Transcript Presentation - Association for Pathology Informatics

APIII 2007
Advancing Practice, Instruction, and Innovation through Informatics
The Challenges of Healthcare Enterprise
Image Data Management
Image Management
Session D1
John S. Koller
Restricted Use – Not for Duplication
©2007 Koller Associates, Inc.
Introduction & Disclosures
John S. Koller - President of KAI Consulting
• 25+ years experience in secure, highly available IT solutions.
• Employment & Consulting relationships
–
–
–
–
Philips Medical Systems
EMC Corporation
Multiple Major Healthcare Imaging & Informatics Vendors
Multiple Major IT Technology Vendors
• Other Relationships
– Contributing Editor
• ITN & AuntMinnie.com
• SCAR – Archiving Primers & Security Primer
– Board of Directors
• HealthSphere Corporation
– Advisory Board Member
• Bycast & ITN
©2007 Koller Associates, Inc.
2
Traditional Healthcare Infrastructure
Hospital Data
Physician Practice Data
Ambulatory Data
Network
Server/OS
E-Medical
Records
Database
Application
Pharmacy, Laboratory
and Ancillary Data
Imaging
Center
Financial
Records
PACS
Financials
HSM to
Tape
ADSM
to
3490s
Operations
Employee Data
Batch
CICS
Billing
e-commerce
OLTP
Storage
Backup
Patient Care
Billing
Backup by
FDR Upstream
STK
Silos
Backup by
tar—8mm
Exchange
Lotus
Notes
HR
DK tape
Cheyenne to
4mm
OmniBack
Test
Disaster
Recovery
Disaster Recovery plan
(mainframe only)
Proactive
Reactive
©2007 Koller Associates, Inc.
Cyclic Process
3
The Digital Healthcare Enterprise (DHE)
• Business Centric
–
–
–
–
• Departmental Systems
Billing
Accounts Payable
HR, Payroll
Supply-Chain Management
• Patient Centric
– Hospital Information System
(HIS)
– Electronic Medical Record (EMR,
EPR, EHR)
– Computerized Physician/Provider
Order Entry (CPOE)
– Networked Intelligent Devices
– Asset Tracking & Management
– Radiology
• RIS (data)
• PACS (images)
– Cardiology
• CVIS (data & waveforms)
• PACS (images)
– Radiation Oncology
• Treatment planning systems
(images)
– Pathology
– Pharmacy
– Laboratory
– Etc.
The Typical Healthcare Enterprise has an average of 100 to 175 Applications.
©2007 Koller Associates, Inc.
4
Clinical Systems Support Pressures
• New Accountability for IT
– Regulatory & Compliance
– Security (CIA)
• Confidentiality
• Integrity
• Availability
• Usability of the Data
• Explosive Growth in Connected Devices
• Lifecycle Data Management Costs
©2007 Koller Associates, Inc.
5
The Side-Effects of
Digital Data in the Healthcare Enterprise
©2007 Koller Associates, Inc.
6
Importance of Standards
• DICOM (Digital Image and
• HL7 (Health Level 7)
– Simple protocol to exchange
information
• Current version of HL7 is Version
2.x
• Version 3.x compliant with XML –
Extensible Markup Language
• CCOW (Clinical Context
Working Group)
– Simplify and improve workflow
and functionality between
applications
Communication in Medicine)
– A true standard dealing with
objects – images > series >
study>patient
– Deals with images, waveforms
and structured reports
• IHE (Integrating the Healthcare
Enterprise)
– Focuses on the interoperability
within & between departments
in the enterprise.
©2007 Koller Associates, Inc.
7
Challenges of Image Management
• Very Large Datasets
– Storage
– Communications
• Very Large Annual Volumes
– Data Protection Requirements
• Variable Retention Periods
– Lifecycle Management
• Variable Locations
– Communication Delays
©2007 Koller Associates, Inc.
8
Growth in Departmental Imaging
• Radiology
– Tens of TBs
• Cardiology
– 2-3X of Radiology
• Pathology
Path
&
Other
– Tens / Hundreds of
TBs to PBs
• Usually no way to age
studies out of storage
Cardio
Rad
©2007 Koller Associates, Inc.
9
Evolving Imaging Architectures
• Film & Processors
• Proprietary Interfaces
• Dedicated Print
Networks
• Closed Mini-PACS
• Enterprise PACS
xIS
©2007 Koller Associates, Inc.
PACS
10
Pathology Imaging Models
Image Acquisition & Workflow Management
Image Acquisition & Display Chain
Analog-Digital
Conversion in
Scanner
(Camera)
Physical
Slide
Digital-Analog
Conversion in
Display (LCD)
Digital Image
Storage &
Management
Diagnostic
Workstation
Image Server
Slide
Scanning
Whole Slide
Imaging Scanner
Digital
Image
Digital
Image
(WSI)
Finished
Slide
WSI Analog to Digital Acquisition
1.
2.
3.
4.
System Calibration
Robotic slide positioner
Special light & filters (Fluorescence)
Camera for Digital Image acquisition
Storage
Archive
Images & Reports
Automated
Preparation
System
Laboratory Information System (LIS)
©2007 Koller Associates, Inc.
11
Lessons & Landmines from Radiology
• Modality <-> PACS - Standards “Grey” Areas
– Limited inter-vendor compatibility in early years
• ROI
– Can we really eliminate film?
• Installation of xIS prior to or with PACS
– The installation of Image Management prior to Workflow
Management introduces potential for errors and lost
studies.
• The 2nd PACS Challenge
– Was it truly stored in a standard format?
– How is all the data migrated when changing vendors?
©2007 Koller Associates, Inc.
12
Technology Obsolesce & Data Migration
• Technology Obsolesce
– 12” & 14” Optical Disk
– 9 - Track Tape
• Data Migration
– Usually Application Vendor Dependant
– Long Migration Times (Can be Years)
How long do you have to keep data?
©2007 Koller Associates, Inc.
13
Enterprise Storage Consolidation
Patient
Accounts
Billing
Radiology and Cardiology
All Ancillary Dept. Data
Financial
Data
E-Medical
Records
Staff Data
and Records
Centralized
Management
Prod
Common
Disaster
Recovery
Standardized
Testing & Change
Control
.
Proactive
Data
Mart
Consolidated
Backup/Restore
Test
High Availability
High Performance
Reactive
©2007 Koller Associates, Inc.
Cyclic Process
14
Backup Vs. Archiving
Backup
Archiving
• When to Use?
• When to Use?
– Variable Content Data
– Transactional w/ Small to
Medium Records
– Commercial Database
• Examples
– HIS / RIS / LIS Databases
– Image Pointer Database
• Backup Management
– Create Regular Backups
– Manage & Rotate Multiple
Generations
– Database Snapshots &
Replication
– Fixed Object Based
Content
– Long-term Retention
Requirements
– Large Volume of Objects
• Examples
– PACS Image Database
– Other Large / Fixed
Content Objects
• Archive Management
– Create Data Protection &
Retention Policies
– Automate Everything
Possible
©2007 Koller Associates, Inc.
15
Image Data Management
How can your image data be managed?
• Let your Image Application (PACS) vendor decide.
– Dependent on vendor for all storage & future migration.
• Use an all disk solution with multiple copies managed by vendor.
– No real intelligence
– Not responsive to conditions affecting individual data objects
– Still potentially dependent on applications vendor for future migration.
• Use an HSM to archive your images.
– Better control, but minimal metadata for policy-based automation.
– No Linkages Back to Application System for Aging of Studies
– Still potentially dependent on vendor for future migration.
• Use an Information Lifecycle Management (ILM) strategy.
ILM – What is it?
©2007 Koller Associates, Inc.
16
Definition of ILM
Information Lifecycle Management is comprised
of policies, processes, practices and tools used
to align the business value of information with
the most appropriate and cost effective IT
Infrastructure from the time information is
conceived through its final disposition…
Source: SNIA Data Management Forum
©2007 Koller Associates, Inc.
17
ILM – What is it?
• What ILM is not!
– Hardware,
– Software,
– A product!
• Depends on the Vendor – DLM? ILM? xLM?
• All Solutions Should be Standards-based (if possible)
• ILM = DLM + Awareness of Business Value of the
Data
• The Entire Life of a Information Object is Predestined at Time of Creation, But …
©2007 Koller Associates, Inc.
18
DLM vs. ILM
• Data Lifecycle Management
– Focus is to make efficient & cost effective
use of storage resources
– Based on basic characteristics of the
data and data access
– Basic Metadata
Application &
Business Process
ILM
Processing
• Date/Time, Last Access, Owner
• Object Size, System Watermark
• Information Lifecycle Management
– Requires information about business
processes (application aware)
– Holistic view of managing information
– Robust Metadata
• DICOM, Derived from HL-7, Other
DLM
Storage
Tier 1
Storage
Tier 2
Storage
Tier 3
– Uses functions of data lifecycle
management
Excerpted from Storage Industry Primer by Dennis Martin
RMWTUG – Fall ‘05
©2007 Koller Associates, Inc.
19
Image Data Management
Clinical ILM
Departmental
Workflow
Policies
• Workflow Aware
• Routing
Business Value
DLM
Clinical
Archive
Policies
Storage
Archive
Policies
(PACS Virtualization)
(Storage Virtualization)
• Application Aware
• Robust Metadata
• Basic Metadata
• Date/Time
• Pre-fetching
•PACS / RIS / xIS
• Access
• Other departmental
specific features
• HL-7
• Size
• DICOM
• Watermark
PACS & xIS
• XDS
• Location
• Web Services
• Manages any storage type
Clinical Archive
©2007 Koller Associates, Inc.
Storage HW & SW
20
DLM vs. ILM – An Imaging Example
Rad
PACS
DLM (Typical HSM)
DLM Policy Engine
30 Days
Fast Disk
1 yr.
Slow Disk
7 yrs.
Long-Term
DR
©2007 Koller Associates, Inc.
21
DLM vs. ILM – An Imaging Example
Rad
PACS
xIS (s)
Clinical ILM Policy Engine
Control
ILM – Negative
DLM Policy Engine
ILM – Positive
Data
7 yrs.
6 mo.
Fast Disk
2 yr.
Slow Disk
+++ yrs.
Long-Term
DR
Intelligent Enterprise Image Archive
22
DLM vs. ILM – An Imaging Example
Cardiac
PACS
Rad
PACS
xIS (s)
Clinical ILM Policy Engine
Data
Control
DLM Policy Engine
ILM – Cardiology
Fast Disk
Slow Disk
Long-Term
DR
Intelligent Enterprise Image Archive
23
DLM vs. ILM – An Imaging Example
Cardiac
PACS
WSI
SACS
Rad
PACS
xIS (s)
Clinical ILM Policy Engine
Data
Control
DLM Policy Engine
ILM – Pathology
7 days
Fast Disk
Key Images – 7 yrs
30 days
Slow Disk
Long-Term
DR
Intelligent Enterprise Image Archive
24
Benefits of an Automated Clinical ILM
• Provides an integrated BC & DR platform
• Maximizes storage infrastructure utilization
• Provides opportunity to address location
based performance requirements
• Addresses current and future technologies
• Can provide application independent data
migration
©2007 Koller Associates, Inc.
25
Summary
• Images are Very Different to Manage
• Departmental Silos are Not Cost Effective to
Acquire or Manage
• New Accountability for Clinical IT Solutions
Requires Enterprise Data Management
• Effective Enterprise Image Management
Requires Intelligence in the Infrastructure
• No Single Imaging or Informatics Vendor
Provides an Effective Single Source of Scalable
True Clinical ILM
©2007 Koller Associates, Inc.
26
Questions?
Contact Information
John S. Koller
Email: [email protected]
Office: 303.681.2854
APIII 2007 Support Page (Private Event Page)
www.kaiconsulting.com/conferences/apiii2007/index.htm
Password: given at conference