Get documents about "Emergency Department Project
Document Sample
AHRQ State and Regional
Demonstration Project Evaluation:
Barbeque,
Blues,
Beneficial Technology
Kevin B. Johnson, MD, MS
Associate Professor, Biomedical Informatics,
Vanderbilt University Medical Center
Nashville, Tennessee
Project Overview
2
Project Drivers
Incomplete information increases
admission rate and ED LOS
Poor communication impacts ED efficiency
Less patient data at the point of care
impacts the rate of test ordering
Less patient data at the point of care
impacts clinical outcomes
3
Data Exchange Has HUGE
Potential ROI
If data is exchanged across all facilities within the three-county region,
the overall savings has potential to reach $48.1 million.
Dollar
Savings
Financial Measures (millions)
Reduced inpatient $5.6
hospitalization
ED communication distribution $0.1
Reduced IP days due to missing $0.1
Group B strep tests
Decrease in # of duplicate $9.0
radiology tests
Decrease in # of duplicate lab $3.8
tests
Lower emergency department $5.5
expenditures
Total Benefit $24.2
Notes:
1 – Core healthcare entities include: Baptist Memphis, Le Bonheur Children’s Hospital, Methodist University Hospital, The
Regional Medical Center (The MED), Saint Francis Hospital, St. Jude Children’s Research Hospital, Shelby County/Health
Loop, UTMG, LabCorp, Memphis Managed Care-TLC, Omnicare 4
System Implementation and Evaluation
Learn,
Get the Build the
Collaborate, Implement Outcomes
Model Team
Design Research
right ID the
settings
Qualitative Research Qualitative Research
5
Key Aspects of Value Proposition
Qualitative Information
Costs
System usability
System use and utility
Clinical value (patient outcomes)
Dollars saved in care delivery process
Workflow efficiency gains
6
Qualitative Questions
Usability (focus groups in ED)
1 month and 1 year after go-live
Barriers to implementing
infrastructure (cognitive artifacts)
Evaluated in year 4
Drivers for adoption (interviews of
governing board and ED staff)
Evaluated in year 5
7
Costs
Personnel
Training
Community Meetings
Sales
Legal agreements
Organizational development
Equipment
Software development
Site-specific customizations and costs
8
Assessing Usability:
Questionnaire for User Interface Satisfaction
The Questionnaire for User Interaction Satisfaction (QUIS) is a tool
developed by a multi-disciplinary team of researchers in the Human-
Computer Interaction Lab (HCIL) at the University of Maryland at
College Park. The QUIS was designed to assess users' subjective
satisfaction with specific aspects of the human-computer interface.
The QUIS team successfully addressed the reliability and validity
problems found in other satisfaction measures, creating a measure
that is highly reliable across many types of interfaces.
QUIS Details
Six scales
Eleven interface factors
Screen
Terminology/system feedback
learning factors
system capabilities
technical manuals
internet access
on-line tutorials, multimedia, voice recognition,
virtual environments, and software installation
10
11
System Usability
Will conduct usability testing of SPL
Vanderbilt as pilot site for face validity
and modifying QUIS
Will modify accordingly
Will survey Memphis ED attendings
and nursing staff 1 month after go
live and again 6 months later
12
System Usage and Epidemiology
Help desk use
Provider enrollment
Patient enrollment (RHIO in versus
RHIO out)
Usage statistics
Latency
Downtime
13
Content Quality
Accuracy
Missing data
Categorization errors
14
Disease-specific Hypotheses
Improved neonatal GBBS
management
Improved asthma controller med use
Improved ACE/ARB use in CHF
Improved immunization rates (flu,
s.pneumo)
?Others
15
ED Administrative Outcomes
Reduce inpatient admissions
Decreased duplicate testing
(radiology and lab)
Decreased ED Expenses
Workflow efficiency
Costs per visit
16
Workflow change
Activity-based costing
Model construction at Vanderbilt
Model validation in Memphis
Use model to construct activity
matrices in EDs under study
Assess how activity matrices change
pre and 1 year post implementation
17
Model Construction: Data Collection
Trained observers will document
Key transition points in information flow:
Eliciting prior medical history
Triage and treatment processes
Disposition/discharge from ED
Data Elements
Activity performed
Agent (RN, MD, Clerk, etc.)
Start-Stop times (hh:mm:ss)
18
Sample of Activity-Based Data
Date Bed Caregiver Start Time Activity Observed (Raw Data) End Time Elapsed Time Elapsed Seconds
7/14/2005 1 A 14:09:35 Look up drug info in Nursing Drug Reference Book 14:10:32 0:00:57 57
7/14/2005 1 A 14:10:32 Decide to call pharmacy re: Rx; check Rx in paper chart 14:11:25 0:00:53 53
7/14/2005 1 A 14:11:25 Enter patient room 14:11:39 0:00:14 14
7/14/2005 1 A 14:11:39 Documentation re: Rx, vital signs 14:12:18 0:00:39 39
7/14/2005 1 A 14:12:18 Check catheter 14:13:11 0:00:53 53
7/14/2005 1 A 14:13:11 Ask pt what he wants to drink to take meds, give options 14:14:35 0:01:24 84
7/14/2005 1 A 14:14:35 Orient pt to place (where are you?); tell pt the answer 14:15:33 0:00:58 58
7/14/2005 1 A 14:15:33 Orient pt to place (where are you?); tell pt the answer 14:15:48 0:00:15 15
7/14/2005 1 A 14:15:48 Ask pt what kind of juice he wants (no answer) 14:16:15 0:00:27 27
7/14/2005 1 A 14:16:15 Exit room to kitchen to get juice for pt to take meds 14:16:28 0:00:13 13
7/14/2005 1 A 14:16:28 Enter patient room & exit again 14:16:44 0:00:16 16
7/14/2005 1 A 14:16:44 Enter patient room w/ straw for drink 14:16:54 0:00:10 10
7/14/2005 1 A 14:16:54 Talk to pt re: taking his pills 14:17:11 0:00:17 17
7/14/2005 1 A 14:17:11 Put pills in cup; tell pt what pills are & what they're for 14:17:43 0:00:32 32
7/14/2005 1 A 14:17:43 Administer meds (pills & drink to patient) 14:18:39 0:00:56 56
7/14/2005 1 A 14:18:39 Juice & pills done 14:18:55 0:00:16 16
7/14/2005 1 A 14:18:55 Administer meds (heparin shot in belly); talk to pt 14:19:41 0:00:46 46
7/14/2005 1 A 14:19:41 Orient pt to time (ask if he knows time of year, president) 14:20:20 0:00:39 39
7/14/2005 1 A 14:20:20 Done w/ heparin 14:20:37 0:00:17 17
7/14/2005 1 A 14:20:37 Discard medication wrappers & syringe 14:20:46 0:00:09 9
7/14/2005 1 A 14:20:46 Rearrange bedding, talk to patient 14:21:36 0:00:50 50
7/14/2005 1 A 14:21:36 Ask pt to scoot up in bed; talk to patient 14:22:35 0:00:59 59
7/14/2005 1 A 14:22:35 Exit room to get help to reposition patient in bed 14:22:50 0:00:15 15
7/14/2005 1 A 14:22:50 Enter patient room, talk to patient 14:24:18 0:01:28 88
7/14/2005 1 A 14:24:18 Talk to pt, perform ROM on upper extremities (w/ 2nd nurse) 14:25:59 0:01:41 101
7/14/2005 1 A 14:25:59 Reposition pt (w/ 3rd nurse, using sheet), attach restraints 14:26:45 0:00:46 46
7/14/2005 1 A 14:26:45 Pt repositioned, restraints in place (4 pt) 14:27:26 0:00:41 41
7/14/2005 1 A 14:27:26 Reattach posey, talk to pt, orient to place (3rd RN out) 14:28:02 0:00:36 36
7/14/2005 1 A 14:28:02 Talk to pt, explain need for restraints (2nd RN out) 14:28:45 0:00:43 43
7/14/2005 1 A 14:28:45 Check vital signs (BP), talk to pt re: school, course of study 14:29:15 0:00:30 30
7/14/2005 1 A 14:29:15 Fix monitor, talk to pt, document on flow sheet 14:30:00 0:00:45 45
7/14/2005 1 A 14:30:00 Talk to pt, explain need for RN to leave, provide reassurance 14:31:22 0:01:22 82
7/14/2005 1 A 14:31:22 Turn on TV, ask for preference re: channel 14:32:08 0:00:46 46
19
Activity-Based Estimates (Aggregate)
Average of Elapsed Seconds
Activity Observed (Aggregate) Total
Assess patient 30
Assist patient 44
Cleaning / Hygiene 9
Communication 46
Documentation 46
Enter / exit location 26
Equipment 50
Medication 34
Multi-task 61
Reference / Research 57
Tubes & Lines 53
Grand Total 41
22
Data Sources
Outcome of
interest
Record Accessed No RHIO record
During Study Accessed
Patient with Patient without
Data in vaults Data in vaults
27
Using the Vault as the Primary Data
Source for Outcomes
Baseline LOS
= LOS of all encounters in
vaults (before go live)
LOS of all LOS of all
Change in LOS
= encounters in
vault whose
records were
vs encounters in
vault whose
records were not
accessed accessed
28
Clinical Outcomes Methodology
Pre-post off
Rollout
stable
on
Easy to implement
Will not impact rollout
or clinic flow
Sensitive to existing
trends
30
Other Approaches
Rollout
off
stable
Assign times of day randomly to
downtime status
Assign patients randomly to control
group (no data for them)
Assign retrieval events randomly to
control (i.e., no result) retrievals
31
Covariate Analysis
ED (site) characteristics survey to
be completed by ED Administration
Readiness survey to be completed
by ED administration and clinical
leadership
32
IRB Approach: Five Approvals
Activity-based costing (approved)
Usability, readiness and demographic
survey (letters of cooperation)
Baseline data for administrative
measures and activity costing
System content quality
Disease-specific hypotheses
33
Thanks!
34
