Upload
others
View
8
Download
0
Embed Size (px)
Citation preview
1
Communicate Value & Build Support for Enterprise Analytics
Session 77, February 12, 2019
Christopher J. Donovan, MBA, Executive Director, Enterprise Analytics
Andrew W. Proctor, MS, Senior Director, Enterprise Analytics
2
Christopher J. Donovan, MBA
Executive Director, Enterprise Analytics
Andrew W. Proctor, MS
Senior Director, Enterprise Analytics
Have no real nor apparent conflicts of interest to report.
Conflict of Interest
3
• Cleveland Clinic Introduction
• Enterprise Analytics Structure
• Program Design
• Executive Support / Communication
• Enterprise Data Vault
• Data Governance / Information Management
• Governing data
Agenda
4
• Describe the four pillars of an Enterprise Information Management
and Analytics strategy
• Explain the story framework necessary to get C-Suite and Board
endorsement
• Discuss use cases that highlight the “why” of analytics rather than
the “how”
• Show hard-dollar savings from investments in data and analytics
platforms
Learning Objectives
5
CCHS Main CampusCCHS Main Campus
6
Organization Structure
• Not for Profit
• Group Practice
• Physician Leadership
• Salaried – 1 Year Contracts
• Annual Professional Review
7
Cleveland Clinic Today
• 51,500 caregivers
• 7.1 million total visits
• 220,000 hospital admissions
• 3,600 physicians & scientists
• 1,960 residents & fellows
8
U.S.News & World Report 2017-18
1 Mayo Clinic
2 Cleveland Clinic
3 Johns Hopkins
4 Mass General
5 UCSF
9
Global Integration
Cleveland Clinic
USACleveland Clinic
Abu Dhabi
Cleveland Clinic
London
10
London
Abu Dhabi
Cleveland
Northeast
Ohio
West Palm Beach, Florida
Weston, Florida
Las Vegas,
Nevada
Toronto,
Canada
Cleveland Clinic Locations
11
Industry Transformation
Advocate, Aurora Health Care to merge, create $11 billion health systemSystems say together they will form the 10th largest nonprofit integrated health system in the country
Optum to buy DaVita
Group for $4.9 billion
CVS Health to acquire Aetna
for $69 billion in year’s
largest acquisition
Amazon, Berkshire Hathaway and JP Morgan Could Disrupt U.S. Health Care and Capitalism As We Know It
Why Apple, Amazon, and
Google are making big health
care moves
Silicon Valley wants to disrupt your health careBy Dylan Scott
12
• Capturing, integrating and interpreting data from all available sources
• Predictive algorithms that surface opportunities for intervention and prevention
• Thoughtful use of augmented intelligence
• Seamless connections to patients and partners as healthcare becomes “borderless”
• New skills, knowledge and pipelines
• Mature data governance
Digital Transformation at CC
13
What is Your Story?
• Understand your audience- Board vs. Executive
Leadership
• Build a storyline and then tell the story
- Problem / conflict
- Solution
14
Different Audiences
• Board Members
– Why and What are we going to do?
– Enterprise strategy
– Risk Management
– Long term viability
• Executive Leadership
- Enterprise strategy
- Operations & performance
- Why, What & How are we going to do it?
- Long & Short term value
• ROI
• Culture
15
Engage in the Vision
• Data is readily available, understood and expected
• Discussions are interactive and supported by data discovery and intuitive visualizations
• Requires investment in technical and human resources
“Without data – you’re just another
person with an opinion”
W. Edwards Deming
16
Building Analytic Capability
How do we act on it?
What’s the best that can happen?
What will happen next?
Why is this happening?
What actions are needed?
What were the key drivers?
How many, how often, where?
What Happened?
Sophistication of Analytics
Com
pe
titive
Ad
va
nta
ge
Std
Reports
Adhoc
Reports
Data
Exploration
Alerts
Scorecards
Statistical
Analysis
Forecasting
Predictive
Analysis
Optimization
Proactive
Decision
Making
Reactive
Decision
Making
Adapted from: Copyright © 2012, SAS Institute Inc. All rights reserved. ANALYTICS MATURITY GOAL: FROM REACTIVE TO PREDICTIVE
17
International Institute for Analytics
2.25
2.88
2.89
2.96
3.05
3.19
3.23
3.29
3.43
3.55
3.58
3.62
3.91
4.56
Healthcare - Providers
Specialty - Retail
Insurance
Healthcare - Insurance
Utilities & Telcom
Automotive
Airlines
Grocery
Manufacturing
Pharma & Med Devices
General Retail
Consumer Brands
Financial Services
Digital Native
Industry Best in Class
Maturity Score
Medians
18
Develop the
ability to act on
insights from all
available data to
improve patient
outcomes,
affordable care,
education and
research
Data Enhance the value of CC data assets
People
Develop an analytically
capable organization
Process
Value analytics over
instinct
Technology
Create a modern analytics platform
CC Enterprise Analytics Vision
19
Dashboards
Enterprise Data Vault
Hadoop
Data LabsExternal Data
Data Asset Management
Enterprise Data Views Analytic Engine
Analytics Spreadsheet Reporting KPIs
Data Integration
Metadata Data Quality Business Glossary Master Data ManagementETL
Analytic Tools
CC Enterprise Analytics Platform
20
• Standardize job descriptions & requirements
• Common entry point for CC analysts
• Leverage rigorous internship programs to seed entry level hiring
• Common early curriculum to develop standard analytical skillset
• Consistent career development timelines, goals and performance measurement
Sophistication
Maturity
Ability to Produce
Make analytics easier to consume
Increase understanding
of analytics
Ability to Consume
Addressing the Analytics “Gap”
21
Program Governance
Executive Steering
Committee
Executive Steering
Committee
Operational Leadership
Team
Operational Leadership
Team
Advisory Council
Advisory Council
Data Governance
Council
Data Governance
Council
Institute Leadership
Council
Institute Leadership
Council
22
• Responsibilities of the data and analytics team:
– Provision of trustworthy, available data
– Deployment of modern, useable analytics tools
– Development of enterprise scale vision, training and support
– Development of expertise and “consulting” around
measurement, performance management, and “math”
• Responsibilities of the enterprise:
– Become facile users of data related insight
– Broadly disseminate an analytically oriented culture
– Commitment to developing required skills
A Shared Effort
23
Cultural Challenges
• 3P’s: Perspectives, Personalities, Politics
• Hoarding of data
• Perceived lack of trust in people
• Perceived lack of trust in data
• Gathering low value/no value data
• Keeping low value/no value data
• Lack of enterprise standards and teamwork
• Need to evolve inspiring leadership in data stewards
• Potential lack of emotional self-accountability
• Potential lack of empathy for others’ data needs
24
Communications Framework
Communication must begin with a clear objective
Message Target Channel Cadence ArtifactWho /
When
Culture
ChangeMeasure
• If the plan was effective, begin final communication
• By establishing effectiveness, you can refine the
communication plan & achieve the results the
organization desires
25
How to communicate what we do
Improved Care Delivery, Operations & Leadership
Quality, Safety & Patient ExpSupport understanding and improvement
of care quality and patient safety in our
journey to high reliability
Enterprise InsightsProvide physicians and leaders with
insights on managing cost and revenue to
ensure our transformation to value
Population HealthProvide information and modeling
capabilities around clinical and financial
risk to support panel management and
performance in risk based agreements
Executive & Operations
Provide a concise view of key KPIs &
statuses for daily operations and
strategic agenda management
Strategic & Financial Planning
Forecast results and understand
financial and market performance to
inform short and long term planning
efforts
Research
Enable data related electronic research
and modeling
Enterprise Capability
Data Governance
Provide appropriate access to well governed
data sets and fulfill our duty as stewards of
our patient & caregiver data
Analytic Culture
Improve enterprise analytic maturity through
the development of new roles, career paths
and focused training and education
Technology & Infrastructure
Provide modern tools and technology to
facilitate data driven decision making and
advanced analytic capabilities including ML
and AI
Serve as Enterprise Resource
Deliver custom analytics in areas outside of
where products are delivered and / or
distributed resources are not available
26
Data Governance Mission
• Enterprise strategic and cultural framework of:
• creating, ensuring, managing high quality, reliable data,
• delivered in a timely and secure manner,
• in compliance with policy, law and regulation,
• while efficiently enabling the organization to meet its datareliant goals,
• at its critical points of need.
27
Data Governance Competency Cntr
Technology
External Data
Access
Strategic
Internal
Data Access
Strategic Strategic
Policy /
Procedure
Strategic
Trust Framework
Data Governance Office
Data
ManagementCompliance
Interlock
28
Data Governance EcosystemHIPAA AND STATE PRIVACY POLICY/REGULATIONS
IT SECURITY INFOSEC POLICY
DATA GOVERNANCE POLICY
ENTERPRISE DATA
GOVERNANCE COUNCIL
E. DOMAIN STEWARD COUNCIL
DATA GOVERNANCE OFFICE
DOMAIN STEWARD COUNCIL
PROVIDER DOMAIN
GOLDEN RECORD
INFORMATICA - MDM
USE CASES
DATA RELIANT NEEDS OF CCF
EXECUTIVE COUNCIL
EXTERNAL DATA
RULES COUNCIL
SECURITY/APPROPRIATE USE
INTERNAL DATA
RULES COUNCIL
PROCEDURE/PROCESS/ACCESS
29
Institute Liaison Program
EA DATA GOVERNANCE OFFICE:
INSTITUTE:ENTERPRISE:
Data Requests?
• Quality?
• Operations?
• Research?
• Priority
• Approval
• Fulfillment
• Collaboration
Administrator
Liaison
Division
Management
Institute
Analysts
Physician
Liaison
Requests,
Policy,
Technology
Adoption &
Oversight
Department
Management
Feedback
30
Primary point of contact regarding initiatives and projects being supported by EA:
• Primary Contact: delegated by Chair between Institutes and EA
• Prioritization: related to analytics requests and projects
• Conduit: for all EA policy and procedure dissemination
• Inter-Institute: contact with other Institutes for Data uses
• (research, validation, operations and analytics collaboration)
• Feedback: contact for collecting feedback/mitigating disputes among Institutes
Institute Liaison Program
31
Enterprise Governance Interlock
32
Established enterprise data governance
structure
– DGO Initiated Councils - 3
– DGO Initiated Task Forces - 6
– Participating in other key workflows across
Enterprise – 11
Developed, approved and implemented key
policies / procedures
– Policies created and approved – 2
– Procedures created and approved – 5
– Guidelines created and approved – 2
– Procedure Implementation Checklists - 3
Establishing DGO as enterprise resource
– Included first of its kind DG questionnaire
section in M&A (Union)
– Initiated discussions with key constituents
related to Appropriate Data Use Internally and
Externally
How to measure progress
EA Executive Council
Institute Data
Access Council
Enterprise
Liaisons
Enterprise Data
Governance Council
33
How Customers Interact/Access Data
Database Tables
Fields Online Glossaries
Business
Glossary
Business Terms,
Name,
description
usage,
categories
policies and
relevant
concepts
Metadata
Management
VarChar 10
View Lineage
Impact Analysis
Data Sharing
Business &
Report
Analysts, Data
Stewards
CONTENT CONTEXT
Who uses this
Information?
CATALOG
Data
Catalog
Central
Repository
stores metadata
from various
sources
View
relationships
between assets
Search Assets
Database
Developers,
DBA’s Data
Architects
34
Data Quality Management (DQM)
Accomplished:
1. 33 Dx rules built and producing error logs
2. Engaged HIM – HIM has been able to utilize output from error logs to identify improvement opportunities
3. Expanding data to validate: Px and CPT rules in development
4. Expanding skillset to add’l team members
5. Utilization of new technology
Goals:
1. Transition data integrity efforts from reactive to proactive/preventative
2. Quantify quality confidence and make available to all users via dashboard
3. Engage and support source areas in prioritization and correction of errors detected via scorecard/error logs
35
Culture Change
Enterprise Patient
Enterprise Data
Data “Care paths”
Better Analytics
Better Patient Care and
Outcomes
36
Show The Work
Show the Math
Workflow Integration
Make it easy to consume
37
• Loaded 996 VCF files from Gene Bank (Cardiovascular Cohort)
• Matched with EPIC derived phenotypic data
61 cases with mutations in colon
cancer related genes
34 cases with no
clinical indication of
colon cancer risk
25 cases with no noted colonoscopy
Opportunity: 15
cases with
increased risk and
no colonoscopy
• Connecting genomic
data with traditional
patient data from our
EMR
• Develop differentiated
care paths and
treatment options
• Identify populations at
risk where early
intervention has
preventive potential
• Historic challenges
regarding size and
complexity of the
datasets
• Ability to blend, explore
and analyze the data
Leveraging Genomic Data
Applying Analytics to Drive Value
• Most risk models will tell you that high cost one year will predict high cost
the next
• At some point, that cost becomes unpreventable
Risk
Pre
venta
bili
ty
• From data mining: 10% of Cleveland Clinic’s attributed Medicare ACO
population had low expenditures in CY1 that tripled in CY2
• 75% of the cohort who had such a significant increase had an
admission in CY2
• In order to identify the “rising risk” cohort, we should be identifying patients at a high risk for preventable admissions (ambulatory care sensitive)Maximizing the skills of care coordinators means identifying
patients who are “rising risk” but haven’t yet “fallen off the cliff”
How can we define the rising risk
population?
Applying Analytics to Population Health
Identifying patients who are at risk for
increased healthcare spend
39
• K-means clustering was used to analyze the
annual expenditures and network utilization
preference (leakage or keepage) of attributed
Medicare beneficiaries.
• Only two variables were used to cluster the
population: annual reimbursement and percent
leakage
• Unique traits of the “leakers”:
- Higher propensity to use home health services
- Tended to be in the far western suburbs
- Twice as many patients suffering from
dementia or other cognitive diseases than any
other cohort
Keepers vs. Leakers
Applying Analytics to Population Health
Net
Re
ve
nu
e
40
Leveraging the enterprise platform
– Revenue cycle project driving $8.5m savings over 5 years
– HIM compliance reporting – discharge audits from 1% to 100% with lower costs &
FTEs ($280k savings over 5 years)
– Cancer center pilot demonstrated an 18% reduction in manual data retrieval and 250
hrs. saved per year
– Captured 2,000+ variant control files for annotation and genotypic / phenotypic analysis
Advanced analytics
– International patient program pricing – projecting $6m of improved collections for 2017
– Applied Monte Carlo simulations to contract negotiations – saving 80 hours per
negotiation
Population Health
– HCC improvement effort – achieved 83% RAF compliance through November (vs.
2016 63%)
AccomplishmentsDelivering Value
41
Are We Building Analytic Maturity?
Source: 2017 IIA Maturity Survey
Stage 1: Analytically Impaired
Stage 2:Localized Analytics
Stage 3:Analytical
Aspirations
Stage 4: Analytical
Companies
Stage 5: Analytical
Competitors
3.06
CC ’17
2.25 3.21
Median Highest
CC ‘14
2.17
Healthcare Domain