Project Title

1
Power in Numbers- Six Sigma in Healthcare Premier
s 2006 Annual Breakthroughs Conference June 22,
2006 Presented by Dr. Mark Vaaler Vice
President of Medical Affairs Michelle
McCray Six Sigma Black Belt
2
BayCare is comprised of nine leading
not-for-profit hospitals in the Tampa Bay,
Florida area and a host of other health services.
We have 17,000 employees - we call them "team
members" - who are dedicated to our common
Mission of improving the health of all we serve.
What that means is we are an organization that
feels very strongly about caring for
our community.
3
The simplest way to describe BayCare is this we
are an organization of people taking care of
people. Our core business begins with the single
relationship between our team member and you, the
customer. Therefore, BayCares Quality Model is
built upon a foundational philosophy of Customer
Needs, Process Focus, and Continuous
Improvement.  Guided by these principles, we
are able to establish a Quality Process, a series
of actions that bring about changes and results.

4
What is Six Sigma?
Six Sigma is a methodology that adds tools and
infrastructure to our Quality Improvement
process.  Six Sigma enhances our ability to apply
the BayCare Quality Philosophy and Process to
problems team members face each day.  To realize
continuous improvement in our work processes, we
must evolve our tools and methods for improving
processes.  Six Sigma is the next step in the
development of BayCare's Quality Philosophy and
Process.
5
Six Sigma is
6
All Six Sigma Teams focus on

• Voice of the Customer
• What does the customer truly want and need?
• How can we most efficiently meet that need?

7
What is different about Six Sigma from
initiatives of the past?

• Six Sigma builds an infrastructure with lines of
  accountability
• running throughout the organization
• Stresses breakthrough improvement!
• Emphasis is placed on producing better, faster,
  and lower cost
• products and services than the competition
• Emphasis is placed on the use of valid data as a
  driver for
• process change and measurable bottom line
  results

8

• Work-Out
• Fast-paced, activity-driven
• workshop to solve problems
• Examples
• Standardization of forms
• Call light response

• Lean
• Projects that eliminate
• waste and increase efficiency
• Examples
• Patient registration
• Emergency Room flow

Types of Six Sigma Teams or Projects

• DMAIC (Define, Measure, Analyze, Improve,
  Control)
• Led by Black Belts or Green Belts
• Team Members have the opportunity to earn Yellow
  Belt certification
• A data driven in-depth project to eliminate
  errors or defects
• Examples
• Reduce length of stay
• Reduce insurance denials

9
What is D.M.A.I.C.?

• DMAIC (Define, Measure, Analyze, Improve,
  Control)
• Led by Black Belts or Green Belts
• It is a structured, disciplined, rigorous
  approach to process improvement consisting of the
  five phases mentioned, where each phase is linked
  logically to the previous phase as well as to the
  next phase.
• A data driven in-depth project methodology to
  eliminate errors or defects
• Examples
• Reduce length of stay
• Reduce insurance denials
• Tools
• Range from SIPOC to complex statistical tools,
  such as Chi Square hypothesis tests

10
Six Sigma the D.M.A.I.C. Methodology

• Practical problem

• Statistical problem

• Statistical solution

• Practical solution

11
Black Belt Project

• ICU Throughput
• St. Josephs Hospital
• Tampa, FL
• Project Start March 2005
• Close of Project December 2005
• Problem Statement
• For the last 3 years, the availability of ICU
  beds in the Adult Medical/Surgical ICU has become
  such a problem that patients wait up to 72 hours
  for a critical care bed based on data reviewed
  from ICU, PACU ER reports.
• .

12
Roles Responsibilities
Dr. Mark Vaaler, VP Medical Affairs Michelle
McCray Margie Butler, RN Director ICU Judy
Paltoo Barbara Pricher, RN Manager ICU Vickie
Miranda, RN Manager Admitting Anne MacMillan, RN
Manager PACU Lynn Dopp, RN Director
Nursing Adrienne Galluppo, RN ICU Kelli
Stephanko, RN Case Manager Toni Bush, Information
Systems Shellia Keller, Environmental
Services Dr. Lee Kirkman, Medical Director ICU

• Champion
• Black Belt
• Process Owner
• Finance Rep
• Team Members

13
SIPOC/Macro Map
Suppliers
Inputs
Process
Customers
Outputs

• Printers
• PCs
• Telephone
• Housekeeping
• Invision
• Physicians
• Nursing
• Bed Briefing
• Unit Clerks
• Bed Tracker

• See Below

• Bed Control
• Nursing units
• Housekeeping
• EC
• PACU
• Physicians
• Direct Admits/ Transfers In
• Interventional RadiologyDietary

• Transfer order (Dr. Order)
• Bed Request/Invision info
• Bed Tracker info
• Patients
• Clinical Info

• Transfer order
• Bed request
• Bed assignment
• Clean request
• Transported pt.
• Open ICU bed

Step 1 Physician writes order
Step 2 Bed requested
Step 3 Bed assigned
Step 4 Bed ready
Step 5 Patient transferred
Step 6 ICU room cleaned
Step 7 ICU bed is available
14
Project Definition

• Defect Definition
• Time from when ICU requests a bed to when the
  patient is on the receiving units census is
  greater than 4 hours.
• Objective Statement
• To improve the cycle time from when the ICU
  requests a transfer bed to when a patient is on
  the receiving units census from 9.52 to 4 hours
  in 6 months.
• Metrics
• Business
• Avoidable days (hours), ICU LOS
• Primary
• Turn-around-time of bed request to patient
  transferred
• S.M.A.R.T. Specific-Measurable, Attainable,
  Relevant, Timebound

15
Process Capability
Sampled 59 patients over 2 weeks with an average
of 9.5 hours
Sigma Level
Customer Need/Target 4 hours
Defects per million opportunities
Translation 75 defect rate!
16
Pour in all possible input variables
Measure (all input variables)

• Process Mapping
• Mind Mapping
• Ishakawa diagrams
• Survey design

30 - 50
10 - 20
Use soft tools to narrow the possibilities

• C E Matrix FMEA

• UUse quantitative tools to further narrow the
  field

• ANOVA
• Correlation
• Multi-vari studies

8 - 10
Analyze

• Logistic Regression
• Survey analysis

(Key input variables)
4 - 8
Implement and validate solutions
Improve

• DOEs
• Surveys

(Critical input variables)
2-5

• IImplement systems to ensure improvements are
  maintained

Control

• SPC
• Poka-Yoke

• Audits
• Control Plans

• (Key leverage variables)

17
Data Collection Plan
Details of completed deliverables available in
Appendix A
18
Analyze Phase - Summary of Critical Xs
What was driving the turn around time to be long?
Y f(x)

• Significant Xs
• Delay in giving bed assignment (Bed request to
  bed assign)
• Patients moved at shift change and by shift
• Delays in transferring patients by certain units
• Delay in moving patient to receiving units
  census
• Delay by day of week bed is requested
• Other The lack of ability to properly measure
  the turn around time
• See Appendix B for details on hypothesis testing

19
Improve Phase - Ideas
20
Improve - Recommendations
Short Term, Little to No Investment

• Change afternoon Bed Briefing meeting time from
  400pm to 200 pm
• Bed Request/Bed Assigned time captured in
  Invision to improve measurement system
• Streamline Invision process for transferring
  patient. Reduce of screens, reeducate staff.
• Look alike screen to print in ICU when bed
  assignment is made to aid in transfer
• Inform ICU that bed is ready when bed is In
  Progress instead of waiting until Clean
  status.
• Transfer Order set
• Simplify process of rewriting physicians orders
  prior to transfer.
• Transfer Med List will print from Invision

21
Improve - Recommendations
Short Term, Little to No Investment

• Create Prioritization Matrix for placing patients
• Review Patient Placement Guidelines and review
  policy for transferring patients from room to
  room within same level of care.
• Add Bed Tracker screens in key areas to improve
  communications.
• Anticipate next-day transfers from ICU and
  communicate bed needs to Patient Access.
• Create Kanban system for isolation room drapes
• Put ladders on every nursing unit for changing
  the isolation drapes
• Allow patients to be moved prior to curtains
  being re-hung

22
Improve - Recommendations
Short Term, Little to No Investment

• Create incentive program for housekeepers to
  improve status accuracy of Bed Tracking system
• Train housekeepers to clean entire room. Look
  into educational ladder for housekeeping.
• Redeployed housekeepers to cover the hours of day
  when the discharge volumes are high (no change in
  FTEs)

23
Control - Sustain the Gains

• Process Control System
• Control Charts
• Daily monitoring of the critical inputs
• Bed Request to Bed Assign
• Bed Assign to Transfer
• Daily monitoring of the key output
• ICU Transfer time
• Accountability for process
• Identify who is responsible for monitoring
• See Appendix C for completed tool

24
Project Transition

• Project Transition Action Plan (PTAP)
• Formal meeting to transition project from Black
  Belt to Process Owner
• Any open action items are noted
• Deployment Leader, Champion, Finance
  representative must sign off on PTAP
• 12 Month Realization Phase
• Monitor data via the Process Control System
• Monitor financial impact
• Report the primary metric on a regular basis

25
Key Project Results

• Primary metric improvement
• Transfer time reduction
• 9.52 to 4.6 hours (as of 6/7/06)
• Sigma Level DMPO
• 0.82 to 1.29 Sigma (ST)
• 750,553 to 583,627 DPMO
• Financial Savings
• 670,084 Net Contribution margin and Direct
  Variable cost savings (8/15/05 through 4/30/06)

26
Key Project Results

• Effect on Secondary Metrics
• Comparing Pre project to Post project data in
  2005, PACU Holds for ICU beds have decreased by
  36
• ED holds continue to decrease (see next slide)
• Data Accuracy
• Changes to measurement system in order to capture
  accurate and timely data
• Transfer request order
• Bed request order

27
Key Project Results
Implemented Process Control System
Manual tracking during pilots
28
Key Project Results

• Average Emergency Department hold time for ICU
  beds decreased significantly in 2005 (average 105
  patients per month)

29
Final Thoughts

• Next Steps
• Continue to monitor via PCS and HOLD the GAINS!
• Transfer knowledge throughout BayCare
• Lessons learned
• Executive support for large scale projects is key
  to success
• Validating improvement pilots with data is new to
  staff ensure they understand that items my be
  rejected post-pilot.
• Validate your data sources Make sure everyone
  is measuring the same thing, the same way
• The Value of DMAIC
• Focus improvements on what you know will fix the
  problem
• Brings results that are sustainable for the long
  run

30
Questions?

• For additional information or further information
  on this project or Six Sigma at the BayCare
  Health System
• Dr. Mark Vaaler
• mark.vaaler_at_baycare.org
• 813.870.4000
• Michelle McCray
• michelle.mccray_at_baycare.org
• 727.519.1794

31
Appendix A

• Fishbone Diagram
• Detailed Process Map
• C E Matrix
• FMEA
• Input Verification Matrix

If you can not open these documents, you can
contact Michelle McCray at michelle.mccray_at_baycare
.org
32
Appendix B

• Appendix B
• Analyze Phase
• Factor Reduction

33
Factor Reduction Appendix B

• Findings
• Variation of median by Day of week bed requested
  (Moods-Median p0.36)
• Wednesday (5.5), Thursday (4.95) Sunday (7.15)
  have highest median TAT

34
Factor Reduction Appendix B

• Findings
• Difference in median by shift (Moods-Median p
  0.60)
• Median 2nd shift is 5.7, while 1st shift is 2.8
• Also, Time of day had similar results

35
Factor Reduction Appendix B

• Findings
• Bed Assigned to Patient moved interval
• Difference in median by Transfer Unit
  (Moods-Median p 0.159)

36
Factor Reduction Appendix B

• Findings
• Patients moved 2 hours before or after shift
  change have a longer overall TAT
• 22 out of 59 patients in sample were moved in
  this 4 hr window
• Median 1 hour greater for these patients (p
  0.087)

37
Factor Reduction Appendix B

• Findings
• While nurse staffing did not seem to be an issue,
  approx ½ of the patients required telemetry to be
  transported, which requires an RNs support
• Over a quarter of the patients request/need
  private rooms
• Housekeeping delays are an issue, but the extent
  of the problem can not truly be known until the
  Bed Tracker system utilization is under control

Would this be higher/lower if tracker was being
utilized correctly?
38
Appendix C Process Control System