This document discusses how predictive analytics and intelligent service platforms can transform healthcare delivery at scale by leveraging data and technology while keeping people at the center. It highlights Accenture's digital health technology vision which focuses on five trends: intelligent automation, liquid workforce, platform economy, digital trust, and predictable disruption. The trends emphasize the importance of people - both patients and healthcare workers. The document also provides a case study of analytics-driven case management in Valencia, Spain that has led to reductions in unplanned hospital admissions and stays through risk stratification, coordinated care, and a holistic patient approach.

1. Digital Health -
Leveraging Predictive
Analytics to transform care
@ scale with Intelligent
Service Platforms

2. Keeping up with changing technology is vital, but it’s just as
important to evolve the consumer experience, care delivery
methods and career development opportunities for the
healthcare workforce.
The Accenture Digital Health Technology Vision 2016 reveals
five trends that prove winning in the digital age hinges on
people – the patients and the healthcare workforce.
Transforming care pathways at scale requires Intelligent Patient
Platforms designed as services into people’s lives
Copyright © 2016 Accenture All rights reserved. 2

3. Copyright © 2016 Accenture All rights reserved.
1. Intelligent
Automation
People do
things
differently and
adopt new
services
2. The Liquid
Workforce
People
learn highly
agile
healthcare
skillsets
3. Platform
Economy
People Connect
at scale across
transformed
ecosystems
5. Predictable
Disruption
People assume
new roles and
incentives in
new care models
5 Trends put People Digital Health
4. Digital
Trust
People
confront risks
and place
premium
value on trust.
3

4. PEOPLE X MACHINES
Data can help clinicians make informed decisions
through computer-assisted algorithms, and it can get
the right supply to the right place at the right time.
Apps can help patients navigate their choices in
healthcare and wellbeing services, enabled by
analytical insights based on their profile.
Copyright © 2016 Accenture All rights reserved. 5
Health plans can use data to help guide
consumers in making a decision about the
best plan for them.

5. This emergent fluidity requires some fundamental
shifts in how the enterprise is structured, how people
are trained and how the culture adapts to new
technology-enabled ways of working. But these
changes come with benefits.
Copyright © 2016 Accenture All rights reserved.
Organizations can launch innovations faster.
Health plans can use crowdsourcing to hear
from innovators and entrepreneurs. And, virtual
care provides greater accessibility and flexibility
for patients.
CROWD CULTURE CHANGE READY+ +
7

6. (CLOUD + APIs + IOT) X OPEN SOURCE X BUSINESS MODELS
Imagine a future when healthcare consumers
will go to one central place to book an appointment,
check their electronic medical record or pay an
out-of-pocket expense.
Providers will track a patient’s activity from
hospital to home.
Copyright © 2016 Accenture All rights reserved. 9
Health plans will connect with consumers
through engagement platforms, collecting
data from wearables and offering rewards or
premium discounts for progress.

7. Disruption can be a game changer if a
business can predict it.
Healthcare organizations must link up with
those outside of the industry to seize new
disruptive opportunities.
Copyright © 2016 Accenture All rights reserved.
If the banking industry has mastered mobile payments,
health plans should explore mobile apps that can make
out-of-pocket payments pain-free for consumers.
If companies like Spotify can successfully deliver music
as a service, healthcare providers should also look at how
care can be delivered as a service.
PLATFORMS + ECOSYSTEMS + STRATEGY
13

8. Intelligent Patient Platforms @ scale
- e.g avoid Heart Failure readmissions
Lower
Avoidable
Readmissions
and costs
Data Integration & Advanced Analytics
Digital Care Coordinator & Patient Engagement
Care
Coordination
Services
People Centric
Care
Remote Care
Coordination
Optimized
discharge
Coordinated remote
care
Inpatient Remote
Consistent
inpatient
coordination:
14

9. 16
Case Study – Valencia La Fe
Analytics driven case management

10. • La Fe Health Department
210.000inhabitants
10% of the Chronic Population in Valencia
Integrated Health Department
University Hospital La Fe (+1.000
beds)
5 Outpatients Centers
6 Primary Care Centers
1 center for addictions
Per Capita reimbursementmodel
500.000 Outpatients visits
385.000 Hospital Stays
50.000 Admission
6.000 Births
Analytics Driven Case Management

11. 18
Our Approach
Analytics Driven Case Management

12. 19
• On average, 20% of patients drive around
60% of the total system health expenditure
Real data from our
experience in La Fe
Case Management
High complexity patients
Case Management /
Disease Management
High risk patients
Selt-treatment
Patients with moderate risk
Healthy Living &
Prevention
Patients with low risk
7% Hospital stays (days)
8% Hospital Admissions
33% Visits to Emergency
7% Hospital stays (days)
24% Hospital Admissions
27% Visits to Emergency
64% Hospital stays (days)
53% Hospital Admissions
16% Visits to Emergency
22.5% Hospital stays (days)
15% Hospital Admissions
24% Visits to Emergency
Analytics Driven Case Management

13. (*) Source: Roger Halliday, Department of Health for England
20
Number of Hospital Days
Before Integrated Case Management
After Integrated Case Management
-4
Years
-3
Years
-2
Years
-1
Year
Intense
Year
+1
Year
+2
Years
+3
Years
+4
Years
+5
Years
• By anticipating
which patients are
evolving to their
‘Intense Years’,
we can maximize
potential
healthcare savings
Analytics Driven Case Management

14. 21
ANALYTICS
Analytics Driven Case Management
CUMMUNITY ASSESSMENT RISK SCREEN (CARS) ACCENTURE PREDICTIONMODEL
Prediction Model Tries to predict Inpatient and E&A visit risk in the next 12 month.
Main Characteristics Group patient by risk
Tries to predict unplanned admissions risk with more than three days length in
the next 12 month.
Predictive variables Number of Comorbidities Defines risk profile per patient.
Number of Pharmacology More than 20 key different variable
Number of E&A visits • Age
Number of Hospital Admissions • Gender (only for some disease)
• Number of Hospital admissions (unplanned): recent (last 1-2-3 months)
• Number of visits to ER
• Number of visits to Specialties NEW!
• Chronic Diseases Flags based on historical diagnoses (using ICD Coding)
NEW!
• Comorbidity Indexes (for instance, Elixhauser and Charlson) based on
historical diagnoses NEW!
TRUE POSITIVES
True Positives and among
predicted patients
True Positives and among
predicted patients

15. 22
Real data from our experience
in La Fe, Valencia
• This Patient Profile
requires
multidisciplinary
coordination and a
holistic approach
applied to each case
Analytics Driven Case Management
• Integrated approach vs. aggregation of
treatments by disease
• Avoid redundancies and inefficiencies
(test, visits, analysis,…)
• Save patient and caregiver time
• Rationalize drugs prescription

16. 23
• Our Case Management Model is a
NEW capability that coordinates
health service provisioning and
helps patients better navigate the
system
Analytics Driven Case Management

17. 24
HospitalConsultant/Nurse
• Hospitalization
• A&E
• Specialists
Primary CareCentre
GeneralPractitioner
• Visits
Nurse
• Deliver Care
• Home Visits
Home Care Unit
Consultant/Nurse
• Home Care
Case ManagementCenter
Case Manager(Nurse)
• Manage patient (remotely)
• Manage relationship with
Home Care and Primary
Care
PatientNavigation
Proactive coordinationbetween
levels of care and barrier
removal
Analytics Driven Case Management
TRADITIONALORGANIZATION NEW ORGANIZATION

18. 28%
72%
45%
55%
From…
To…
Planned Hospitalizations
UnPlanned Hospitalizations
Planned Hospitalizations
UnPlanned Hospitalizations
28
SOME PRELIMINARY RESULTS AND ITS POTENTIAL IMPACT
Analytics Driven Case Management
100%
(89.400)
72%
(64.457)
39%
(34.554) 25%
(22.570)
0%
20%
40%
60%
80%
100%
TOTAL HOSPITAL
NIGHTS
UNPLANNED
HOSPITAL NIGHTS
UNPLANNED
PREDICTABLE NIGHTS
AVOIDABLE HOSPITAL
NIGHTS
• Reduction 9% unplanned admissions and 16% unplanned stays
equivalent 21 beds daily
• Clinical study on quality of life impact conducted – to be published shortly

19. (*) Based on initial savings estimated with provisional real data in La Fe. 2.8% Direct Costs – 3.2% Indirect Costs
(**) Accenture estimate assuming that savings with this cohort are 50% of those achieved with the first cohort of patients
29
-6%
-3%
100%
TotalHealthExpenditure
0%
91%
-9%
(*)
(**)
POTENTIAL FINANCIAL IMPACT (PRELIMINARY ESTIMATES)
Analytics Driven Case Management

20. David Champeaux
David.champeaux@accenture.com
@ChampeauxDavid
Copyright © 2016 Accenture All rights reserved. 30
Follow us @AccentureHealth
Digital Health Tech Vision 2016
Fjord Trends 2016
Fjord Fido Diabetes video
2016 Consumer Survey on Patient
Engagement
Virtual health: The untapped opportunity to
get the most out of healthcare
Losing patience: Why healthcare providers
must up their mobile game
For more information: Link to Sources: