Developing a Learning Health System

HIMSS Session ID: 226 Larry Sitka, MS, BS

James Forrester, MS

Agenda • Introductions • The Learning Health System Vision

– Larry Sitka, Founder of the Lexmark Acuo VNA • Putting the Learning Health System into Action

– Jim Forrester, University of Rochester Medical Center

• Closing Comments • Q/A

The Learning Health System Vision

A Healthcare System

5

Proactive/Suggestive Health System Reactive Healthcare System

Today the U.S. system has a healthcare focus

The world is teaching vendors we must focus on a health system

The entire system is procedural-based

Procedural-based reimbursement model

Vendors build products based on procedures and not patient outcomes

Healthcare means the patient is already sick. Are we too late in the process?

Reimbursements based on patient outcomes

Departments go from revenue generating to capital costs

Roadblocks to entry are being eroded across localities

We still have vendor lock and vendor block

A Health System vs

Proactive/Suggestive Health System is a Learning Health System

A Proactive/Suggestive Healthcare Content Management Platform for a Learning Healthcare System must exist

Patient’s Outcome is the Center of the Universe, Not the Product in a Learning Health System

6

vendor neutral archive enterprise content management

EMR

Cloud NAS SAN CAS

What it is… • An ONC proposed ecosystem

comprised of providers and controlled by the patient.

• Creates new knowledge. • Is supported by a wide variety of

systems. • Lexmark surrounds a chaotic HIT

environment.

Characterized by… Continuous learning cycles

A broad array of stakeholders

Extends beyond the enterprise to external care contributors

Supportive LIVE Infrastructure

The Learning Healthcare System maps well to Mature VNAs

• Can install under/over traditional PACS

• Interoperable now and into the future; Canonical data models

• Reduces complexity, eliminating vendor lock and vendor block

• Consolidates storage and improves its management

• Modular scalability; Adopts new technology and virtualization; Additional benefits emerge

• Savings helps offset investments

• Interoperability through certified standards

Learning Healthcare System

• Build upon the existing HIT system

• One size does not fit all

• Empower individuals

• Leverage the market

• Simplify

• Maintain modularity

• Consider the current environment and support multiple levels of advancement

• Focus on value

• Scalability and universal access

Mature VNAs

A Learning Health System and the Mature VNA

The foundation of a Learning Healthcare System is a Mature VNA

2015 2016 2017 2018 2019 2020 2021 2022 2023 2024

3 Year Agenda | 2015-2017

Send, receive, find and use a common clinical data set to improve health and healthcare quality

6 Year Agenda | 2018-2020

Expand interoperable health IT and users to improve health and lower cost

10 Year Agenda | 2021-2024 Achieve a nationwide learning health system

Figure 10. Connecting Health and Care for the Nation: A Shared Nationwide Interoperability Roadmap – Version 1.0

Ability to acquire, send, find, receive. Canonicalization of content metadata Provides a means of perceived

centralization of content (both storage and application) Data ownership by the healthcare

organization. Ownership of the content on behalf of the patient

Canonical data mapping not just inbound but outbound. Interoperability through

standards Interoperability using

secure and authorized Open Image Exchange for sharing Leverage the existing HIE and

private networks All built around this new patient-

centered health system PIX Manger Services

Today’s Mature VNA

Industry Driver #1: Unprecedented Demand for Information

10

Patient Referring Physician

Surgeon

Oncologist

HIM

Radiologist

Physician

Terabytes

Petabytes

Exabytes

Zettabytes

Analytics

Required by a Learning Health System

The next 1000+ concurrent users ALWAYS ONLINE

Define outcomes Radiology/Cardiology Imaging

Supporting Content

Digital Pathology

Genomics

Scanned Documents

B I G D A T A

Viewing Component – Diagnostic

– Clinical

– Specialized

• Multiple sub specialties

Industry Driver #2: PACS Redefined

The Separation of the PACS Components (IHE Actors) into Enterprise Imaging

Workflow Component Modality worklist

Physicians worklist

Exam state

Dictation

Credentialing

Physician load balancing

Archiving (Content Manager) Secure storage, distribution,

routing, canonical data models

Migrations, PID resolution, SOA access

Data security, data preservation, data separation

All content location services (DICOM and other content)

Multiple if any DR plan? PHI Exposure? $$$?

Today’s Environment

Silos of vendor locked and blocked information

Cardiology/Radiology PACS

Other ‘ologies

Individual Departmental Storage Silos

ECM PDF, JPG, TIFF, TXT, .RAW

Custom Interfaces

Proprietary File Access

Vendor Lock? Vendor Block?

EMR

Healthcare Content Management System

EMR

Radiology PACS

Interfaces Custom

Cardiology PACS ECM

MAIN FACILITY SCSI SAN

Radiology PACS

Cardiology PACS

Mature VNA Storage Virtualization

Mature VNA Enterprise Image Management

Proprietary File Access CIFS, NFS, API

SATA

DISASTER RECOVERY SITE TAPE OTHER CAS/COS NAS CLOUD

HTTP/REST iSCSI

EMR

RIS/HIS/EMPI Image Enabled UniViewer

XDS Web Services

XDS.b Reg/Rep (XDS)

ECM CONTENT (JPG, TIFF, PDF,

.RAW, ETC.)

SOA Service Bus Architecture Vendor Neutral Archive

Enterprise Content Management

DICOM Assisted Migration Utility

ECM Verification Extraction Utility

Store – “Storage Virtualization” – Data Lifecycle Management

Workflow Services – Web Services Federation and Morphing

“DICOM Virtualization” – Clinical Information Lifecycle Management

VNAMed DICOM

VNA Semantix HL7

VNA IHE Audit Supplement-95

VNA HA Business Continuity

XDS/XDS-I DICOM &

non-DICOM Support IHE ITI

Dat

abas

e

Inte

lligen

ce

Laye

r

Use

r Bas

ed

ACL

MSA

D

Archive Integration (api) Managed Shares

Open Image Exchange PIX Manager (IHE)

Secure Access Protecting PHI

DICOM World Other Content World

ECM CAPTURE

Perceptive Search

DICOM / WADO / QUIDO / STOW RESTful Webservices MINT Dynamic Encrypted URL via HL7 HL7 DICOM DICOM

What’s Next for Interoperability in Healthcare

“21st Century Cures Act” • Data Locking and Data blocking • What are the consequences ?

– Lost MU certifications for the Vendor – Provider will have reduced reimbursements

and ultimately complete loss.

14

SMART on FHIR as the platform to link apps to the EMR “SMART HealthIT is the interface between healthcare data and innovation. The goal of SMART is audacious and can be expressed concisely: an innovative app developer can write an app once, and expect that it will run anywhere in the health care system. Further, that one app should be readily substitutable for another. When apps are substitutable, they compete with each other which drives up quality and down price. SMART and the “App Store for Health” model….”

Open Content Exchange Leverage Healthcare Content Management Apps to enable health information exchange between facilities

Content Upload

Service Content Download

Service

HCO A (Un)Affiliated HCO B

VNA + ECM

Content Upload Service

Content Download Service

HCO A (Un)Affiliated HCO B

Enterprise Viewing +

Image Sharing

Content Upload

Service Content Download

Service

HCO A (Un)Affiliated HCO B

Multifunction products (MFP)

Media Writer PACS Scan

Sending PACS Receiving PACS

Enterprise Image Connectivity

Healthcare DURSA Governance Agreement (Data Use Reciprocal Support Agreement)

Secure Open Image/Content Exchange

The Learning Health System

Implementation at the University of Rochester Medical Center

Monthly Image Volume • Provides a metric that is understandable by C-Suite execs • Useful for monitoring and trending –

– Used to trend average images per study as compared to patient outcomes. As average image count per study increases, does diagnostic accuracy increase? Does turn around time of diagnostic read improve?

Monthly Image Volume

0

200,000,000

400,000,000

600,000,000

800,000,000

1,000,000,000

1,200,000,000

1,400,000,000

1,600,000,000

Jan-14 Apr-14 Jul-14 Oct-14 Jan-15 Apr-15 Jul-15 Oct-15 Jan-16

Total Images in Archive (VNA)

Image Access from EMR • Provides a metric that is understandable by C-Suite execs • Useful for understanding clinical workflow (referring provider) • Useful for capacity planning

Foreign Image Ingestion • Useful for understanding confirmation of “gatekeeper”

functionality • Useful for understanding trends and resulting implications

• Such as consult load on radiologist for foreign studies

Foreign Image Ingestion

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000

Jan-

14

Feb-

14

Mar

-14

Apr

-14

May

-14

Jun-

14

Jul-1

4

Aug

-14

Sep

-14

Oct

-14

Nov

-14

Dec

-14

Jan-

15

Feb-

15

Mar

-15

Apr

-15

May

-15

Jun-

15

Jul-1

5

Aug

-15

Sep

-15

Foreign Film Studies

Received

Push to PACS

Linear (Received )

Linear (Push to PACS)

Foreign Image Ingestion

FY11 FY12 FY13 FY14 FY15Foreigns 25307 32577 39250 49323 57631Rad Dx 518728 526028 525051 548342 581220

0

100,000

200,000

300,000

400,000

500,000

600,000

700,000

Enterprise Imaging Studies

VNA Image Study “Prefetch” • Useful for understanding customer\end user experience • Useful for finding anomalous usage patterns

VNA Image Study “Prefetch”

0

100,000

200,000

300,000

400,000

500,000

600,000

Images “Retrieved” from Archive (VNA) to Viewer Cache

Ad Hoc RetrievalPreFetch OrthoPreFetch RadLinear (Ad Hoc Retrieval)Linear (PreFetch Ortho)Linear (PreFetch Rad)

VNA Image Study “Prefetch”

0

20,000

40,000

60,000

80,000

100,000

120,000

140,000

160,000

180,000

200,000

5/2…

5/2…

5/2…

5/2…

5/3…

5/3…

6/1/

…6/

2/…

6/3/

…6/

4/…

6/5/

…6/

6/…

6/7/

…6/

8/…

6/9/

…6/

1…6/

1…6/

1…6/

1…6/

1…6/

1…6/

1…6/

1…6/

1…6/

1…6/

2…6/

2…6/

2…6/

2…6/

2…

Images Retrieved from Archive (VNA) to Viewer Cache

Ad Hoc Retrieval

Linear (Ad Hoc Retrieval)

Create Historic Imaging Order Set Create historic order load from VNA image study archive meta data • Imaging studies migrated in to UR Medicine VNA from foreign or

internal archive • Studies are modified for demographic information upon migration to

match UR Medicine Epic demographics • Upon migration a data extract is queried from UR Medicine VNA SQL

database which includes study information such as modality type, exam code, exam date.

• This extract is converted to HL7 and fed in to imaging downstream systems so that they can now readily access the imaging studies from VNA. Can also be fed in to EMR as reportable order.

• This is a powerful tool for imaging study migration and consolidation

Image Study Viewed by Hour of Day • Image study views in viewer

• Grouped by hour of day and day of week • Average over 52 weeks • Diagnostic and enterprise combined data set

• Useful for understanding image study viewing patters • Used to plan scheduled downtimes • Used to understand impact of unscheduled downtimes

Image Study Viewed by Hour of Day

0

500

1000

1500

2000

2500

Sun

0:0

0S

un 4

:00

Sun

8:0

0S

un 1

2:00

Sun

16:

00S

un 2

0:00

Mon

0:0

0M

on 4

:00

Mon

8:0

0M

on 1

2:00

Mon

16:

00M

on 2

0:00

Tues

0:0

0Tu

es 4

:00

Tues

8:0

0Tu

es 1

2:00

Tues

16:

00Tu

es 2

0:00

Wed

s 0:

00W

eds

4:00

Wed

s 8:

00W

eds

12:0

0W

eds

16:0

0W

eds

20:0

0Th

urs

0:00

Thur

s 4:

00Th

urs

8:00

Thur

s 12

:00

Thur

s 16

:00

Thur

s 20

:00

Fri 0

:00

Fri 4

:00

Fri 8

:00

Fri 1

2:00

Fri 1

6:00

Fri 2

0:00

Sat

0:0

0S

at 4

:00

Sat

8:0

0S

at 1

2:00

Sat

16:

00S

at 2

0:00

Image Study Views By Hour

Study Completed • Imaging studies tech complete timestamp

• Grouped by hour of day and day of week • Average over 52 weeks

• Useful for understanding imaging study acquisition patterns

Study Completed

0.0

50.0

100.0

150.0

200.0

250.0

Sun

0:0

0S

un 4

:00

Sun

8:0

0S

un 1

2:00

Sun

16:

00S

un 2

0:00

Mon

0:0

0M

on 4

:00

Mon

8:0

0M

on 1

2:00

Mon

16:

00M

on 2

0:00

Tues

0:0

0Tu

es 4

:00

Tues

8:0

0Tu

es 1

2:00

Tues

16:

00Tu

es 2

0:00

Wed

s 0:

00W

eds

4:00

Wed

s 8:

00W

eds

12:0

0W

eds

16:0

0W

eds

20:0

0Th

urs

0:00

Thur

s 4:

00Th

urs

8:00

Thur

s 12

:00

Thur

s 16

:00

Thur

s 20

:00

Fri 0

:00

Fri 4

:00

Fri 8

:00

Fri 1

2:00

Fri 1

6:00

Fri 2

0:00

Sat

0:0

0S

at 4

:00

Sat

8:0

0S

at 1

2:00

Sat

16:

00S

at 2

0:00

Completed Studies By Hour

ED CT Data • Similar to prior dataset but specific to trauma center ED CT

acquisition • Useful for understanding implications of planned and unplanned

downtimes for ED trauma center • Demonstrates\confirms known ED encounter and usage

patterns • Useful to plan staffing for 3D\advanced visualization processing lab

ED CT Data

0

50

100

150

200

250

300

350

400

Sun

0:0

0S

un 4

:00

Sun

8:0

0S

un 1

2:00

Sun

16:

00S

un 2

0:00

Mon

0:0

0M

on 4

:00

Mon

8:0

0M

on 1

2:00

Mon

16:

00M

on 2

0:00

Tues

0:0

0Tu

es 4

:00

Tues

8:0

0Tu

es 1

2:00

Tues

16:

00Tu

es 2

0:00

Wed

s 0:

00W

eds

4:00

Wed

s 8:

00W

eds

12:0

0W

eds

16:0

0W

eds

20:0

0Th

urs

0:00

Thur

s 4:

00Th

urs

8:00

Thur

s 12

:00

Thur

s 16

:00

Thur

s 20

:00

Frid

0:0

0Fr

id 4

:00

Frid

8:0

0Fr

id 1

2:00

Frid

16:

00Fr

id 2

0:00

Sat

0:0

0S

at 4

:00

Sat

8:0

0S

at 1

2:00

Sat

16:

00S

at 2

0:00

SMH ED CT Studies 2014 by Begin Time

ED CT Data

0

500

1000

1500

2000

2500

7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 0 1 2 3 4 5 6

Total EDCT Volume 2014 Grouped by Hour of Day

Questions

Larry Sitka, MS, BS, Principle Solutions Architect, Founder Acuo VNA Lexmark Healthcare James Forrester, MS, Director of Product Management and Imaging Informatics University of Rochester Medical Center