Healthcare institutions are aggressively moving towards meeting compliance with MU1 and MU2 with the implementation of full-featured Electronic Health Records. Concomitantly, there will be a massive increase in the amount of clinical data captured electronically. Business intelligence (BI) which traditionally has focused on financial data can be leveraged to use clinical data to support providers in delivering high quality, efficient care. In addition, BI coupled with population health analytics can help meet many Accountable Care Organization needs. This presentation will discuss the Denver Health journey in using BI in a variety of was to facilitate the attainment of high quality care.
Case Study “Business Intelligence: Supporting Delivery of High Quality Care and Attainment of ACO Goals”
1. Business Intelligence:
Supporting Delivery of High
Quality Care and
Attainment of ACO Goals
iHT2 Summit in Atlanta
Co
April 25th, 2012
Andy Steele, MD, MPH, MSc
Director, Medical Informatics
Denver Health, Denver, CO
2. Learning Objectives
• Identify the impact of business
intelligence (BI) on clinical areas
• Understand unique ways to leverage BI
for supporting ACO goals
3. Denver Health
Integrated public safety net
institution
5,300 employees
Closed medical staff
500 bed hospital
Extensive primary care
network
Level I Trauma Center
Public Health Department
4. Denver Health
Over 160,000 patients
25% of Denver population
Payer mix
– 35% Medicaid
– 28% Uninsured
– 10% Medicare
– 27% Other
$2B in unsponsored care
since 1992
~$400M in 2011
5. Clinical Technology Strategy
Dashboard
Single Enterprise
Clinical
Sign-on Master Patient
Documentation
Medication Index
Results
Administration
Repository
Check
Analytics / BI Patient and
Workflow
Dashboard Provider
Data PACS/Imaging
Warehouse Systems
Enterprise CPOE and
Document Immunization Clinical Rules
Management Tracking
6. Centers for Medicare and
Medicaid Services: ACO
"an organization of health care providers that
agrees to be accountable for the quality,
cost, and overall care…
7. ACO Original Core Principals
Provider-led organizations
– Strong base of primary care
– Accountable for quality and total per capita costs
– Provide full continuum of care for a population of
patients
Payments that are linked to quality
improvements that also reduce overall costs
Use sophisticated performance
measurement
– Support improvement
– Show savings via improved care
8. Payment Reforms Will Motivate and
Reward Innovation at a Whole New Level
-Todd Park, Chief Technology Officer,
U.S. Department of Health and Human Services
IT Innovations Needed:
Accountable • Shared savings; redesigned care
Care processes for high quality, efficient
Organizations delivery Timely Clinical Data,
Decision Support
Patient • Organized outpatient care,
Centered coordination and team-based Care Integration Tools
Medical Homes approaches
Technology to Extend
Bundled • Pilot program for episodes of care; Physician Reach
incentivizes reduced costs around
Payments eight conditions
Consumer Engagement
Tools/Platforms/Apps
Readmission • Motivates hospitals to engage with
Reduction care coordinators and better
Programs organize delivery systems Data Mining/Analytics
10. Big Data: 3 “V’s”
• Selective data retention
Volume • Offload “cold” data
• Outsourcing
• Data caches
• Point-to-point data routing
Velocity • Balance data latency with decision
cycles
• Inconsistency resolution
Variety • Data access middleware and ETLM
• Metadata management
http://blogs.gartner.com/doug-laney/files/2012/01/ad949-3D-Data-Management-Controlling-Data-Volume-
Velocity-and-Variety.pdf
11. Goals for Enterprise Business
Intelligence Strategy
Baseline, documented strategy that includes the
standards, processes, definitions, and approach that
can be developed over time as business needs
change
– Organization wide consistency and coordination for
business intelligence, analytics, and reporting efforts
– Lower costs (people, systems, and software) by reducing
redundancy and unbeneficial activities
– Have an architecture that supports the Enterprise BI
Strategy
– Include plan for Governance of the BI environment
– Communicate consistent vision across the entire
organization
12. High-level Vision: Data Integration
Integrated Reporting, Registries and Analysis
EDW Financial
Claims&
Eligibility Data
Single source for
complex data analysis
Clinical
and reporting
Data
13. Data Warehousing: Denver Health
1998 Data
Warehouse
Financial Demographics
2007
Pathology Pharmacy Pulmonary GI Lab Radiology Laboratory Encounter
2008
Orders Ultrasound EDM Forms Med Administration Custom
Interfaces
2009
Med Recon ED Fetal Monitoring OR
2010
Scheduling Nursing Documentation Workflow Wait List/Referrals Vaccinations
15. Data Warehouse Model
Web Rpts
External Rpts
Internal Rpts
End-user Value Executive Reporting
Portal Design & Implementation
Decentralized Reporting / Training
Patient Value Disease Management / Registries
Clinical Data Validation & Rpt Development
Quality Clinical Interface Development & Testing
Financial Data Validation & Rpt Development
Financial Interface Development & Testing
Maintenance, Upgrades and Support
Foundation Cubes / Data Structures Development
Security and Auditing Tool Implementation
Basic Application Structure / Reporting Tool Implementation
Network & Hardware Infrastructure
Foundation
16. Data Facing Methods
Excel “spreadmarts” and Data Cubes
Crystal Reports/Data Cubes in Web
Publishing
Microsoft SQL Server Reporting
Services (Microsoft SharePoint
Integrated mode)
VPSX delivery
Microsoft Performance Point
Dashboards
Geo-coded Maps via ArcGIS
Microsoft Report Builder ad hoc
reporting model
Microsoft Power Pivot
17. Financial,
Quality,
Safety
Reports
Ad-hoc
mHealth
Reports
Clinical
Registry
(DW & EHR) Point of
Research Care
Support
Employee Outreach
Evaluation Programs
18. Financial,
Quality,
Safety
Reports
Ad-hoc
mHealth
Reports
Clinical
Registry
(DW & EHR) Point of
Research Care
Support
Employee Outreach
Evaluation Programs
21. Quality Scorecard & Registries
2010
– Electronic interface
– 102 measures – all with trend lines
– Ability to drill down to clinic level
– Most measures updated automatically from the
data warehouse (others inputted into intranet
site)
– Much broader audience for most measures
– Ability to secure access to sensitive metrics
22. Data Warehouse-
Medical Quality and Safety
Registries completed for:
– Colon Cancer
– Hypertension
– Diabetes
– Amiodarone
Registries in progress for:
– Breast Cancer
– Cervical Cancer
– Narcotic Users
27. Registry Reports CHS Colorectal Cancer Screening Indicator
All Clinics
Colorectal Cancer Screening is defined as having a colonoscopy in the last 10 years or a flexible sigmoidoscopy in the last
5 years or a fecal occult blood test in the last 15 months.
(Eligible Patients with visits in last 18 months) (Eligible Patients with visit to SGU < 6 months)
Summary By Clinic
Eligible Patients
Eligible Patients (50 - 75 years old) with visit to SGU < 6
months
Site of Greatest Use
(SGU) % Current Total % Current
Total % Current with % Refused % Current
Number with with Number with
Colonoscopy Colonoscopy
Screening FOBT < 15 screening
<10 years
months
Webb FIM 3,390 49 26 1 28 2,169 57
Westside Adult IM 2,977 51 26 1 29 2,017 57
Eastside Adult IM 2,599 50 19 1 36 1,828 56
La Casa/Quigg 1,699 26 20 0 8 1,036 30
Newton
Lowry 0 22 949 44
1,501 37 19
DHMP 1,197 47 42 1 8 687 50
Park Hill 1,140 49 21 0 33 749 55
Westwood 945 41 19 4 25 611 45
Montbello 569 42 12 0 34 337 53
SGU Unassigned 23 30 13 0 17 0 0
Others 9 33 11 0 22 8 38
Total 16049 45 24 1 26 10391 51
Report validated by DSS Development Data Current As of: 08/15/2009
28. Colorectal Cancer Screening
Registry
CHS Colorectal Cancer Screening Indicator
All Clinics
Colorectal Cancer Screening is defined as having a colonoscopy in the last 10 years or a flexible sigmoidoscopy in the last
5 years or a fecal occult blood test in the last 15 months.
(Eligible Patients with visits in last 18 months) (Eligible Patients with visit to SGU < 6 months)
Summary By Clinic
29. Colorectal Cancer Screening
Colorectal Cancer Screening Registry
Registry
Summary By Clinic
Eligible Patients
Eligible Patients (50 - 75 years old) with visit to SGU < 6
months
Site of Greatest Use
(SGU) % Current Total % Current
Total % Current with % Refused % Current
Number with with Number with
Colonoscopy Colonoscopy
Screening FOBT < 15 screening
<10 years
months
Webb FIM 3,390 49 26 1 28 2,169 57
Westside Adult IM 2,977 51 26 1 29 2,017 57
Eastside Adult IM 2,599 50 19 1 36 1,828 56
33. Financial,
Quality,
Safety
Reports
Ad-hoc
mHealth
Reports
Clinical
Registry
(DW & EHR) Point of
Research Care
Support
Employee Outreach
Evaluation Programs
38. Data Warehouse-Medical Quality
and Safety
Examples of clinical informational
queries:
Return to ED and Admit within 7 days
Unexpected transfers to Critical Care
Hypertensives on HCTZ who develop
Acute Gout
39. Data Request Process:
Outcomes
548 requests in 6 months
40% quick strike
30% critical priority
Average report completion
– 6.3 days for quick strike requests
40. Financial,
Quality,
Safety
Reports
Ad-hoc
mHealth
Reports
Clinical
Registry
(DW & EHR) Point of
Research Care
Support
Employee Outreach
Evaluation Programs
44. Financial,
Quality,
Safety
Reports
Ad-hoc
mHealth
Reports
Clinical
Registry
(DW & EHR) Point of
Research Care
Support
Employee Outreach
Evaluation Programs
45. Navigator Report
Community Health outreach workers contact
patients on our hypertension or diabetes
registries in an effort to improve their
preventative care and disease management
Desire for patient lists:
– Need to be contacted
– Already contacted
– MOGED or Opt out
Need ability to “write back” to DSS
46. Navigator Encounter Report
Brings forward patient
demographics
Displays clinical
characteristics for this
patient’s registries
Shows registry statuses
for this patient
Allows the Navigator to
log contact and activity
with the patient
47. Patient Outreach letters
Letters sent to patients if they need to be
screened for breast, cervical, or colorectal
cancer based on national guidelines
English or Spanish version mailed based on
patient’s preferred language
48.
49. Financial,
Quality,
Safety
Reports
Ad-hoc
mHealth
Reports
Clinical
Registry
(DW & EHR) Point of
Research Care
Support
Employee Outreach
Evaluation Programs
50. Neonatalogist Competency - Length of Stay in
Premature Births (33-36 weeks)
For First Quarter 2008
Premature births between 33 and 36 weeks included (ICD-9 codes 765.27 and 765.28).
A B C D E
n n n n n
ia ia ia ia ia
sic sic ys
ic
ys
ic sic
Phy Phy Ph Ph Phy
Medical Discharge Length of
Patient ID Patient Name Admit Date
Record Date Stay (days)
Number
Physician A BETSEY
CHAMBERS,
2940827 000105032718 PARAMO-TERRONES ,KARLA M 01/24/2008 01/29/2008 5
2941282 000105056055 MENENDEZ ,JOSE ALEJANDROD 01/25/2008 02/08/2008 14
2941288 000105056188 MENENDEZ ,ANTONIO MIGUELD 01/25/2008 02/14/2008 20
2944325 000105203970 GANO ,BOY D 02/01/2008 02/08/2008 7
2951097 000105518252 ARELLANO ,GIRL 02/16/2008 02/18/2008 2
Physician B
JONES, M DOUGLAS
2931037 000104560032 CERRILLO-ZAPATA ,ANDY D 01/02/2008 01/15/2008 13
2934945 000104757307 BUSTOS-ARAIZA ,YOSAJANDID 01/11/2008 01/24/2008 13
2936290 000104812250 MONZON-GARCIA ,ADRIAN EMD 01/15/2008 01/26/2008 11
2940709 000105024517 PORTALES-MARZO ,JESSICA D 01/24/2008 01/28/2008 4
LANGENDOERFER, SHARON
Physician C
2929548 000104504386 GONZALEZ ,GIRL 12/29/2007 01/04/2008 6
2931034 000104560024 CERRILLO-ZAPATA ,EMILY D 01/02/2008 01/16/2008 14
2949559 000105437925 RUBIO-GUTIERREZ ,ELIZABED 02/13/2008 02/27/2008 14
2955130 000105704316 DOMINGUEZ-CEBAL ,LIZBETHD 02/26/2008 03/19/2008 22
Certified by DSS Data Warehouse Page 1 Report Date: 06/30/2008
Denver Health CONFIDENTIAL - DO NOT copy, disseminate or distribute this document.
51. Financial,
Quality,
Safety
Reports
Ad-hoc
mHealth
Reports
Clinical
Registry
(DW & EHR) Point of
Research Care
Support
Employee Outreach
Evaluation Programs
52. Data Sharing/Comparative
Effectiveness Research
HVHC: High Value Healthcare Collaborative
(HVHC)
UniversityHealth Consortium (benchmarking)
SAFTINet: Scalable Architecture for
Federated Translational Inquires Network
HMO Research Network
CCTSI - Colorado Clinical & Translational
Sciences InstituteHRSA Collaborative
AHRQ “ACTION” (accelerated research)
53. Financial,
Quality,
Safety
Reports
Ad-hoc
mHealth
Reports
Clinical
Registry
(DW & EHR) Point of
Research Care
Support
Employee Outreach
Evaluation Programs
55. Financial,
Quality,
Safety
Reports
Ad-hoc
mHealth
Reports
Clinical
Is It All Worth It?
Research
Registry
(DW & EHR) Point of
Care
Support
Employee Outreach
Evaluation Programs
56. Clinical Quality Indicators
80%
Denver Health
71%
70%
HEDIS (50th 64%
percentile)
60%
56%
54%
52% 52%
50%
40% 39%
35%
30%
20%
10%
0%
Diabetes Blood pressure < Diabetes LDL < 100 mg/dL All Hypertension BP < 140/90 Breast Cancer Screening
130/80 mm HG mm HG
58. “Obvious” Lessons Learned
DSS can improve efficiency and provide easily
accessible data for quality and safety initiatives
Executive staff must be fully engaged and
supportive
Clinical leadership needs to believe that IT efforts
will improve patient safety and quality
Patience is required to develop and maintain
appropriate infrastructure
Developing clinical registries is a challenging
iterative process
Integrated strategy needed to avoid silo solutions
59. “Obvious” Lessons Learned
Gain physician, financial and administrative buy in
Allocate appropriate funding
Clinical development takes much longer then
financial
Primary care is multi-factorial, solutions need to be
multi-pronged
“Model” is better
– The more model the source is, the easier it is to validate
DSS
– Customizations should be done outside the DSS database
60. “Surprise” Lessons Learned
Start with small wins at high levels
Determine type of BI model the organization can
support
Getting end users involved to early can cause loss of
interest and support
Grab as much data as possible
Look for seed/grant money to start
Data Warehouse data is e-discoverable (Litigation)
and must be in compliance with HIM policy
Physicians don’t know what they want until they see
it
61. “Surprise” Lessons Learned
Almost every “project” can be leveraged
Registry “engine”
Data Management “engine”
Business Intelligence “engine”
“These reports are wrong”
Data is wrong/different at the source
The report is defined incorrectly
The data doesn’t mean what you think it means
Not all Super Users are “super”
Training does not imply proficiency
More difficult the more data that is available
62. Future
New hardware and software platform to
leverage the advancements in BI tools
Extensible data model to support new and growing
data sources
Predictive and “google-like” analytics
Migrate from static reports to self-service BI
tools
Transition “reports” team to BI tool development and
expansion
Revise governance model
More visionary role
Transfer data warehouse functions into EHR