himss clinical & business intelligence community of...

HIMSS Clinical & Business Intelligence

Community of Practice

September 24, 2015

Welcome

Shelley Price, MS, FHIMSS

C&BI Community Organizer

Director, Payer & Life Sciences, HIMSS

Nancy Devlin

C&BI Community Organizer

Senior Associate, Payer & Life Sciences, HIMSS

Arthur Panov, MPH, CPHIMS

C&BI Community Co-Chair

HIT Architect IBM

Michael Berger, PE

C&BI Community Co-Chair

Chief Analytics Officer

Affinity Health Plan

Agenda • Welcome

• HIMSS C&BI Community Updates / Announcements

• Presentation & Discussion:

“Big Data Analytics to Drive Actionable Insights” o Anil Jain, MD, FACP, Senior VP and Chief Medical Officer, Explorys, an

IBM Company

• Wrap-Up / Next Steps

C&BI Community Updates / Announcements

DRAFT - HA3

HIMSS Analytics Adoption & Assessment ModelSM Stage Overview of Key Cumulative Capabilities

Stage 7

Personalized Medicine &

Prescriptive Analytics

• Mass customization of care through analytics expands to wellness management, physical and behavioral functional health

• Natural Language Processing (NLP) of text and/or verbal interfaces and CDS

• Point of care prescriptive analytics to improve patient specific outcomes

• 7x24 biometrics data, broad social determinants of health data, genomic data

Stage 6

Clinical Risk Intervention &

Predictive Analytics

• Analytics identifies and supports high volume diagnosis-based per-capita cohorts

• Focus expands from management of cases to collaboration between stakeholders for the management of episodes of care

• Advanced analytics with outreach, education, population health, triage, referrals, admin avoidance, and readmissions

• Patient engagement is profiled with flagging in registries for those that are unable or unwilling to participate in protocols

Stage 5

Triple Aim & Suggestive

Analytics While Managing

Financial Risks

• Analytics are available at the point of care to support the Triple Aim of maximizing the quality of individual patient care,

population management, and the economics of care

• The oversight shares in the financial risk and rewards that are tied to clinical outcomes

• Data content expands to include bedside devices, monitoring data originated in the home care setting, external pharmacy

data, and detailed activity based costing

Stage 4

Clinical Effectiveness &

Population Health

• Analytic activities are focused on measuring adherence to clinical best practices, minimizing waste, and reducing variability

• Analytics governance expands to support care management teams focused on improving the health of patient populations

• Precision of registries is improved by including data from lab, pharmacy, and clinical observations

• Enterprise Data Warehouse (EDW) content is presented as evidence-based, standardized data marts that combine clinical

and cost data and have associated links with patient registries

Stage 3

Automated Operational,

Planning, and Performance

Management Reporting

• Enterprise Data Warehouse (EDW): An EDW with an enterprise wide data model and orientation that brings in data from all

critical internal data sources and some critical external data sources

• Consistent, efficient production of reports supporting basic management and operations of the healthcare organization

• Key performance indicators are easily accessible from the executive level to front line managers

• Data governance expands to increase data literacy and develop a data acquisition, stewardship, and management strategy

• Centralized data governance exists, reviews/approves externally released data via standard release policy and procedure

Stage 2

Relating and Organizing

Core Data

• Data Warehouse Build-out: A formal primary database is acting as an enterprise repository of centralized and managed

data. Includes historical patient, clinical, and financial data, managed with master data management approaches

• Patient registries are defined at least by ICD billing data

• Data governance is accelerating exercising master data management and creating patient registries

Stage 1

Foundation Building

• Data Warehouse Foundation: A central repository of managed and integrated data from one or more disparate sources

• Searchable metadata repository is available across the enterprise

• Data sources are updated within one month of source system changes

• Data governance is forming around the data quality of source systems

Stage 0

Fragmented Point Solutions

• Distributed Isolated Silo Solutions: Multiple fragmented business and clinical data presentation and management solutions

are not architecturally integrated

• Overlapping ungoverned data content leads to multiple versions of the truth

• Report development is labor intensive and inconsistent

Population Health Summit

Population Health Management: Evolving the Provision of Care to

Accelerate Healthcare Transformation

• Summit Focus:

– Describe how healthcare organizations are meeting the

challenges of healthcare transformation;

– Identify ways that healthcare organizations are replacing

value with volume by offering patient-centered quality

care across an entire population;

– Illustrate the critical success factors organizations need

to thrive in this new era of pophealth management.

Population Health Pavilion

Your experience as a Company:

• 10-12 companies

• Speaking sessions

• Packages start at $3,500

• Reception Sponsorship

• Overall Summit Sponsorships

• Discount offered for participation in the Summit and HIMSS16

www.himsspopulationhealth.org/population-health-pavilion

For more information, please contact:

Tia Peterson

312.915.9232

[email protected]

Your experience as an Attendee:

• The Population Health

Pavilion will demonstrate tools

and resources focused on

helping providers build the

sophisticated clinical and

financial capabilities needed to

mitigate financial risk and

demonstrate increased quality

outcomes.

http://www.himsspopulationhealth.org/about-exhibition

http://www.himsspopulationhealth.org/population-health-pavilion






mailto:[email protected]








HIMSS Big Data+Healthcare Analytics Forum

The two-day forum brings together leading providers, payers, researchers, and

government officials to provide

• best practices,

• cases studies,

• peer-to-peer learning, and

• expert insights into how healthcare organizations are currently using analytics

to mine data to improve clinical care and enhancing financial performance

and administrative decision-making.

REGISTER TODAY: http://www.bigdatahitforum.com

http://www.bigdatahitforum.com/

http://www.bigdatahitforum.com/

C&BI Community

Guest Speaker

© 2015 IBM Corporation

Overcoming Challenges and Finding Opportunity

Anil Jain, MD, FACP

Senior VP & Chief Medical Officer, Explorys

IBM Watson Health

HIMSS Clinical & Business Intelligence Community Webinar

September 24, 2015

Big Data Analytics to Drive Actionable Insight

12 © 2015 IBM Corporation | NOT FOR DISTRIBUTION

Objectives

• Review big data and the impact on the clinical and

business intelligence community

• Demonstrate challenges of leveraging data across health

systems (harmonization, sharing, skills, etc.)

• Describe how data can be organized to support actionable

analytics for net new knowledge discovery and value-

based care


About the Presenter

Anil Jain, MD, FACP

Co-Founder, Senior VP

Chief Medical Officer

Explorys

(an IBM Company)

Dr. Anil Jain is Senior VP and Chief Medical Officer of Explorys, an IBM Company,

a Cleveland-based BIG DATA healthcare analytics company formed in 2009 based

on innovations that he developed while at the Cleveland Clinic. He leads

Informatics and Analytics, Product Development and Product Management, and

Secondary Use.

Dr. Jain began his career at the Cleveland Clinic in 1995, most recently as Senior

Executive Director of IT until July 2011 where he led several Health IT innovations,

including programs to support research and quality informatics and created

interactive dashboards to monitor the “meaningful use” of the Electronic Health

Record. He continues to practice medicine and teach medical residents as

Consulting Staff at Cleveland Clinic’s Department of Internal Medicine. He was

formerly co-Chair of the Information Management Committee of Better Health

Greater Cleveland (BHGC), a Robert Wood Johnson Foundation Aligning Forces

for Quality (AF4Q) Community. In addition, Dr. Jain had previously served as co-

Director of the Biomedical Research Informatics core of the Clinical & Translational

Research Collaborative (CTSC) at the Case Western School of Medicine and

Instructor at Cleveland Clinic Lerner College of Medicine.

Dr. Jain also serves on several advisory boards of Health IT companies in

California and New York. He has authored more than 100 publications and

abstracts and has given numerous talks at national and international meetings on

the benefits of Health IT and how BIG DATA analytics can support quality

improvement and biomedical research. He is a Diplomat of the American Board of

Internal Medicine (ABIM), a Fellow of the American College of Physicians (ACP),

and an active member of both the Health Information Management and Systems

Society (HIMSS) and the American Medical Informatics Association (AMIA). Finally,

he serves as a reviewer for several biomedical journals and national meetings.

| CONFIDENTIAL | Anil Jain, MD, FACP

About Explorys • Spun out of the Cleveland Clinic in

October 2009, based in Cleveland, OH

• Big data aggregation, analytics and applications platform for population health and cohort discovery

• 55 M patients across 27 IDNs, 360 hospitals include EMR, Billing & Claims

• Acquired by IBM, April 13, 2015

• Part of Watson Health, the first newly formed unit


March 4, 2013

Health Care Challenges over the Years

December 1, 2008 August 10, 2009 January 8, 2015 January 22, 1996


Value & Science-Driven Healthcare Strategy

http://www.iom.edu/, 2012 & 2014

http://www.iom.edu/


Drivers influencing analytics investments

Deloitte Center for Health Solutions

2015 US Hospital and Health System

Analytics Survey, n=50


*The value of analytics in healthcare: From insights to outcomes, IBM Institute for Business Value

Healthcare Analytics Goals and Objectives


Healthcare Analytics has different data priorities


Gartner Hype Cycle &

The evolution of Explorys Gartner, 2013


What is Big Data?

Volume

Velocity

Variety

Veracity

Any data that is not readily

amenable to traditional

tools for storage or analysis

can be considered Big Data.


“Perfect Storm”: The “rise” of Big Data in Healthcare

Big Data

2010 ACA

Promote new Delivery and

Reimbursement Models

2009 ARRA/MU

Foster meaningful adoption of EMRs and HIEs (Health

IT) Exogenous Data

Wearables.

Devices, Imaging and “omics”, social

media


Adoption of Electronic Health Records


Increases in Health Information Exchanges since 2008

SOURCE: ONC/American Hospital Association (AHA),

AHA Annual Survey Information Technology Supplement


New data about individuals, used in new ways

6 Terabytes Per lifetime

60% Exogenous

Factors

1100 Terabytes Generated

per lifetime

0.4 Terabytes Per lifetime

30% Genomics Factors

10% Clinical Factors


Growth of Wearables and Social Media

http://www.wearabletechworld.com/t

opics/from-the-

experts/articles/323855-wearable-

technology-next-mobility-market-

booming.htm

http://blog.himss.org/2014/04/03/

himss14-by-the-numbers/


Growth of “Omics” – DNA Sequencing

Keshavan M. “Genomics: Our most data-hungry industry”, MedCity News, July 7, 2015


The How: Driving Knowledge from BIG DATA…

Alert

Drill-

Down

Query

Report What happened?

How many, how often, when?

What exactly is the problem?

What actions are needed?

Optimize

Predict

& Model

Forecast

Analyze

What’s the best that can happen?

What will happen next?

What if these trends continue?

Why is this happening?

Predictive Analytics

(the “so what….

And the “now what”)

Future-oriented and source

of competitive advantage

Descriptive Analytics

(the “what”)

“Rearview mirror” –

Provides foundation and

insight

Com

pe

titive

Ad

va

nta

ge

Degree of Intelligence


Gartner Hype Cycle &

The evolution of Explorys Gartner, 2014


Socio-Technical Challenges of Leveraging Data

• Data Governance

• Privacy & Security related

regulatory considerations

• Standardization and

Harmonization

• Data Quality & Validation

• Interpretation and Analysis

• Sustainability

Sittig D & Singh H. Qual Saf Health Care 2010;19:i68.


Primary Obstacles to Widespread Analytics Adoption, 2012


Results of the 2014 National Survey of ACOs*

• More than 50% of respondents on

the recent survey listed the

following as some challenges to

leveraging HIT Infrastructure:

Interoperability

Cost/Funding/ROI

Workflow Integration

Lack of Provider Engagement

Training

Lack of consensus on quality

measures & benchmarks

*Premier Inc. & eHealth Initiative; https://www.premierinc.com/aco-interoperability-survey-9-24-14/

• More than 50% of respondents on the

recent survey* listed the following as

some challenges to leveraging HIT data

and analytics:

– Access to external data

– Integration of disparate data

– Workflow Integration

– Data liquidity & quality

– Applying analytics to actual practice

– Training


Barriers to Adoption, 2015

Out of 50 responders – rated as high influencers

– Culture & Politics (31)

– Fragmented Ownership (29)

– Access to Skilled Resources (27)

– Funding Process (26)

– Data Quality (23)

– Sponsorship (20)

– Tools and Technology (18)

– Data Access (15)

Deloitte Center for Health Solutions 2015 US Hospital and Health System Analytics Survey


Data Quality, e.g., Validating diagnoses

Navaneethan SD, Jolly SE, Schold JD, Arrigain S, Saupe W, Sharp J, Lyons J, Simon JF, Schreiber MJ Jr, Jain A, Nally JV Jr.

Development and validation of an electronic health record-based chronic kidney disease registry. Clin J Am Soc Nephrol. 2011

Jan;6(1):40-9. Epub 2010 Nov 4.


HIPAA & Trade-off between Privacy Protection and Value

Value of Information

Dis

clo

sure

Pro

tections

Less Value

Safe Harbor

De-Identified

DS

Less

Protection

Full PHI

Balanced

Statistically

De-identified

DS or

Limited DS


Data Science

http://practicalanalytics.files.wordpress.com/2012/01/b

ig-data-value-chain.png

Data science is the study of the generalizable extraction of knowledge from data (including big data) incorporating many fields, including signal processing, mathematics, probability models, machine learning, statistical learning, computer programming, data engineering, pattern recognition and learning, visualization, uncertainty modeling, data warehousing, and high performance computing with the goal of extracting meaning from data and creating data products. A practitioner of data science is called a data scientist.


MGMT.design / Source: EMC (survey of 497 data professionals)

Data Science Talent

Source: McKinsey Global Institute


Conceptual Architecture for Big Data Analytics

From Raghupathi and Raghupathi Health Information Science and Systems 2014 2:3


Applications

Acquisition &

Storage Layer

Inpatient

EHRs

Patient

Portal Pharmacy

Billing &

Accounting

Satisfaction

Surveys Health Plans

Data

Warehouses Laboratory HIE / HL7

Ambulatory

EHRs

Care

Mgmt.

Systems

Explorys HDG Health Data Gateway

Over 1000 live connectors

Data Curation Engine

Patient Matching Engine

Data Governance Engine

Risk & Prediction Engine

Measure and Gap Engine

Registry & Work-List Engine

Unified Patient View Engine

Engage/Outreach Engine

Report & Analytics Engine

Transformation

Layer

Health Services & Clinical Research

Health & Wellness Population Health Management

Value & Efficiency Optimization


Examples of Descriptive Analytics

Population

Assessment Know their past and future utilization, their risks, and which are can be mitigated given your

network’s capabilities.

• Population Profile

• Demographic distribution

• Disease profile

• Geographic footprint

• Historical Utilization

• Utilization break-down

• High cost claimants

• E&M distribution

• Risk and Projected Utilization

• PMPM by condition and type

• Drivers of projected utilization

• Clinical Performance

• Overview of program

performance (key measures

and benchmarks)


Examples of Risk Stratification Analytics & Insights


Examples of actionable insight through advanced visualizations

Performance

Measurement Up-to-date network-wide reporting and measures relative to performance targets, program return-on-investment, and pinpointed opportunities for continued improvement.

• Role Specific Dashboards for…

• Leadership

• Care coordinators

• Providers

• Program Specific Measure Libraries & Scorecards

• ACO (commercial and Medicare)

• Medicare Advantage

• Employee Health plan

• Physician-based HEDIS

• Inpatient quality and efficiency

• Utilization

• Pre-built Reports & Data Marts

• Support for GPRO reporting

• Provider scorecards and performance plans

• HCC and proper coding opportunities

• Contract performance


Need for better risk models

• Current State Future State

– Risk-adjustment models typically rely upon demographic factors and

medical conditions.

– Models should include treatments (medications and procedures) in addition

to diagnoses.

• Current Performance

– Existing risk-adjustment models all perform similarly in predicting future

costs (SOA).

– Typical Models with diagnosis and demographics exhibit modest

performance

• Bottom Line

– Customers want better clinically-transparent models.


Factors to consider in the evaluation of Risk Models

• Ease of use

• Availability of the model

• Cost of using the model

• Access to data of sufficient quality

• Ability to leverage for a specific

application

• Provides both useful clinical as well

as financial information

• Used for payment or for care

management

• Reliability of the model across

settings, over time, or with

imperfect data

• Current adoption in the market or

organization

• Susceptibility to gaming or up-

coding


How Big Data improves Predictive Analytics

• Volume

– More patients across more time

– More data will generally improve the performance characteristics by looking

at more samples

• Variety

– Adding additional data to the typical demographics and diagnoses can

improve the performance characteristics

– Traditional models may treat poorly controlled diabetics (A1c >9) the same

as well controlled (A1c < 7) is laboratory data is absent

• Velocity

– Frequent addition of data may allow for self-tuning or learning models

• Other

– Parallel computing can allow for advanced machine learning algorithms

such as neural networks and random forest models


Predicting Episode-Based Care Costs following Coronary Intervention

Cleveland Clinic Authors: Bunte M, Cacchione J, et al.

Explorys Authors: Miller J, Pohlman M, Gilder J, Jain A

Background: Bundled payment models for healthcare offer cost reduction

opportunities, although such models have yet to be applied to percutaneous

coronary intervention (PCI)-related treatment. Episode-based cost prediction

may advantage providers taking on additional financial risk in an emerging

landscape of accountable care.

Results: Cost bundling allowed categorization of patients into low, medium,

and high price classes. Relevant acute care costs determined the full episode

price for 72% of patients at a median of $10,946. Ten percent had high post-

procedural costs. Full episode price was predicted within $5000 for 75% of

patients.

Conclusion: As proof of concept, bundled payment modeling can predict

relevant costs based on pre-procedural data. This cost bundling framework

may be used to identify resource-intensive patients early in the course of

care, prompting strategies to reduce overall cost.

http://content.onlinejacc.org/article.aspx?articleid=1664790


Objective To demonstrate the potential of de-identified clinical data from multiple healthcare systems using different electronic health records (EHR) to be efficiently used for very large retrospective cohort studies. Results Comparing obese, tall subjects with normal body mass index, short subjects, the venous thromboembolic events (VTE) OR was 1.83 (95% CI 1.76 to 1.91) for women and 1.21 (1.10 to 1.32) for men. Weight had more effect then height on VTE. Compared with Caucasian, Hispanic/Latino subjects had a much lower risk of VTE (female OR 0.47, 0.41 to 0.55; male OR 0.24, 0.20 to 0.28) and African-Americans a substantially higher risk (female OR 1.83, 1.76 to 1.91; male OR 1.58, 1.50 to 1.66). This 13-year retrospective study of almost one million patients was performed over approximately 125 h in 11 weeks, part time by the five authors. Conclusions With the right clinical research informatics tools and EHR data, some types of very large cohort studies can be completed with minimal resources.

Predicting Patients with Thromboembolic Events


Concluding Thoughts

• Focus on building a big data strategy before building big data

technology

• The clinical and business intelligence community needs to

help shape the Big Data strategy

• Cultural alignment and a governance structure is just as

important as access to data

• The use cases should be developed with “crawl, walk and

run” phases to drive continuous incremental progress

• Identification of required skills and current gaps early is

essential to ensure appropriate staffing

• Use ROI models to drive scarce resources to your analytic

projects

• Reach out to your network for help


Questions and Contact Information

Anil Jain, MD, FACP

Senior VP & Chief Medical Officer

Explorys, an IBM Company/Watson Health

[email protected]


• Want to get involved?

Speaker or topic ideas

Key note presenter

Blogger, twitter

Contact Nancy Devlin

• Community Website

www.himss.org/ClinBusIntelCommunity

Wrap-Up

http://www.himss.org/ClinBusIntelCommunity

JOIN US!

• Next meeting: Thursday, November 19, 2015

TBA

Speaker: TBA

Next Steps

FY15 Leadership and Contact Information Co-Chairs: Mike Berger, PE Chief Analytics Officer Affinity Health Plan [email protected] Arthur Panov, MPH, CPHIMS HIT Architect IBM [email protected] HIMSS Community Organizers: Shelley Price, MS, FHIMSS Nancy Devlin Director, Payer and Life Sciences Sr Assoc., Payer and Life Sciences HIMSS HIMSS [email protected] [email protected]





Thank You

Appendix

John Middleton, MD, MSc FY16 C&BI Committee Chair

VP/CMIO

SCL Health

David Butler, BSME, MBA, FHIMSS FY16 C&BI Committee Vice-Chair

President

Heartland Innovations, LLC

Cheryl Bowman, CPHIMS* Data Manager

University of Wisconsin Hospital and Clinics

Raj Lakhanpal, MD, FACEP* CEO

Spectramedix

Sharon Lynn Morley, RN/CNS* Client Manager

Humedica

John S. Moses, MA Director of Enterprise Architecture, The University

of Chicago Medicine

Ravi Narayanan, MS Director, Research Data Management and

Analytics

Medica Research Institute

Stuart Rabinowitz, MBA, BC* Director Federal Markets - Socrata

Socrata

Chester H Robson, DO, FAAFP* Medical Director, Clinical Programs and

Quality

Walgreen Co.

Deborah Jane Rupe, RN, MS,

FHIMSS Clinical Analyst, Shriners Hospitals for

Children - Tampa Hospital

Ahmad Sharif, MD, MPH, SCPM Chief Medical Information Officer, Resolute

Health

Louise Sulecki, MBA Systems Analyst, Cleveland Clinic

J.D. Whitlock, MPH, MBA,

CPHIMS* Vice-President, Clinical & Business

Intelligence

Mercy Health

* Indicates a

returning

committee

member

2015-2016 C&BI Committee Members

C&BI Community of Practice The goal of the C&BI Community is to bring together thought leadership and share knowledge that will support the future success of our members by improving their ability to understand and form partnerships to manage C&BI as a part of doing business and providing accountable and quality care to their members. The Community will support activities that promote peer-to-peer networking, problem solving, solution sharing, and education.

Topics of focus may include:

• Storage and Management of Data and Supporting Technologies

• Knowledge Management to Support Accountable and Quality Care

• Case, Risk & Cost Management

• Best Practices Clinical & Business Analytics

• Clinical Decision Support

• Research Data Warehousing/EDW

• Data Lifecycle Management

C&BI Community of Practice

• Open to all HIMSS members (current membership: approx 6,700 people)

• Will meet virtually 6 times/year

• Agenda for the meetings may include:

• Commencing with a short series of 2-Minute Drills presented various Community members

• Topical discussion with key note presenter

The ‘2-Minute Drill’ is based loosely on the sports analogy, and in this case

is a fast-paced (short in length) presentation on a hot, emerging, or timely topic, news event (e.g. research paper, game-changing market or technology news), or recent and relevant event (e.g., federal public meeting, legislative/federal/judicial news, critical conference or educational event).

2-Minute Drills foster greater peer-to-peer networking, member engagement, problem solving,

solution sharing, and education. If you are interested in presenting any drills, please contact Nancy or Shelley.

http://www.clker.com/clipart-12908.html




C&BI Task Force

NEW! C&BI for Population Health Task Force

CO-CHAIR: Karen Golden Russell, FHIMSS, MA, MBA | Vice President, Population Health | Verisk Health

CO-CHAIR: Michelle Vislosky, M.B.A., FACHE | Zone Sales Executive – East Region of Canada & the United States | Caradigm

This group creates resources and tools that employ practical guidance and unbiased information to help healthcare organizations (providers, hospitals, integrated delivery networks, health plans and other stakeholders) use C&BI to harness, use and analyze data captured in the healthcare setting to execute population health management initiatives and improve care and health outcomes.

Meeting times: 3rd Tuesday of the month, 3:30-4:30pm ET

himss clinical & business intelligence community of...

Documents