Data, Digital and AI strategic event slides

Welcome
Strategic Event
Digital, Data and AI
Tuesday 19th November 2024
Axis Conference Centre
Southampton Science Park
Chilworth
SO16 7NP

Matthew Guy
Voice of a carer: diabetes focus

Voice of a carer:
diabetes focus
Matthew Guy
Technology Co-Lead for Wessex CYP Network
Clinical Scientist and Healthcare Designer
Digital Diabetes Research Leaders Programme Fellow,
University Hospital Southampton
Diabetes Technology Manager, SE CYP Diabetes
Transformation Programme, NHSE
Honorary Consultant Clinical Scientist, Southern Health
Honorary Academic Lecturer, University of Bristol

Voice of a carer:
diabetes focus
Matthew Guy
Father of Janki

Voice of a carer:
diabetes focus
Matthew Guy
Carer of Janki

Data, Digital and AI strategic event slides

Over ten years:
~500 glucose
sensor insertions
~1,500 cannulas
1,000s of finger
prick glucose and
ketone checks
1,000s of insulin
pen injections

Over ten years:
~500 glucose
sensor insertions
~1,500 cannulas
1,000s of finger
prick glucose and
ketone checks
1,000s of insulin
pen injections
Over 500,000
additional health
decisions
(~180 per day)

Hybrid Closed Loop Systems
SDM Tool: Courtesy of UHS https://www.england.nhs.uk/publication/decision-support-tool-making-a-decision-about-managing-type-1-diabetes/

Learning
Tandem T-Slim IQ
(and Tandem Mobi)
Medtronic 780G Insulet Omnipod 5 CamAPS FX
Predicts 30 minutes ahead Predicts up to 2 hours ahead Predicts 60 minutes ahead Confidence based prediction
Does not “learn the patient” Does “learn the patient”* Does not “learn the patient” Does “learn the patient”
Uses TDD and body weight to
moderate adjustments
Learns user profile using TDD over past
2-6 days
Learns your total daily dose updated
every new pod
Learns complete user profile: tries to
learn responses to insulin & carbs

What does that mean?
Tandem T-Slim IQ
(and Tandem Mobi)
Medtronic 780G Insulet Omnipod 5 CamAPS FX
If your parameters are close to
“correct” it can do a great job of
steering 24/7
“Bullet-proof”
Ignores chaos and keeps trying to get
back to target
Only cares about total daily dose - give
insulin if you’re high
Super brainy with multiple strategies it
can rely on 24/7
Especially effective overnight Learning is limited and based on
several days, so doesn’t like big
changes… use modes to adjust target
If it doesn’t learn fast enough, record
TDD, reset, and update parameters.
Or use temp manual mode
Fake carbs will mess with its mind…
Update your parameters Update parameters for big changes Gives the same response at midday
and midnight
Boost or ease off to give hint that
changes are needed, keep weight
updated

https://breakthrought1d.org.uk/what-a-cure-feels-like/

Dr Alec Price-Forbes
Supporting sustainable digital transformation
in health and care

Supporting
sustainable digital
transformation in
health and care
Dr Alec Price-Forbes
National CCIO NHS England
November 19th 2024

Optimization of the parts does not optimize the whole

25
Financial sustainability
growing demand will require increasing
funding to deliver care in its current
form.
Technology offers opportunities to
deliver care in completely new ways, at
lower cost.
Ageing population
Demand will increase as the population
ages.
Meeting this demand will be challenging.
We need to meet and manage the
complex and acute demands more
efficiently.
Patient outcomes
A stretched NHS leads to longer waiting
lists and worse patient outcomes.
Greater productivity means we can treat
more patients with the same resources,
improving lives, while using data to focus
efforts more effectively.
Staff satisfaction
Overworked staff who cannot deliver the
care they know is the best will not stay in
the health service.
Better technology cannot solve the
problem but can make their working day
better and increase morale and
efficiency.
Community care
We need to shift care out of acute
settings and into the community.
Achieving this requires integrated
services underpinned by common data
and digital tools to enable a person-
centred view of care.
Prevention
We need to intervene earlier, tackling
problems proactively, based on data, to
enable patients to identify and tackle
health issues before they become
problems.
Technology-driven transformation essential to respond to the
challenges services face today

Reality
• Recent diagnosis of metastatic
cancer awaiting MDT outcome
• Sudden deterioration at home
• Refused hospital admission
• Wanted to die at home
• All family and health care
professionals in agreement
• No meds available at home as
unanticipated event

Anticipatory medications needed to
manage acute distress and ongoing
symptoms
Four medications prescribed as standard for management of end of life
symptoms (pain, nausea, agitation, secretions)
• Who is going to prescribe? GP/Palliative Care Dr/ACP?
• Which pharmacies will have them available? Only able to find out by phoning
around or going to local pharmacies (+ waiting in queue). No guarantee that any
one of the pharmacies will hold all medication required
• Who will collect them? Health care professional/ family?
• How long will this process take? May required waiting in queues or driving
between pharmacies.

Prescription
• Community staff handwrite four separate prescriptions to ensure
ability to go to different places to get all that is required
• Pharmacy staff often busy and unable to be distracted from task
in hand to answer phone

Collection
• If family member available, leave the patient in distress for an
unknown period, potentially needing to drive to unfamiliar places
to find pharmacies that hold the drugs. Fear of patient dying
whilst away or in extreme distress + risk to family member of
driving when stressed and exhausted.
• If no family member able to collect meds, community staff taken
away from other responsibilities to source medication.
Significant impact on other patients needing care

Ultimate risk
Patient dies in
distress alone

Electronic Health and Care Record (5 Years)
“You will never solve a problem with the mindset
that created it”
Coventry
and
Warwickshire
ICS
Organisations
Primary Care
Hospices
TPP
Prison
TPP
Social Care
Urgent Care
CCC (adults) WCC
Care Director Mosaic
Integrated EPR
Mental Health Acute EPR
CWPT
CareNotes
SWFT GEH
UHCW
Cerner
GP Practices
EMIS
Compliance,
Risk and
Security
Authentication and
Authorisation
Data and Information
Governance
Audit and Logging
Privacy, references
and Opt-out
Business Continuity
and Disaster Recovery
Security Monitoring
Cyber Security
Clinical Coding /
Terminology
Security Standards
Governance
Standards
Data Standards
Messaging
Standards
Healthcare
Interoperability
Exchange
Standar
ds
UTC
Adastra
WMAS
OOH
Adastra
NHS 111
Regional
ICS
West Midlands
Shared Care
Record
National Services
SPINE PDS
National Care
Record
NHS App
Integrated Care
Record
Employee Engagement
Human Resources
Estate Management
Asset and Resources
Management
Rostering
Supplier / Partners /
Procurement
Training and Knowledge
Management
Finance and Payments
Collaboration and
comms tools
RPA
PMO
Enterprise, Resource and
Operations Enablers NHS Login
Service Management
Service Desk
Community Services
CareNotes
EMIS
Network
Diagnostics
Services
Collaborative
care record
Cleric
Information
Access Layer
User
Experience
products
Interaction
channels
Consumers
Virtual Health and Care Platform
Remote Monitoring
Virtual Health and Care
Citizen / Patient Portal
CRM
Digital workforce tools
Hybrid Working
Employee Portal /
App
Patient Citizen
Portal / App
Staff
Patients Things (IoT)
ICS-wide APIs Gateway
Personal Health
Record
Master
Patient Index
Data
Orchestration
Data Quality
Strategy and
leadership
ICS Vision and
Strategy
Transformation
roadmap
Governance and
Assurance
Enterprise
Architecture and EA
Governance
Virtual Consult
Virtual MDT
Virtual Wards
Integrated
Care Record
Integration and
Data Platform Clinician
Index
Integration
Management
Framework
Clinicians
Citizen
Data
Segmentation
Data
Visualisation
/ Reporting
Secondary
uses
Data
Analytics
De-id Data
Store
Direct
Care
Identified Data
Store De-Id/Re-id
Services
PH
M

Research
Pop
Health
Shared
Care
Commissione
r
Organisation
Acute EPR Community EPR Mental Health Primary
Care
Ambulance
SUS
Data
Management
Environment
Integrated
Care Records
Population
Health
System
SDR/TRE
Federated
Data
Platform
Other Data
Sources
X 120
Local Authorities

1.
Rethink
care
delivery
3.
Redefine
ICS
operating
model
2.
Redesign
ICS
enterprise
architectur
e
4.
Redesign
ICS digital
architectur
e
Core
processes
Pathways
Multiple
pathways
PROCESS REQUIREMENTS
defines
enables
FUNCTIONAL REQUIREMENTS
DATA ARCHITECTURE PRINCIPLES
Functionality generates
data
Data quality depends on
functionality and usability
TECHNICAL + USABILITY
STANDARDS
One platform per ICS
One user interface
defines
supports
defines
enables
Integrated Care Enterprise Architecture - summary
5.
Measure
patient
experienc
e
Patient
experience of
outcomes,
relations and care
integration
Pathways
functionality
Multiple
integrated
pathways
functionality
Core
processes
functionality

Dx:
meta-
static
cance
r
Sudden
deterio
r-ation
Hospital
GP
Community
Pharmacy
Wants
to die
at
home
Clinica
l team
agrees
Meds
not
availabl
e at
home
Delay
deciding
who to
prescribe
meds
They
phone
multiple
pharmacie
s to check
stock
Communit
y team
phones
pharmacie
s
Communit
y team
handwrite
s 4 scripts
Family
members help
to collect
drugs -
stressed
Refuses
hospital
admissio
n
Multiple
hours to
collect
drugs
Patient
dies in
distress
alone
Reality
Patient
Awaitin
g MDT
outcom
e

Dx:
meta-
static
cance
r
Sudde
n
deterio
r-ation
Hospital
GP
Community
Pharmacy
Wants
to die
at
home
Clinica
l team
agrees
Meds
not
availabl
e at
home
Delay
deciding
who to
prescribe
meds
They
phone
multiple
pharmacie
s to check
stock
Communit
y team
phones
pharmacie
s
Communit
y team
handwrite
s 4 scripts
Family
members help
to collect
drugs -
stressed
Refuses
hospital
admissio
n
Multiple
hours to
collect
drugs
Patient
dies in
distress
alone
Patient
Awaitin
g MDT
outcom
e
Reality – solution per ICS
2. Centralised drug stock
control and delivery per ICS
Single ePrescribing system
Single drug record
1.
Single
data
and
functio
n-ality
platfor
m
per ICS

Multiple
pathways
per
patient
Shared care processes per patient
Interactions
between
pathways
patient
Integrated Care Enterprise Architecture – the defining
use case for each ICS

IT Infrastructures
EMI
S
TPP
Epic
Meditec
h
Nerve
-
centre
Oracle
Cerner
System
C
Ri
O
TP
P
Epic
RiO
TPP
Civica
System
C
GPs Hospitals
Communit
y
Mental
Health
Social
Care
Vertical
standardisation
Horizontal integration
Home
Building the bridge for integrated care
DIGITAL PLATFORM: Integrated data and functionality layer (complete data, real-time, unified user
interface)
CORE PROCESSES: Shared care (e.g. refer, prescribe, plan, coordinate, manage multiple care pathways)
NHS
app

39
A vision for tech-enabled health and care services to addresses those challenges
and build for the future
Real-time risk
stratification of waiting
lists to get the sickest
seen soonest / make
the best use of
resources
People being notified of
high CVD risk, enrolled
in exercise programmes
and given the right
medicine

40
NHS App used
to get help for
symptoms using
111 online.
Suggests
appointment
with a specialist.
Proxy access
used to book an
appointment.
Confirmation
notification in
app and digital
pre-appointment
checklist and
relevant forms
sent, including
appointment for
calendar.
While in app,
reviews Digital
Personal Health
Record.
This record is
consistently
updated in real
time.
Prescribed
medication and
notified when
ready for
pickup.
NHS app
provides
medication
schedule,
including
reminders to
take.
Uses NHS app
to log any
medication side
effects and track
symptoms.
Information
accessible to
the clinician for
remote
monitoring as
required.
Clinician
reviews pre-
appointment
information to
note any
concerns.
All recorded in
the Patients
Health Record.
Clinician
reviews side
effects and
symptoms
remotely
Can initiate a
follow-up,
advising on
necessary
adjustments to
care plan or to
bring in if
symptoms
worsen.
User Journey #1:
Digital Personal Health Record

41
Maximise the power of our national services
Proposals to achieve these goals in draft
Need to work with health care leaders to set priorities and
deliverables
Digital, data, tech and innovation priorities
That deliver the following
benefits
And an annual
productivity saving
of
Need to be clear on the vision and priorities, ensure we make the most of the assets we already have and change management expertise is
embedded
~ 0.65% - 0.85%
productivity gains
annually
extensive non-
productivity and
wider
economic benefits
Unlocking up
to £35bn in total
benefits
Reduce operating costs
Save staff time
(clinical and administrative)
Reduced system demand through
prevention and self-management
Reduced length of stay
Support flow into community
Improved theatre utilisation
Systemise improvement
Fix the digital infrastructure
Modernise our data platforms
Transform the patient experience through the
NHS App
Create a thriving innovation ecosystem

42
These proposals would provide holistic transformation of NHS
services
A single vision of an
interoperable data-led
health and care
system.
Innovation
and Research
One Digital
Estate
Transformatio
n through
data
Releasing
time for the
workforce
Transforming
Patient and
People Facing
Services
And reducing the administrative burden on
staff…
Staff will be able to spend more time focusing
on care with administrative tasks automated.
Mature digital health records…
Clinicians have access to all the information they
need about patients and are supported in making
the safest decision for the person they are
treating.
Engaging each patient individually…
Patients will feel like we know them better when
using our services and be empowered to manage
their health and care journeys, and those of
others, easily and effectively.
Feeding into a single source of data…
Data fed into the ‘Federated Data Platform’
to be a single source of NHS data for
clinicians, managers and planners to be
able to make data led decisions.
01
02 03
04
05
The next generation of technology…
Foundational electronic systems and data
environments will enable innovation and
research on a far larger scale.

Dr Malte Gerhold
Supporting sustainable digital transformation
in health and care

Supporting sustainable digital
transformation in health and care
Dr Malte Gerhold, Director of Innovation & Improvement
19 November 2024

1. Invest in the change,
not (just) the tech

< 15%
Source: https://www.health.org.uk/news-and-comment/blogs/a-complex-patchwork-of-programmes

2. Your next improvement will
(most likely) not come from the
next tech

Source: https://www.health.org.uk/publications/long-reads/which-technologies-offer-the-biggest-opportunities-to-save-time-in-the-nhs

3. What success looks like is not
always what you think it is

< 1%
Source: https://www.health.org.uk/publications/long-reads/how-would-clinicians-use-time-freed-up-by-technology

< 1/3
Source: https://www.health.org.uk/publications/long-reads/how-would-clinicians-use-time-freed-up-by-technology

Health Foundation publications
Is Innovation being Squeezed out of the NHS?
https://www.health.org.uk/news-and-comment/blogs/innovation-is-being-squeezed-out-of-the-nhs
Priorities for an AI in health strategy
https://www.health.org.uk/publications/long-reads/priorities-for-an-ai-in-health-care-strategy
What do tech and AI mean for the future of work in health care?
https://www.health.org.uk/publications/long-reads/what-do-technology-and-ai-mean-for-the-future-of-work-in-health-care
https://www.health.org.uk/publications/long-reads/which-technologies-offer-the-biggest-opportunities-to-save-time-in-the-nhs
https://www.health.org.uk/publications/long-reads/how-would-clinicians-use-time-freed-up-by-technology
The Spread Challenge
https://www.health.org.uk/publications/the-spread-challenge
Scaling Innovation in the NHS
https://www.health.org.uk/publications/against-the-odds-successfully-scaling-innovation-in-the-nhs
Harnessing the Potential of Automation and AI
https://www.health.org.uk/news-and-comment/blogs/harnessing-the-potential-of-automation-and-ai-in-health-care
The Patchwork of Innovation Programmes
https://www.health.org.uk/news-and-comment/blogs/a-complex-patchwork-of-programmes
The Case for Improvement
https://www.health.org.uk/publications/a-guide-to-making-the-case-for-improvement
Agility: the missing ingredient for NHS productivity
https://www.health.org.uk/publications/long-reads/agility-the-missing-ingredient-for-nhs-productivity

Thank you
www.health.org.uk
@HealthFdn
malte.gerhold@health.org.uk

James Woodland
Realising the potential of digital innovation to
transform health and care across integrated
care systems: Dorset and HIOW

Realising the potential of
digital innovation to transform
health and care across
integrated care systems
James Woodland
Chief Digital Information Officer – NHS Dorset

Innovation in Data & Analytics
Targeted
Prevention
Hub
DiiS
Integrated
Analytics
Programme
DACOE
data &
analytics
centre of
excellence

Targeted Prevention Hub
“ICBs will have the primary responsibility for ensuring the delivery
of neighbourhood health, identifying population health needs and
acting on reversible risk factors to improve healthy life expectancy
and reduce utilisation of secondary care. This vital work must
continue at pace for us to deliver a neighbourhood health model.
NHSE - Evolution of our operating model - Nov 2024”

The Targeted Prevention Hub will be a clinical decision-support toolbox leveraging
machine learning to assist health & care professionals in preventive care.
The aim is to support prevention at scale with real patient impact, to do that
we need to:
• Co-design a suite of products direct with the people and teams that work
with patients
• Collaborate and listen to users, designing iteratively as guided
• Embed machine learning and software engineering as standard
• Integrate with other clinical systems, teams and re-id seamlessly

Expected Outcomes
• Improved patient care through identifying needs earlier
• Identifying the underlying causes of risk to help identify
appropriate interventions
• Reduce duplication by showing patient record view dates,
last contact points across the ICS and any previous
interventions.
• Be able to monitor the effectiveness of interventions and
remove duplication across organisations
• Streamline similar tools in one place and create a simple to
use interface
Predictive risk models
currently available:
• Falls
• Frailty
• Mortality
• End of Life
• Risk of hospitalisation while
on an elective waiting list
• Risk of readmission
• Risk of Social Care
• Risk of high cost acute care

ICAP is a programme that will enable ICS partners to codesign and develop
data & analytics capabilities and leverage opportunities when working at scale
• Cloud based data hosting (One Lake)
• A content management solution to allow more integrated System
analytics into Insights reporting (supporting neighbourhood Teams)
• Improving data security and controlled access
• Unlocking data science and advanced analytics capabilities
• Seamless data interoperability from EHRs using shared data standards
• Single version of the truth
DiiS - Integrated Care Analytics Programme (ICAP)

DiiS - Integrated Care Analytics Programme

Currently….
562 Members
15 Partner
Organisations
170+ Followers on
LinkedIn
48 Skill & Insights
sessions held so far
100+ views on average
of monthly comms
13,000+ site visits to
DACOE SharePoint site
since its launch

This DACOE Year so far….
103 training spots filled
4 Virtual Events with an
average participation of approx.
150 attendees
1 Summer Spectacular with
approx.
145 attendees
15 members of our new
DACOE Voices group
3 new training courses
launched
50 attendees on average at
our monthly Skills & Insights
sessions

Caroline Morison and Andy Eyles
Realising the potential of digital innovation to
transform health and care across integrated
care systems: Dorset and HIOW

Analogue to Digital
Caroline Morison, Chief Strategy Officer
Andy Eyles, Director of Digital, Data and
Technology

Our Ambitions
We have reviewed our Digital, Data and
Technology Strategy building on national
and local digital and clinical priorities.
Lord Darzi’s review has further highlighted
how integral digital, data and technology is
to patient outcomes and staff experience.
There are challenges though, for example
poor digital infrastructure, information being
inaccessible and the lack of investment.
As partners we know the strategic
importance of digital, data and technology
in delivering our collective ambitions.

Priority to upgrade the Digital, Data and Technology
estate for reliable, safe and resilient care.
• Productivity could improve by better addressing
digital maturity
• Integration of digital tools to transform clinical and
operational workflows, providing better patient
experience.
Total Triage and the NHS App
Single Acute EPR and Digital Diagnostics
Digital tools to manage demand
Automation
Shared Care Record

Frieda’s journey (Patient view)
6
4

Frieda’s journey (Professional view)

Integrated Care Partnership Working
• ‘Digital Solutions, Data and Insights’ is one of five key areas of focus
• Integrated Care Partnership Digital Assembly
• Local Maternity & Neonatal System Exclusion Project
• Digital Inclusion Five actions
• Digital Inclusion Survey

Mark Heffernan and Dr Matthew Stammers
Accelerating the adoption, spread and scale
of innovation using digital and data

77
77
Mark Heffernan – Wessex SDE Operations Director
Dr Matt Stammers - Consultant Gastroenterologist, Data
Scientist & Theme Lead for SETT: Data & AI
Wessex SDE and
SETT Talk @
“Accelerating the adoption, spread and
scale of innovation using digital and
data”

The national mandate for change to
Secure Data Environments (SDE)

85
OMOP and Data Interoperability

Free-Text OCR & Redaction
DISCLAIMER
This slide contains dummy data only. No personal identifying information or personal medical information has been used.

88
Wessex SDE (2024+)
Wessex Research Data Marts (WRDMs)
Prostate
Prostate
Prostate
Prostate
Prostate
Prostate
Prostate
Prostate
Prostate Study 1
NOTE: Default is
that data only
moves into the SDE
on a Study Basis.
Prostate
Prostate

Advanced Patient Cohort Identification

Advanced Warehousing
SDE
Technical
Approval
DAC
Approval
Data Linking
and
Interoperation ETL
Base Cohort
Identification
FiFi
Analyst
FiFi
Results
3 Days
EOI
Comes
Back
DAC
2-4
Weeks
Analyst
Analyst
2-4
Weeks
Background
Process
Analyst
2 Weeks
Analyst
1 Week
Analyst
1-2 Days
DAC Committee
+
Commercial
Algo
+
Local Model
Local Model
+
Background
Warehousing
4 Weeks
Inventory
Ribosome
SQL-Assist
Clinician
Agrees

91
AMD Pre-Screening using the SDE
Federated model-to-data patient
pre- screening has already been
demonstrated at Moorfields using
Bitfount.
Manual pre-screening by clinicians
can take up to 30 minutes per OCT
vs 10- 20 seconds for the Altris
model.
Hybrid scalable template for
future workflows
Source: Williamson DJ, Struyven RR, Antaki F, Chia MA,
Wagner SK, Jhingan M, Wu Z, Guymer R, Skene SS,
Tammuz N, Thomson B. Artificial intelligence to facilitate
clinical trial recruitment in age-related macular
degeneration. Ophthalmology Science. 2024 Jun
19:100566. Proposed Reproducible Data/Model Flows

92
Inferring genetic mutations from cellular
pathology images using the SDE

Dr Tom Brown, Deputy Chief Research Officer
Portsmouth Hospitals
Accelerating the adoption, spread and scale
of innovation using digital and data

Accelerating the adoption, spread and
scale of innovation using digital and data
Dr Thomas Brown MBChB FRCP PhD
Deputy Chief Research Officer, Portsmouth Hospitals University NHS Trust & Isle of Wight NHS Trust
Consultant Respiratory Physician and Portsmouth Severe Asthma Service Lead
Deputy Programme Director for King’s College London MBBS in partnership with the University of Portsmouth
Honorary Reader, University of Portsmouth
Thomas.Brown@porthosp.nhs.uk

Introduction
• Large fragmented data-sets held by the Organisation but we remain
information poor
• Congested innovation pathways
• Emphasis on prevention and population health management
• Ambition to harness digital technologies and data analytics
• Goal to improve integration and collaboration across the system
Alignment with clinical strategy
Our current reality
However, as an Organisation, we are striving for an enabling culture and
good examples of digital and data driven innovation are emerging

VitalPAC Early Warning Score (The Learning Clinic)
• Implementation enabled deployment of a digital EWS across the Organisation
which is estimated to have reduced hospital-wide mortality by 15.5%
• The use of vitalPAC data allowed subsequent refinement of EWS scores in
specific clinical environments
• Evolved into the CORE-D database capturing routine electronically recorded
patient data to model adverse clinical outcomes and healthcare resource
usage
• With the new ED opening, CORE-D has been
modified to capture data/routine samples from
patients with potential sepsis to develop a
platform to assess innovative point of care tests to
predict risk

VitalPAC Early Warning Score (The Learning Clinic)
• Highlighted value of well curated data in assessing risk but issues of non-live
data with limitations around accessibility
• Clinical utility and engagement supported embedding of the innovation
• Challenge of modifying existing digital solutions/platforms to address new
problems
• Challenge for innovators around commercialisation vs academic endeavour
• From ‘push’ to ‘pull’ harmonising a route for innovations in this space

Digital Consent (CONCENTRIC)
• Recognition that traditional paper-based consent process no longer fit for
purpose, leading to medicolegal risk, system inefficiencies, whilst not
supporting shared decision making
• Digital consent has been shown to improve the process in these three areas.
Less risk, more efficient, better decision making.

• Rapid innovator-led implementation supported by an SBRI grant
• Partnership working – innovators, funders, patients, staff and academics
• Implemented Trust-wide over 12 months crossing many traditional service
barriers to become business as usual
Digital Consent (CONCENTRIC)
• Importance of NHS recognising the role of innovators with clarity around
governance processes
• Clarity over benefit to all parties including the Organisation
• Largely ‘push’ with energy maintained by innovator – would this have
happened without this individual at the Organisation?
• Importance of training staff and patients to operate in a digital world

Digitising Long-Term Condition Management
(MyRenalCare)

(MyRenalCare)
• Moving care into the community to increase outpatient capacity
MyRenalCare increased outpatient follow up capacity by
33%
Actual + forecast outpatient follow up activity January to May 2023
• Virtual consultations take <5mins vs 15mins
traditional face to face / phone consultations
• Delivered 424 appointments in 30% less time
• With the remaining capacity could deliver a
further 185 appointments in the same amount
of time (609 vs 444 in total)
• Increases outpatient follow up clinic capacity by
37%
• Extra capacity could be used to
- Eliminate waiting times for new referrals
- Manage more patients
- See emergency follow ups
- Patient Initiated Follow Ups (PIFU)

(MyRenalCare)
• A digital innovation capable of transforming the way we deliver care
• Service focus and innovator-led with challenges of ‘buy-in’ from workforce
• Attitudes towards innovator benefit and importance of transparency over
finances
• Challenge for the Organisation as the employer of the innovator, test-bed for
the innovation and ultimately purchaser of the product
• Risks of double-running particularly with innovations offering remote
monitoring/PIFU
• Method of introduction (word of mouth, by consultant/developer,
hierarchy/buy in for product) can determine success

Improving Population Health and Optimising Care
Pathways
• Identifying high-risk patient groups for earlier review and intervention using
primary care data
• Development of a pathway management
system to streamline regional care and
ensure inter-operability between
systems:
• Reducing duplication, accelerating access
to treatments and providing a
mechanism for the implementation of
outcomes-based commissioning

Improving Population Health and Optimising Care
Pathways
• Need for well curated live data that is easily accessible to managers and
clinical staff but also to our academic colleagues and innovators
• Inter-operability between data sets is essential
• ‘Lift and shift’ of innovations challenging where we have different systems,
patient populations and governance processes. A framework to support the
adoption of digital innovation is needed
• Challenge of using data positively with a balance between commercial
opportunity, academia and providing the necessary information to support
our system ensuring we gain the most value from our data for our patients
and communities

Haris Shuaib, CEO, Newtons Tree
Innovator: AI market place

Newton's Tree
Building an AI-driven healthcare
system

There is a deployment blockage and no
one is fixing it
Data curation
Clinicians, academics
AI Development
academics, companies
AI regulatory approval
notified bodies, evaluation centres e.g. KiTEC
Local evaluation of AI
?????
Routine clinical deployment of AI
?????

AI presents unique clinical risks
Finlayson et al. Science 22 Mar 2019:Vol. 363, Issue 6433, pp. 1287-1289. DOI: 10.1126/science.aaw4399

AI presents unique risks
Finlayson et al. Science 22 Mar 2019:Vol. 363, Issue 6433, pp. 1287-1289. DOI: 10.1126/science.aaw4399

AI presents unique benefits
Sim et al. Radiology Nov 12 2019. DOI:10.1148/radiol.2019182465
Normal CXR - 3
radiologists

AI presents unique benefits
Primary
adenocarcinoma
Sim et al. Radiology Nov 12 2019. DOI:10.1148/radiol.2019182465

Exponential growth in FDA-approved
AI products
115
130
160
220
300
2020 2021 2022 2023 2024

The world’s first
vendor-neutral AI
marketplace for
healthcare
1. App Discovery
2. Managed Procurement
3. Supplier management

https://www.nature.com/articles/s41746-024-01270-x/figures/4
No public information on safety and efficacy
Series1
13
679

AI Sandbox:
to evaluate
multiple AI products
on your data
1. Automated product evaluations
2. Comply with best practice as it develops
3. Head to head comparison

IT departments are out of capacity
6 to 18 months to
deploy

AI Deployment: to
embed AI
products within your
workflow
1. Hosting
2. Integration
3. Continuous Monitoring

3 solutions for AI adoption
Marketplace Sandbox Deployment

One platform for
selection,
evaluation,
and
implementation
of in-house and
3rd party AI
products

Are you AI ready?
Email
haris@newtonstree.com
Website
www.newtonstree.com

Dr Heather Mitchell
Developing effective Population Health
Management and meeting the challenge of
reducing health inequalities

December 2, 2024
Dr Heather Mitchell, Director of Population and Health Equity
Developing effective Population Health Management and
meeting the challenge of reducing health inequalities –
From a Healthcare Provider Perspective
Health Innovation Wessex Strategic Event – 19th
November

124
Make Health Equity a priority

Community & Mental Health Trusts have a valuable role to play in addressing prevention
and inequalities
• Based in the community – opportunity to reach people not in contact with
services and link with community assets
• Reaching people in own homes (0-5’s and older adults) = opportunities to
reach households / carers
• Offer preventative care to key groups, including addressing wider
determinants and inequalities where community health services are in
contact with individuals
• Recognise the impact of delivering messages from healthcare professionals
to promote health and wellbeing
• Join up physical and mental health and wellbeing between services /
programmes
• Use population health data or take a population health management
approach
• Workforce are local residents (role as anchor organisation)

127
Trust Health Equity Priorities (2024/25)
Mental Health, Learning Disability and Neurodiversity
• Data – Improve data capture for ethnicity, Housing status & Employment status (for people with serious mental illness)
• ‘Don’t just screen intervene’ – Improve uptake of healthchecks and support with interventions for people with SMI
• Patient Carer Race Equality Framework – Anti-racism strategy
• Reasonable Adjustments – recording reasonable adjustments and improvement work to ensure these are accommodated (focus on
learning disability and autism)
Children and Young People
• Looked after children & care leavers
• School aged children – prevention and early intervention
• CAMHs – equitable access (focus on LGBTQ+ & neurodiversity)
• Obesity and physical activity
Core20 Neighbourhoods & Inclusion Groups
• Neighbourhood team development in prioritised geographies
• Homelessness, Refugees/asylum seekers, Sex workers, Traveller communities
Anchor Institution
• Staff health and wellbeing
• Staff representative of the communities we serve at all grades
• Climate change and sustainability

5
PLUS
Carers – (282)
Homeless / Vulnerably Housed (538 - HHC)
People with English as a Second Language (5999 –
28%)
Black, Asian and White not British populations (17746 –
83%)
LGBTQ+ communities (Unknown – Poor data quality)
Asylum Seeker / Refugees (73 HHC / 200 GP)
People with Learning Disabilities (LD) (110)
People with Physical Disabilities (98)
People with Severe Mental Health (SMI) (210)
Sex Workers (13 HHC)
CORE20
53% of Patient Population (11375 /
21521)
Key Wards / LSOA’s / Core 20 LSOA’s
Bevois – 6 of 7 (86%)
Bargate – 2 of 7 (29%)
Redbridge – 8 of 8 (100%)
Millbrook – 4 of 6 (67%)
71% of the Key Wards = CORE20
63% of the CORE20 population of
Southampton
Chronic Respiratory
Disease
Uptake of vaccinations
(Flu / COVID /
Pneumonia) to reduce
exacerbations (1200)
Early Diagnosis
Cancer
75% of cases diagnosed
at stage 1&2
Hypertension Case
Finding
Optimal Management
(335 – Undiagnosed
Hypertension)
Diabetes
Increase real time
continuous glucose
monitoring – focus on
IMD and ethnicity
(8)
Epilepsy
Increase access to
epilepsy specialist
nurses – increase
access in first year of
care for LD and Autism
(17)
Oral Health
Address backlog in tooth
extractions for U10s.
(21)
18+ GP SURGERY
Dedicated
smoking
cessation
support has an
impact on all
key clinical
indicators
(3340
Current)
PLUS
CORE2
0
Maternal Health
Reduction in maternal
mortality for Black and
Asian parents.
(96 Non-White
Pregnancies 01/08/24)
Severe Mental Illness
Annual Health Checks
for SMI to national set
targets
(210)
Asthma
Address over reliance
on reliever meds /
Reduce Asthma attacks
(120)
Mental Health
Improve access rates –
focus on ethnicity, age,
gender, deprivation
(16)
Children

Identifying Interventions that link with CORE20PLUS5
Core 20
• Integrated Neighbourhood Teams
• Health Campaigns
• PHM – Identification of communities in CORE20 Populations
Plus
• Homeless Healthcare
• RSI Outreach
• Ladies Night
• Trans Health
• Carers Programmes
• Supporting people that speak English as a Second Language
• LD Health Checks
Serious Mental Health
• Supporting specific parts of communities – Asian Men over 40
• SMI Health Checks
Maternal Health
• Co-Location with Family Hubs to provide Co-Located care.
• PHM – Identification of pregnant people who need more support
Hypertension Case Finding
• RAG Rating of High Intensity Users
• Community BP testing
Cancer Diagnosis
• Smoking Cessation
• Targeted Lung Health Checks
Chronic Respiratory Disease
• Flu and COVID vaccination programme
• Long Term Conditions reviews
Diabetes • Long Term Conditions reviews
Epilepsy • Long Term Conditions reviews
Oral Health • Co-Location with Family Hubs to provide Co-Located care.
Asthma
• Flu and COVID Vaccination
• Smoking Cessation
• Asthma Checks

Case Studies Southampton Homeless Healthcare Team have been
working with partners (including VCSE organisations)
to provide a targeted early evening session for sex
working and homeless women – providing a safe
space for women to access care, advice and
information. This has seen an increase in women from
this group accessing cervical smear testing and
supported a safe space to discuss mental health
issues.
Integrated Neighbourhood Teams work is focusing on
CORE20 Communities. In Southampton the focus is
Frailty. The early adopter programmes are supporting
development of models that reflect some of the drivers
of CORE20PLUS5 including focus on specific
communities.
A programme supporting the development of Blood
Pressure testing in local communities has targeted
Black African and South Asian communities who are at
increased risk of hypertension and are less likely to be
diagnosed and treated.
The programme is supported by the ICB and delivered
at a local level by community and statutory providers.
Interventions target CORE20 communities.

132
Principles we can all work to
Accountability
and
leadership
Patient-
Centered
Care
Holistic
Approach
Data-Driven
Approaches
Cultural
competence
Training and
education
Community
Engagement
Access for all
Prevention
and early
intervention
Feedback
and
continuous
improvement

133
Culture / team practices– what can I do?
 Ask and record protected characteristics
 Be curious about your data
 Work with partners & undertake quality improvement to
improve health equity
 Identify team champions and drive your own learning
 Make service users/peer workers part of your team
 Make equity part of BAU (business as usual)

134
The Power of Numbers
Ask and record: We cannot understand the differentials in access
experience or outcomes without the data on protected
characteristics.
Sexual orientation – 1.6% recorded
Almost one in four LGBT people (23%) have witnessed
discriminatory or negative remarks against LGBT people by
healthcare staff.
42% of LGBT+ school pupils have been bullied in the past year,
double the number of non-LGBT+ pupils (21%).
(Data from Stonewall)

Eating Disorder Calculated Prevalence
• Young People Mental Health Survey (Table 5.4 Prevalence of Eating
Disorders)
* Rounding means that 0.00 is shown, but the figure is >0
11- to 16-year-olds 17- to 19-year-olds
2017 2023 2017 2023
Eating Disorder % % % %
Boys
/
Young
Men
Anorexia nervosa 0.00 0.00* 0.00 0.57
Bulimia nervosa 0.00 0.22 0.00 0.00*
Other eating disorders* 0.21 0.73 0.00 4.54
Any eating disorder
0.21 0.95 0.00 5.11
Girls
/
Young
Women
Anorexia nervosa 0.15 0.41 0.30 6.43
Bulimia nervosa 0.09 0.81 0.28 3.62
Other eating disorders* 0.64 2.93 1.02 11.45
Any eating disorder
0.87 4.29 1.60 20.83
* Other eating disorders could include: ARFID, PICA, emotional
over-eating, rumination disorder, OSFED, selective eating
disorder or orthorexia nervosa
-
500
1,000
1,500
2,000
2,500
Currently in ED Service Calculated Prevalence Currently in ED Service Calculated Prevalence Currently in ED Service Calculated Prevalence
11-16 yrs 17-19 yrs 20-25yrs
Hampshire: North and Mid 42 902 23 1,743 17 1,827
Hampshire: South East 40 805 20 1,743 10 1,453
Hampshire: South West 24 604 18 1,291 14 1,037
Isle of Wight 30 236 8 546 3 434
Portsmouth - 421 3 1,059 5 1,289
Southampton - 574 12 1,590 12 2,415
30.9%
25.5%
27.4%
21.9%
27.9%
21.6%
29.4%
22.7%
23.8%
21.9%
16.4%
17.2%
17.6%
17.1%
21.4%
16.2%
23.0%
12.3%
22.1%
6.7%
9.5%
6.8%
4.9%
5.1%
0.0%
11.9%
3.6%
13.3%
8.2%
15.2%
0.0%
16.2%
14.3%
19.9%
19.7%
28.6%
Actual vs Calculated Prevalence by ICS Place (11 – 25yrs)
Hampshire: North and Mid Hampshire: South East Hampshire: South West Isle of Wight Portsmouth Southampton

0%
6%
6%
5%
6%
10%
10%
11%
19%
25%
IMD Score (1 Most Deprived - 10 Least Deprived) Of MSOA Home Address Of Under 18
Yr. Olds Accessing Eating Disorder Services
IMD Decile: 1
IMD Decile: 2
IMD Decile: 3
IMD Decile: 4
IMD Decile: 5
IMD Decile: 6
IMD Decile: 7
IMD Decile: 8
IMD Decile: 9
IMD Decile: 10
13% of patients live in
the three most
deprived deciles
55% of patients live in
the thee least deprived
deciles
3%
7%
8%
11%
7%
10%
8%
13%
14%
17%
IMD Score (1 Most Deprived - 10 Least Deprived) Of the Home Address Of People
(All Ages) Accessing Eating Disorder Services
IMD Decile: 1
IMD Decile: 2
IMD Decile: 3
IMD Decile: 4
IMD Decile: 5
IMD Decile: 6
IMD Decile: 7
IMD Decile: 8
IMD Decile: 9
IMD Decile: 10
18% of patients
live in the three
most deprived
deciles
44% of patients
live in the three
least deprived
deciles

Population health analytics in HIOW
Using insight from combined health and care data to improve patient-centred care, reduce inequalities, target
interventions and make evidence–based decisions which improve outcomes for people and communities
Data sources
Individual level community service data
Individual level mental health service data
Individual level acute hospital data
Individual level social care data
Insight data (police, fire, environmental)
Reference information (deprivation, finance)
Individual level GP patient data
Planned
Future
HIOW population health platform
Easy to use dashboards where linked data can be
used for analysis, risk stratification and
identification of patients for targeted actions
Management Visualisations Prediction
Current analytics
Population profile
Preventing chronic disease
Severe mental illness
Self-service tool
Reduce health inequalities
Use data-driven insights to
inform targeted, proactive
interventions
Make informed judgements
Make the best use of collective
resources
Act together – NHS, councils,
VCS and communities
Achieve practical, tangible
improvements
Improve the health and
wellbeing of specific
populations
System utilisation
Population health
analytics service
Taking a population health management approach enables us to:
Use data on the wider
determinants of health

138
Small change, big impact – Citizen’s Advice in
Inpatient Wards
Areas of stress often revolve around:
• Debt
• Issues at work or with family
• Relationship breakdown
• Problems with landlords
• Legal disputes.
Cycles of stress-related admission
• Ward colleagues not equipped to deal with these issues
• Evidence shows that signposting doesn’t work.
Piloted in Winchester, before replicating it to Basingstoke, Havant & Southampton.
Citizens Advice (CA) Case Worker for 3 days a week (2 days on the ward and 1 day supporting people in the community)
Citizens Advice drop-in (24 months data from 1 hospital (pilot) + 5-6 months data from all 4 hospitals)
• Worked with 286 service users
• Each had, on average, 5 distinct areas of advice need
• Addressed total of 1,523 distinct advice needs
• Finance and Housing are top advice need areas
• Advice resulted in total of c.£556,000 of direct financial benefit to service users
• Positive impact on colleagues and ward environment.

139
Creating the evidence base
• National innovation
• No other Citizens Advice offices or NHS Trusts working to address mental health stress in this way.
• Research study concluded in September 2024 by external Health Economics partner
• Study focused on first year of operation – in-depth analysis of 50 service users
• Small sample size and short period of time – further research potential
 Data re-sampled 5,000 times to give 95% confidence rating
• Study shows significant positive impact on health system
 £14.06:1 Return on Investment
 Cost avoidance of £244,850
• Research paper to be published and shared
• Inform future investment models
• Expand model to work more preventatively
• Share learning nationally – potential impact is phenomenal!
• NHS Parliamentary Awards 2024 national finalist, in the Excellence in Mental Health Care category
• Case Study in NHS Providers Annual report 2023/24

Dr Mark Ratnarajah and Vladimir Ljubicic, C2-Ai.com
Innovator: Predict and prevent - C2-Ai.com

142
Elective Care Digital
Learning Event
PTL Risk Stratification
Mark Ratnarajah
MD C2-AI

PTL RiskTriage accuracy
Figure 1: Confusion matrix for mortality risk predictions at NHS Trust Figure 2: ROC curve for NHS Trust
UKCA marked with MHRA as a medical device
Performance of the PTL model is evaluated for the site NHS TRUST on 3 years historical data through to June 2023.
Mortality rate: 0.57 %
ROC: 0.9455

145
Widespread backing across the NHS evidenced by outcomes
https://informatics.bmj.com/content/30/1/e100687.full

Patients waiting longer are more frail,
have worsening comorbidities and
underlying condition
AMBITION
–
Reduce
patient
deterioration
awaiting
surgery
Example – Here is one group of patients you could treat electively if triaged effectively. 16% will remain in
hospital with complications, blocking beds, increasing mortality and delaying clearing the backlog.

147 |
Urgency priority matrix score
Urgency Matrix
Trust Inputs
Subset viewed within pivot
table
Priority Points Score (4-
100)
C2-Ai
RiskTriage

Interactive dashboard (demo)
Allow
clinicians and
operational
teams to
identify
outliers
C2-Ai
RiskTriage

Individualised and dynamic patient risk summary
record
C2-Ai
RiskTriage

Right patient, in the right place, at the right time,
with the right team to ensure the right outcome
RiskTriage - capacity planning at scale
Supporting trust-level capacity planning and to
plan capacity at system level – mutual aid.
For example:
• Theatre-lite – creating new capacity based on
patient need and resource availability
• Super-Saturdays – micro-scheduling of
weekend lists and independent sector
capacity for HVLC patients with assurance of
case-mix complexity and risk
• Prehabilitation as part of the waiting well
agenda – identify patients with modifiable
risks and optimise their perioperative care
ahead of surgery
• Mutual Aid – ICB-wide risk stratification and
matching to critical resources such as ICU bed
capacity
150 World Congress of Prehabilitation poster 2024 – C2-Ai/ Surgery Hero

Regional surgical hub and mutual aid
Non-Admitted pathway Admitted pathway/ Waiting Well
Enhanced Recovery/
Rehabilitation
CRITICAL/HIGH
RISK
MEDIUM
RISK
LOW RISK
Clinical Harm
Review
COMPASS
PRE-OP Risk
Assessment
Shared
decision
making
No surgery/
alternative therapy
Preop Clinic +/-
CPET
Regional/Spinal
Anaesthesia (option)?
Surgery Hero
Prehabilitation
Prehab
enrolment
threshold met
SURGERY
HOT SITE
SURGERY
HOT SITE
SURGERY
COLD SITE
Shared
decision
making
Shared
decision
making
Pre-op education
Patient on
boarded to PTL
Clinical
Deterioration
C2-Ai
Observatory
– SDoH
equity/ Risk
adj outcomes

Emergency admission avoidance
Patient Care
Coordinator

Biopsychosocial patient risk stratification
WSIC – primary care data
Direct patient Q
oL
Q
uestionnaire
(PEP)
Dynam
ic
risk
prediction
–
C
2-A
i
Integration
engine
(FD
P) Foundry
E
N
T
s
u
r
g
i
c
a
l
t
e
a
m
Post op
surveillance
–
H
um
a
• Onboarding questionnaires
and recurring quality of life (QoL)
and red flag questionnaires
completed by patients
• Part of urgency score matrix
input to C2-Ai/ thresholds agreed
by specialist clinical teams
• C2-Ai risk platform assimilated
data from WSIC, EPR and
patient questionnaires
• Raw data processed using the
C2-Ai algorithms to stratify
waiting list patients by risk
• Foundry pass real-time HES and
PTL data to C2-Ai for risk
analysis
• C2-Ai send risk-stratified waiting
list viaTrust’s instance of
Foundry • A high-risk cohort will be
onboarded to Huma’s digital
preoperative surveillance app
• Nurse observers will monitor these
patients through the clinician-
facing web-portal and escalate to
the ENT team as needed
• The clinical team will use the
smart triaged waiting list to
prioritise high-risk and HVLC
patients for surgery
• Clinical staff can review the post
op cohort and remove need for
outpatient follow up - PIFU
• Patient level primary and
pathology data will be pulled
regularly from regional
integrated data warehouse
and ingested by the C2-Ai
algorithm
EPR
- C
erner
• Real-time PTL and HES data
• Automated write back to Cerner
of C2-Ai risk prediction and
clinical harm review
• Automated and auditable
scheduling based on priority
FDP
C2-Ai
RiskTriage

154
CONFIDENTIAL - All Rights Reserved C2-Ai 2019
Case-mix based avoidable harm benchmarking
Impacts of
multiple
triggers
C2-Ai
OBSERVATRORY
–
case
mix
adj.
clinical
outcomes

Quantifying the ‘equity gap’ in care
C2-Ai
OBSERVATRORY
–
case
mix
adj.
clinical
outcomes
155
A
B
Split by
Ethnicity
CONFIDENTIAL - All Rights Reserved C2-Ai 2019

156
Health economic benefits
Clinical and operational savings – independently evaluated by NHS England and Health Innovation Network (HIN):
• >1 bed-days freed up per patient on PTL = £626 productivity saving per patient listed on PTL
• >8% reduction in emergency admissions
• 95% reduction in avoidable cancellation rate in independent sector referrals
• 27% reduction in highest urgency patients (within 6 weeks of deployment)
• 66% reduction in highest risk patients requiring inpatient admission and ICU dependency
• >65% reduction in all surgical complications and no post op deaths or ICU admissions compared to control group
• No post operative chest infection as primary outcome measure, compared to 5% in the control group
• >4 bed days reduction in length of stay for patients with >10% complication risk compared to control group (accounts for 15-20% of PTL cohort)
• >15% conversion from inpatient to day case procedure in intervention group
• 5 mins saved in consultant administration time per patient per month in reviewing the PTL that can be used for clinical
commitments/ patient care
Scalability:
• PTL risk stratification and risk adjusted outcomes Observatory deployments in over 25 NHS Trusts
• 2 ICS regional rollouts - Cheshire and Merseyside and BNSSG ICSs deployed across all acute Trusts for PTL risk triage. Supporting
risk stratification for hot/cold site allocation and mutual aid/ regional surgical hubs by CRG
Publications:
• GIRFT Best Practice (
https://www.gettingitrightfirsttime.co.uk/wp-content/uploads/2023/06/GIRFT_HVLC_Guide_Edition_2_updated-June-2023.pdf )
• King’s Fund report 2024 (https://www.kingsfund.org.uk/publications/health-inequalities-nhs-waiting-lists )
• BMJ ( https://informatics.bmj.com/content/30/1/e100687.full ) ;
• Imperial paper – Cureus:
https://www.cureus.com/articles/115527-a-pilot-study-of-augmented-intelligence-risk-based-stratification-for-endocrine-surgical-waiting-li
st-prioritisation
)
• Waiting Well – risk stratification and targeted prehabilitation support (Royal College of Surgeons funded analysis (2022/23)) and
presented at the World Congress for Prehabilitaiton ((EBPOM - https://ebpom.org/abstract-submission/ ) abstract 115):
156

Dr Keith Gomes Pinto
Paul Wyman (video)
Supporting the workforce to meet the demands
of the future

Digital, Data, AI
Supporting the workforce to meet the demands of
the future
Dr. Keith Gomes Pinto
Paul Wyman (video)
19th
November 2024

Introduction
• Ex-surgeon
• Currently GP in Weymouth
• FMLM National Medical Director’s Fellowship 2019-2020
• ICB CMIO/CSO and Clinical Lead for DTI
• MSc Healthcare Leadership and Management 2024

Patient – Betty
• 75-year-old
• Normally in good health + independent
• Recent fall (hip #) – health has deteriorated
• Now on multiple medications + care package in place
• Unable to see her GP for weeks
• Feels let down by the NHS

Patient Survey 2023
The Kings Fund
2023
Three MOST common
reasons:
1) Takes too long for a GP or
hospital appointment
(71%)
2) Not enough staff (54%)
3) Government does not
spend enough money on
the NHS (47%)

Doctor – Omari
• 48-year-old
• Has worked as a GP for 20 years (partner for 15)
• Used to love the job but it has changed
• Works 8:00 – 18:00 (but logs in late for a couple of
hours to keep on top of admin)
• Stressed and exhausted

GP workforce
• 2,273 patients per FTE GP 17% since 2015
BMA
2024
BMA
2024

Health in 2040
Health Foundation 2023

Three most common
conditions:
1) Chronic Pain
2) Diabetes
3) Mental Health -
Depression
Health Foundation 2023
Health in 2040

• NHS Long Term Workforce Plan 2023
Workforce Plan
TRAIN
RETAIN REFORM

Workforce Plan
TRAIN
Medical school
places
60-100% by 2031/32
GP training places
50% by 2031/32
Nurse training places
92% by 2031/32
Dentistry training
places 40% by
2031/32
NHS Long Term Workforce Plan
2023
Reduction in
workforce being
recruited overseas

Workforce
• Lord Darzi report 2024
 Workforce shortfall across the NHS 260k – 360k FTEs by
2036/37
 Projected shortfall in qualified GPs 15k FTEs by 2036/37
• National Audit Office 2024
 optimistic future assumptions
 modelling outputs that could not be fully replicated

Workforce Plan
RETAIN
Culture
Leadership
Wellbeing
“Institutional
inadequacy in the way
that leadership and
management is trained,
developed and valued”
– Sir Gordon Messenger
and Dame Linda Pollard
2022

MSc
• Enablers and barriers to clinical leadership and management in the NHS –
scoping review
• Key findings:
 Leader rather leadership development
 Barriers at organisation level
 Stage of career
 General practice

Recommendations
Number of
articles
Self Team Organisation System
Need for formal mentoring and
coaching + positive role modelling
14
Early involvement and development of
clinicians
11
Support from senior clinicians and
management
10
Succession planning + talent
management
9
Development of career structures +
supportive career planning
8
Role-specific and corporate skills 8
Focus on personal growth and building
resilience
8
Inclusivity in hiring, providing
opportunities – multidisciplinary
approach
7
Establish networks externally – locally,
regionally, nationally
6

Workforce Plan
REFORM
Enabling
Innovation
Preventative
proactive care
PRODUCTIVITY
Flexible
workforce
Digital +
Technology

• Lord Darzi report 2024
 Lack of capital (£37 billion shortfall)
“The last decade was a missed opportunity to prepare the NHS
for the future and to embrace the technologies that would
enable a shift in the model from ‘diagnose and treat’ to ‘predict
and prevent’ – a shift I called for in High Quality Care for All,
more than 15 years ago.” – Lord Darzi 2024
Digital and Technology

• Utilisation of AI and RPA
• Investment in self
• Investment in staff
What does this mean…
• Patient app – manage long-term
condition
• Self-book appointments
• Better relationship with GP practice
Empowered
Supported
Safe
Time to Care
Improved
targets
Increased
satisfaction

• NHS Dorset (video)
 RPA and AI
• Tortus AI
 Ambient voice technology for clinical documentation
• Abtrace
 Optimisation of GP clinical workflow
Digital Technology

Inès de Lestapis, Tortus.ai
Innovator: Ambient voice technology for
clinical documentation - Tortus.ai

Introduction to
TORTUS
Growth Manager
TORTUS AI
Ines de Lestapis

60% of clinician’s time is spent on administrative tasks
Doctors spend 60% of their time on
computers, clicking the mouse 4,000
times per shift.
EHR use has gone from <10% in 2010 to
now >90%
Every patient-clinician interaction now
involves a computer, and that is slowing
down healthcare, burning out staff, and
reducing productivity.

Yet systems need data …
and with AI that need is only
going up

TORTUS is automating the end-to-end patient-
clinician workflow
We are an AI agent,
starting with
speech-to-text AI
All documentation is
automatically created
TODAY
Clinical Notes
Letters
Codes – ICD-10, SNOMED QOF,
SNOMED, HRG
Referrals, prescriptions,
ordering - in any EHR
Q1-25
Referrals
Prescriptions
Ordering
NICE guidelines

We believe clinical evaluation of AI is imperative
for safe use
TORTUS pioneered a new evaluation framework
of LLMs accuracy in healthcare.
Technical Evaluation Clinical Evaluation
New hallucination
definitions adapted to
healthcare
Validated at scale
TORTUS has worked with Great Ormond Street
Hospital and other NHS partners in a phased
research approach
Phase I, II, III and IV for clinical evaluation
Handles accents
Tested in noisy environments Works offline
Portable – desktop and
mobile

We are building the largest
patient trial study for ambient AI
in Europe
£1m NHS study
5,000 patients,
8 clinical sites,
live in 4
Tested across specialties
Epic-integrated & Enterprise
bespoke automations for
new EHRs e.g. CERNER, Rio,
Lorenzo
Currently working with Kent and Medway across Primary, Secondary and Social care, and ICB Cambridge across CPFT and primary care.
Outpatient Clinic
A&E / Emergency Dpt
Mental Health
Ambulance Service Adult Social Care
Primary Care
Tested across care settings
Paediatrics
Gynaecology
Paeds Dietetics
Endocrinology Physiotherapy
Rheumatology
Illustrative non-exhaustive

With strong initial results – a win for clinicians,
patients and healthcare trusts
TORTUS users saw a 20% increase in
EPR utilization “in-clinic”, reducing
post-clinic documentation time.
1 in 4 clinicians spent over 25% more
time directly speaking with or
treating patients (as opposed to
using the computer in clinics)
Direct Care time
Increased
Clinicians used
their EPR in clinic
20% more
Total consult
time decreased
Top users saw average time savings
of 6.8 minutes per consultation, or
an hour for an average 10 patient
clinic.
82% said “TORTUS
made clinic better”

We do not store patient data nor train our models
on patient data
Full DCB-160 provided
TORTUS has 3 internal CSOs and 1
external CSO
Accuracy date provided every
three months to customers in an
evolving technology test
Bias and fairness testing built in
All data servers in the UK
DPST compliant
We don’t train our models.
No data is retained
All local data is deleted on app
closure
No data retained,
no model trained
A culture of clinical
safety
Built for the NHS
1st
ambient scribe to achieve NHS
assurance - DTAC compliant
Custom coding (inc. HRG coding)
5 clinicians previously worked at the
NHS
Cybersecure, pen
tested
CREST-approved penetration test
Cyberessentials Plus Certified
Double-layer redundancy fallbacks
on all AI models

Keen on having a demo?
We have a dedicated space near the
conference center.
• UK first DTAC compliant
ambient scribe
• GDPR / CE PLUS certified
• Stores no data
• Crown Commercial Supplier
(RM6200/ G-Cloud 14)

Thank
You!
Growth Manager
TORTUS AI
Ines de Lestapis

Check it out here
Earlier this year we published
a review paper on the
cognitive load of using EHRs
and mid-year published the
results of the Phase II study at
GOSH,
We have published our work on
defining clinical safety
frameworks in hallucinations and
omission detection for LLMs in
healthcare.
Clinical Research Technical Research
Check it out here
Check it out here
We continuously publish research in the field of clinical AI
with our partners

Phase I
Phased
Evaluation
TORTUS studied the initial
development and roll-out
in phases, I-IV
We evaluated and tested the
system in a test sandbox,
while simultaneously taking
the system through data
governance, cybersecurity
and ethical reviews.
Noisy environments test
Multi-party consultations
Word-error rate

Phase I
Phased
Evaluation
We evaluated and tested the
system in a test sandbox,
while simultaneously taking
the system through data
governance, cybersecurity
and ethical reviews.
Phase II
48 simulated patients, 8 real-
world clinicians, time-in-
motion observers and
qualitative assessment of
experience and
documentation.
development and roll-out in
phases, I-IV, vs usual care
26.3% time saving
2x documentation quality
Improved clinician experience

Phased
Evaluation
Phase IIIa
Real-world clinician and
written consented patient
study - 200 patients.
Phase II
48 simulated patients, 8 real-
world clinicians, time-in-
motion observers and
qualitative assessment of
experience and
documentation.
Phase IIIb
Ongoing - business case
creation with additional
patients
Feasible and safe
Clinician and patient
acceptance
Signal to time saving

Phased
Evaluation
Phase IIIa
Real-world clinician and
written consented patient
study. 200 patients.
Phase IIIb
Ongoing - business case
creation with additional
patients
Phase IV
£1m NHS backed study
across 8 sites in London -
time in motion, qualitative
and quantitative
measurements across
multiple different sectors.
£1m NHS study
5,000 patients, 8 clinical sites
Multi-site evaluation
Primary care
Secondary care
Mental health,
Emergency dept
Ambulance (paramedics)

Pricing for primary
and secondary care
40p per patient in primary
care for PCNs/ICBs
£200k per year* in
secondary care
Bespoke ICB wide
partnerships
EMIS integrated, S1 next year, SNOMED
QOF codes.
*Up to 100,000 hours of annual
computer audio use, including up to 50
recommended microphones supplied.
Epic integrated, ICD-10 and HRG
coding.
Currently working with Kent and
Medway across Primary, Secondary
and Social care, and ICB Cambridge.
Enterprise bespoke automations for
new EHRs e.g. Rio, Lorenzo etc

Professor Age Chapman and
Professor Michael Boniface
Future of digital health care – what’s on the
horizon - Local

Future of digital health care
- what’s on the horizon
Hampshire and the Isle of Wight - Digital, Data and AI Workshop
18th
November 2024
Philip Brocklehurst - Health AI Lead, Deloitte
Professor Age Chapman - Head of Digital Health and Biomedical Engineering Research Group, University of Southampton
Professor Michael Boniface - Director of the IT Innovation Centre, University of Southampton

Integration of AI into existing Digital and Data solutions has already begun
“Oracle Health builds out
generative AI tools in its quest to
'eliminate clicks' for clinicians”
Oracle creates new capabilities in its healthcare
data platform, including a generative AI service
to simplify care management, prebuilt clinical
quality analytics and automated alerts
“Microsoft and Epic expand AI
collaboration to accelerate generative
AI’s impact in healthcare”
Microsoft and Epic announced a partnership to
integrate generative AI into the Epic’s EHR to
support note summarization, automate clinical
documentation, optimize revenue cycle
management, and enable data analysis
“Generative AI makes diagnosis
easier in radiology”
Siemens Healthineers partnered with German
Hospital University Hospital Essen to develop
large language models to support providers
in interpreting imaging results and patient
diagnosis.
“Medtronic to boost AI innovation
with new platform introduction”
Medtronic – in partnership with Cosmo
Pharmaceuticals and Nvidia – announced a plan
to integrate generative AI capabilities for cancer
diagnosis and a platform to host additional AI
solutions in their flagship colonoscopy tool.

What to expect in the next 5 years?
Consumers are the
CEOs of their own
health
Intelligent healthcare
and the
democratisation of
health data
Interdependent innovations
in science and technology
are reshaping treatment
paradigms
80%
Of healthcare professionals
will work alongside AI-driven
tools across clinical and
administrative activities2
75%
Of healthcare markets will have
established regulations
governing the use of GenAI in
health2
$148bn
The smart hospital market is
estimate at US$60bn in 2024 but is
expected to grow to $148bn in 2029,
at an annual rate of nearly 20%1
$78bn
The global market for remote
patient monitoring software and
services is expected to reach
US$78bn by 2032, up from $6.7bn
in 2022, at an annual rate of 29%3
Sources
https://www.deloitte.com/uk/en/Industries/life-sciences-health-care/collections/life-sciences-and-health-care-predictions.html
1. Smart Hospital Market Size & Share Analysis – Growth Trends & Forecasts (2024 – 2029), Mordor Intelligence
https://www.mordorintelligence.com/industry-reports/smart-hospital-market
2. Global Remote Patient Monitoring Software and Services Market, Market.us, https://market.us/report/remote-patient-monitoring-software-and-services-market/
3. Generative AI in HealthCare Market Trends, Grand View Research, Generative AI In Healthcare Market Size, Share Report, 2030

Professor Age Chapman
Head of Digital Health and Biomedical
Engineering Research Group
University of Southampton

202
AI in Personalized Cancer Treatment
YOU ARE UNIQUE
AI in Personalized Cancer Treatment
Professor Adriane Chapman Head of Digital Health and Biomedical Engineering
Adriane.Chapman@soton.ac.uk

203
Accelerating Medical Research
TRANSLATIONA
L PATHWAY
TOWARD
EVOLUTIONARY
PERSONALISED
CANCER
TREATMENT
Cancer Research UK (http://www.cancerresearchuk.org)
PI: Z. Belkhatir in
collaboration with
Memorial Sloan
Kettering Cancer Center

204
SUPPORTING
AND
OPTIMISING
CANCER DRUG
THERAPY
PI: Z. Belkhatir
Closed-loop
artificial pancreas
“For a given cancer patient, with particular
stage and cancer type (drug), can we have a
co-optimized adjustment of cancer therapy
dosage and frequency? analogue to
‘controlled’ insulin pump”

205
Functional
Near-Infrared
Spectroscopy
(fNIRS) to
understand
brain activity
PI: E. Vidal Rosas

206
Precision
Control for
Rehabilitation
PI: C Freeman

207
Roberta Project
PI: Wendy Hall, Michael
Boniface, Ying Cheong,
Chris Kipps, Adriane
Chapman

208
Roberta Project
PI: Wendy Hall, Michael
Boniface, Ying Cheong,
Chris Kipps, Adriane
Chapman
3.
Data
dona
tion
to
Trus
t
4.
Predi
ction
mod
el(s)
5.
Alert
indiv
idual
sugg
est
chan
ges
1.
Indiv
idual
focu
s
speci
ficati
on
2.
Indiv
idual
Devi
ce
usag
e
8.
Pr
e
di
ct
io
n
m
o
d
el
(s
)
9.
In
fo
r
m
at
io
n
to
cl
in
ic
ia
n
C
o
m
bi
n
e
wi
th
Cl
in
ic
al
d
at
a
10.clinical
11.findings
6.
Inform
care /
public
health
Community Involvement and Power

209
Digital Health and
Biomedical Engineering
Driving research in individual care and
treatment, as well as creating health insights.
We are eager to work with our clinical
colleagues to ensure the technology is
appropriate and usable within the care setting.
From nano-drug
delivery techniques
to AI for health

Professor Michael Boniface
Director of the IT Innovation Centre
University of Southampton

AI opportunities for
understanding
patient burden

Multiple long-term conditions (MLTCs) commonly develop
across the lifecourse
People with MLTCs often experience significant impact on
their lives or ‘burden’, e.g symptoms, treatment burden
© 2024 University of Southampton

Source: Holland E, Matthews K, Macdonald S, Ashworth M, Laidlaw L,
Cheung KSY, Stannard S, Francis NA, Mair FS, Gooding C, Alwan NA,
Fraser SDS. The impact of living with multiple long-term conditions
(multimorbidity) on everyday life – a qualitative evidence synthesis.
BMC Public Health 2024 (In press)
from qualitative analysis to formal (computer-readable) knowledge graph
Source: Smart, Paul, Fair, Nic and Boniface, Michael (2024) Modeling
biomedical burdens in basic formal ontology. 15th International
Conference on Biological and Biomedical Ontology, University of
Twente, Enschede, Netherlands. 17 - 19 Jul 2024. 12 pp
© 2024 University of Southampton

Using knowledge graphs can help understand patient burden
Representation of
Burden in EHRs
Burden assessment
with Large Language Models
Semantic enrichment
of population data analysis
Class A?
Class B?
Class C?
© 2024 University of Southampton, IT Innovation Centre

Tools for automatic
assessment of AI
compliance risk

Software (including AI) as a medical
devices brings risks and regulation
Safety, security, robustness, transparency, explainability, fairness,
accountability & governance, contestability & redress

from document checklists to computational models of socio-technical risk

Protocols for
algorithm (incl AI)
trials at scale

Digital platforms with patient facing
interfaces can have significant reach
to communities

from site-based to digital platform-based engagement and recruitment
in trials (750+ participants)
A person-centred clinical model for COPD
management using algorithms
Proactive medical treatment review
recommendation based on self-reported
symptoms and assessment tests

Philip Brocklehurst, Deloitte
Future of digital health care – what’s on the
horizon – National / International picture

Health Innovation Wessex
Innovation Centre
Southampton Science Park
2 Venture Road
Chilworth
Southampton
S016 7NP
E: enquiries@hiwessex.net
@HIWessex
T: 023 8202 0840
healthinnovationwessex.org.uk

Data, Digital and AI strategic event slides

More Related Content

Similar to Data, Digital and AI strategic event slides (20)

More from Health Innovation Wessex (20)

Recently uploaded (20)

Data, Digital and AI strategic event slides