The potential of technology for self-management of long term self management of long term

conditionsThe SMART Consortium

www.thesmartconsortium.orgwww.thesmartconsortium.org

VisionVision• “over the next few years we will give 100,000 people

with long-term conditions the opportunity to manage g pp y gtheir care as 'expert patients’. And during 2008 we will bring forward a patients' prospectus that sets out how we will extend to all 15 million patients with a chronic por long-term condition access to a choice of 'active patient' or 'care at home' options - clinically appropriate to them and supported by the NHS”, pp p pp y ,

Prime Minister 2008

• Whole System Demonstrators –> SMART-2

Technology Landscape Technology Landscape

• HealthSpace NHS Choices – patient • HealthSpace, NHS Choices patient information Map of Medicine evidence based • Map of Medicine – evidence based content, treatment pathwaysH lth V lt • Health Vault

• Continua Alliance • Existing platforms

– doc@HOME, Philips Motiva, Intel Shimmerdoc@HOME, Philips Motiva, Intel Shimmer

HealthSpaceHealthSpace

Motiva / DOCOBO

Personal Self Personal Self s fs fManagement System Management System

(PSMS)(PSMS)

Policy and PracticePolicy and Practice

• “there is considerable national and international evidence to show that supporting self care results in health benefits for the people and therefore overall gain for the care system.” Department of Health g y p

• Over 400 studies worldwide report self management can lead to dramatically improved outcomes for patients British Medical Association patients. British Medical Association

• Expert Patient Programme (1000 participants), 4-6 months after the course (a) GP consultations decreased by 7% (b) Outpatient visits decreased by 10% (c) A&E by 7%, (b) Outpatient visits decreased by 10%, (c) A&E attendances decreased by 16%, (d) Pharmacy visits increased by 18%.

Policy and PracticePolicy and Practice

• “the evidence supporting the impact of self-pp g pmanagement and other chronic disease management initiatives on heath service utilisation is more equivocalthan policy statements often imply.”

• “Lay-led self-management education programmes may lead to small, short-term improvements in participants'

lf ffi lf t d h lth iti t self-efficacy, self-rated health, cognitive symptom management, and frequency of aerobic exercise. There is currently no evidence to suggest that such programmes improve psychological health symptoms or programmes improve psychological health, symptoms or health-related quality of life, or that they significantly alter healthcare use.” Cochrane review, 2007

Behavioural changeBehavioural change

• 20% ready to change 40% plan to change 40% in• 20% ready to change, 40% plan to change, 40% in denial/resistance

• Model of behavioural change• Stage of change will impact on information provided,

how presented etc.• PSMS feedback loops to validate change would be p g

helpful• Motivational aspects• Track knowledge absorption retainment and• Track knowledge absorption, retainment and

establish stage of readiness => changing intervention

Key Research Questions

• Can the impact of long term conditions be

Key Research Questions

p geffectively monitored, modelled and analysed using technology?

• Can an assistive technology solutions be identified to deliver self management interventions in key target high volume interventions in key target, high volume, chronic conditions?

Can technology remote from a therapist • Can technology, remote from a therapist, promote health behaviour change?

Can technology which situates behaviour • Can technology which situates behaviour change in everyday life improve traditional self-management strategies?

C 3 M ‘ t k ’Case 3: Mary – ‘stroke’

"Y 't gi ' g t t t h g f d i g thi g Patient's

story

• "You can't give up, you've got to prepare to change your ways of doing things and you cannot afford to dwell on it."

• Poor control of upper limb, reduced i ti l t Clinical

needsproprioception, neglect

Patient's • Improving dressing

Patient s life goal

p g g

Therapy

• Restorative and adaptive, Education, Neuro-developmental Techniques Weight-bearing

i i t d h d i f db kpyintervention exercises on wrist and hand using sensory feedback

Technology• Furniture sensors, clothing sensors

ec ology

Evidence base - Search Strategy Terms –‘ h i i ’‘chronic pain’

• Search terms were produced with guidance from previous

Overhead Term Search Terms

Technology "distance" OR "remote" OR "internet" OR with guidance from previous systematic reviews (e.g. Murray, Burns, See Tai & Nazareth, 2005) and by

Technology distance OR remote OR internet OR "on-line" OR "online" OR "technology" OR "telemedicine" OR "telecare" OR "e health" OR "e-health" OR "ehealth" OR "computer" OR "multimedia" OR "interactive" OR "software" OR "cd-rom" OR "medical

dissecting the main components of the research questions.

so twa e O cd o O ed cal informatics"

Behaviour "behaviour" OR "behavior" OR "behavioural" OR "behavioral"

Ch

• Databases PubMed, PsycInfo, MedLine, and EMBase were searched.

Change"change" OR "changes"

Therapy "therapy" OR "outcome" OR "treatment" OR "program" OR "programme" OR "intervention" OR "assistance" intervention OR assistance

Health Condition "health" OR "illness" OR "disease" OR

"condition“ OR "pain"

Levels of AgreementLevels of Agreement

• There was 91% agreement between initial ratings

Level of Agreement Between Researchers

L l f T t l T t l P ibl N P t between initial ratings• Articles with discrepant

ratings were discussed and agree upon by both

Level of Agreement

Total Agreement [Include]

Total Agreement [Exclude]

Possible Inclusion

No Agreement

Percentage Agreement

Number of Articles

138 1,415 180 79 91%

g p yresearchers

• Two categories of articles with 100% agreement were h l d

Second Filtering of Discrepant Ratings –Both Researchers in Agreement

then compiled:– Definite articles– Articles dependent on

h ti

Definite Articles 194

Dependent on Question 94

research question

Monitor Daily Activities

Breathlessness

Fatigue

Sleep PatternsBlood Pressure

Fatigue

Vital Signs

Stroke

(CHF)

Congestive Heart Failure

(CHF)

Chronic Pain

(CHF)(CHF)

Weight

GPSFunction & Balance

Upper Limb Position

QuestionsUptime/Downtime Measurements

Lower Limb Position

Gait

Stoke CHF Pain Technology AvailMonitor Daily Activities

Stoke CHF Pain

Breathlessness -- CHF --Blood Pressure -- CHF --Sleep Patterns -- -- Pain Bed Sensors

Fatigue -- CHF --Vital Signs -- CHF -- Bluetooth ECG Monitor

W i h CHF Bl t th S lWeight -- CHF -- Bluetooth Scales

Function & Balance Stoke -- -- Wii Matt | Smart Shoes

Upper Limb Position Stoke -- -- Accelerometers | Mulle

GPS Stoke CHF Pain Mulle

Questions Stoke CHF Pain PDA | Touchscreen

Uptime/Downtime Stoke -- Pain Pedometer | Motion SensorsUptime/Downtime Measurements

Lower Limb Position Stoke -- -- Accelerometer | Smart Shoe

Gait Stoke -- Pain Smart Shoe

Example Sensorsp

Bed SensorI b d ith 4 l l f

Sensor Proposed Use in Project

F ll D t t F ll d t t • In‐bed sensor with 4 levels of sensitivity

• Up to 6 monitored periods per dayI t l 169 48125 & 433 92 MH

Fall Detector Fall detectors

Flood Detector Useful if we wish to determine if someone is doing the dishes or measuring water flow

PIR Will be useful to detect movements in different areas of the room• Integral 169.48125 & 433.92 MHz 

transmitters• 1 year battery life• Out of bed sensor or bed in bed 

different areas of the room

Bed Occupancy Mats

Useful to detect if someone is sitting on a chair or standing in a specific position.

Pill Dispenser Used to manage the dispensing of medication

Client Wandering Detects if a user has left the housesensor• Interface dimensions: 35x85x170 mm (HxWxD)

• Interface weight: 220 grams

Client Wandering Keypad

Detects if a user has left the house

Door Contacts Useful to detect the opening of cupboards, doors, fridge, etc.

Panic Button Useful to generate warning messagea c utto Use ul to ge e ate wa g essage

Sayphone 21 Home based hub system

Personal Pendant See panic button

Personal panic button

As abovebutton

Remote controlled Mains

Can be used to turn on and off equipment in the environment

Accelerometers

MTi‐G (Xsense):MTi G (Xsense): • real‐time computed GPS‐enhanced attitude/heading and 

inertial enhanced position/velocity data 

• GPS integration overcomes typical IMU/AHRS challenges 

• high update rate (100 Hz) 

i di id ll lib d f i li• individually calibrated for temperature, 3D misalignment and sensor cross‐sensitivity

Philips wireless accelerometersPhilips wireless accelerometers

(smart target)

Smart EnvironmentSmart Environment

PSMS – Conceptual DesignPSMS Conceptual DesignApplication Layer

Service Layer

Feedback

Healthcare Stationary MobileHealthcare Professional

Stationary Device

Mobile Device

Life Goals

Service

Service Service

Service

Service

MySQL Database / Server

Sensor Platform Layer

Pain Stroke CHF

Physical Layer

Sensor SensorSensor Sensor Sensor Sensor SensorSensor Sensor Sensor

Application LayerWhat feedback?

Service Layer

Feedback

What information should the carer 

Healthcare Professional

Stationary Device

Mobile Device

Life Goals

see?What services and how many?

Service

Service Service

Service

Service

MySQL Database / ServerAutomatic or therapist?

Sensor Platform Layer

Pain Stroke CHF

Which sensors and

Physical Layer

Sensor SensorSensor Sensor Sensor Sensor SensorSensor Sensor Sensor

Which sensors and how many?

Information PathwaysEnters Details

UserProfile

Life Goal Library

Therapist and Patient enter life goals

System assigns life goals based on profile matching

Patient enters own life goals

System Interaction

Application Layer

FeedbackAudio – Visual ‐ Sensory Life Goal

Attend Church

User Interaction

Service Layer

Healthcare Professional

Stationary Device

Mobile Device

Life Goals

MySQL Database / S

MySQL Database / Server

IF Day == Sunday AND patientLocation!= Outside THEN sendFeedback

Sensor Platform

Sensor Platform Layer Stroke

Physical Layer

Gait Server

No

Smart ShoeUpper‐limb Bracelet

Mobile Device PIR Sensor

It’s Sunday, wouldn’t you like to go to church?

Yes

Design prototypesg p yp

Decision Support SystemDecision Support System

• Objectives• Objectives– Support decision making in justification of

life goal/care planlife goal/care plan

• TasksMining patterns/associates from a large – Mining patterns/associates from a large dataset of patient activity sensor data/vital sign/self reports (datasets required)sign/self reports (datasets required)

– Making suggestions to users

Information Flow of Decision Support System

Identify Value

Life Goal pain/stroke/CHF

Justify

Vital SignsActivity Decision

Support

Def

Over a period of time

MonitoringSelf Report

Support System

fine

Real time

Reminder/Feedback

Users

Real time

Decision Support - activityDecision Support activity• Investigating basic framework for Decision Support

Techniques for data analysis feature selection data mining and data – Techniques for data analysis, feature selection, data mining and data classification.

• Study 1: investigating Pain data:Th t g t t t t t t f ll– Three stages: pre-treatment, treatment, follow-up

– Current work is investigating the data patterns associated with these three stages

• Study 2: analysing activity data• Study 2: analysing activity data– Activity recognition

AcknowledgementsAcknowledgements

• University of Sheffield: Simon Brownsell, Annette y ,Haywood

• Sheffield Hallam University: Gail Mountain, Sue Mawson, Peter Wright, Nasrin Nsar, Silvia Torsi, Ruth Mawson, Peter Wright, Nasrin Nsar, Silvia Torsi, Ruth Moore, Jack Parker, Ben Heller

• University of Bath: Chris Eccleston, Ed Keogh, Kevin Vowles Lance McCrackenVowles, Lance McCracken

• University of Ulster: Norman Black, Chris Nugent, Paul McCullagh, Huiru Zheng, William Burns

• Advisory Group: Peter Lansley, William Maton-Howarth, Paul Gee, Chris Clark, Louise Tillman, Nigel Barnes, Julia MacLeod, Peter Moore

Extra SlidesExtra Slides

1. Is technology effective at achieving behaviour change?

Research Questions

Adjunct Proxy

What types of behaviour changes have been achieved through technology?

2. What technology has been employed to change behaviour?

Are specific features of technology associated Is the technology transferable to a rangeAre specific features of technology associated with specific behavioural outcomes?

3. Is technology more successful, less successful, or of equal success in comparison with traditional/non technological techniques?

Is the technology transferable to a range of conditions or limited to specific conditions?

traditional/non-technological techniques?

Pros Cons

4. What variables mediate the effectiveness of technology?

Features of technology:e g training req irement diffic lt of se

Features of the individual:e g demographic info beliefs e perience

Location of use:Clinic or home basede.g. training requirement, difficulty of use e.g. demographic info, beliefs, experienceClinic or home based

Additional Support

Potential Data to be C ll t dCollected

Pain (Accommodative) Stroke (Restorative) CHF (Preventative)

pain measurement activity diary reporting relating breathlessness p{categorical data};

sleep and rest{categorical data};

compliance to medication{categorical data};

y y p g gto specific life goals

{categorical data & free text description};

{categorical data }; fatigue {categorical data};activity diary reporting{categorical data and free{categorical data};

activity diary reporting{free text};

{categorical data and free text description};

timing of activities of daily living

gross motor activity{accelerometer or

gross motor activity{accelerometer orliving

{numerical & categorical data};

{accelerometer or categorical data };

wrist sensor for upper limb activity{accelerometer data};

b d f l it i

{accelerometer or categorical data };

sleep patterns{categorical data};

activities of daily living{ t i l d t }bed sensor for sleep monitoring

{sensor};{categorical data};

weight {electronic scales};

blood pressureblood pressure{monitor};

heart rate{monitor};

Electronic

Healthcare Professionals

Electronic Health

RecordAdvice & feedback

Client P fil

Life Plan & G l

Professionals

Single Assessment

Client

Profile& Goals Assessment Process

LMS Medical WWSClient specific monitoring Carers

Expert SYSTEM Expert management

system STRUCTURE

Touch Screen Devices for D t E tData Entry

HTC P3300• 130 g• Touch screen

(finger/stylus)

iiyama ProLite T1730SR-B 17

• Touch Screen• 17 LCD Panel

• Few buttons• Mobile phone(GSM)• GPRS/WLAN• Built‐in GPS (SiRF III)

• 5 ms• 800:1 CR• USB and Serial• DVI and VGA

• Windows Mobile 5.0 PocketPC

• 2 x 1w Speakers• Black