation phase, each pain assessment is to be documented on a Pain Management Record designed for the purpose of this study. Resident’s pain will be reviewed using electronic Pain Assessment Tool (ePAT) unless the Assessor wishes to validate the result obtained against a standard tool such Abbey Pain Scale (APS). Following the ePAT assessment, the standard practice will be followed. Weekly outcome review for pain management for each resident will also be documented

Total number of ePAT assessments for resident sample, number of ePAT assessments per resident, number of ePAT assessments at rest, number of ePAT assessments after a movement or activity will also be recorded.

Ease of use and staff satisfaction are expressed in percentages. Responses (scores of entire tool and subscales) of Clinical Utility Attribute Questionnaire (CUAQ) and Mobile Application Rating Scales (MARS) will be presented as percentages, mean +/-SD and median with interquartile ranges.

Intervention code [1]

Treatment: Devices

Comparator / control treatment

Participants will be served as own control and will be subjected to the ePAT as an intervention of measuring pain. Pain assessments will be conducted during rest (no activity) as a reference point, and upon movements (activity) during the implementation trial. In addition, overall improvement in pain management and secondary outcomes will be compared before, during and subsequent to the introduction of the ePAT as an interventional method of measuring pain.

Abbey Pain Scale (APS) may be used (as a comparator for validation) sometimes by staff during some ePAT data collection points to validate the ePAT score. The use of APS is carried out at the discretion of care staff, when they feel that the ePAT score may not be accurate.

Each pain assessment is to be documented on a Pain Management Record Form designed for the purpose of this study. Resident’s pain will be reviewed using electronic Pain Assessment Tool (ePAT) unless the Assessor wishes to validate the result obtained against a standard tool such Abbey Pain Scale (APS). Following the ePAT assessment, the standard practice will be followed.

Control group

Active

Outcomes

Primary outcome [1]

There is a composite primary outcome in this study which is identifying presence and severity of pain using the ePAT or Abbey Pain Scale or Self -rating (Verbal Descriptor Scale). The choice of scale to use is dependent on the self-rating capacity of residents and the on the adherence of staff to use ePAT or Abbey Pain Scale.

Timepoint [1]

Time points for clinical data collection
1)From the clinical audit 8 weeks before implementation/introduction of the ePAT
2) At baseline (Day 1,Week 1)
3) During implementation phase at Week 4-6 as comparators for change with the ePAT (i.e. change against baseline data) only for residents who are likely to have an attrition due to hospitalisation or death.
4) At Week 9 post-implementation to compare with previous time points of data collection (i.e. comparative change with Week 4-6)
Overall, these data are collected to review changes that occurred in residents’ outcomes over time as a result of using the ePAT in participating care homes and to establish whether practice improvement (if any) has been maintained after withdrawal of the ePAT. Data collected outside of the normal range of data routinely collected by the home will be the conducted by a research assistant at baseline, 4-6 and 9 week periods.

Secondary outcome [1]

Medication (analgesic & psychotropic) use (assessed by Medication Quantification Scale)

Timepoint [1]

1)From the clinical audit 8 weeks before implementation/introduction of the ePAT
2) At baseline (Day 1,Week 1)
3) During implementation phase at Week 4-6 as comparators for change with the ePAT (i.e. change against baseline data) only for residents who are likely to have an attrition due to hospitalisation or death.
4) At Week 9 post-implementation to compare with previous time points of data collection (i.e. comparative change with Week 4-6)
Overall, these data are collected to review changes that occurred in residents’ outcomes over time as a result of using the ePAT in participating care homes and to establish whether practice improvement (if any) has been maintained after withdrawal of the ePAT. Data collected outside of the normal range of data routinely collected by the home will be the conducted by a research assistant at baseline, 4-6 and 9 week periods.

Secondary outcome [2]

Agitation (measured by Cohen-Mansfield agitation inventory)

Timepoint [2]

1)From the clinical audit 8 weeks before implementation/introduction of the ePAT
2) At baseline (Day 1,Week 1)
3) During implementation phase at Week 4-6 as comparators for change with the ePAT (i.e. change against baseline data) only for residents who are likely to have an attrition due to hospitalisation or death.
4) At Week 9 post-implementation to compare with previous time points of data collection (i.e. comparative change with Week 4-6)
Overall, these data are collected to review changes that occurred in residents’ outcomes over time as a result of using the ePAT in participating care homes and to establish whether practice improvement (if any) has been maintained after withdrawal of the ePAT. Data collected outside of the normal range of data routinely collected by the home will be the conducted by a research assistant at baseline, 4-6 and 9 week periods.

Secondary outcome [3]

Aggression (rated by neuropsychiatric inventory-nursing home version (NPI-NH))

Timepoint [3]

1)From the clinical audit 8 weeks before implementation/introduction of the ePAT
2) At baseline (Day 1,Week 1)
3) During implementation phase at Week 4-6 as comparators for change with the ePAT (i.e. change against baseline data) only for residents who are likely to have an attrition due to hospitalisation or death.
4) At Week 9 post-implementation to compare with previous time points of data collection (i.e. comparative change with Week 4-6)
Overall, these data are collected to review changes that occurred in residents’ outcomes over time as a result of using the ePAT in participating care homes and to establish whether practice improvement (if any) has been maintained after withdrawal of the ePAT. Data collected outside of the normal range of data routinely collected by the home will be the conducted by a research assistant at baseline, 4-6 and 9 week periods.

Secondary outcome [4]

Depression (scored by Cornell Scale for Depression in Dementia (CSDD)

Timepoint [4]

1)From the clinical audit 8 weeks before implementation/introduction of the ePAT
2) At baseline (Day 1,Week 1)
3) During implementation phase at Week 4-6 as comparators for change with the ePAT (i.e. change against baseline data) only for residents who are likely to have an attrition due to hospitalisation or death.
4) At Week 9 post-implementation to compare with previous time points of data collection (i.e. comparative change with Week 4-6)
Overall, these data are collected to review changes that occurred in residents’ outcomes over time as a result of using the ePAT in participating care homes and to establish whether practice improvement (if any) has been maintained after withdrawal of the ePAT. Data collected outside of the normal range of data routinely collected by the home will be the conducted by a research assistant at baseline, 4-6 and 9 week periods.

Secondary outcome [5]

Cognition (as rated by Dementia Rating severity Scale)

Timepoint [5]

1)From the clinical audit 8 weeks before implementation/introduction of the ePAT
2) At baseline (Day 1,Week 1)
3) During implementation phase at Week 4-6 as comparators for change with the ePAT (i.e. change against baseline data) only for residents who are likely to have an attrition due to hospitalisation or death.
4) At Week 9 post-implementation to compare with previous time points of data collection (i.e. comparative change with Week 4-6)
Overall, these data are collected to review changes that occurred in residents’ outcomes over time as a result of using the ePAT in participating care homes and to establish whether practice improvement (if any) has been maintained after withdrawal of the ePAT. Data collected outside of the normal range of data routinely collected by the home will be the conducted by a research assistant at baseline, 4-6 and 9 week periods.

Secondary outcome [6]

Quality of life (measured by the Quality of Life in Alzheimer’s Disease (QOL-AD)

Timepoint [6]

1)From the clinical audit 8 weeks before implementation/introduction of the ePAT
2) At baseline (Day 1,Week 1)
3) During implementation phase at Week 4-6 as comparators for change with the ePAT (i.e. change against baseline data) only for residents who are likely to have an attrition due to hospitalisation or death.
4) At Week 9 post-implementation to compare with previous time points of data collection (i.e. comparative change with Week 4-6)
Overall, these data are collected to review changes that occurred in residents’ outcomes over time as a result of using the ePAT in participating care homes and to establish whether practice improvement (if any) has been maintained after withdrawal of the ePAT. Data collected outside of the normal range of data routinely collected by the home will be the conducted by a research assistant at baseline, 4-6 and 9 week periods.

Secondary outcome [7]

Care workers and support staff burden (rated by Care Burden Index).

Timepoint [7]

1)From the clinical audit 8 weeks before implementation/introduction of the ePAT
2) At baseline (Day 1,Week 1)
3) During implementation phase at Week 4-6 as comparators for change with the ePAT (i.e. change against baseline data) only for residents who are likely to have an attrition due to hospitalisation or death.
4) At Week 9 post-implementation to compare with previous time points of data collection (i.e. comparative change with Week 4-6)
Overall, these data are collected to review changes that occurred in residents’ outcomes over time as a result of using the ePAT in participating care homes and to establish whether practice improvement (if any) has been maintained after withdrawal of the ePAT. Data collected outside of the normal range of data routinely collected by the home will be the conducted by a research assistant at baseline, 4-6 and 9 week periods.

Secondary outcome [8]

Length of training time required, assessed by review of training session logs.

Timepoint [8]

Immediately prior to start of implementation phase

Secondary outcome [9]

Usability and quality of the software, assessed by staff using Mobile Rating Scale (MARS) questionnaire

Timepoint [9]

Post-implementation phase

Secondary outcome [10]

Staff acceptability of the ePAT assessment system, assessed by semi-structured interviews with pain champions and using Clinical Utility Attribute Questionnaire (CUAQ ) for all centre staff.

Timepoint [10]

Post-implementation phase

Eligibility

Key inclusion criteria

(1) are 60 years or older; (2) have been diagnosed with moderate-to-severe dementia with a current MMSE score equals 9 or PAS equals 10 and/or HDS less than 160; (3) are living in a high-level care designated dementia unit; (4) are dwelling greater than 3 months in the care home; and (5) possess a documented history of any of the followings: (a) chronic pain conditions such as, but not limited to, joints pain (osteo- and rheumatoid arthritis), osteoporosis, chronic back pain, compression fracture, neuropathic and post-herpetic pain; (b) have had a recent or history of falls; (c) presence of skin infections e.g. shingles, active or chronic wounds e.g. pressure sores and skin tears; and/or (d) have a long-lasting infection(s) such as cellulitis and osteomyelitis.

Minimum age

60 Years

Maximum age

No limit

Sex

Both males and females

Can healthy volunteers participate?

No

Key exclusion criteria

If the resident is clinically unwell, deemed ineligible or if it is inappropriate to include as decided by clinical or aged care staff.

Study design

Statistical methods / analysis

Test-retest analysis pain scores are based on the Pain app during Familiarisation Phase. The Cohen’s Kappa statistic will be used to assess the level of agreement between measurements based on a categorical scale. If the measures are on a continuous scale, then the Intra-class Correlation Coefficient (ICC) will be used. The ICC can also be used to assess agreement between more than 2 raters. The ICC returns a value between zero and 1, with values near 1 indicating good agreement.

Non-probability purposive and convenience sampling will be used to recruit subjects. A target sample of 20-30 subjects per aged care home is aimed for during this study. A subject-to variable ratio of 5:1 appears to be the most reported approach of sample size in the literature. Using this approach (i.e. 5 subjects per one item) and bearing in mind that the ePAT is composed of 42 items, a sample size of at least (5 subjects x 42 items) 210 residents is the target. This number is the minimum value needed before reaching an adequate power of 80% and subsequent statistical significance of less than 0.05. However, failing to retain a subject due to attrition associated reasons (e.g. delirium, hospitalisations, and/or other clinically unstable conditions as well as death) is anticipated during the course of this study. To account for these unforeseen circumstances which may contribute to the reduction of sample size, we have made the choice to recruit an additional number (n=40) of subjects to arrive at a final sample size of 250.

All analyses will be performed by a statistician using the Statistical Package for the Social Sciences (SPSS) software (V.17.0, Chicago, Illinois, USA). Descriptive data of central tendencies measures (such as mean, median and standard deviations) and inferential data such as odd ratio (OR), confidence interval (CI) and regression analyses will be reported in the findings. Regression analyses will be calculated to assess the impact of using the ePAT on secondary outcomes (cognition, behaviour, depression, anxiety, medication use and quality of life). Adjusted and unadjusted ORs for the secondary outcomes will also be reported and assessed for significance using 95% CI.

Sample characteristics (e.g. age, gender, ethnicity, and years of employment in the care home) of Assessors involved in the project will be reported in descriptive statistics: mean, median, mode, standard deviation (SD) and range. Training length, administration time and observation period are expressed in units of time (e.g. minutes, seconds), which will also be reported in descriptive statistics e.g. mean time taken to do ePAT assessment. Other feasibility data such as total number of ePAT assessments for resident sample, number of ePAT assessments per resident, number of ePAT assessments at rest, number of ePAT assessments after a movement or activity will also be presented. Ease of use and staff satisfaction are expressed in percentages. Responses (scores of entire tool and subscales) of Clinical Utility Attribute Questionnaire (CUAQ) and Mobile Application Rating Scale (MARS) will be presented as percentages, mean +/- SD and median with interquartile ranges. A receiver operating characteristic (ROC) analysis will be used to establish cut-off pain severity scores .

Recruitment

Recruitment status

Recruiting

Date of first participant enrolment

Anticipated

Actual

6/06/2016

Date of last participant enrolment

Anticipated

1/03/2017

Actual

Date of last data collection

Anticipated

1/07/2017

Actual

Sample size

Target

250

Accrual to date

Final

Recruitment in Australia

Recruitment state(s)

WA,VIC

Recruitment postcode(s) [1]

6052 - Inglewood

Recruitment postcode(s) [2]

3106 - Templestowe

Recruitment postcode(s) [3]

3058 - Coburg

Recruitment postcode(s) [4]

3088 - Greensborough

Recruitment postcode(s) [5]

3550 - Bendigo

Recruitment postcode(s) [6]

3564 - Echuca

Recruitment postcode(s) [7]

3444 - Kyneton

Recruitment postcode(s) [8]

3442 - Woodend

Recruitment postcode(s) [9]

3162 - Caulfield

Recruitment postcode(s) [10]

3690 - Wodonga

Recruitment postcode(s) [11]

6102 - Bentley

Funding & Sponsors

Funding source category [1]

Charities/Societies/Foundations

Name [1]

Alzheimer's Australia Dementia Research Foundation

Country [1]

Australia

Funding source category [2]

Commercial sector/Industry

Name [2]

EPAT PTY LTD

Country [2]

Australia

Funding source category [3]

Commercial sector/Industry

Name [3]

Brightwater Aged Care Group

Country [3]

Australia

Primary sponsor type

University

Name

Curtin University

Country

Australia

Secondary sponsor category [1]

Commercial sector/Industry

Name [1]

EPAT PTY LTD

Country [1]

Australia

Ethics approval

Ethics application status

Approved

Ethics committee name [1]

Human Research Ethics Committee (HREC), Curtin University

Ethics committee address [1]

Curtin University
GPO Box U 1987
Perth, Western Australia 6845

Ethics committee country [1]

Australia

Date submitted for ethics approval [1]

07/01/2014

Approval date [1]

18/03/2014

Ethics approval number [1]

HR 10/2014

Ethics committee name [2]

Brightwater Research Centre-Ethics Committee

Ethics committee address [2]

Level 3, 355 Scarborough Beach Road
Osborne Park WA 6017

Ethics committee country [2]

Australia

Date submitted for ethics approval [2]

24/04/2016

Approval date [2]

02/05/2016

Ethics approval number [2]

Summary

Brief summary

Pain management amongst patients with cognitive dysfunction is problematic, particularly in those cases where the carer is given the responsibility of determining whether or not pain relief is required.  Numerous pain scales have been developed to assist carers assess patients’ pain status; most of which include items derived from widely accepted guidelines of persistent pain .  This study aims to develop and assess the feasibility of using face recognition software linked to those pain indicators on a Samsung Note 3 to enhance pain management amongst a population of cognitive impaired residential aged care residents.  Surrogate markers of pain control will be used including cognition and behaviour.

Trial website

Public notes

Contacts

Principal investigator

Name

Prof Jeff Hughes

Address

School of Pharmacy
Curtin University
GPO Box U1987
Perth Western Australia 6845

Country

Australia

Phone

+618 9266 7367

Email

[email protected]

Contact person for public queries

Name

Mustafa Atee

Address

School of Pharmacy
Curtin University
GPO Box U1987
Perth Western Australia 6845

Country

Australia

Phone

+618 9266 2527

Email

[email protected]

Contact person for scientific queries

Name

Mustafa Atee

Address

School of Pharmacy
Curtin University
GPO Box U1987
Perth Western Australia 6845

Country

Australia

Phone

+618 9266 2527

Email

[email protected]

Trial Review

Documents added manually

Documents added automatically