Transcript Palm-Based Extensions to the New York Presbyterian

Palm-Based Extensions to the
New York Presbyterian Hospital
Clinical Information System:
PalmCIS
James J. Cimino, M.D.
Department of Medical Informatics
Columbia University College of
Physicians and Surgeons
Questions:
• How can hand-held wireless technologies be
used to reduce medical errors?
• How can we study the impact of such
interventions?
Medical Errors
• Slips vs. mistakes
• Proximal causes of mistakes:
– inadequate information during decision-making
– ineffective communication for coordination of
patient care team members
Preliminary Studies
• Studied nurses and physicians
• Three modalities
– Survey questionnaires
– Focus groups
– Direct observation
Physician Surveys
• Patient-specific information
–
–
–
–
–
A list of current medications and time administered
Problem lists
Outpatient notes (esp. subspecialty consultations)
List of current providers for patient (consults, nurses)
Laboratory and other test results
• Institution-specific information
– Current providers on-call and how to contact them.
• Domain-specific information
– Disease management information
– Prescribing information
– Medical formulas linked to patient data
• Communication difficulties
– Identifying/contacting other providers (esp. consults)
Nursing Surveys
• Patient-specific information
– Patient diagnoses
– Laboratory and other test results
• Institution-specific information
– Policies and protocols (IV care, blood bank)
– Census reports
• Domain-specific information
– Drug information (dosage, teaching info)
– Diagnostic definitions
– Educational materials (e.g. colostomy care)
• Communication difficulties
– Identifying and contacting other care providers.
Observation of Information Needs
• Don’t Know
HS1: …medically he’s fine. Actually, we’re doing a follow-up
ESR today.
A2: Good.
HS1: Um, I don’t know what we’re going to do if it’s still high.
Because we’re still doing the standard treatment anyway.
• Type: Medical Science - Use: Resource- Outcome: Pending
HS1: …actually, we discussed putting him on sub-q heparin.
And, I didn’t, I don’t…
HS2: Yeah, I’ll have to, go and get a paper on this, because this
actually seems to be a discussion often. But, you know, is
there an existence of this self-anticoagulation…
Observation of Information Needs
(cont.)
• Type: Administrative - Use: Human - Outcome: Success
(A2 tells story of patient suing hospital)
A2: So then, a few years later, we were in court, and it was,
because it was (hospital name), they sued (hospital name),
and (another hospital name), the attending, and all the
housestaff. Any housestaff that had written a note. So all the
(2nd hospital name) housestaff who were rotating through…
HS1: Are, um, are we protected?
A2: Yes. So, actually, yes, you are protected, that’s right.
Observation of Information Needs
(cont.)
• Type: Patient - Use: Record - Outcome: Failure
A1: How about his chest x-ray?
MS1: His chest x-ray showed… it showed an enlarged heart,
but no obvious signs of CHF.
A1: And no pleural effusions?
HS1: No…
HS2: I’m not sure what the timing of the chest x-ray was, in
relation to the Lasix.
HS1: See, yeah that may have been after the Lasix. Usually it
doesn’t clear up that quickly.
Observation of Coordination
• Don’t Know
HS2: We were discussing whether she should really go to the
Neuro service, because it’s, it’s sort of, they were asking us
for q4 hour neuro tests, which is not something we’re really
set up to do here. And the nurses aren’t really prepared to
do it, either. But, I don’t know.
• Pending Contact[re: Patient with liver failure]
HS1: It looks like this is going to be a pretty rapid course.
You know, in the next few months.
HS2: So, talks, you know, discussion, surrounding what he
feels about how this should go, should be entered into,
obviously. But first is the breaking of the bad news, that…
HS3: Well, yeah, we should talk to GI and Oncology, and see
what they’re going to do. And see what they want.
Observation of Coordination (cont.)
• Pending Outcome
A1: Is he going to get an upper GI series?
HS3: Uh, it depends what, um, it depends what the Surgery
people want to do.
• Success
HS1: …we treated him for diverticulitis, starting Friday
night. We gave him iv-[unintelligible] and fluids. And he’s
doing, he did spike in the ER, he had like one, a hundred
and one, um, but he’s been doing great. He’s fine. His
abdominal exam is totally benign, he’s been afebrile, um,
we talked to GI, we switched him to PO antibiotics last
night, and in fact he should be able to go later today.
According to GI, and then follow up with them.
Observation of Coordination (cont.)
• Failure
HS1: As far as the abscess, Ortho was not too impressed with
his clinical exam, or with the MRI films, really. So they
said, you know, you can call IR, and see if they want to do
any further drainage. He’s been afebrile since the day we
drained it, so I think that that’s a good thing. Um, so in fact,
I did call IR yesterday, but being a holiday, the people who
do the actual procedures weren’t in. But I talked to them
and had them review the films, supposedly, so I called back
today, and they’re really reluctant to; they want a follow-up
study. They want a follow-up Doppler. But. But, when we
drained it, it was by Doppler. So we have a film of us
draining it, and what it looks like afterwards. So, I called
radiology to set him up for a new Doppler. They are like,
“No, that’s crazy, cuz IR’s gonna need to do a Doppler
anyway, when they drain it.” So they don’t want to do it.
Improving Access to Information
• Study information needs
• Anticipate information needs in context of
clinical information system
• Build "infobuttons" to respond to
anticipated needs
Improving Communication
• Synchronous vs. asynchronous communication
• Identify common communication tasks
• Build "virtual whiteboard" to manage care team
tasks
Virtual Whiteboard
Patient
Intern
John Smith
b5151
Resident
Mary Jones
b1234
Sandiego,
Carmen
Check IV:
Will do after
rounds
Talk to GI consult
Billings,
John S.
Paroo,
Marian
Attending
William Osler
b4321
Nurse
Jan Gomez
X4567
Discuss DNR with
patient:
DONE
Call intern when
changing dressing
Extension to Wireless Palm
• Access at point of care
– rounds
– bedside
• Local information resources
Technical Challenges
•
•
•
•
•
Platform
Integration with CIS
User interface
Security concerns
Evaluation
Platform
• Off-the-shelf technology
• Web client
• Secondary features (beeper? phone?)
New York Presbyterian Hospital
Clinical Information Systems Architecture
Medical Logic
Modules
Clinical Database
Alerts & Reminders
Database Monitor
Results Review
Database
Interface
Medical Entities
Dictionary
Administrative
Research
Reformatter
...
Radiology
Reformatter
Discharge
Summaries
Reformatter
Laboratory
...
Security Issues
• Authentication
– ID, password, device ID
• Confidentiality
– SSL
– no caching on device
Evaluation
• Triangulation on information needs and
communication tasks
– questionnaires
– focus groups
– direct observation
• Nurses and physicians
• Phased roll-out of PalmCIS, with control group
• Repeat three-pronged studies
Next Steps
•
•
•
•
Modify WebCIS for Palm-based browsing
Adapt patient list to handle whiteboard functions
Develop infobutton manager
Observations to determine specific whiteboard
functions and infobuttons
• Seek funding to equip housestaff and carry out
clinical trial