Presentation

Clinical Informatics Clinical Informatics

W. Ed Hammond. Ph.D., FACMI, FAIMBEDirector, Duke Center for Health Informatics

Professor , Department of Community and Family MedicineProfessor, Department of Biomedical Engineering

Adjunct Professor, Fuqua School of BusinessDuke University

Vice-chair HL7

Nothing to Disclose

Enabling Model

2

Output

Application of HIT

3

Feedback Loop

OutcomeUnderstand the problem to be

solved.

Understand causes, factors,

issues

Understand measurements

and data required to

monitor & solve

Implement proposed solution.

Evaluate outcome

What problem does the EHR solve?

• Documents patient care

• Provides data for decision making

• Provides for continuity of care

• Provides data for billing

• Can be mined for new knowledge

• Can be used to measure performance

• Contributes to patient safety

• Source for reporting, audits4

HIT – Stimulus target

• Inclusion of $36.5 billion to create nationwide network of EHRs

• Can HIT improve health and health care?

• Can HIT save $77.8 b each year through universal use of HIT?

• What is meaningful use of EHR?

• Will the goal be met: will every citizen have an EHR by 2012?

5

Rising CostsRising Costs

• For the past 4 decades, per capita health care spending grew much more rapidly than per capita GDP

• Most experts agree that the key factor driving the long-term growth of health care costs has been the emergence, adoption, and widespread diffusion of new medical technologies and services.

• Technological advances are likely to yield new, desired medical services in the future, fueling further spending growth and imposing difficult choices between health care and other priorities.

6

CBO Testimony, 2008

7

Australia $2,876 18.2 88 34 80 8.8 0.4Canada $3,165 17.7 92 26 82 12 0.5France $3,159 18.4 75 n/a 79.7 8 0.5Germany $3,005 17.5 106 28 78 11.9 0.6Japan $2,249 19.6 81 n/a 79 10.3 0.2New Zealand $2,083 17.1 109 38 79 10.9 n/aUK $2,546 16.9 130 13 80 11 0.5US $6,102 16.6 115 51 88.9 14.8 0.7

Exp

endi

ture

s pe

r ca

pita

Life

exp

ecta

ncy

at a

ge 6

0

Pre

vent

able

dea

ths

per

100,

000

pop.

Acc

ess

Pro

blem

s

5 Y

r. s

urvi

val

brea

st c

ance

r

MI

30 d

ay

hosp

ital m

orta

lity

Dea

ths

from

sur

gica

l or

med

ical

mis

haps

pe

r 10

0,00

0 po

p

Source: BMJ, 2007

Healthcare Indicators

Reducing Disparities in Cost

• Why are costs for health care so different around the world when outcomes are particularly different? Some countries have better outcomes at half the cost. – Costs are also related to administrative and fiscal

structure of healthcare system.– New technologies frequently increase the cost without

significant improvement in outcomes

• What role does appropriate and effective care play? Is more better?

• What should be evaluation criteria for the introduction of new technologies?

8

Technology

• Technology has been one of the major factors that has increased human life expectancy about 3 months for every 1 year for the past century.

• Does technology save money?– Technology costs money!– Longevity increases cost of healthcare. – We can now treat diseases that were previously

untreatable.

• Imaging increases cost – is the resulting value worth it?

9

Value

• Evidence has not shown that increasing costs for health care has resulted in commensurate gains in health

• Further, excessive spending may result in steps not being taken that could prevent the onset of disease – even when clear evidence exists about the benefits of such steps

• What is the value evidence for the widespread use of CIS in healthcare?

10

Value of HIT

• Efficiency in data capture

• Continuity of care – aggregation of data from all sites of care

• Increased consistency of care – for an individual and across populations

• Increase in quality of care by timely access to data and knowledge by both provider and consumer

• Reduced cost through elimination of redundant testing and unnecessary testing

• Reduced cost through elimination of medical errors

11

Value of HIT

• Improved accuracy of diagnoses and precision of therapeutic interventions

• Through use of geocoding, better understanding of environmental and social factors impact on cause and course of disease

• Identification of all factors involved in impacting disease and quality of life

• Provide national statistics on prevalence of disease• Better health through delivery of higher quality care• Higher quality of life and longer life through instilling

healthy behavior

12

Value of HIT

• Data and outcomes available for the understanding of the effects of treatment and for the extraction of knowledge

• Through measurement, a better understanding of cause and effect

• Creating models that will better predict the cost of health care

• More rapid identification of candidates for clinical trials

• Quicker determination of global adverse drug events

• Quicker awareness of disease outbreaks

13

Value of HIT

• Data is reusable; provides value for multiple secondary uses

• Analysis of patient conditions, treatments, outcomes, demographics and environment produce new knowledge that is automatically fed back into the care process.

• Models of care are produced that permit projections for improved outcomes, reduced costs, higher quality.

14

Why is HIT a high priority?

• Patient safety– Increased concerns related to safety of prescription drugs– New emphasis on eliminating Healthcare Affiliated Infections

• Efficient and effective health care delivery a must

• Health surveillance, bio-defense and natural disaster; increased mandatory reporting requirements

• Demand for quality; pay related to performance and outcomes

• Cost containment in face of increasing costs of healthcare

• Accommodating an aging and mobile population

• Effective management of chronic disease

15

Why is HIT a high priority?

• Equal access to care• Consumer sophistication and knowledge in health;

mobility• Increasing importance of multiple uses of data –

translational medicine• Changes in doctor’s information gathering skills• Increase in options for testing and treatment• Limited connectivity among providers with multiple

providers involved in care• The Healthcare Gamble – who calls the play?

16

Why is HIT a high priority?

• Practice of medicine that is predictive, personalized and pre-emptive

• Resources are becoming limited– Decreasing number of providers– Smaller hospitals disappearing– Long waits for appointments– Few walk-in appointments available

• Changing models for healthcare– Consumer driven health care– Health savings accounts– Shopping mall clinics, Doc in Box clinics– Wal-Mart, Google and Microsoft movement into healthcare

17

Why is HIT a high priority?

• Volume of data about a patient has increased tremendously over the past decades

– Increasing number of diagnostic tests– Increasing numbers and modality of images– Genetic testing– Access to data at place and time of decision making is

critical– Informed decision requires data– Data must be used for multiple purposes– From bytes to kilobytes to megabytes to gigabytes to

terabytes to petabytes to …

18

Why is HIT a high priority?

• Sources and amount of knowledge have increased exponentially over the past decades

– Amount of new knowledge introduced each year would take more than 200 years to assimilate into one’s practice reading and understanding two papers each night

– Undergraduate and graduate education is based on out of date concepts

– Continuing medical education is inadequate

• We can’t learn fast enough to be effective

• New knowledge requires new skills and new understanding

19

Focus

• Improving quality, safety, efficiency, and reducing health disparities

• Engaging patients and their families

• Improving care coordination

• Ensuring adequate privacy and security precautions for personal health information

• Improving population and public health

20

Meaningful Use

Preamble: “We recommend that the ultimate goal of meaningful use of an Electronic Health Record is to enable significant and measurable improvements in population health through a transformed health care delivery system. The ultimate vision is one in which all patients are fully engaged in their healthcare, providers have real-time access to all medical information and tools to help ensure the quality and safety of the care provided while also affording improved access and elimination of health care disparities. “

A “meaningful use” vision

• A ubiquitous infrastructure that permits the creation of an Electronic Health Record in which all relevant data about an individual is aggregated across regional, state and national boundaries.

• This EHR serves all sites, views, presentations and purposes relating to health and health care.

• With a single data entry, the EHR meets all data reporting requirements as well as health care. Its every thing for every body.

• The EHR becomes an active partner, not just a passive data repository.

22

A “meaningful use” vision

• Comprehensive data for patient care• Integrates data with knowledge for cognitive

support of both providers and patients• Accommodate heterogeneity of many sites of

care• Empower persons to become involved in the

healthcare• Provide operational value through aggressive

interaction with patient and provider • The vision depends on understanding what

problems you are trying to solve at the moment and at that location.

23

Meaningful Use• Primary objectives – stage 1

– Capture health information in a coded form– Use data to track key clinical conditions and

communicate for care purposes– Implement clinical decision support tools– Report clinical quality measures and public health

information to relevant government officials.

• Primary objectives – stage 2– Heighten quality management requirements – Move to the most structured format for information

exchange

24

Meaningful use for providers (1/4)

1. Use CPOE

2. Implement drug-drug, drug-allergy, drug-formulary checks

3. Problem list of current and active diagnoses [ICD9-CM or SNOMED]

4. Generate and transmit prescriptions electronically

5. Maintain active medication list

6. Maintain active allergy list

25

Meaningful use for providers (2/4)

7. Record demographics

8. Record and chart changes in vital signs

9. Record smoking status [>13 yo]

10. Incorporate clinical lab test results as structured data

11. Generate lists of patients by specific conditions for use in quality improvement, reduction of disparities, research, outreach

12. Report ambulatory quality measures to CMS

26

Meaningful use for providers (3/4)

13. Send patient reminders for preventive follow-up

14. Implement 5 clinical decision support rules

15. Check insurance eligibility electronically

16. Submit claims electronically

17. Provide patients electronic copy of health information

18. Provide patients with electronic access to health information

27

Meaningful use for providers (4/4)

19 . Provide patients clinical summaries for each office visit

20. Exchange key clinical information

21. Perform medication reconciliation at relevant encounters

22 Provide summary care record for each encounter and referral

23 Submit immunization data to immunization registries

24 Provide electronic syndromic data to public health agencies

25 Have certified EHRS

28

29

StakeholderInteroperability

Security/Privacy

Interoperability

FunctionalInteroperability

SemanticInteroperability

BusinessInteroperability

Technical & Syntactical

Interoperability

eHealtheHealthSystemicSystemic

InteroperabilityInteroperability CommunicationsInteroperability

Legal, ethical and societal

Interoperability

Human/computer Interface

Interoperability

All impact patient safety and quality.

International Interoperability

Clinical Information System

• Computer-based system that is designed for collecting, storing, manipulating, and making available clinical information for healthcare delivery process.

• May be limited to a single area (laboratory system, pharmacy system, imaging system, etc.) or they may be widespread and include virtually all aspects of clinical records (e.g. electronic medical records)

• Provide a clinical data repository that stores clinical data such as patient’s history of illness and interactions with care providers.

• Includes service functions such as Hospital Information Systems (HIS), functional systems (ADT, scheduling), departmental systems (LIS, RIS, PIS), Computerized Physician Order Entry systems (CPOE), ePrescribing systems

30

CIS is about …

• Data and data management– The right data and only the right data– Complete, aggregated, timely, trustworthy, unambiguous, reusable,

logically accessible– Event driven displays, logically driven

• Knowledge and knowledge management– Evidence-based, up-to-date, appropriate, integrated into work flow,

human and computer useable

• Processes and work flow– Effectively and efficiently combines data with knowledge to enable

optimum human decision-making– Monitor decisions and outcomes and provide safety checks,

feedback and recommendations– Integrate data collection, presentation and decision support

transparently into care delivery process

31

32

Building the pieces

EHRSCPOEe-Prescribing

EHRS

CPOE

Functional Components for CIS

PlanningPhase

Data Transfer StorePresen-tation

Collection

DecisionSupport

EHR Functions

Data Exchange

Other Systems

ReportsExternalSystems

QueryWhat is required to provide the interoperable connectivity?

A suite of standards!

33

Planning

• Defining and understanding problem to be solved

• Defining and understanding requirements

• Understanding data flow

• Understanding work flow

• Understanding linkages

• Understanding data requirements

34

What are the tools for planning?

• Start with defining use cases, story boards or scenarios in order to understand actors, interactions, activity diagrams, required data elements, data flow, trigger events, work flow, and decision support required.– Use cases are created by many groups

including HITSP, HL7, IHE, CDISC, caBIG, VA, DOD, FDA, CDC, …

• Package as Domain Analysis Model

35

Establish a common base

• Need a common base so different groups can work independently yet still maintain interoperability

• Start with a common Reference Information Model on which all data items, entities, acts, roles and relationships are defined. [ISO/HL7 RIM is a global standard.] [CEN 13606]

• Data models [BRIDG], [CDISC], [caBIG]

36

Data elements

• Fundamental component for data interchange• Key attributes include:

– Unique persistent identifier code – ISO OIDs, UMLS, other ?– Precise definition validated by domain experts– Single terminology assigned to data element derived from

controlled vocabulary– Data types (HL7/CEN/ISO)– Standard units (ISO/HL7)– Classifications– Defined value set – Synonyms– Other attributes

• Centrally maintained; globally distributed; free to users; dynamically maintained; appropriate tools

• ISO 11179 underlying standard

37

The Terminology Dilemma

• SNOMED CT (clinical terminology)

• LOINC (laboratory tests)

• RxNorm (orderable clinical drug codes and formulations, mapped in UMLS)

• Structured Product Labeling – dosing, potential interactions, etc.

• VA NDF-RT (therapeutic classification, components, mechanism of action, physiologic effect, diseases treated)

• FDA terminology sets (dosage form, packaging, routes, application methods)

• MedDRA – adverse events

• ICD-9 or ICD-10 for reimbursement and clinical

• CPT – procedures

• ICPC – primary care

• MEDCIN – clinical terminology

• IEEE medical devices

• Local terminologies38

Mapping is a workable solution but costs extra, is never synchronized, and loses information.

Approach to Semantic Interoperability

39

Terminology

Master Data Element Set

Value Sets

NCI ThesaurusSNOMED-CT

LOINCRxNormICD9/10Others

Attributes including code, name, definition, units, datatype, class, …

Leveraged in interchange standards, data models, ontologies, etc.

Data Elements collected in this

facility

Data Elements collected in this

facility

Healthcare facility A

Healthcare facility B

Business Agreement

Compound data elements

• Attributes similar to data elements

• Examples include blood pressure, heart murmur, titers

• Expressed as – Templates (HL7)– Archetypes (openEHR)– Common Message Element Type (CMET)

• Use XML syntax

• Clinical Example– Blood Pressure

• Systolic• Diastolic• Arm• Position of patient• Cuff size

40

Complex data elements

• Attributes similar to data elements

• Examples include drug sensitivity, microbiology results, body mass index, pulmonary functional tests

• Administrative such as name, address, telephone

• Extended into data groupings– patient admit profile– TB screen– Well-baby workup– Clinical trial component

• Trigger-driven data transport profiles– Data required when a patient is transferred from hospital to nursing

home

• Disease Management Profiles

41

Document standards

• Clinical Statements

• Clinical Document Architecture (CDA)– Radiology reports– Patient summary– Discharge summary– Referrals– Claims attachments– Infectious Disease Reports

• Continuity of Care Document (CCD)

• Structured Documents

42

<ClinicalDocument> ... <structuredBody> <section> <text>...</text> <observation>...</observation> <substanceAdministration> <supply>...</supply> </substanceAdministration> <observation> <externalObservation> ... </externalObservation> </observation> </section> <section> <section>...</section> </section> </structuredBody></ClinicalDocument>

DOCUMENT

BODY

Header

SECTIONS

Narrative Block

ENTRIES

ExternalReferences

Major Components of a CDA Document

<section> <code code="48765-2" codeSystem="2.16.840.1.113883.6.1" codeSystemName="LOINC"/> <title>Allergies and Adverse Reactions</title> <text> <list> <item><content ID="A1">Penicillin - Hives</content></item> <item>Aspirin - Wheezing</item> <item>Codeine - Itching and nausea</item> </list> </text> <entry> <observation classCode="OBS" moodCode="EVN"> <code code="247472004" codeSystem="2.16.840.1.113883.6.96" codeSystemName="SNOMED CT" displayName="Hives"> <originalText><reference value="#A1"/></originalText> </code> <entryRelationship typeCode="MFST"> <observation classCode="OBS" moodCode="EVN"> <code code="91936005" codeSystem="2.16.840.1.113883.6.96" codeSystemName="SNOMED CT" displayName="Allergy to penicillin"/> </observation> </entryRelationship> </observation> </entry></section>

CDA, Release Two

44

<Results> <Result> <CCRDataObjectID> 2.16.840.1.113883.19.1 </CCRDataObjectID> <DateTime> <Type> <Text>Assessment Time</Text> </Type> <ExactDateTime> 200004071430 </ExactDateTime> </DateTime> <Type> <Text>Hematology</Text> </Type> <Description> <Text>CBC WO DIFFERENTIAL</Text> <Code> <Value>43789009</Value> <CodingSystem>SNOMED CT</CodingSystem> </Code> </Description> <Status><Text>Final Results</Text></Status>

<section> <templateId root="2.16.840.1.113883.10.20.1.14"/> <code code="30954-2“ codeSystem="2.16.840.1.113883.6.1" codeSystemName="LOINC"/> <title>Laboratory results</title> <text> CBC (04/07/2000): HGB 13.2; WBC 6.7; PLT 123* </text> <entry> <organizer classCode="BATTERY" moodCode="EVN"> <templateId root="2.16.840.1.113883.10.20.1.32"/> <id root="2.16.840.1.113883.19" extension="1"/> <code code="43789009" codeSystem="2.16.840.1.113883.6.96" codeSystemName="SNOMED CT" displayName="CBC WO DIFFERENTIAL"/> <statusCode code="completed"/> <effectiveTime value="200004071430"/>

Continuity of Care Document (CCD)

• CCD maps the CCR elements into a CDA representation.

45

Continuity of Care Document (CCD)

• CCD sections include:– Payers– Advance Directives– Support– Functional Status– Problems– Family History– Social History– Alerts (e.g. Allergies, Adverse Events)– Medications– Medical Equipment– Immunizations– Vital Signs– Results– Procedures– Encounters– Plan of Care

46

Data Collection

• Machine interfaces

• Personal health devices

• Xforms

• Clinical Guidelines

• Electronic Data Collection (EDC)

• Decision Support Algorithms

• CDA

• Integrated collection for multiple purposes

47

Communications standards

• W3C, OMG, OASIS, IETF, HL7 others

• Include XML,TCP/IP, HTTP, Web services, SOAP, CCOW

• Supported by schemas, XSL, OCL, OWL, GIS

• Web services – Web 2.0, Web 3.0

• WIFI standards

• RFID standards – GS1

48

Data Interchange

• Messages– Content– Trigger events– Routing– Timing and synchronization– Corrections– Business agreements– Security– Authentication

49

Data interchange standards

• HL7 V2 and V3 for data and document transport

• DICOM for imaging

• HL7, NCPDP for prescription and related data

• HL7, X12N for claims data

• IEEE for medical device, sensors

• OASIS for business data

• IHE XDS

• Document standards (CDA, CCR, CCD)

50

51

HL7 v2.x Syntax

MSH|^~\&|REGADT|MCM|IFENG||199112311501||ADT^A04^ADT_A01|000001|P|2.4|||EVN|A04|199901101500|199901101400|01||199901101410PID|||191919^^GENHOS^MR~371-66-9256^^^USSSA^SS|253763|MASSIE^JAMES^A||19560129|M|||171 ZOBERLEIN^^ISHPEMING^MI^49849^""^||(900)485-5344|(900)485-5344||S^^HL70002|C^^HL70006|10199925^^^GENHOS^AN|371-66-9256||NK1|1|MASSIE^ELLEN|SPOUSE^^HL70063|171 ZOBERLEIN^^ISHPEMING^MI^49849^""^|(900)485-5344|(900)545-1234~(900)545-1200|EC1^FIRST EMERGENCY CONTACT^HL70131NK1|2|||123 Industry Way|Shepeming^MI^49849^””^||(900)545-1200|EM^EMPLOYER^HL70131|19940605||PROGRAMMER|||ACME SOFTWARE COMPANYPV1||O|O/R||||0148^ADDISON,JAMES|0148^ADDISON,JAMES||AMB|||||||0148^ADDISON,JAMES|S|1400|A|||||||||||||||||||GENHOS|||||199501101410|PV2||||||||199901101400|||||||||||||||||||||||||199901101400ROL||AD|CP^^HL70443|0148^ADDISON,JAMESOBX||NM|3141-9^BODY WEIGHT^LN||62|kg|||||FOBX||NM|3137-7^HEIGHT^LN||190|cm|||||FDG1|1|19||0815^BIOPSY^ACODE||00|GT1|1||MASSIE^JAMES^""^""^""^""^||171 ZOBERLEIN^^ISHPEMING^MI^49849^""^|(900)485-5344|(900)485-5344||||SE^SELF^HL70063|371-66-925||||MOOSES AUTO CLINIC|171 ZOBERLEIN^^ISHPEMING^MI^49849^""|(900)485-5344|IN1|0|0^HL70072|BC1|BLUE CROSS|171 ZOBERLEIN^^ISHPEMING^M149849^""^||(900)485-5344|90||||||50 OK|

52

Model-based DevelopmentModel-based Development

HL7 FrameworkHL7 Framework HL7 SpecificationHL7 Specification

RIMDatatypesData elementsVocabularyTemplatesClinical Statements

Co

re S

tructu

red

Co

nte

nt

•V3 Messaging

•CDA Specifications

•System Oriented Architecture

Decision support

• Knowledge reference framework and knowledge representation– Arden Syntax, Protégé, PRODIGY, vEMR, GELLO

• Clinical Guidelines– Guideline Interchange Format (GLIF) [HL7]– Guideline Elements Model (GEM) [ASTM]

• Disease Management Protocols

• Evidence-based Care Plans

• Infobutton [HL7]

53

EHR standards

54

• Functional Requirements– HL7 Draft Standard for Trial Use -2004

• Site variation – functional profiles

• Content standards

• Structure/architecture

• Screens/presentation/icons

• Document Registration, IDs, and Naming

Presentation

• Independent of collection

• Depends on event or circumstances, logically driven

• Priority-based and filtered

• Dashboards

• Wireless devices

• Smart phones

55

Identifiers

• Provider identifier– National Provider Identifier [HIPAA]

• Facility Identifier• Employer Identifier

– IRS tax identifier

• Person Identifier– Debate: Unique identifier vs identifying parameters

• Patient registries - Master Patient Index (local, regional, national)

• Record Locator Service

56

Medical and personal health devices

• Interface standards for medical devices– Cable connected– Infrared– Wireless

• Simple to sophisticated devices

• Sensors

• Terminology

• Safety

• Integration of SDOs – IEEE,CEN,DICOM, HL7, ISO

57

Supporting standards

• Implementation manuals (HITSP, HL7)– Standard developers– System designers– Implementers– Users

• Service-oriented architecture (HL7/OMG)• CCOW• Tool sets (HL7)

– Data element/terminology servers– Transitional tools– Message/document creators

58

Other standard requirements

• Repositories– Master data element set– Messages– CDA applications– DAMs– Clinical guidelines– Decision Support Algorithms

59

Application profiles

• Integration profiles (IHE)

• End to end information flow

• Defined domain by domain

• Includes concepts of persistent objects

• Examples– Integrating the Healthcare Environment [IHE]– EHRVA Roadmap– ELINCS– CHI

60

Security/privacy Interoperability

• Security, Privacy and Confidentiality– Authentication– Authorization– Role Based Access– Access logs– Audit Control– Digital signature– PKI– Integrity– Non-repudiation– Encryption– De-identification standards– Probability of risk vs value

• ISO, CEN, HL7, others outside of health care

61

Privacy

• No person should be harmed by the release of health-related data

• No person should be denied proper treatment as a result of privacy concerns

• Unique personal identifiers do not mean privacy is invaded; may be the opposite.

• There is a social responsibility to share data.

62

63

PlanningPhase

DataData

ExchangeDataStore

Reposi-tories

Collec-tion

DecisionSupport

EHR Applica-

tions

Presen-tation

ISO, CEN, HL7, CDISC, DICOM, IEEE, IHTSDO,LOINC, GS1, IHE, NCPDP, X12, openEHR, WHO/ICD, …

ReportsAudits

Story boardsUse cases

Data modelsRIMMDFDAM

BRIDGSOA

Data Elements

Data typesTerminology

UnitsCMETS

TemplatesArchetypesTemplatesClinical

StatementsCDACCDCTS

CDASH

XFORMTemplates

CDA

HL7 v2HL7 v3CEN 13606IEEE

DICOM ISO

EHR FMPHR FMCCOWGeno-micsISO

Arden SyntaxGELLO

GLIFInfo

buttons

Privacy and Security StandardsPrivacy and Security Standards

IdentifiersIdentifiers

ProfilesProfiles

QueriesFilters

Medical Devices

PHD

New Needs

Supply ChainSupply Chain

DemLabMedsProbsAE

CDW

What have we missed?

• We have designed systems that mimic the paper based system; we have not taken advantage of technology; we have not stated or understood the problems we want to solve.

• We have yet to answer the simple question: “What is the purpose of the Electronic Health Record, and how can it most effectively be used?”

• Legacy is overpowering. We are dominated by the past; we have not been bold enough to tempt the marketplace with new vision.

64

65

Creative use of HIT

• Rather than using technology to identify medical errors, use technology to prevent medical errors.

• Real time analysis of data to direct safe and quality care.– Dashboard displays at each level to focus on priority

interventions.– Stop errors before they happen– Order timely and effective testing for disease

• Proactive presentation of data with understanding of next event.

• Not just show but inform.

Requirements

• Comprehensive data on patients’ conditions, treatments and outcomes that will lead to safe, high quality, less expensive, and more efficient health care

• Cognitive support for health care professionals & patients to help integrate– Patient-specific data– Evidence-based practice guidelines & research results

• Accommodation of growing heterogeneity of locales for provision of care

• Empowerment of patients and their families in effective management of health care decisions and their implementations

66

And more …

• Effective and timely disease management with a personalized approach

• Knowledge driven filters for presentation and exchange of data

• Proactive recommendation of actions for both providers and patients

• Determine factors that impact health including social, economic, and environmental situation. Focus on reducing impact of negative factors.

• Influence the way providers practice medicine; it must change.

67

68

What is the Potential?

• Comprehensive data for patient care• Integrate data with knowledge for cognitive

support of both providers and patients• Accommodate the many and heterogeneous sites

of care, understanding how they fit together• Provide universal access to care• Empower personal involvement in healthcare• Provide operational value through aggressive

interaction with patient and provider • The vision depends on understanding what

problems you are trying to solve at that moment and at that location.

• Technologists – more appropriate use of technology; understanding the problems that need to be solved; better coupling with the clinical community

• Clinical community – recognize what technology can do to significantly enhance health care; become the drivers for the use of eHealth; understand value of team approach thst includes the patient

• Patient – Accept responsibility for one’s own health; become engaged in decision-making related to one’s own health; enhanced awareness of personal risk factors; practice prevention

Requires paradigm shifts by stakeholders

69

The Enablement

EHR

Patient CarePersonalized CareCommunity Care

Public Health

EHR – The Centerpiece of HIT

70

Data CreationData Collection

Data InterchangeData Aggregation

Real-time integration of knowledge to direct and control collection of data.

Proactive interpretation of data to direct behavior to enable quality care.

Includes the service functions: HIS, CPOE, CDS, ePrescribing, billing

71

Home HealthHome Health Emergency Emergency CareCare

Intensive Intensive CareCare

Primary Primary CareCare

Nursing Nursing HomeHome(LTC)(LTC)

InpatientInpatient

Electronic HealthElectronic HealthRecordRecord

Electronic HealthElectronic HealthRecordRecord

What is common and what is different among these sites of care? How do we effectively put this package together?

Specialist Specialist CareCare

Centralized vsFederated

72

The Electronic Health Record• Architecture designed for fast and varied retrieval and

presentation; independent of collection modality; anticipates query

• Purpose is to enhance and enable the care of the individual; reusability of data is also a goal

• Content focused on informational value; contains only data contributing to current and future health of person; store only what varies with patient; data warehouse satisfies complete and permanent storage for legal, other purposes

• Structured for unambiguous clarity, understanding and interoperability

• Support common core throughout varied sites of care

• Patient-centric = one person, one EHR

• Rich in functionality; varies with site

• Includes workflow and process management

73

EHR

• Demographics– Why important?– Includes demographics, living environment, occupation,

family history, preferences, quality and nature of life, stress and pain, geocoding

• Clinical Data– Why important and what content?– Includes laboratory tests, diagnostic tests, diagnostic

images, subjective and physical findings, diseases, behavior

74

EHR

• Treatments– Why important?– Understand why choices made; define expected results

• Outcomes– Why important?– Feed back to influence treatment and enhance

knowledge and understanding of disease and related factors

• The EHR must contain everything anyone might need as part of the healthcare process.

75

Structured Data Storage (1)

• Demographics (including living situation, occupation, …)

• Summary Problem List

• Procedures

• Studies (laboratory, radiology, diagnostic tests, …)

• Therapies (prescription drugs, immunizations, blood products, alternative medicines, over-the-counter drugs, …)

• Allergies, adverse events, drug reactions

Structured Data Storage (2)

• Subjective & Physical Findings (history, physical exam)

• Encounter (place, date_time, providers, problems, procedures, studies, therapies, S&P, supplies, disposition)

• Providers involved in care

• Scheduling

• Protocols and disease management, trigger event flags

• Patient education record

76

Patient-centric Electronic Health Record

• Contains all data related to patient’s present and future care from all sites of care using a standard structured architecture with standard data elements.

• Content– Structured Architecture

• Data elements (compound, complex, templates)

• Defined and mapped location for each and every data component

• Organized by category of data

– Organization independent of collection and presentation

77

78

EHR

• Architecture of EHR that can support at variety of uses.– Requires independence of data from

application set• Data must be interoperable; it must be

automatically reusable; it must be capable of integrating with new data to produce new value and understanding.

• Granularity of data must start at lowest levels to permit effective computer analyses and understanding

• Reevaluate patient care and treatment as new data enters incorporating old data

Electronic Health Record

• Cognitive support provides intelligent interaction with content

• Access requires meeting privacy requirements and a master patient registry and record locator.

• Is an active partner with human in awareness, evaluation and decision making

• Supports push, pull, interactive queries, packaged queries, event-based queries

79

80

Download Process

HL7 Message

ID

DoubleEncryption

SiliconEncoder

Sensitive Demographic

Data

Encrypted IDIdentifying Data,

name, address, etc.

Identifying Data, Translated (e.g. Zip). Aggregated

SummaryPopulation

EHR

Summary Data

ID

81

EHR Interoperability Diagram

PatientEncounter

Provider EHR

Database

PersonalEHR

SummaryEHR

Profile

Profile

ProfileResearchDatabase

EnterpriseData

WarehousePersonal

EHR

Profile

Disease Registry

Profile Billing/Claims

Institution EHR

Database

ResearchDatabaseResearch

Database

Disease RegistryDisease

Registry

Derived from master data element registry

Possible Scenario

82

InstitutionalData

RepositoryData

Clinical Data

Warehouse

Service Applications

(CPOE, ePrescribing,

etc.)

Filter

ThePatient-centric Essential EHR

Data Mining

Other systems

Serves Population Linkage

Push, pull or query based

Knowledge Database

Contained in Regional HIE

Mining the Data Warehouse

• Compliance and performance measures

• Patient care/safety

• Revenue

• Research– Provides evidence for EBM– Outcomes monitoring shows effectiveness of

treatment– New knowledge is translated quickly into practice

83

84

Creation of aggregated EHR

Institutional EHR

Institutional EHR

Institutional EHR

The PatientEHR

Personal Health Record

Component of Population Health Record

Why 3 databases?

• Institutional EHR– Permits site variation in data kept in record– Supports institutional business requirements– Used with service applications

• Clinical Data Warehouse– Supports institutional research, quality and performance

measures, compliance, patient safety

• EHR– Is the essential EHR that serves as the patient-centric

EHR and linked creates the population health record

85

86

What is behind this model?• Standards

• Infrastructure to support aggregation of data into a single patient record which requires …

• Infrastructure to support a regional network

• Infrastructure and linkage of regional networks to provide a virtual national network

• A business case based on supported facts and includes a financial model that balances rewards with costs

• A workable process that permits us to reach the destination in doable chunks

• Understanding and creating the necessary linkages among stakeholders

• State efforts blended into a common process that will support interoperability among states

87

These views must serve …

• Real-time connectivity to provide appropriate and controlled access to aggregated patient data.

• Disease registries permit the monitoring and assurance of high quality care.

• Research databases are derived for specific purposes and for specific periods of time.

• Reimbursement is derived from clinical data, ideally in real time.

• Accreditation, credentialing and statistical reporting are derived products.

• The data warehouse contains all data for legal and archiving purposes.

• Support consumer driven healthcare

• Mandatory reporting such as immunizations, HAI, etc.

88

Institutional EHR

• Presents a provider-centric view of the patient’s record

• Record contains patient data as well as intellectual content from provider of care; serves multiple purposes; may contain notes of provider only.

• Patient should have access to data and be able to identify errors and require correction of errors

• Content is driven by the institutional needs and varies from site to site

• Should contain all data important for patient’s care plus other data required by the institution

• Constitutes the billing record

• There is a need to interchange data with other sites at which patient receives care

• EHRS support full functionality for site care and management

89

The Inpatient View

• Deals with acute events and data; has mostly immediate value for decision making and intervention. After intervention occurs, data has less value. (Short persistence)

• Required functionality deals primarily with service activities – ordering, results review, admission and discharge

• CPOE systems particularly valuable to support services

• Real-time decision support valuable

• Inpatient version of ePrescribing, unit dose

• Patient monitoring, medical device component of IT support

• More tolerance for additional time required for IT activities

• Administrative support provides value to physicians – rounding data

• Intensive care even more acute. High payoffs for decision support; very short persistence of data

90

The Inpatient View

• Presentation of data for direct patient care– Ease documentation requirements– Evidence-based clinical pathways/guidelines/protocols– Multiple views of data, usually time-oriented– Automatic creation of discharge summaries

• Task and workflow management– Automatic linkage to task management– Coupled to scheduling for radiology and diagnostic tests, e.g.– Patient location, patient status

• Asynchronous communication among healthcare providers and workers

91

Outpatient view

• Data vary over a longer period of time. Data have a longer persistence.

• Trend analyses are important.

• Outpatient ePrescribing valuable; medication history key.

• Focus on prevention and management of chronic disease.

• Primary responsibility, with patient, for the patient’s overall health, i.e., medical home

• Educational and behavioral modification

• Depends on integration of data from multiple sources

• Provider overhead for IT critical

92

Personal Health Record Permitting the patient to view an institution’s EHR is NOT a PHR

PHR has three components Clinical data that will be similar to the summary health record

plus data that is entered directly or by sensor into the PHR. Clinical data is downloaded from sites of care. Data may be uploaded to site EHRs.

Management of a person’s health including prompts for appointments, medication refills, screening tests, immunizations, etc. Decision support algorithms suggest what provider should be doing in terms of frequency of visit, tests, etc.

Access to knowledge that is tailored to a person’s needs and is driven by clinical data.

May be located at a site of care, at a PHR provider or on the person’s personal computer. Backup issues are important.

Personal Health Record

• EHR data, including clinical data, demographic, images, genomic data, and biomarkers, is downloaded into PHR.

• Decision Support algorithms analyze data to do risk assessment and create a personal health plan. With the entry of each new data, the risk factors and personal health plan are re-evaluated, and patient is advised of changes.

• Manages health-related activities

• Encourages behavioral modifications for better health

93

Personal Health Record

• PHR includes data from personal health devices including sensors and hand-entered data. Examples include:– Exercise – Food intake by coded entry– Pain monitoring– Attitude and mood– Travel– Health journaling and health concerns

94

Personal Health

• Disease Management accomplished as team effort with provider, health workers and patient. Disease management is personalized to the individual. Feedback to patient is important.

• Medical Home concept provides a primary focus and specific responsibility for a patient’s health status and care.

95

96

Risk Assessment

Personal Health Plan

PATIENTPHYSICIAN

Home Enhancement Tools

HEALTHCARE TEAM

Lifestyle

Treatments

Procedures

Other Investigations

Clinical/Images

Biomarkers

Genomics

Demographics

Outcome Tracking

Personalize HealthFeedback loop

Measurements

97

Community or Public Health View

• A summary record (essential EHR) from all sites and sources of care; RHIO EHR

• Linkage of data for new sites of care as well as base for population surveillance, research, quality, analysis

• Data arrives as identified data, available as de-identified• Data source for authorized providers; provides

connectivity• Provides

– Utilization data– Accurate and timely statistics about health and disease in

population– Accurate reporting of events, disease and outcomes– Early discovery of outbreaks, new diseases, bioterrorist attacks– Immunizations, infectious disease tracking– Creation of “on-the-fly: randomized clinical trials– With geocodes, permits understanding statistics of health, spread

of disease

Population Health

• Enhanced understanding of the prevalence of disease by many categories– Using geocoding, by location to very detailed levels– By social and economic categories– By occupation– By race or ethnicity

• Automatic reporting of public health data including immunization, infectious disease, other

98

Population Health

• Health surveillance– Automatic reporting– Search robots looking locally, regionally, and

nationally– Design models for predicting spread of disease

and, more importantly, design processes to control spread of disease

• Evaluate impact of health care strategies including immunization, good health behavioral programs, others

99

Population Health

• Disaster Management– By knowing dependencies, appropriate

interventions can occur before events– Data can be transported to alternative points of

care– Disaster planning becomes easier and more

effective

100

101

Regional Center

• Accommodates 3 to 5 million persons

• Contains summary, aggregated data

• Is a local database

• Available 24/7

• Contains linkages to other centers so patients crossing boundaries of regions can be aggregated

• De-identified data available for local or global queries and analysis

Enabling Model

EHREHR

EHREHR

EHREHR

EHR

EHREHR

EHREHR

RHIO

EHREHR

EHREHR

EHREHR

EHR

EHREHR

EHREHR

RHIO

EHREHR

EHREHR

EHREHR

EHR

EHREHR

EHREHR

RHIO

POPULATION

HEALTHNETWORK

Serves as NHIN

102

103

Regional Linkages

Estimate approximately 100 such centers at implementation cost of $2m each and maintenance of $0.5M.

104

RHIOs/HIEs

Estimate approximately 100 such centers at implementation cost of $3m each and operational cost of $1 M.

RHIO

RHIORHIO

•Each RHIO provides backup for other RHIOs.

•MPI identifies home RHIO

105

Regional Health Information Organizations

• Regional collaboration of multi-stakeholder organizations working together to connect healthcare communities with the goal of improving quality of care, safety and efficiency

• Typical objectives– Develop community-wide health information exchange

– Create healthcare portal with interoperable applications

– Create a training and support infrastructure to ensure adoption of applications

– Engaging payers in programs that align incentives appropriately

106

National Healthcare Information Networks (NHIN)

• To provide a secure, nationwide, interoperable health information infrastructure that will connect providers, consumers, and others involved in supporting health and healthcare.

• E-health information to follow the consumer, be available for clinical decision making, and support appropriate use of healthcare information beyond direct patient care so as to improve health

• De-identified regional data can be analyzed nationally in aggregate. There is a national MPI which permits authenticated and authorized access to RHIOs for legal health-related purposes.

• Security and privacy are top priorities.

Why Global?

• Maximize use of available resources; common effort and share; amplification of productivity

• Enhances understanding of the problems• Share in creation and use of knowledge; clinical

trials should be ubiquitous• Funding for research should be global and shared;

cost should not limit availability• Mobility of disease; disease knows no borders• Mobility of people• Preserve culture• It just makes sense – the world is one!

107

Seamless Care Environment

• What is it?

• What is its value?

• How do we get it?

• When might we get it?

• The art of the possible

• Redefine possible

108

109

Interoperability Gap

Siloed Efforts

Federal AgenciesSome progress

The final word

• The ultimate criteria for success is not the number of patient records, is not the response time, and is not any technical characteristic.

• The real criteria should be is it affordable, is it accessible, is it convenient, and is my life better. I want a high quality of life and then, I’d like a long, comfortable life.

110