Achieving Clinical Transformation with an Interoperable ...s3.amazonaws.com/rdcms-himss/files/production/... · Fast Healthcare Interoperability Resources (FHIR) profiles, that enable

Achieving Clinical Transformation with an Interoperable Health IT

InfrastructureApril 12, 2015

Stanley M. Huff, MD CMIO, Intermountain Healthcare

DISCLAIMER: The views and opinions expressed in this presentation are those of the author and do not necessarily represent official policy or position of HIMSS.

Conflict of Interest

Stanley M. Huff, MD

Has no real or apparent conflicts of interest to report.

© HIMSS 2015

Learning Objectives

• Describe how quality measures, accountable care, and other models of advanced healthcare enable clinician transformation.

• Illustrate standard based infrastructure models, including Fast Healthcare Interoperability Resources (FHIR) profiles, that enable the development and use of interoperable applications in healthcare.

• Demonstrate real projects underway at a ‘healthcare transformation lab’ that are supporting an interoperable health IT infrastructure.

Acknowledgements• Lee Pierce• R. Scott Evans• Brent James

Intermountain Healthcare ProfileAn Integrated Health System

5

1975 1983 1994

• 22 hospitals • 33,000 employees• 22 hospitals • 33,000 employees

• 600,000 members• 25% market share• 600,000 members• 25% market share

• 200 clinics• 1,000 employed

physicians

• 200 clinics• 1,000 employed

physicians

MEANINGFUL USE

Positive Effects of MU• We got money – a net of ~$29 million so far• More physicians used the system

– From 6% to 89% of physicians meeting all measures

– From 10% to 86% with electronic orders– From 61% to 92% with coded problems

• CPOE and ePrescribing development were accelerated

• A more robust infrastructure for health information exchange

Unexpected consequences•Formulary checking•Problems with “no problems”• Immunization interface•Finding denominators•Delay of other more valuable projects

A Retrospective Opinion• Knowing what we know now…• And if we were not under the threat of increasing penalties…

• We probably would not have pursued the MU incentives…

• And our patients would be receiving better care and our clinicians would be happier

From the Office of the National Coordinator

“We are not the office of meaningful use.”

Beyond Meaningful Use

Suggested Strategy• Develop truly interoperable data exchange

standards– We need development to get to truly

interoperable standards; everything we need does not exist today

• Certify messages and services (not applications)• Mandate the use of the standards• Hold people accountable for outcomes (not process

measures)

Intermountain’s Core BusinessPerfecting the Clinical Work Process

Intermountain’s Core BusinessPerfecting the Clinical Work Process

Foundational LeadersDr. Homer Warner

– Medical informatics founder– 1950s – computer assisted

CV decision support– 1970s – HELP system

developed

Dr. Brent James

– CQI - Standardization of clinical care through data analysis

Intermountain’s Clinical Programs

• Aligns care team in the process of care (9 Clinical Programs)

– Physician specialists– Nurses– Data experts– Administrators– IT– Other caregivers

• Develops, implements, and sustains evidence-based care using information systems and data

• Goal is to deliver the best care to every patient every time

16

Transforming Patient Care

Decision Support

Improved CareIdeas

Research

DataDataPatientCare

PatientCare

ActionAction

InsightInsightEDWEDW

ClinicalClinical FinancialFinancial

ClaimsClaims Pt. SatisfPt. Satisf

DeviceDevice & More& More

Data IntegrationData Integration

OTHERS

EDW Conceptual Architecture

Data SourcesInternal

External

Data AccessEnterprise Data Warehouse

BI ToolsEMR

Pharmacy

Claims

Lab

Finance

State/Federal

SO

UR

CE

Dat

a M

arts

SU

BJE

CT D

ata Marts

Master Reference Data

EMREMR

Patient AcctPatient Acct

ClaimsClaims

Primary CarePrimary Care

Women & NewbornWomen & Newborn

AHRAHR

Surgical ServSurgical Serv

Supply ChainSupply Chain

CardiovascularCardiovascular

Patient Sat.Patient Sat.

Data Warehouse Profile 15 TB Oracle Data Warehouse ~9000 queriable tables ~150,000,000 queries per month 95,000,000,000 rows of data

40 FTEs 29 data architects (data input) 11 business intelligence

developers (data for reports) Coordination with clinical and

business areas

Case Studies

Elective Induction CPM

Per

cent

<39

Wee

ksElective Labor Induction <39 Weeks

Elective Induction CPM

Elective Induction CPM

Per

cent

<39

Wee

ks

0%

5%

10%

15%

20%

25%

30%

35%

J1999

FMAMJJASONDJ2000

FMAMJJASONDJ2001

FMAMJJASONDJ2002

FMAMJJASONDJ2003

FMAMJJASONDJ2004

FMAMJJASONDJ2005

FMAMJJASONDJ2006

FMAMJJASONDJ2007

FMAMJJASONDJ2008

FMAMJJASONDJ2009

FMAMJJAS

Month

Elective Delivers <39 Weeks

Colon Surgery: Evidence Based Interventions and Associated Measures

Intervention MeasurePatient Education EnrollmentEarly Mobilization After Surgery

Activity – PT / Nursing walking, transfers, etc.

Appropriate IV Fluid Admin Fluid AdministrationNarcotic Sparing Analgesia Med Administration, Morphine

EquivalenceEarly enteral nutrition Diet Administration

Bowel/Emesis/Flatus Financial Measures

26

Colon Surgery Care Process Report

27

Colon Surgery Care Process - Financials

28

Colon Surgery

• Results: $1.2 million annual savings, LOS decreased from 8.44 to 6.75, while maintaining or improving clinical quality

• 2010 Computerworld Business Intelligence Award – Driving Process Change with BI

• Lack of evidence-based medicine

• Hospitals and physicians are paid for

volume

• Patients are (far too often) not

engaged

Care Delivery in the US: Three Problems…

Massive over utilization

Massive over utilization

Lead to…

Shared Accountability Goals

Goals: 1.Better care (for patients); 2.Better health (for the population we

serve);3.Sustainable costs (for patients and

other payers).

Pro

vide

Evi

denc

e-ba

sed

Car

e

Alig

n In

cent

ives

Eng

age

Pat

ient

s

© Intermountain Healthcare, 2013

FFS PHMFFS PHM

Business Model Implications

Payment Model ImplicationsProviders will be paid for “value”

• Efficient delivery• Efficient utilization• Prevention • Wellness

These become “part of the job”

Physician Payment Model Beta

Current Model – FFS

Beta Payment Model

Service

Qua

lity

Total Cost o

f Care

Prod

uctiv

ity

Quality: Performance Measures

Quality: Patient List

Cost: Population-Based Measurement

Service

Population Health: Patient-Level Detail

STANDARDS BASED SERVICES FOR SHARING DATA, APPLICATIONS, AND ADVANCED DECISION SUPPORT

Decision Support Modules

• Antibiotic Assistant• Ventilator weaning• ARDS protocols • Nosocomial infection

monitoring• MRSA monitoring and

control• Prevention of Deep

Venous Thrombosis• Infectious disease

reporting to public health

• Diabetic care• Pre-op antibiotics• ICU glucose protocols• Ventilator disconnect• Infusion pump errors• Lab alerts• Blood ordering• Order sets• Patient worksheets• Post MI discharge meds

We can’t keep up!• We have ~150 decision support rules or modules

• We have picked the low hanging fruit• There is a need to have 5,000 decision support rules or modules

• There is no path from 150 to get to 5,000 unless we fundamentally change the ecosystem

HSPC Mission Statement

Improve health by creating a vibrant, open marketplace for

healthcare applications

CommercialEHR

CommercialEHR

Heterogeneous Systems

Home GrownSystem

Home GrownSystem

SystemIntegratorSystem

IntegratorVA

SystemVA

System

Applications Being Built• Neonatal Bilirubin Protocol• Pediatric Growth Chart & BP Centiles• Pulmonary Embolism Protocol• Problem Management

What other kinds of Apps are likely to appear?• Decision support

– Complex or evolving logic– Visualization

• Patient -- Provider data sharing– Simultaneous provider’s view & patient view

• Integration of external data into EHR workflow– Population Health – bilateral data flow– “Real time” HIE integration

• National scale services– Genomics (Smarter ordering, PGX, etc.)

• mHealth / mobile apps– Connecting consumer apps to their EHR data!– Counterpart to Apple’s HealthKit?

• Informatics Research– Clear IP rights (vs. source code approaches)– Local or multi-site

Apps that address specific focused problems…

• Provider-facing services– Focused decision support– Visualization– Disease management– Specialty workflows

• National Shared Services– Genomic testing & CDS– Pharmacogenomic screening– CDC Ebola screening?– CDC immunization forecaster– Prior Authorization / Appropriateness

App 1

EHR

App 2 App 3

Like Google Maps…

Apps that enable data sharing…

• Next-gen Interoperability– Population Health integration– HIE integration– Data capture for research– Clinical Trial recruiting

EHR2

App 1

EHR3

EHR1

Like Facebook…

Apps that empower patients / consumers…

• Apps as Prescriptions– Chronic disease

management– Pt-Provider Communication– Remote monitoring– Outcome capture & Clinical

Effectiveness Monitoring

SMART Phone App

Pop HealthEHR

Like ???? …

Questions

Stanley M. Huff, MDChief Medical Informatics OfficerIntermountain HealthcareSalt Lake City, UTstan.huff@imail.org

Appendix

Current Situation• Each EHR vendor uses proprietary models and terminology to

represent clinical data– Some standardization of codes is now occurring, but– Data is not consistent vendor to vendor, or even organization to

organization within the same vendor• This means that:

– Sharing of data is difficult– Sharing of executable software across vendors is impossible– Each useful application is created or re-created on each

different platform– There are unmet needs for health care applications and

decision support– Software costs are higher than they need to be

Characteristics of a new Ecosystem• Consistent and unambiguous data collection• Data stored and accessed through truly semantically interoperable services

• Sharing of data for direct patient care, population based analytics, and research

• Sharing of applications, executable clinical decision support and knowledge

IsoSemantic Models – Example of Problem

e.g. “Suspected Lung Cancer”

(from Dr. Linda Bird)

Data Comes in Different Shapes and Colors

Finding – Suspected Lung Cancer

Finding – Suspected CancerLocation – Lung 

Finding – CancerLocation – LungCertainty – Suspected(Let’s say this is the preferred shape) 

Data Standardized in the Service

Shape and color of data in the local database

Shape and code translation

Application

Data in preferred shape

Applicationand User

Partial Interoperability

TermTranslators

Standard Terms(Non-standard Structure)

Applicationand User

Application

Local databases,CDA, HL7 V.2, etc.

Preferred Strategy – Full Interoperability

Local databases,CDA, HL7 V.2, etc

Term andStructureTranslators

Application

Standard StructureAND Standard Terms

(As defined by CIMI Models)

Applicationand User

Req

uire

men

ts

Reasons to do it on the server side• Person writing the translation is most likely to understand the

meaning of the data in their own database.• The person writing the translation only has to understand their own

data and the preferred model.– They can optimize query execution for their own system

• The query for the data is simpler. If the application has to write a query that will work for all shapes, the query will be inefficient to process by every system.

Different Physical EHR Implementations

Services based onFormal, Logical

Models

Applications, including advanced decision support,

protocols and guidelines using

Standard Services

http://smartplatforms.org/smart-on-fhir/

FHIR – The “Public API” for Healthcare?

FHIR = Fast Health Interoperability Resource– Emerging HL7 Standard (DSTU 2 soon)– More powerful & less complex than HL7 V3

ReSTful API– ReST = Representational State Transfer – basis for Internet Scale– Resource-oriented rather than Remote Procedure Call (nouns > verbs)– Easy for developers to understand and use

FHIR Resources– Well-defined, simple snippets of data that capture core clinical entities– Build on top of existing HL7 data types– Resources are the “objects” in a network of URI reference links

FHIR: Core Resources (99 in DSTU2)

• AdverseReaction• CarePlan• Condition• Device• DiagnosticOrder• DiagnosticReportt• Encounter• FamilyHistory• ImagingStudy• Immunization

• MedicationAdministration• MedicationDispense• MedicationPrescription• Observation• Order• Organization• Patient• Practitioner• Procedure• More….

# 64

HL7 FHIR Resources and Profiles

Observation

Lab Obs Patient Obs Family Hx Obs

Qn Lab Obs Titer Lab ObsQual Lab Obs

Hematocrit Serum Glucose Urine Sodium

FHIR Resource

FHIR Profiles

Invariant Profile Structure – CIMI Leaf Node Content

The Risk – competing proprietary FHIR profilesWomen, FHIR, and other Dangerous Things

EHR as Platform: Market Forces

“EHRs are becoming commodity platforms. The winner will be the EHR vendor that provides the best platform for innovation – the most open and most extensible platform.”

--- CEO of a major IDN

• Self determination – ability to meet own needs• Desire for vendor independence• Don’t want to rely on proprietary extensions or process• Need clean separation of IP rights (commercialization)

EHR as Platform: Government Forces

• Use “atomic” data elements, not just documents• Require EHRs (vendor/provider) to expose “open”

APIs

• Dismissive of current (MU1 & MU2) efforts• Design systems for research uses, not just clinical care• Focus on “Apps” not monolithic solutions• Give the patient more control over uses of his data

JASON Report: “A Robust Health Data Infrastructure”

JASON Task Force –Recommendations

• A Coordinated National Architecture– Modeled on Internet principles (loose coupling) for scale

• Data Sharing Arrangements (DSA) for governance• All EHRs should deploy a “Public API”

– Implement “Core Data Services & Profiles” – FHIR– Expectation to deploy the API – Permit non-discriminatory access to the API

• API becomes part of CEHRT• Measures and transparency for usage of the APIs

The Healthcare Services Platform Consortium (HSPC)

Sample of Participants

• HL7 FHIR – Grahame Grieve• SMART – Josh Mandel• Cerner – David McCallie, Marc

Overhage• Epic – Janet Campbell• VA – Jonathan Nebeker, Paul

Nichol• openEHR – Thomas Beale • Open Health Tools – David

Carlson• Harris – Vishal Agrawal• Intermountain Healthcare• Systems Made Simple – Viet

Nguyen• LSU – Frank Opelka, Wayne

Wilbright, John Couk

• Center for Medical Interoperability – Todd Cooper

• RelayHealth – Arien Malec• NLM – Clem McDonald• Infocare Healthcare – Herb

White• Mayo Clinic – Cris Ross• Clinical Architecture – Shaun

Shakib• Cognitive Medical Systems –

Doug Burke• IBM – Jeff Rogers, Dennis Leahy• ASU – Aziz Boxwalla, Robert

Greenes• Regenstrief Institute – Douglas

Martin

Essential Functions of the Consortium

• Select the standards for interoperable services– Standards for models, terminology, security, authorization, context sharing,

transport protocols, etc.– Modeling: SNOMED, LOINC, RxNorm – FHIR Profiles – do it together– Publish the models, and development instructions openly, licensed free-for-use

• Provide testing, conformance evaluation, and certification of software– Gold Standard Reference Architecture and its Implementation– We will work with an established company to provide this service– Fees that off set the cost of certification will be charged to those who

certify their software

• Implementation of the standard services by vendors against their database and infrastructure

– Everyone does not have to do every service– There must be a core set of services that enable a marketplace

HSPC wiki• https://healthservices.atlassian.net/wiki/display/HSPC/Healthcare+Se

rvices+Platform+Consortium

72

73

Open Is Happening

Boston Childrens: SMART Growth Chart

SMART Growth Chart – Parent’s View

Intermountain: SMART Neonatal Bilirubin Alerts

77

Commercial: VisualDX

78

Commercial: VisualDx