Transcript Global Healthcare Practice

Consulting – National Health Team
Hospital IT Benefits –
Modeling EMRAM Value
Overview and Discussion Deck
Strictly Private
and Confidential
May 2013
Klaus Boehncke
eHealth Lead Asia-Pacific
[email protected]
+61.414397541
Agenda
PwC in Healthcare
1. Health IT is hard
2. Hospital IT / HIMSS EMRAM Overview
3. PwC Benefits Methodology
4. Insights
5. Discussion
PwC
2
Overview - PwC Asia/Pacific and Healthcare Focus:
(1) Strategy & Benefits, (2) System Performance, (3) Change Mgmt
161,000
Overall Partners
& Employees
Three Core Areas of Expertise:
1.
Health IT Strategy,
Benefits Realization and
Implementing a Value
Focused Program
2.
Improving Hospital
Performance
3.
Successfully
implementing actual
Change in the Ward
10,000+
Technology
Specialists
5,000+
Health Experts
900+
Health IT
Specialists
PwC
August 15, 2012
45
1
Health IT is Hard
PwC
4
Billions of dollars are being invested worldwide
into Health IT
Examples of worldwide national / regional EHR & EMR programs
Source: Public Sources
PwC
4
Not all Programs are successful...
Source: Public Sources
PwC
6
5
What are the Chances of Success?
Project Success Probability
The Standish Group, CHAOS Summary 2009
Success defined as on-time and on-budget and with planned functionality
Success2
Rate
(%)
75
70
65
60
55
50
45
40
35
30
25
20
15
10
5
0
71
38
Data unavailable
“Projects over
$10 M have only
a 2% chance of
coming in on
time, on budget,
[with adoption &
use], and
represent a
statistical 0 in
the success
column”*
2
<$750,000
PwC
By Size of Labour Cost (USD)
$750,000 –
$3 M
$3 M - $10 M
* and in Healthcare, probably less!
>$10 M
7
6
What are the Reasons of Failure?
1| Lack of Senior Executive Engagement
2| Health System Complexity and Doing too Much:
Lack of Clear Objectives and Focus
•
What Capabilities?
•
Which Stakeholders?
•
To achieve What?
Addressed by clear
Benefits-Focused
Approach
3| Lack of Stakeholder Adoption and Change Management
Source: PwC Analysis
PwC
8
7
2
Hospital IT and HIMSS EMRAM
Overview
PwC
9
The HIMSS clinical EMR Adoption Model (EMRAM) is
defined by 7 levels
Digital
Hospital
Paper-based
Hospital
August 15, 2012
PwC
9
HIMSS EMRAM is only one building block of the Digital
Hospital - clinical systems need to be complemented by ICT
and the Back Office
Digital Hospital Systems Building Blocks
Clinical Systems
Corporate Systems
HIMSS EMR Maturity Model, levels 1-7
from a paper-based to a paperless
hospital. Includes all systems used for
clinical processes, such as basic imaging,
pathology, etc., decisions support,
medication management, data warehouse
and analytics, etc.
Back office business operations systems,
including e.g. HR, payroll, finance,
workforce, billing, procurement etc. –
culminating in a full ERP system
Information and Communications ICT Infrastructure
This includes e.g. virtualized platforms for enterprise and desktop, mobility platforms and
area wireless networks, mobile device mgmt, unified communications, RFID
Management, video and other collaboration tools, real-time location services and RFID
management, troubleshooting
August 15, 2012
PwC
10
Australia is in the bottom 25% of IT adoption
compared to US Hospitals
Q2 2011
Stage
Stage 7
1.1%
Q4
2012
1.9%
4.0%
8.2%
6.1%
14.0%
Stage 5
Cumulative Capabilities
Complete eMR, CCD transactions to share data; Data warehousing;
Data continuity with ED, ambulatory, OP
Physician documentation (structured templates), full CDSS (variance &
compliance), full R-PACS
Closed loop medication administration
Stage 4
CPOE, Clinical Decision Support (clinical protocols)
12.3%
14.2%
Stage 3
Nursing/clinical documentation (flow sheets), CDSS (error checking),
PACS available outside Radiology
CDR, Controlled Medical Vocabulary, CDS, may have document
Imaging; HIE capable
46.3%
38.3%
13.7%
10.7%
Stage 1
Ancillaries – laboratory, radiology, pharmacy – all installed
6.6%
4.3%
Stage 0
All three ancillaries not installed
10.0%
8.4%
Stage 6
Stage 2
N = 5319
PwC
N = 5458
11
3
PwC Benefits Methodology
PwC
13
... and it comes
from saying no to
1,000 things to
make sure we don't
get on the wrong
track or try to do
too much. We're
always thinking
[...] but it's only
by saying no
that you can
concentrate on
the things that
are really
important.
PwC
14
Say no? – What to Concentrate on?
Subset of Elements that are Part of the Australian PCEHR Program
Medication Management
Vendor Systems
GPs
Portals
Imaging
Self Management
Allied Health
Hospitals
Patient
Health Summaries
eDischarge
Messaging
Event Summaries
eReferrals
Pilot Sites
PwC
Pharmacy
Pathology
14
A modern Benefits Approach – Building on a
Decade of Pioneering Thinking
• Germany ca. 2006 – National eHealth Program – used to build incentive
mechanisms to redistribute benefit, also it identified a major program risk early that
was not actioned on, thereby derailing the entire project.
• Singapore ca. 2009 – National Electronic Health Record – used for the
business case
• Australia – 2011/2012 Personally Controlled Electronic Health Record –
potential use e.g. to define metrics for “success”, craft key messages to stakeholders
for communications and training material, identify possible incentive structures,
future program focus etc.
• Australia – 2012 Rural Health Initiative – rural health benefit model
• Australia – 2013 Melbourne Health – hospital benefit model
PwC
and several others – non publicly available information
15
Enabling a Benefits-Focused Program
Program Execution from Inception through to Change & Implementation
Strategy
Cost/
Benefit/
Funding
Program
Execution
Change
Management
Evaluation &
Monitoring
Initial Benefit
Estimates
Build Detailed
Value Model
PMO Setup to
Focus on
Benefit Drivers
and
Corresponding
Stakeholders &
Functionality
iterate
Change to
Focus on
Benefit Drivers
and
Corresponding
Stakeholders
Metrics to
Evaluate
Achieved
Adoption and
Value
August 15, 2012
PwC
16
What is the Benefit of moving up the EMRAM scale?
Who knows?
A robust benefits evaluation of a full scale EMR implementation is
next to impossible.
•
o
EMR implementations typically take a long time (24 months – 10+ years)
o
During this time variables are not kept constant (management changes, processes
are improved, policies change, and systems implementation progresses at an
uneven rate).
o
As different components are introduced, they work together in unexpected ways and
again make it difficult to assign value to one particular capability
o
Because of this, the EMR impact cannot be isolated from everything else that is
going on, and thus cannot be measured accurately.

The PwC approach circumvents this problem by focusing on individual
components of the EMR, such as medication ordering, medication
administering, decision support, creating discharge summaries, etc.

These components have often been evaluated in academic peer
reviewed articles, using robust methods and scientific evidence.
PwC
26
38
We have developed and refined a unique benefits
methodology
1. The methodology is unique in its ability to display results in
multiple dimensions, namely by stakeholder, capability and
benefits group
2. It is scalable from a single stakeholder to a full national
system, e.g. in healthcare being able to evaluate and estimate the
effect on a single nurse, a ward, a hospital, or a nation as a whole
3. It can capture flow-through benefits effects, taking into account
where they are generated and where they are received
4. It is based on transparent individual discrete modular EMR
components and associated capabilities and assumptions
based on robust academic evidence, which can be tested with
clinician workshops, thus being an important part of the
change journey
PwC
26
17
The Five Steps of the Benefits Framework
Identify the Capability
1
What new
Capability
will generate
Value?
Sourcing the
Capabilities
EMRAM.
Vendor
Documentation,
Client Experts
Identify the Stakeholders
2
Which
stakeholders
are impacted
by this
Capability?
Sourcing the
Stakeholders
Team
Discussion,
Workshops
Identify the Benefit
3
What proof
is there of
benefits and
how can we
quantify
them?
iterate
Estimate the Impact
& Build Validation Model
4
What is
therefore the
expected
impact of this
capability?
Sourcing the Benefit &
Impact
• Academic literature
• PwC Benefit Library
• Team to generate logic
assumptions
• Clinician validation
PwC
Measure the Real Life
Baseline and Result
5
What are we
actually
seeing on the
ground?
Benefits Management
Plan: Baseline and
results
15
18
Modeling the results creates a Value Cube in the dimensions
of Benefits, Capabilities and Stakeholders
Benefits generated by
one combination of
capabilities and
stakeholders may be
received by other
stakeholders which is
mapped in the cube
Benefit Group
Quality
Safety
...
LHD/Cluster
Efficiency
Nurses.
Patients
Access
Each cube contains the
academic evidence and the
logic flow model for that
stakeholder and capability
combination, as well as a
mapping to benefits
Doctors
...
Capabilities
eMR Benefits
PwC
21
Each Value Cube “lego-block” will be a quantification
element based on volume and evidence of impact
Logic Flow Module
Capabilities
Total
Benefit
Volume
Impact
“How often”
“What changes”
Academic Evidence
$ / lives / wellbeing
indicators
Volume Data per
Episode
Cost Data
Examples
Waste
Elimination
• Elimination of
duplicates
• Less time
searching for
information
Shorter Stay
(Reduced Hoteling
Cost, Increased
Throughput)
New Process
per Episode
• Faster turnaround
for tests leading to
earlier treatment
and discharge
• Automated coding
process
• Business
intelligence/data
analytics
Episode
Disbenefit
• Time spent to enter
information digitally
vs. scribbling on
paper
 and… map to Stakeholders, Capabilities
eMR Benefits
PwC
22
Understanding the patient journey & volume
19.14
18
Average frequency per patient hospital stay
(episode)
16
14
12
10
8
6
4
2
3.19
3.19
2.76
1
1
1
0.38
0.35
0.21
0.02
0
PwC
21
21
Understanding the Impact by EMRAM Level:
Example for Stage 6 - Medication Administering
50 – 90% of IV medications have errors (wrong patient, drug, concentration, drip)
20 - 40% of drugs have errors from ordering (wrong patient, drug, or dose)
Reduced medication administration errors leading to a
reduction in ADEs (43% reduction in errors)
Reduced medication dispensing errors leading to a
reduction in ADEs (64% reduction in errors)
Reduction in turnaround time of discharge scripts
leading to a reduction in length of stay (23% reduction
in turnaround time)
Avoided deaths from ADEs due to improved
medication management (4.3% of ADEs)
Reduced length of stay from avoided ADEs due to
improved medication management (10 additional days
per ADE)
Increased staff time spent on medication tasks
(between 1% and 8% increase)
PwC
Source: Research by the University of Sydney, Johanna Westbrooke, presented at the National Health Informatics Summit
Singapore 2009, various other sources in public domain
26
23
Example Logic for Administering Errors
Logic Module
Reduced medication
administering errors
leading to a
reduction in ADEs
Calculation Logic
Number of prescriptions (XYZ) x
Results
$NN (lower)
Number of administering events per
script (6) x
$NN (average)
% of scripts with medication errors that
cause an ADE (0.45%/0.73%/1%) x
# of avoided ADEs
(average)
$NN (upper)
Sources
Assumption
(number of orders
per script)
Bates et al (1995)
Seeley et al (2004)
% of errors at administering stage (46%)
x
Bates et al (1998)
Reduction in admistering errors
(27%/43%/59%) x
Leape et al (1995)
Additional cost per episode from ADE
($XYZ)
Frances et al (2008)
Ehsani et al (2006)
Pirmoha et al (2004)
Howard et al (2003)
Lagnaoui et al
(2000)
Winterstein et al
(2002)
Gurwitz et al (2003)
PwC
26
24
Excerpts from the Evidence Base
Author
Title
Bates et al
(1995)
Relationship between
Medication Errors and
Adverse Drug Events
Bates et al
(1998)
Effect of Computerized
Physician Order Entry
and a Team Intervention
on Prevention of Serious
Medication Errors
Ehsani et al
(2006)
The incidence and cost of
adverse events in
Victorian Hospitals 200304
Jensen
(2006)
The Effects of
Computerized Provider
Order Entry on
Medication Turn-around
Time: A Time-to-first Dost
Study at the Providence
Portland Medical Center
PwC
Journal
Abstract
Journal of General Internal
Medicine
OBJECTIVE: To evaluate the frequency of
medication errors using a multidisciplinary
approach, to classify these errors by type, and to
determine how often medication errors are
associated with adverse drug events (ADEs) and
potential ADEs.
Journal of the American
Medical Association
Objectives.— To evaluate the efficacy of 2
interventions for preventing non-intercepted serious
medication errors, defined as those that either
resulted in or had potential to result in an ADE and
were not intercepted before reaching the patient.
Medical Journal of Australia
Objectives: To determine the incidence of adverse
events in patients admitted in the
year 2003–04 to selected Victorian hospitals; to
identify the main hospital-acquired
diagnoses; and to estimate the cost of these
complications to the Victorian and
Australian health system.
AMIA Symposium Proceedings
Paper
As the Providence Health System is phasing in its
Computerized Provider Order Entry (CPOE) system
at the Providence Portland Medical Center, we
conducted a study to demonstrate the effects of
CPOE on medication turn-around time.
Retrospectively, we tracked and compared
medication orders that were placed via the existing
paper-based system and the CPOE system. The
results of this study coincide with, and confirm,
previous research that has been performed at large
academic medical centers.
25
More than 40 Logic Modules have been researched
and modeled for the EMRAM to date
PwC
26
26
The evidence is based on >200 studies, >60 interviews,
multiple business cases, site visit reports and discussions
with HIMSS
Hypotheses validation
through external sources
Create key assumptions
First-Pass
Literature
Review
First-Pass
Benefits
Hypotheses
PwC research
team sourced
>200 peerreviewed and
other studies
> 60 Interviews
and meetings
with leading client
clinicians and
departments
PwC
Business
Cases from 7
Other Large
Scale EMR
Programs
US Site Visit
Reports from
Client Team
Interviews
with HIMSS
PwC Experts
from Asia,
the US and
Australia
Build model and test/check
Detailed
Quantification
and Scenarios
>40 Logic
Models to quantify
benefits based on
peer-reviewed
literature, expert
assumptions, client
data
Clinician
Validation
Validated by
17 senior
clinicians and
steering
committee
members
28
Two Results: a Benefits Model, and a Management
Tool
1.
The Benefits Model is a useful tool to estimate
•
the aggregate impact of the investments for funding bodies (e.g. treasury)
especially for a business case
•
the flow – through effects
•
the prioritization of benefits and hence of program focus
•
the ability to estimate impact on wards, modules or the whole hospital
2. The Benefits Management Plan (and taking appropriate action) does NOT require
the Benefits Model and is a strategic element in operational performance
management and accountability. This is a key change management enabler
for management to commit to benefits and program delivery early on
PwC
•
The Logic Flow Modules will identify the key indicators, and the target value
•
If the indicators are below target value, measures can be taken, e.g. improved
training, added incentives, better change management, system re-design, or more
detailed research and evaluation
25
27
4
Example Insights
PwC
30
The following results are based on the PwC HIMSS EMRAM
value modelling and include some data and assumptions
from our collaborative benefits work with Melbourne Health
• Top Ten Benefits
• Top Capabilities mapped to EMRAM
• Benefits by Stakeholder
• Benefits by Type
• Impact on Bed Days & Episodes
PwC
31
% of total cost base
40.0
30.0
20.0
10.0
50.0
1.9
1.8
Source: PwC Benefits Modelling with results based on peer reviewed academic research,
case studies and some client data
1.5
Benefits
1.4
1.0
0.7
Other Benefits
Reduction in ED length of stay for existing patients
Reduced length of stay due to improved
prescription turnaround
Reduced time spent transcribing prescriptions
Reduced medication order errors
Productivity improvement due to access to tablet device
Reduced radiology turnaround time
Reduction in readmissions
Increased DRG coding accuracy
Reduction in time spent looking for information
Reduction in medication administration errors
Illustrative Modelling Results: Top ten Benefits
60.0
4.0
26.0
0.7
3.0
3.0
7.0
Top 10 Benefits
Net Benefits Impact as % of Hospital Annual Cost Base
-
Note: Confidence intervals represent the upper and lower bounds
of achievable benefits as indicated by the peer reviewed literature
30
Top 7 Capabilities (mapped to EMRAM)
Relative Benefits by Stage
(as % of total hospital cost base)
and component
Data Warehouse & BI
CDSS,
Closed Loop
Meds Admin &
Physician Doc
Not currently modeled
CPOE and CDSS
Nursing/clinical documentation
CDR
Not currently modeled
Not currently modeled
18.0
16.0
14.0
12.0
10.0
8.0
6.0
PwC
4.0
2.0
Source: PwC Benefits Modelling with results based on peer reviewed academic research,
case studies and some client data
Note: Confidence intervals represent the upper and lower bounds of
achievable benefits as indicated by the peer reviewed literature
31
Analysis by generating Stakeholder
Benefits as % of Total Hospital Costs
20.0
18.0
16.0
14.0
13.5
11.4
12.0
10.0
8.0
6.0
4.0
2.0
1.0
Doctors
Nurses/Allied health Administration staff
Source: PwC Benefits Modelling with results based on peer reviewed academic research,
case studies and some client data
PwC
0.1
Pharmacists
Note: Confidence intervals represent the upper and lower bounds of
achievable benefits as indicated by the peer reviewed literature
34
Analysis by Type of Source Benefit
Benefits as % of Total Hospital Costs
25.0
20.0
14.9
15.0
8.9
10.0
5.0
2.4
(0.2)
Better Quality of Care
Enhanced Safety of
Care
Improved Efficiency
Disbenefits
(5.0)
Source: PwC Benefits Modelling with results based on peer reviewed academic research,
case studies and some client data
PwC
Note: Confidence intervals represent the upper and lower bounds of
achievable benefits as indicated by the peer reviewed literature
35
Significant operational impact
Comparison of Current vs. To Be Episodes
Base Case: current beddays, current episodes, current costs
With EMR: 66% of beddays, 73% of
episodes, cost reduction approx. 30%
•
•
•
•
Faster results
Faster treatment
Fewer errors
Elimination of
duplicates
• Less time searching for
information
 Earlier discharge
 Lower cost per order
events
New capacity:
34% beddays
Actual impact depends
on how well benefits &
implementation is
managed. Hospitals
may only be able to
achieve a fraction of
this potential if there is
no accountability and
ownership.
Source: PwC Benefits Modelling with results based on peer reviewed academic research,
case studies and some client data
PwC
36
In addition, the PwC methodology also allows for
non-monetary results
Sample modelling of a 500-bed hospital
•
Improving the capacity and throughput of the Emergency Department
through a reduction of 31,100 ED patient hours
•
A saving of 34% bed days
•
Improving medication compliance and reducing Adverse Drug Events
(ADEs) as well as Intensive Care services potentially saving over 200 lives
per annum
•
Reducing the readmission rate by 11%
•
A 12 % reduction of laboratory tests (approx. 80k)
•
Avoiding more than 500 preventable hospital admissions
•
Reducing the time spent by clinicians searching for information by
between 30-60 mins per shift.
PwC
26
35
5
Discussion
PwC
38
How much of these benefits are achievable? (1)
The approach is conservative...
1. Ancilliary systems, such as laboratory, pharmacy and imaging were
not valued (as these systems were in place at the client)
2. Full radiology PACS was not valued (as this was in place at the client)
3. Only those capabilities that were well represented in peer-reviewed
academic articles were quantified (so several other EMR capabilities
also included in the EMRAM were not valued, e.g. data continuity
between ED, OR)
4. The mid point of literature ranges was used, giving a considerable
upside to the dollar impact of almost 50% for a well-managed
program
PwC
26
36
How much of these benefits are achievable? (2)
But achieving benefits in a healthcare systems is always hard...
1. The peer-reviewed evidence base for having achieved these benefits is real
and has been identified – this value has really been generated
elsewhere
2. It does not mean that your institution will also achieve this, nor achieve it
from day 1. Introducing many components at once means that the challenge
to manage for benefits will be greater
3. And managing variable costs (and long term variable costs) is
difficult. For example, if no staff changes are implemented, a reduction in
duplicate testing can result in higher cost per order rather than a reduction in
testing costs
4. Each hospital is different – and the benefits models and expectations
should ideally be developed with your expert clincians, assisted by
our experienced team
5. For example, one of our clients used a 50% discount on the benefits identified
by this methodology in their business case
PwC
26
36
Resulting Business Intelligence from the Benefits
Strategy and Modeling
1. What to spend the marginal $ on (e.g. rollout, training,
functionality, provider adoption)
2. Which stakeholder(s) to focus on for change and adoption
3. How the benefit equation looks for each stakeholder (and
where to place incentives)
4. The value generated and received by each individual stakeholder
5. The impact of introducing capabilities earlier or later
6. How benefit flows to the different funding organisations
including both public and private!
7. What to measure out of the hundreds of metrics that might be
identified
8. The interpretive value of the measurements (what it means)
PwC
26
38
PwC now has the capability to define, measure and evaluate
the continuum of care and the exchange of health related
information
Within a Hospital
(our discussion
today)
Primary,
Community and
Intermediate and
Long Term Care
Based on hospital
clients
Based on the shared
Personally Controlled
Electronic Health
Record in Australia
Rural Care
including
Telehealth
Based on work with
our rural clients
Working closely with our clients and their clinicians to determine their
specific variables and benefits management model / implementation plan,
and even more importantly, bring them along on the change journey.
PwC
26
38
Potential Applications of the Model
1. Health IT Strategy (what to focus on, how to implement)
2. Business Cases
3. Enabler for Change Management, including accountability for
the management team and commitment to performance
improvement (focusing on benefits realisation)
4. Monitoring implementation success
5. Rolling out new health IT capabilities (what to focus on)
6. Expanding existing health IT capabilities
(e.g. to a new ward)
7. Extending existing health IT capabilities
(e.g with enhanced functionality)
8. Non-IT projects, such as business process re-engineering
PwC
26
39