Transcript CPOE Future 2 way interface

Computer Provider Order Entry
CPOE
The First Step in Computerized
Decision Support
Bruce Slater, MD, MPH 263-8242 - Pager 9015
DGIM Primary Care Conference 4/14/2004
Learning Objectives
• Be able to
– describe what CPOE is
– describe proven advantages of CPOE
– describe what is expected of attending
physicians using CPOE
– articulate advantages of CPOE for physicians at
UWHC
Financial Support
• No outside support
Introduction
• Definitions
–
–
–
–
Clinical Decision Support (CDS)
Not Continuing Medical Education (CME)
Computer-based Patient Record (CbPR)
Clinical Data Repository – CDR (WISCR-IT)
Background
•
•
•
•
CPOE is the infrastructure for CDS
Why aren’t we there yet?
Security, Standards and Simplicity
What it can’t do
The CbPRS as Car
Cockpit
Lubricants
Driver
Fuel to Engine
Transmission
Brake
Accelerator
Car Part Function CbPR Part Function
Dashboard
Status
Display Screen
Status
Steering/Pedals
Control
Keyboard/Mouse
Control
Fuel&Engine
Propulsion
CDR
Clinical Data
Accelerator
Move forward Reminders
Do things
Brake
Stop
Alerts
Don’t do things
Transmission
Transfers power
Standards
Transmits
messages
Lubricants
Reduces friction
Terminologies
Reduces
misunderstanding
Driver
In Charge
Provider
In Charge
How it works – 1
Paper Orders
A2K terminal
(HUC)
MD-paper
orders
Action 2000
Orders
CPOE
CPOE
MD-on Web
Browser
WISCR-IT CDR
Med Orders
Intermediate Document
RxTFC
Accuscan
How it works – 2
Personal Computer
Web Browser
Java Virtual Machine
CPOE
Screen Painter
Form Printer
Dialog Boxes
The Network
WISCR-IT
http
How it works – 3
How it works – 4
How it works – 5
How it works - 6
Evidence-based Informatics - 1
• Tierney, et al. JAMA 1993
– Financial
• Total Charges – $6964 v $6077 12.7% less (p=.02)
• Test Charges - $1852 v $1621 12.5% less (p=.006)
• Drug Charges - $1181 v $1001 15.3% less (p=.008)
– Time
• Minutes writing orders 25.5 v 58.5 more (p<.001)
• 5.7 minutes less writing “scut cards” (p=.02)
– Acceptance
• 52% of housestaff users thought it made work easier
Evidence-based Informatics - 2
• Kuperman, et al. Annals 2003
– Review article
• Time entering orders – 3 studies summarized
– Shu 2001 – (9% v 2.1%) minus 2% = 5% more time
– Bates 1994 – (10.5% v 5.3%) minus 2.7% = 4.5% more time
– Overhage 2001 – 6.2% overall p=NS. Experienced users neutral.
• Drug monitoring and preventive care
– Overage 1997 – 46.3% v 21.9% p<.001
• Lab Orders
– Tierney 1990 – 13-14% fewer tests and charges p<.05
Evidence-based Informatics - 3
• Kuperman, et al. Annals 2003 (continued)
– Medication Errors
• Bates 1998 – 55% fewer non-intercepted serious
med error, p=.01 17% less ADEs p>.2
• Shojania 1998 – 32% fewer Vanco orders p=.04
• Bates 1999 – 81% fewer non-missing dose errors
• Evans 1998 – 86% fewer Antibiotic ADEs, 94%
fewer mismatches, also fewer excess doses
Evidence-based Informatics - 4
• Mekhjian 2002
– Process variables
•
•
•
•
•
Med turn-around time 64% reduction p<.0001
Radiology completion time 43% reduction p<.05
Lab result reporting time 25% reduction p=.001
Un-countersigned orders reduced 34%
Length of stay decreased .2 days in acute hospital,
no change in cancer hospital
The Leapfrog Group
•
•
•
•
•
145 large healthcare purchasers
“Safety Leap” forward in health care quality
Pay (more) for (higher) performance
Consumer education and report cards
First 3 Safety standards
– CPOE
– Evidence-based hospital referral
– ICU physician staffing
The Institute Of Medicine
•
•
•
•
•
•
To Err is Human – 48,000-98,000 lives/year
Exact figure is controversial, but not idea
17-29 Billion in cost per year
Under-use, Overuse and Misuse
System problem NOT “bad apples”
Systems can be designed to make it easier
to do the right thing than the wrong thing
CPOE at UWHC
• Workflow, workflow, workflow
• “Wetware” more important that hardware or
software.
• Rollout schedule – pilot until all bugs out
• No firm schedule for subsequent units
• Swarming support 24/7 for 6 weeks
• Daily status meeting, weekly feedback
“Microscope Effect”
•
•
•
•
•
Nursing sign-off and acknowledgement
Multiple wrist band printing
Nurses feel “monitored”
Height not measured
Verbal order policy ‘misunderstanding’
Inpatient before Outpatient
•
•
•
•
•
•
Error prevention more crucial
Costs higher
Costs accrue to bottom line
Hospital infrastructure ready
Less diverse infrastructure
More experience around the country
What we have found so far
• Positive Issues for
Attendings
– Orders available
quicker
– Able to see all orders
– Able to check orders
offsite
– Able to check if certain
labs ordered instead of
paging houseofficer
• Positive Issues for
Housestaff
–
–
–
–
Speeds up intern work
Charts more available
Less call back
Antibiotics hung
quicker
Challenges
– Some decrease of communication between
physicians and nurses
– Mixed environment (paper/CPOE) unsatisfying
to HUC and RN, they prefer all CPOE!
– CPOE process will force difficult issues to be
addressed before implementation can proceed
– Some errors introduced during transition
– Pharmacy work load has increased
Obstacles expected but not seen
• Response time was not too slow
• Software was not difficult to
understand
• There were not excessive warnings
• Orders were not hard to find
• CPOE does not prevent any order
being written
Cedars-Sinai Medical Center
• 877 Bed hospital in Los Angeles
• 1800 physicians – mostly attendings
• CPOE part of system removed after 4 months due
to “revolt” of 400 physicians
• 2 week pilot in OB in July, 2 weeks per floor
• Human factors, workflow not adequately
considered.
• Software had “functionality issues”
• Inadequate education of physician users
What is Expected of Attendings
• Ask residents about non-formulary medications
written
• Ask residents about DNR/DNI order changes
• Encourage residents to use CPOE as intended and
not “work around it”
• Listen to residents comments and concerns about
CPOE and let me know
• Let me know of your own comments or concerns
Questions and Comments
Bruce Slater 263-8242 - Pager 9015