HIMSS National Capital Area (NCA)

1
HIMSS National Capital Area (NCA)The Role
of Informatics in Improving Quality and Safety in
the Health Care Delivery System Key Bridge
Marriott Hotel, Rosslyn, VA 16 March 2011,
530-800 PM 

• Robert E. Connors, FACHE, PMP
• MS in Information Systems Technology
• MHA, Healthcare Administration
• Executive Healthcare IT Manager and IPA
• Henry M. Jackson Foundation for the Advancement
  of Military Medicine
• U.S. Army Medical Research and Materiel Command
• Telemedicine and Advanced Technology Research
  Center (TATRC)
• Fort Detrick, MD
• Robert.Connors_at_tatrc.org

2

• The Clinical Looking Glass Project
• A Civilian-Military Partnership to
  Revolutionize Clinical Intelligence
• http//exploreclg.montefiore.org/

3
How Empowered Are Your Practitioners?
4
Clinical Looking Glass What is it?

• Break-Through Clinical Intelligence Application
  Developed By Emerging Health IT, Inc./Montefiore
  Medical Center and In Use There
• Loading of Clinical Looking Glass data repository
  from EHR and other data
• Ability to create user defined patient groups
  quickly, i.e. All Medicare patients with CHF
  with a diagnoses of diabetes within the last 6
  months
• User defines clinical outcomes of interest at
  will (e.g., lab value within normal limits,
  readmission, mortality) of a patient group
• CLG searches for user-defined outcomes in
  clinical data repository, sends data to
  statistics engine, then presents user with study
  results
• CLG fully documents all criteria, statistical
  methods and output

• QI studies are now executed in minutes, not
  weeks.

5
Clinical Looking Glass Story Themes
6
Information Must Be Patient-Centric
Business-Centric Patient-Centric
How many admissions in 2008? How many individuals admitted in 2008?
How many empty beds? Whats average wait time for a bed?
How many hypertensive patients? How long does it take us to control hypertension?
How much do we spend on drug X? How many improved when on drug X?
What is our estimate of lost business? Who did we lose to follow-up?
7

• Timeframes must be relative to an individuals
  experience.

8
Cohort Paradigm Patient-Centric

• Each individual starts therapy at different
  timesrolling enrollment
• Period of observation (e.g. for good lab value)
  are different for each individual
• Individual contra-indications taken into account
• Stop looking for outcome when patient is no
  longer at risk.
• Example Patients who die should not be followed
  for good lab value.

9
Practitioner Thinking Is Already Patient-Centric

• Doctors and nurses are trained to think from this
  perspective
• For example epidemiology studies, clinical
  literature
• Medical institutions need to nurture and leverage
  their creative intellectual capacity
• Clinicians need the flexibility to explore with
  ad hoc questions for Quality Improvement,
  Research, and Education.

10
Accountable Care Demands

• Evaluate success for each patient using metrics
  achieved within a defined time after the
  patients own start time
• Create Patient Cohorts with a unique index date
  (start time) for each patient when he became
  eligible for intervention
• Evaluate institutions and practitioners by how
  they manage their cohorts over time

11
Cohort Paradigm Patient-Centric

• Each individual starts therapy at different
  timesrolling enrollment
• Period of observation (e.g. for good lab value)
  are different for each individual
• Individual contra-indications taken into account
• Stop looking for outcome when patient is no
  longer at risk.
• Example Patients who die should not be followed
  for good lab value.

12
Practitioner Thinking Is Already Patient-Centric

• Doctors and nurses are trained to think from this
  perspective
• For example epidemiology studies, clinical
  literature
• Medical institutions need to nurture and leverage
  their creative intellectual capacity
• Clinicians need the flexibility to explore with
  ad hoc questions for Quality Improvement,
  Research, and Education.

13
Analytic Steps

• Create a cohort
• Patients seen in three outpatient clinics with a
  HgbA1c of gt9
• Define patients w/ Good and Bad outcomes
• Identify as Good those patients who achieve a
  HgbA1c of 0-7
• Identify as Bad those Patients who achieve of
  HgbA1c of 9-25
• Compare the hospitalization experience of the two
  groups over the ensuing year

14
Overall Vision of Healthcare Market - Direction

• Movement away from Fee For Service to
  Longitudinal Responsibility for the Patient
• CMS - Pay for Performance (P4P)
• Stimulus Funds for Research, HIE/RHIOs and
  EMRs
• Movement toward Universal Health Care Insurance
  Coverage
• Patient Centric Medical Home

15
Who are the Constituencies? The Ultimate Market?

• Classically
• Insurance companies classic utilization review
  (so 1980s)
• The Future
• Clinicians
• Patient Remediation Management (a new concept and
  core to cost effective care),
• Physician Education
• Administrators
• Concierge oversight of delivery of care
• Nurses
• Quality Improvement Professionals
• Health care delivery systems to supervise and
  improve longitudinal care delivery and
  coordination
• Immediate customers DOD, VA, Kaiser, HHS?, FDA?
• Ultimate customers all patients in single payer
  system.
• Patients themselves might query what is going
  to happen to a patient like me (mortality
  complications what if scenarios bp,
  cholesterol..)

16
Why Is Montefiore confident that the market is
moving in CLGs direction?

• CMS pay for performance (primitive but a start)
• JCAHO standards speak to this longitudinal
  oversight responsibility
• American Board Internal Medicine expects MD to
  review their own patient charts and learn from
  errors required part of Recertification
• American College of Graduate Medical Education
• Obama administration emphasis on effectiveness
  research and importantly continuing the Bush
  administrations emphasis on the centrality of
  the electronic medical record to obtain value at
  an affordable cost.

17
What are the nuances between healthcare
information technology management, clinical
informatics, and biomedical informatics?

• My perspective We dont need to dwell on this
  too much, but.
• Healthcare Information Technology Management is
  not the same as Clinical or Biomedical
  Informatics
• Information Technology Hardware and Software
• Clinical Informatics is about using information
  technology to help improve healthcare access,
  availablity, acceptability, cost-effectiveness,
  continuity and quality. (3A2CQ)
• Clinical Informaticists must first understand the
  practice of medicine and business of healthcare
  delivery, then understand how to apply technology
  to improving that business
• Healthcare CIOs have typically failed because
  they dont understand the practice of medicine,
  and have been too focused on technological
  issues. Successful CIOS act as true clinical
  informaticists and must understand what outcomes
  are possible, given an investment in IT
• Hence, the evolution of new roles CIO versus
  CTO versus CKO versus CMIO, etc. CAIO? CNIO?
  CFIO?

18
What are the nuances between healthcare
information technology management, clinical
informatics, and biomedical informatics?,
continued

• Traditional clinical informatics involves the
  application of technologies to improve the
  capture, use, distribution, and storage of
  inpatient admissions and clinical encounters
• Biomedical Informatics is more involved with the
  application of technologies to improving life
  sciences processes, proteomics, genomics,
  pharmacovigilance, post-marketing drug
  surveillance, and the research surrounding the
  fundamental causes of disease
• Given the move to personalized medicine,
  proteomic and genomic information will merge with
  traditional clinical encounter and inpatient
  observation documentations in the Electronic
  Health Record, demanding new data models that
  account for combined data
• Semantic web technologies may be able to help
  mediate terminology differences between
  traditional EHR and life sciences data (i.e.
  Mirhaji/Kashyaps work)
• Predictive modeling and data mining are important
  technologies used by both clinical and biomedical
  informaticists

19
Patient Trajectory

• Business Intelligence Tools (BI) use Timeframes
  such as calendar year, quarterly
• BI tools Slice and dice events in calendar time
  but give no sense of the patients temporal
  trajectory experience

20
Patient-Centric Cohort is Key Concept
1/1/2005
1/1/2007
1/1/2006
Patient
0
1
Diabetes Control
0
2
0
index date
0
3
(EG start therapy)
0
4
0
5
0
6
0
7
0
8
0
9
0
10
4 / 10 40
What of new diabetic patients were controlled
in the year 2005?
21
Cohort Concept (cont)
Enrollment
2 Years
1 Year
Patient
Diabetes Control
0
3
0
index date
0
8
(EG start therapy)
0
9
0
1
outcome
0
4
(EG achieve lab value)
0
7
0
patient experience
0
5
0
10
0
2
(same data, re-sorted)
0
6
5 / 10 50
What of new diabetic patients were controlled
within 1 year?
22
(No Transcript)
23
(No Transcript)
24
(No Transcript)
25
(No Transcript)
26
(No Transcript)
27
(No Transcript)
28
(No Transcript)
29
(No Transcript)
30

• We have more than
• a report card

31

• We have a tool for
• targeted remediation

32
(No Transcript)
33
Identify Failing Patients and their Providers
34
Object Oriented Design Philosophy

• The Output for the first analysis becomes the
  input for the next supporting rapid chains of
  sequential discovery
• Each analysis creates reusable objects for reuse,
  modification, and sharing with colleagues
  supporting an adaptive intellectual ecosystem

35
Comparing Hospitalization Rates for Diabetics

• Diabetics brought under good control
• vs.
• Diabetics who achieve awful control

36
Analytic Steps

• Create a cohort
• Patients seen in three outpatient clinics with a
  HgbA1c of gt9
• Define patients w/ Good and Bad outcomes
• Identify as Good those patients who achieve a
  HgbA1c of 0-7
• Identify as Bad those Patients who achieve of
  HgbA1c of 9-25
• Compare the hospitalization experience of the two
  groups over the ensuing year

37
(No Transcript)
38
(No Transcript)
39
Impact of Good Diabetes Management

• A reduction of almost 1/3
• in hospitalizations

40
Benefits of Clinical Looking Glass

• Monitor efficiency of our processes
• Monitor the quality of the outcome
• Provide report Cards
• Provide a tool for ongoing QI
• Identify areas of focus
• Track those patients over time
• Analyze processes and results
• Target remediation for improvement of care
• Over longer term,
• Rationalize the evaluation of care and
• Bring down true costs

41
CLG SOA Integrated Architecture
42
The DOD-Clinical Looking Glass Project

• Congressional Special Interest Project and DOD
  Joint Program Committee for Healthcare IT
  Research Project
• Research Objectives
• Technological Assessment to determine if Clinical
  Looking Glass clinical intelligence tool in use
  at Montefiore Medical Center, can be used with
  Military Health System (MHS) data
• Test scalability
• Evaluate clinician response to usability for
  quality assurance and patient remediation issues
• Use in actual QA and research studies (Medical
  Home)

43
The DOD-Clinical Looking Glass Project, continued

• Phase I (complete)
• Developed Clinical Looking Glass Platform with
  de-identified clinical data from National Capital
  Area
• Demonstrated platform to hundreds of clinicians
  and executives with very favorable clinician
  reviews
• Worked through Institutional Review Board, Data
  Use Agreements, HIPAA requirements

44
The Clinical Looking Glass Project, continued

• Phase II (underway)
• Establishing platform with live PHI data supplied
  from new USAF Health Data Services Warehouse
• Re-engineering application to make it comply with
  all DOD Security Requirements
• CAC-enabling access to the application
• Working on alternative bolt-on application to
  existing Military Health System (MHS) data
  warehouses, versus mapping of data from warehouse
  to Clinical Looking Glass
• Assessing scalability of application to the size
  of the MHS using Teradata, Netzeeza, and cloud
  computing
• Will make application available for continued
  evaluation by clinicians at Walter Reed National
  Military Medical Center

45
Project Funding

• Total Research Funding To Date is 5.8 million
• FY 06, 08, 09, and 10 Congressional Special
  Interest Funding
• FY 10 Department of Defense Joint Program
  Committee for Healthcare IT Research Funding
• Oversight by U.S. Army TATRC, USAF, USN SPAWARS,
  and DOD JPC for Healthcare IT

46
What is the IT Impact on 3A2CQ?

• Seek technologies that can help improve
  healthcare
• Availability
• Acessibility
• Acceptability
• Cost-Effectiveness
• Continuity
• Quality
• If no potential for3A2CQ, dont invest in it!

Clinical Looking Glass is About Promoting All
Of These !
47
Points of Contact

• Robert E. Connors, FACHE, PMP
• Executive Healthcare Manager and IPA
• Henry M. Jackson Foundation for the Advancement
  of Military Medicine working at U.S. Army
  Telemedicine and Advanced Technology Research
  Center
• Robert.Connors_at_tatrc.org
• [phone] (Google Voice Forwards
  Automatically To Apple iPhone, Work, and Home)
• Eran Bellin, MD
• Vice President
• Clinical IT Research and Development Emerging
  Health Information Technology Department Outcomes
  Analysis and Decision Support
• Montefiore Medical Center
• ebellin_at_emerginghealthit.com

48
Questions?
49
Backup
50
Clinical Looking Glass Was Developed By
Montefiore Medical Center and Is In Use There

• Large Urban Medical Center (Bronx, New York)
• 4 Hospitals, 1491 beds
• 94,000 Discharges
• 280,000 Emergency Department Visits
• 21 Clinics , 2 Million Clinic Visits
• 150,000 Covered Lives Through IPA
• Chronic Disease Treatment and Prevention
  Challenges
• Asthma
• Diabetes
• HIV
• Heart Disease

51
Clinical Looking Glass At Montefiore Medical
Center

• More than 500 users 700 trained
• 2800 analytics run per month
• 25 QA projects
• More than 50 IRB projects
• Training in CLG is now a mandatory part of a MMC
  Medical Resident
• Education/Research - Publishing of papers around
  asthma, HIV and ER treatment of the elderly

52
CLG Track Record

• Hundreds of active users
• Dozens of manuscripts citing CLG
• Grants
• Required QI rotation for medical students and
  residents
• (Albert Einstein College of Medicine)
• US patent
• Dept of Defense

53
Clinical Looking Glass At Montefiore Medical
Center, continued

• Examples and Success Stories
• CLG used to monitor success of use Chronic Care
  Model in treatment of diabetics.
• CLG used to monitor effectiveness of use of
  hospitalists to reduce Length of Stay of
  Inpatient admissions. Study found use of
  hospitalists could reduce LOS while not adversely
  affecting readmission of mortality.
• CLG used to provide mandatory reporting to State
  agencies and to drive performance improvements in
  treatment of chronic disease in a large
  outpatient clinic in the South Bronx.

54
Differentiators

• Longitudinal cohort
• Multi-group comparison
• Transparent study criteria for replication/refinem
  ent
• Statistics to guide inference
• Sophisticated group and outcome definition
• unlimited temporal relationships
• mixed logical operators (and/or/not)
• event combinations as outcome
• Analytic methods borne of the years of need
• HIPAA safe discovery / audited re-identification

55
The Military Health System (MHS)

• One of the worlds largest integrated healthcare
  delivery systems
• Responsible for delivering high quality,
  cost-effective healthcare to 9.6 million eligible
  beneficiaries
• Operates 720 fixed medical and dental treatment
  facilities
• Pays for civilian care delivered through TRICARE
  Prime, Extra, and Standard
• Coordinates care with VHA

56
U.S. Army Medical Research and Materiel
Command Telemedicine and Advanced Technology
Research Center (TATRC) Fulfilling the Vision
Thru RD
Mission
Explore science and engineering technologies
ahead of programmed research, leveraging other
programs to maximize benefits to military medicine
Vision
Be the DoD model of government enablement of
technology transfer to use
2010
57
TATRC Portfolios
U.S. Army TATRC Portfolios
Joint Technical Coordinating Group
58
TATRC Health Information Technology GroupProject
Categorization

• Acquisition and Access to Data
• Analysis-Execution (CLG Fits Here)
• Distribution-Portability
• Representation and Visualization
• Archive and Retrieval
• Interoperability and Standards

59
TATRC Health Information Technologies (HIT) Group

• Past Three Years
• 81M in Emerging HIT Research Projects
• 31 Congressional Special Interest
• 18 Small Business Innovative Research and TATRC
  Augmentation-Funded
• 8 Army Advanced Medical Technology Initiatives
• 3 Cooperative Research and Development Agreements
  (CRADAs)

CLINICAL LOOKING GLASS PROJECT HAS SIGNIFICANT
POTENTIAL FOR IMPROVING MILITARY HEALTHCARE
OUTCOMES
60
TATRC Health Information Technology Group
Partnerships (Partial)

• Google
• Microsoft
• McKesson/Relay Health
• IBM
• GE
• Mysis/Allscripts
• Unisys
• Metadata, Inc.
• Nuance, Inc.
• Moberg Research
• DocBox
• Braun Systems, Inc.
• InRange Systems
• Intersystems, Inc.
• Open Health Tools
• VA and VHA
• HHS ONC for Healthcare IT
• CDC, NIH, Homeland Security

• Kaiser Permanente
• Regenstrief Institute
• Partners Healthcare
• CIMIT
• Cleveland Clinic
• Mayo Clinic
• Montefiore Medical Center
• Intermountain Healthcare
• University of Utah
• University of Texas Medical Center, Houston
• Grady Memorial Hospital, Atlanta, GA
• Childrens Hospital, Oakland
• Spartanburg Regional Health System
• CareFx
• Parsons The New School of Design
• Collexis, Inc.

61
Many TATRC Research Projects Support Advanced
Concept Demonstration for the Nationwide Health
Information Network (NHIN)