1. Tagging Data Governance Issues in a
Hospital by Using a Logical Business Data
Model
EnterpriseDataWorld2013
San Diego
Erik van der Velde
Program Manager Data Governance
University Medical Center Groningen
2. A short introduction
• Started career as a nurse
• Nursing Informatics
• CBIP
• 25 years experience in healthcare
• 15 years in IT and data
• Developed the “UMCG
information factory”
• Member executive committee
congres All about data
• Program manager
Datagovernance since June 2012
3. Agenda Presentation
• Healthcare in The Netherlands and Groningen.
• A useful reference model for healthcare in a disorderly world. Wouldn’t it
be nice ?
• Benefits of a logical business data model without losing the “real” world .
• Developing a data lifecycle strategy.
• Setting up a data governance organization in a politically delicate
environment .
4. Key elements of Dutch healthcare system
All adults are compulsorily
insured
Children
(under 18)
are insured
for free
Low income groups receive financial
compensation
Insurer must
accept everyone
and get
compensated
from risk
equalization
fund
Insurers have the role of
prudent purchasers of health
care (value for money)
additional optional health
insurance packages
5. Hospital density in the Netherlands
• 16,6 milj. residents
• 120 hospitals
• 8 academic hospitals
• +/- 50.000 beds available
• 8,500 beds in Academic
Hospitals
6. University Medical Center Groningen
• 10.571 employees
• 1.339 beds (including cradles)
• 35.412 admissions
• 324.954 nursing days
• 509.867 visits to outpatient clinics
• 1.219 trauma helicopter flights
• 25.010 day care treatments
Facts and figures
7. Profile UMCG
• 1 of the 8 University Hospitals
• Founded in 1819
• Focus on Healthy Ageing
(growing older in a healthy and active way)
• The transplant center of the Netherlands
• Strong regional function and partnership other
hospitals
8. Starting Point
Datagovernance
program
Best of
breed policy
Business
Intelligence
Implementation of a
new HIS
Risk management
Legislation
Indicators
Dataownership
Data maturity
organisation
A reference model for healthcare
9. A reference model for healthcare, wouldn’t it
be nice ?
Problems usability
• Most transaction driven
• No business language
• Too Immature
• Too detailed
• Doesn’t cover total hospital
environment
10. A reference model for healthcare in a
disorderly world : RDH 2.0
Describes
process and
information
objects in
unambiguous
domains
11. A reference model for healthcare : RDH 2.0
• Advantages
─ One model for plotting applications, processes and
information
─ Understandable for different professions in organization
─ Explained in Laymans terms
─ Not only focused on cure or care
• Challenges
─ Not every domain evenly developed
─ Relationships between information objects not defined
─ Business glossary incomplete
12. Benefits of a logical business data model
without losing the “real”world .
• Decision to use RDH 2.0 as base for LBDM
• LBDM base for tagging DG elements
• LBDM base for migration strategy
• DG <-> Migration
• Complicating factor: two hospitals ErasmusMC and
UMCG
• Scope +/- 500 applications
13. Benefits of a logical business data model
• Management afraid for
Sagrada familia effect
• Developed the LBDM in
9 month
• Replenish it with a
GBDM for validation
• A RBDM for reference
data to avoid
redundanty
• A TBDM as the ideal
world of datasets and
base for migration
16. Benefits of a logical business data model
The elevator Pitch
17. Developing a data lifecycle strategy
Goal: Qualitative correct, complete and valid data
Quality is:
1. Filtering for end use
2. Cleaning the lake
3. Using a filter drain for
the factory
4. Preserve the factory
process
18. Benefits of a logical business data model
“The Data governance arena”
19. Developing a data lifecycle strategy
• Analyse sources
• Define Dataownership
• Set up Data delivery agreements
• Making Extractions
• Analyse DQ extractions
• Setting up a definition list
• Spec. Technical business datamodel
• Spec. DQ-rules
• Spec. Mapping-rules to TBDM
• Setting up audit trails
20. Developing a data lifecycle strategy
Extractions
ETL
Mapping
LBDMDatamodel
source
Issue
list
Audit
list
check
Issue
list
Mapping
Audit
list
Migration Database
New
HIS
Sources