Part 2 PM - Metadon.net

Report
ดำรงศักด์ ิ บุลยเลิศ
•
•
•
•
•
•
•
•
B.Sc. (Med. Sc.), Chiang Mai University, 1979
Doctor of Medicine, Chiang Mai University, 1981
Doctor of Philosophy, Emory U, USA, 1989
Medical Informatics, Stanford U, USA, 1995
Total Quality Management, TU, Thailand, 2000
Hospital Accreditation for Top Executives, HA, 2000
Medical School Executive, COTMES, Class of 2002
Information Security Policy, MISTI, USA, 2005
1
ICT Impacts on Healthcare Sector
• Objectives
– Patient outcomes remain the same over the past 2
decades
– Processes are being improved but without directions.
– Your job is to provide recommendation to healthcare
providers (hospitals) on what to do to mitigate the
situation
2
Healthcare XPD % GDP (World Bank, 2012)
20
18
16
East Asia & Pacific (all income levels)
14
East Asia & Pacific (developing only)
European Union
12
10
North America
World
Japan
8
Malaysia
Myanmar
6
4
Philippines
Singapore
Thailand
2
0
3
GDP (in Billion US$)
ผลิตภัณฑ์มวลรวมในประเทศ (มูลค่าทางการตลาดของสินค้าและการบริการทีผ่ ลิตได้)
Thailand
400,000,000,000.00
350,000,000,000.00
300,000,000,000.00
250,000,000,000.00
200,000,000,000.00
Thailand
150,000,000,000.00
100,000,000,000.00
50,000,000,000.00
0.00
1993199419951996199719981999200020012002200320042005200620072008200920102011
4
Common Causes of Illness (per 1,000)
600.0
500.0
400.0
2007
300.0
2008
2009
2010
200.0
100.0
0.0
Respiratory D
Circulatory D
Gastrointestinal Musculoskeletal
D
D
Endocrine and
metabolism D
Infection,
infestation
5
Missions of Healthcare Facility:
Public and Private
1. Health promotion (Poor outcome)
– Main function receiving little attention. Exercise and diet.
2. Disease prevention (Poor outcome)
– Environment, personal hygiene and city sanitation
including early disease detection.
3. Treatment (Very poor outcome)
– Too much focus due to interest and gain (nepotism and
cronyism).
4. Rehabilitation (unacceptable outcome)
– Basic obligation for any illness.
6
• Organizations in Public Sector exist because of
government mandate and not because of any vision
• Mission begets Vision and not the other way around
– HA, TQA, etc. on vision vs mission
• Practically almost all hospitals have been
accredited during the past 20 years, but the
situation remains the same.
• The authority and hospitals are lost. It is obvious.
7
• The concept of production line applies to hospital
operation.
• Input > Processes > Output > Outcome > Impact
• Input = a patient with high blood pressure
• Process = diagnosis and treatment
• Output = blood pressure is normalized
• Outcome = patient health is restored
• Impact = healthy individual is productive, etc.
8
• A lot of processes (mostly campaigns) have been
publicized and implemented.
• Yet, the outcomes are unchanged or worsening.
• How do we know that the situations are not
improving?
• The health statistics data says so.
– Mortality, Complication, Length of Stay, Cost
• Composite indices are hard to monitor. We need
simple indicators.
9
Basic Indicators Required
A. Patient safety indicators
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
Foreign Object Retained After Surgery
Air Embolism
Blood Incompatibility
Stage III and IV Pressure Ulcers
Falls and Trauma
Manifestations of Poor Glycemic Control
Catheter-Associated Urinary Tract Infection
(UTI)
Vascular Catheter-Associated Infection
Surgical Site Infection
Deep Vein Thrombosis (DVT)/Pulmonary
Embolism (PE)
B. Drug-related indicators
C. Lab-related indicators
D. Waiting times for
outpatient services and
elective interventions
E. Personnel indicators
F. Customer satisfaction
rates and complaint rates
10
• In order to run a hospital we need an information
system largely known as Hospital Information
System
• Basic requirement
–
–
–
–
Operation
Planning
Communication
Documentation and reporting
11
Hypothetico-deductive Approach
Patient presents
with a problem
Initial hypotheses
ID, CC, HPI
Ask questions
Patient is cured;
no further care
required
More questions
PI, PH, FH,
Social, ROS
Patient
dies
Observe
Results
Laboratory
tests
Refine
hypotheses
Radiologic
studies
Chronic
diseases
Treat patient
accordingly
PE
Examine
patient
Select most
likely diagnosis
EKG,
etc.
12
Medical Informatics: Computer Applications in Healthcare and Biomedicine, 2nd edition.
• To serve a patient, information recording requires
an equivalent of 1,600 tables (in relational database
approach)
• Many issues to be considered in developing a
hospital information system
13
Relational Concept and the Real World
14
15
Systems Design
Some issues to be considered
1.
Quality and style of interface
2.
Convenience
3.
Speed and response
4.
Security
5.
Integration
16
Systems Design: Network
• Network infrastructure
–
–
–
–
How many subnets in a hierarchical network? VLAN?
How many functional subnets (intranets)?
Public vs private IP address scheme (IP V4 vs V6)
Intranet bandwidth requirement: core, distributive and
access
– Wireless for mobile computing?
17
• Internet bandwidth requirement
– Personal access, e.g., online journals, Google, etc.
– Personal communication: email, Skype, etc.
– Public access: webpage presence
18
Systems Design: Hardware
• Server: mainframe, mini or microcomputer
• Input device: keyboard, mouse, barcode reader,
RFID
• Output device: monitor, pooled vs stand alone
printer, etc.
• Storage device with RAIDs (what RAID level)
– Speed vs redundancy
– Dedicated device such as SAN
19
• RAID 0: block striping, no parity
• RAID 1: mirroring without parity or striping (duplexing with
separate controller)
• RAID 10: mirroring and striping (many 1ry and 2ry)
• RAID 2: bit level striping
• RAID 3: byte level striping with dedicated parity
• RAID 4: block level striping with dedicated parity
• RAID 5: block level striping with distributed parity
• RAID 6: block level stripping with double distributed parity
20
Systems Design: Software
• Operating system: proprietary v.s. open source
• Applications:
– Buy proprietary products (ผูผ้ ลิตทรงสิทธิ ์) off the shelf
with/without add-on programming
– Build in-house application from scratch
– Implement a public domain software, e.g., VistA
(Veterans Health Information Systems and Technology
Architecture)
– Outsourcing completely: Theresa system of Grady M.
21
Systems Design: Software
• Source code management:
– Software escrow (สัญญาซอฟท์แวร์) option
• Information exchange functionality
– Information useless if not exchangeable
– HL7 protocol should be standard and built-in
22
Systems Design: Computing Model
Centralized, modular and distributed computing
• Centralized computing: all functions in a single
homogeneous application.
• Modular computing: subsystems working together in
one application, each can be enabled/disabled.
• Distributed computing: many stand-alone systems
working together in an integrated environment
23
• Database model: relational vs. non-relational
• Computing architecture:
– Data management, business logic and user interface
model
– 2-tier client/server: thick client vs thin client
– 3-tier client/server: scalability
– N-tier client/server
• Any role of mobile computing?
24
Depressing Fact
• Almost all (99.99%) of healthcare personnel in
charge of ICT in public sector are ICT illiterate.
– Self improvement, probably little reading
– Most books read are on programming but not systems
designs and development.
• Most of them are computer hackers (geeks).
• They don’t easily take consultancy.
25
For hospital information system: 6 major functions.
1.
2.
3.
4.
5.
6.
Patient management
Departmental Management
Care delivery and clinical documentation
Clinical decision support
Financial and resource management
Managed care support
26
Basic Functional Modules
•
•
•
•
•
•
•
•
Medical record system
Outpatient system
Laboratory system
Radiology system
including PACS
Pharmacy system
Patient financial system
Physician workbench
Nursing workbench
• Inpatient system
• Operating theater/room
system
• Labor room system
• Anesthesiology system
• Dietetic system
• Central Sterile Supply
System
• Laundry Service System
27
Two Integration Requirements
• Data integration: a single piece of information is
generated once and then shared amongst various
healthcare personnel, e.g., drug prescription
– Patient arrives at the hospital and is registered by medical record
department
– Patient arrives at the outpatient clinic
– Physician enters clinical information
– Through computerized order entry system, physician places an
order for a drug
– Cashier system receives the payment
– Pharmacy system dispenses and releases the drug
– Nurse could view what the patient receives
28
Two Integration Requirements
• Process integration (process improvement)
– Only a few ways to get a job done, not indefinite ways.
• Patient registration: by MR personnel, nurses, physicians
using standard operating procedures. Everyone will do the
same approach, same user interface.
– Transactions can be carried out by more than one
authorized individual.
• Cashier functions may be performed by financial/accounting
personnel as well as authorized users such as management.
29
MoPH Report Requirement
• All stakeholders perform data entry according to
individual roles.
• All stakeholders get relevant data from the pool.
• Each hospital sends a set of data to the MoPH and
stakeholders retrieve relevant data.
– One change request per year from a central location
and each hospital responds/submits information to that
body. Other Department retrieve relevant data from this
central location.
30
Dept. A
Hospital X
Dept. B
Hospital Y
Dept. C
Hospital Z
Dept. A
Dept. B
Dept. C
Current mode of
operation
(Poor system, no
audit)
Hospital X
MoPH
Data
Center
Hospital Y
Hospital Z
31
• Change management inefficiency coupling with lack
of integration knowledge may lead to a complete
implementation failure
– Legacy pharmacy system story from KKU
– Legacy lab system story from Ramathibodi Hospital
32
Software Issues
• Build: can be 100% customized but time
consuming, failure rate approaching 100%
• Buy: customization almost impossible, but ready to
serve. Major problem in patch management.
• Best approach is to use integration engine to solve
legacy issues, e.g., clearinghouse, rhapsody,
ensemble, etc.
33
Uses of Medical Data
1. Create the basis of historical record
– Medical record for life long service provision but patient
health record for patient self-management and
interagency exchange
2. Support communication among providers, internal
and external (coding standards and standard for
information exchange)
3. Anticipate future health problems
4. Record standard preventive measures
34
Uses of Medical Data
5.
6.
7.
8.
Identify any deviations from expected trends
Provide a legal record
Support clinical research
Serve as a source for indicator calculation
– Mortality rate, complication rate, length of stay,
expense
9. Etc.
35
Information Audit
• International Standard
– ISO 27001, 27002
36
ISO 27002
•
•
•
•
•
•
•
•
security policy: 99% of public hospitals have no policy
organization of information security: ditto
asset management: ditto
human resources security: minimal
physical and environmental security: partial
communications and operations management: minimal
access control: minimal and mostly password scheme
information systems acquisition, development and
maintenance: ditto
• information security incident management: none
• business continuity management: emergency power mostly
• Compliance: none
37
Buying Issues
• Procurement
– How to establish a sound TOR as systems analysis
takes forever.
– Competition and e-auction requirement
• Acquisition and maintenance
– The project does not end at contract signing. It is just
the beginning. Most agencies falsely mark this
milestone as an indicator (process).
• Project management
38
Project Management
Quality PIC of CHRiST = 9 areas
• Quality
• Procurement, Integration, Communications
• Cost, Human resources, Risk, Scope, Time
• I Saw the Car Quickly Hit Charlie’s Rear Plate
39
Waterfall Model of SDLC
40
Implementation Issue
• Scanning of medical records
41
Medical Record (paper vs. electronic)
•
•
•
•
•
•
Availability
Remote Access
Simultaneous Access
Legibility
Information organization
Completeness
• Timely analysis
• Research & education
support
• Initial Investment
• Security issues
• Reliability
• Human errors
42

similar documents