Developing a Learning Health System - HIMSS20 · 2017-07-20 · Canonical data mapping not just...
Transcript of Developing a Learning Health System - HIMSS20 · 2017-07-20 · Canonical data mapping not just...
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