Transcript A Bucket Full of Numbers?

A Bucket Full of Numbers?
An Insight to Hemodialysis Data
Management Systems
Kevin Lyons
Fresenius Medical Care, Bad Homburg / Germany
Topics
• Evolution of Hemodialysis Data Management
Systems
• Potential Pitfalls
• A case for automatic data acquisition
• The Current ‘State of the Art’
– What should a modern IT system provide?
• What about the future?
Evolution…
• The early days….
– Paper record.
- Only option available.
- Manual.
- Difficult to evaluate trends.
- Prone to tpyographical error.
- Very short term use.
- Limited contribution to
patient care.
Life Cycle of the Paper Record
Evolution over 25 years
1973
Number of
Patients
Treatment time
Dialysers
1998
27000
325000
18-24 hours per week
9-12 hours per week
Hollow fibre just a prototype
Hollow fibre dominates
QB/QD
200-300/500
Up to 500/800
Dialysate
Dialysate delivery
Volume control
Staffing
Acetate
Bicarbonate
Central
Individual
Kuf and TMP
Volumetric
Few actions required
Multiple complex actions
Machines
Relatively simple
Highly complex with
physiological or bio-feedback
Reimbursement
Straightforward
Increasingly complex
Computers
Mainframes and Character terminals
PC’s with Pentium processors
Adapted from Ronco C 2000
A case for a Database?
• A dialysis unit may collect several GB of data per year
• Device & machine data
• Pressure, conductivities, flows, temperatures, physiological data
• Other treatment information
• Symptoms, access documentation, hospitalizations …
• Laboratory results
• Diagnosis
• Demographic information
• Drug prescription & intake
• Reference data (medications, diagnosis catalogues, …)
It is hard to store/maintain this information without databases
Driving Forces towards Data Management
Downward
pressure on
healthcare costs
Steady
increase in
patient
numbers
Sharing & Exchange of
information
Sheer Volume of
data to manage
COMPUTERISE!
Compliance to
Standards E.g.
DOQI
Optimisation
of Resources
Desire to
become
paperless
Keeping abreast
of technology
Industry Responds…..
Sample for illustrative purposes only
Evaluation of Clinical Computing Systems as
Used by DOPPS Participants
Nurse
Manager
Medical
Director
Pleased with current system
41%
34%
Need to add additional functions
75%
73%
Need a new system
45%
55%
37%
33%
44%
38%
Current system is very good at
Patient care….
Current system is very good at
Monitoring quality and outcomes
Sargent J 1999
Factors Limiting the Effectiveness of
Computer Systems
• Purchasing system before needs are defined
• System does not meet the requirements
• Data collection and entry
• Management of collected data is complex
• Lack of a friendly interface
• Reporting not targeted to improving patient care
levels
• Not upgradeable
• Lack of an ongoing training program / poor training
Adapted from Ronco C 2000
… modern dialysis machines are able to provide much
more treatment information...
T [°C]
38
38
T [°C]
37
37
Tbody
36
Tbody
36
Tven
35
0.0
0.5
E[kJ]
35
0.0
0.5
E[kJ]
0
1.0
-100
0.0
1.0
1.5
0.5
1.5
Tven 2.0
2.0
1.0
1.5
t[h]
t[h]
2.0
t[h]
0
-100
0.0
Online plus™ System
0.5
1.0
1.5
Online
Time1[min] v RBV [%]
Time1[min] v RBV [%]
RBV Upper Specification
RBV Upper Specification
Time1[min] v RBV [%]
Time1[min] v RBV [%]
Time2[min] v UFR [l/h]
RBV Upper Specification
102
1.25
96
1.00
96
0.75
RBV [%]
0.50
94
86
84
92
0
50
100
90
150
200
250
100
1.50
98
1.25
96
92
88
0.25
86
0.00
300
84
1.00
0
50
0.75
100
150
time [min]
time [min]
0.50
88
200
94
250
86
0.00
300
84
150
time [min]
200
250
1.50
1.25
0.75
0.50
1.00
0.00
300
0.75
0.50
88
84
100
1.00
90
86
50
1.75
1.50
0.25
92
0.25
0
1.75
1.25
94
90
90
88
RBV [%]
92
98
1.75
0.25
0
50
100
150
time [min]
200
250
0.00
300
UFR [l/h]
98
UFR [l/h]
RBV [%]
96
102
100
1.50
100
UFR [l/h]
98
UFR [l/h]
1.75
100102
RBV [%]
102
94
Time2[min] v UFR [l/h]
RBV Upper Specification
Time2[min] v UFR [l/h]
Time2[min] v UFR [l/h]
2.0
plus™
t[h]
System
Other Medical Devices Can Provide
Information too…
…in principle, any device with an
RS 232 or similar interface can
pass on its measured parameters
The Missing Link…
File Servers
Workstations
Workstations
Manual -v- Automated Measurements (1)
Patient
Time taken (secs)
Patient
Time taken (secs)
HD
DD
NB
BB
SD
MU
JC
MG
VP
NP
50
55
55
40
68
35
60
35
63
80
HD
QF
GP
AD
BB
PC
MU
JR
SD
ER
180
255
360
320
240
230
285
195
245
290
Range
35 – 80
Range
180 – 360
Mean
54.1
Mean
260
Table 1: Time needed to produce dialysis prescription using automated
data acquisition compared to manual data entry
Source: Rourke E et al. “Man or Machine”
Manual -v- Automated Measurements (2)
Blood Pressure Pre-Finesse
Blood Pressure Post-Finesse
150/90
191/110
140/92
212/89
150/110
196/99
150/80
210/104
140/90
188/89
150/90
214/105
202/100
191/89
150/90
169/84
150/90
175/84
150/85
190/110
Table 2: Comparison of 10 Blood Pressure readings taken by a patient immediately
before the installation of Finesse and 10 automated readings produced immediately
after the installation of the system
Source: Rourke E et al. “Man or Machine”
Manual -v- Automated Measurements (3)
10
9
% patients
8
7
6
5
4
3
2
1
0
Headache
Pre Finesse
Hypotension
Post Finesse
Cramp
Nausea
Chest Pain
Figure 4 : Comparison of reported symptoms 3 months prior to, and
3 months after installation of Finesse
Source: Rourke E et al. “Man or Machine”
Benefits of a Data Acquisition System
• Real time treatment information
• Accuracy of recorded data
– Patient reported
– Typographical
• Therapy given based upon patient status
• Burden of data collection greatly reduced
• Facilitates Audit and Analysis
Acquisition alone is not
enough…
It can only be successful when
implemented with the ‘right’
database.
What Characterises an Outstanding
Data Management System?
• Functionality
– Data presentation & analysis
– Integration of dialysis machines & medical devices
- Advanced features (decision support, bio-feedback)
– Integration with other systems
Time1[min] v RBV [%]
Time1[min] v RBV [%]
RBV Upper Specification
RBV Upper Specification
Time2[min] v UFR [l/h]
Time2[min] v UFR [l/h]
102
1.75
1.75
100
1.50
98
1.50
96
1.25
96
1.25
94
1.00
92
UFR [l/h]
98
RBV [%]
RBV [%]
100
94
1.00
92
0.75
0.75
90
90
0.50
88
86
84
0
50
100
150
time [min]
200
250
0.50
88
0.25
86
0.00
300
84
0.25
0
50
100
150
time [min]
200
250
0.00
300
UFR [l/h]
102
What Characterises an Outstanding
Data Management System?
• Acceptance: The user likes to work with it, …
– Simplicity, usability and workflow
– Don’t have to drastically change their way of working
– Training
Where do I have
to enter what??
Without workflow:
The worker looks
for the work
Please
Confirm
F60
With workflow:
The work looks
for the worker
Acceptance: Training is an ESSENTIAL element
• NEVER underestimate the
importance of training
• Create local experts via the TTT
concept
• Ensure ALL relevant disciplines
are involved
• Bilateral responsibilities
What Characterises an Outstanding
Data Management System?
• Cost effectiveness
– Benefit from changing technologies and
competition
– Scalability
Cost Effectiveness
• Profit from competition, be independent of
– Platform (AS400, PC, SUN, …)
– Database (DB2, Oracle, MS-SQL, …)
– Operating System (OS400, NT, Win9x, Solaris, …)
• Profit from changing Telecoms market, choose
– LAN, Frame-Relay, Internet, Dial-up, …
• Profit from global developer community and use open
and (license) free standards
– Java, HTML, XML, Servlet, Internet (www, ftp, …)
XML: Extended Mark-up Language
– THE data exchange format
– Non proprietary, widely supported
(Microsoft, SUN, IBM, oracle, …)
– Standard also in the medical field (HL7)
XML
Standard-Interface
150220001947211926954877840000415300324331093de34ar153
150220001947260000000035290581822500deas01808de34ar154
...
Vendor specific
Human readable
<machine_data id="Fresenius_4008">
<patient id=Q1299817212 />
<row>
<data name="time">15.02.2000 19:47:21</data>
<data name="ven. pressure">1926</data>
<data name="art. pressure ">9548</data>
<data name="Temperature">7784</data>
<data name="UF-Volume">4153</data>
<data name="UF-Rate">1093</data>
</row>
<row>
<data name="time">15.02.2000 19:47:26</data>
<data name="Temperature">3529</data>
<data name=“Dialysate-flow">681</data>
<data name="UF-Volume">8225</data>
<data name="UF-Rate">1808</data>
</row>
...
</machine_data>
Cost Effectiveness: Scalability
1000+ PCs
Technical
Scope
10 PCs
1 PC
Single
doctor
100 PCs
Warehouse
Workflow
Machine integration
Dialysis
unit
Dialysis
clinic
Disease state
management
Entire medical
documentation
Functional
Scope
Hospital
Chain
National
Healthcare
Internet….. What’s so Special?
For the very first time in history we have:
– Option to share information and to communicate
without any restrictions by borders and locations
- Send: text (letters), pictures (x-ray), movies
(echo), sounds
- Discuss: chat-rooms,
via clinical
chat-rooms,
news-groups
news-groups
State-of-the-art
IT
systems
have
- Retrieve
information:
on-line databases &
to utilize
these opportunities!
journals
– Standardised technology
– Chance to share common data:
consistent, non-redundant, without lavish
interfaces, shared by larger community, e.g.
Several clinics
– Affordable by nearly everybody
Thank you
for your
attention
Kevin Lyons
Service & Training Manager
Fresenius Medical Care
Regional Office Middle East
PO Box 3264
Dubai UAE
Tel +9714 332 9317
Fax +971 332 9316
Mob +971506250443
E-mail: [email protected]