Improving Patient Flow and Reducing ED Crowding: Findings from 10 Hospitals

1
Improving Patient Flow and Reducing ED Crowding
Findings from 10 Hospitals July 1, 2004 Bruce
Siegel, MD, MPH
2
Urgent Matters

• A national program of The Robert Wood Johnson
  Foundation
• Housed at The George Washington University
  Medical Center, Washington, DC.
• Support and funding for this webinar is provided
  by the Robert Wood Johnson Foundation

3
The 10 Hospital Learning Network

• Grady Health System Atlanta, GA
• Boston Medical Center Boston, MA
• Henry Ford Hospital Detroit, MI
• Inova Fairfax Hospital Fairfax County, VA
• BryanLGH Medical Center Lincoln, NE
• The Regional Medical Center at Memphis Memphis,
  TN
• St. Josephs Hospital and Medical Center
  Phoenix, AZ
• Elmhurst Hospital Center Queens, NY
• University Health System San Antonio, TX
• University of California at San Diego San
  Diego, CA

4
Urgent Matters Input/Throughput/Output Model
5
Patient Flow Improvement Principles

• Executive Management Buy-In
• This is a hospital, not ED, problem
• Inpatient staff must be involved
• Metrics and data collection are critical
• Cant fix a problem we cant measure
• Rapid cycle change is an effective vehicle
• Test on a small scale
• Collect data over time
• Build knowledge sequentially with each cycle
• Dont Forget Implementation
• Transparency
• Communicate results and spread success

6
Learning Network Elements

• Tool kit of best practices on patient flow
• Data collection methods
• Strategies for work flow re-design
• Organizational buy-in strategies
• Implementation strategies
• Training and Reinforcement on Rapid Cycle Change
• Plan, Do, Study, Act
• Core Metrics
• 17 key performance indicators (weekly and monthly
  reporting)

7
Factor Key Performance Indicator Key Performance Indicator Reporting Interval
I. Inpatient Flow 1. Time from inpatient bed assignment to bed placement 1. Time from inpatient bed assignment to bed placement Weekly
I. Inpatient Flow 2. Time of day of discharge 2. Time of day of discharge Weekly
I. Inpatient Flow 3. Bed turnaround time 3. Bed turnaround time Weekly
II. ED Throughput 1. Total ED throughput time 1. Total ED throughput time Weekly
II. ED Throughput 2. By treatment pathAdmitted,Fast Track, other ED discharged a. Time from arrival to bed placement Weekly
II. ED Throughput 2. By treatment pathAdmitted,Fast Track, other ED discharged b. Time from bed placement to examination Weekly
II. ED Throughput 2. By treatment pathAdmitted,Fast Track, other ED discharged c. Time from disposition decision to departure Weekly
III. Other ED 1. Hours on diversion 1. Hours on diversion Monthly
2. Percent incomplete treatment 2. Percent incomplete treatment Monthly
3. Patient Satisfaction 3. Patient Satisfaction Monthly
IV. Clinical Process 1a. Time to heart treatment 1a. Time to heart treatment Monthly
IV. Clinical Process 1b. Time to pain management 1b. Time to pain management Monthly
8
Learning Network Structure

• Web-based management
• Tool kit, new developments and other resources
• Action plans and data submission
• Run-charts
• Monthly project reports
• Celebrating Success
• Shared data shared outcomes
• Peer Collaboration
• 3 group meetings
• 3 consultant site visits to each site, with phone
  consultation
• Monthly conference calls with topic presentation,
  update from each site
• Listserv

9
(No Transcript)
10
(No Transcript)
11
Sample Strategies

• Boarding and Inpatient Bed Assignment
• Pull System
• Early Discharge
• Initiate preliminary discharge
• Re-design rounding and discharge processes
• Create a discharge room/lounge
• Establish a discharge coordinator
• Implement physician and nurse incentives

12
Sample Strategies

• Patient Flow Coordination and Facilitation
• Implement Bed Czar or patient flow manager
• Dedicate a nurse with admission/discharge/transfer
  duties
• Utilize care teams
• Accelerated triage and registration
• Diversion Management and Reduction
• Establish new protocols and monitoring systems
• Hospital-wide diversion response
• Coordinate a community diversion plan

13
Sample Team Structure

• Chief Operating Officer
• Vice President for Nursing
• Chief Medical Officer
• Chief of Emergency Medicine  

Steering Committee
Emergency Department Team
Inpatient Team

• Chief of Emergency Medicine  
• Director, Emergency Care Services  
• Information Systems Coordinator 
• Clinical Manager  
• Clinical Supervisor  
• Quality Management  
• Clinical Nurse Specialist (2)
• Business Analyst
• Asst. Medical Director

Vice President for Nursing Chief Medical
Officer Nursing Director Nurse Manager Housekeepin
g Supervisor Admitting Director Inpatient
Attending Inpatient Medical Director Director of
Patient Access Services
14
Process Improvement

• What are you trying to accomplish?
• How will you know that a change is an
  improvement?
• What change can we make that will result in an
  improvement?

15
date RCT initiative Responsible Data collection Summary Results KPI Improve
8 6/30 2003 Met with Pharmacy to develop par level and add Td to current EC pyxis. Td was removed some time back because of national shortage. Shortage has improved but Td is expensive and often wasted. Staff educated. Td protocol added to standing orders for EC. Kathy Haddix, triage nurses Manual tracking pre RCT Mon, Tues, Wed post RCT Thurs, Fri, Sat. RCT was successful. Td added to pyxis. Meeting planned with Pharmacy to create additional list of meds to be added to pyxis to reduce LOS for pts and decrease work for Pharmacy. Thru-put Time 69
11 7/21 2003 For this RCT, a registration specialist will be designated to register all pedi and PA Triage patients. During the RCT, nursing staff in triage will be asked to put triaged patient charts in a bin designated for pedi and PA Triage patients. This will ensure that these patient charts will not be included with other EC charts during the RCT. The designated specialist will be continuously monitoring the bin. Registration will be done continuously, without delay due to other charts. Rocio Garcia Pre RCT DataAverage Reg Time for Pedi 15.08 minutes (Random Audit 2003) Average Reg Time for PA Triage patient 23.24 minutes (Random audit 2003) Although we believe change should be beneficial, the data collected from three different sources was discongruent. We will reschedule this RCT for the week of 8/11/03 using a differrent data collection method. Arrival Process N/A
14,11 8/11 2003 See RCT 11 Rocio Garcia See RCT 11 Post RCT data for this project was collected manually. The time stamp indicated the actual time registraion was requested, the actual registration time was taken from IDX. This was in fact a very successful RCT. With a registration specialist dedicated to Fast Track and Pediatric patient there was no delay due to other charts. Post RCT Data for Pediatric Patients was 4.4 minutes and for PA Fast Track was 2.6 minutes. Arrival Process 70 89
University Hospital, San Antonio
16
Rapid Cycle Test Sample

• RCT preliminary (night before) discharge order
• Process and Results
• W1 tested part of the Neuro floor
• control group1822, RCT group1312 5 hour
  difference
• W2 expanded to the entire Neuro floor
• control group1732, RCT group1425 3 hour
  difference
• W3 all Neuro floor and expanded to Medicine
  teams 6 7
• control group1655, RCT group1345 3 hour
  difference
• W4 expanded to all Emory Medical Services
• control group1622, RCT group1525 1 hour
  difference
• Impact
• 57 minute change in the overall average time of
  day of discharge, from 1622 to 1525
• Next Steps
• Improve communication to ensure consistent
  implementation
• Continue to implement house-wide and maintain
  momentum

17

• IMPACT

18
(No Transcript)
19
(No Transcript)
20
(No Transcript)
21
Black linesMaximum and Minimum values, Red
lineMedian values Time Period 1Weeks 1-10,
2weeks 11-20 3weeks 21-30 4weeks 31-40
5weeks 41-50
22
(No Transcript)
23
(No Transcript)
24
(No Transcript)
25
Black linesMaximum and Minimum values, Red
lineMedian values Time Period 1Weeks 1-10,
2weeks 11-20 3weeks 21-30 4weeks 31-40
5weeks 41-50
26
(No Transcript)
27
(No Transcript)
28
(No Transcript)
29
Impact Hospital View

• Small scale trials in a highly controlled area.
  If you try to make monumental changes, it will
  take a monumental amount of time without the data
  to prove that the change is necessary and
  worthwhile. Inova Fairfax Hospital
• The RCT process continues to foster a culture of
  change management and team development.  Clinical
  Nursing Directors have placed suggestion boxes in
  units to encourage suggestions for new RCTs from
  employees. - The MED
• Our numbers are coming reliably on a weekly
  basis for all the KPIs. We have never had better
  data to make management decisions. Boston
  Medical Center
• From the data we are collecting we have learned
  the implications for the varying patterns of
  diversion and what it takes to recover from these
  episodes in terms of patient throughput and
  personnel resources. We will be developing a
  series of RCTs based on these patterns.
  University Hospital, San Antonio

30
Closing Thoughts

• Rapid Cycle Change Rapid Is Key
• Early Results More Important Than Consensus
• Communications are critical for system diffusion
• Data Systems Lag
• But its doable
• Executive Management Buy-In Is Key
• Collaboration? Or Competition?
• Productivity May Increase Dramatically
• Hospitals Can Improve Their Performance
• With minimal new resources
• This is no longer optional JCAHO

31
(No Transcript)
32
(No Transcript)