Improving Healthcare Workshop Brittany Hagedorn
Apr 22, 2015
Improving Healthcare
Workshop
Brittany Hagedorn
SIMUL8 Corporation | SIMUL8.com | [email protected]
Introductions
Brittany Hagedorn is SIMUL8’s new Healthcare Lead for North America.
Brittany’s mission is to promote the use of process simulation and related tools within healthcare. The role will include: 1. Supporting existing users. 2. Publicizing the great work already being done. 3. Fostering growth of the simulation community. 4. Pioneering new applications within healthcare. 5. Developing tools and training.
SIMUL8 Corporation | SIMUL8.com | [email protected]
Clinical Quality and
Patient Safety
Management
Consulting
Introductions My experience has been in project-oriented roles, first as a Six Sigma Black
Belt within a hospital system, then as an external consultant. Through these roles, I have had the privilege to work on a wide variety of challenges.
Lean and Six Sigma
(Process Improvement)
My favorite projects include: • Reducing the lead time for pediatric sedated procedures from six weeks to seven days. • Addressing bottlenecks in nursing workflows. • Eliminating 70% of duplicative “double checks” for physician documentation. • Constructing a clinical quality scorecard that could be easily managed and integrated into
executive compensation. • Developing a primary care compensation plan for 150+ physicians to incentivize their
transition toward a value-based, accountable clinical care model. • Creating an integration strategy for a newly formed cardiology medical group. • Building a business case for post-acute care services. • Supporting preventable harm interventions.
SIMUL8 Corporation | SIMUL8.com | [email protected]
Agenda
I. Project Overview
II. Results
III. Recommendations
IV. Discussion & Next Steps
SIMUL8 Corporation | SIMUL8.com | [email protected]
Project Overview – Goals
A local hospital was constructing a new bed tower. They wanted to know often they would need a medical/surgical bed for post-surgical observation patients.
We recommended a simulation.
The executive team’s request was for an Excel analysis that would produce: • An average number
of patients. • An average number
of beds.
After discussions, we recommended a project charter for a simulation that would produce: • The range for the
expected number of beds.
• Identification of any downstream effects.
SIMUL8 Corporation | SIMUL8.com | [email protected]
Project Overview – Process
The process to be modeled was fairly simple, with a few routing decisions. Each step had a variable time duration, which included both random
variation and patient-specific factors such as specialty and acuity.
Inpatients
Outpatients
Add-ons
Pre-Surgery Prep
Surgery Post-
Surgery Recovery
Home
Observation
Return to Unit
Entry Points Post-Surgical Routing
SIMUL8 Corporation | SIMUL8.com | [email protected]
Project Overview – Model Building This process translated into a SIMUL8 model quickly, but there was some additional work to build the OR schedule into the simulation.
Entry Points
Resources
Post-surgical routing
SIMUL8 Corporation | SIMUL8.com | [email protected]
Project Overview – Excel Interface
By utilizing a unique identifier for each patient entering the simulation, we obtained individual-level data and results that were like-real-life.
Patient MRN
Characteristics Scheduled Actual Time Stamps
SIMUL8 Corporation | SIMUL8.com | [email protected]
-
0.50
1.00
1.50
2.00
2.50
3.00
Monday Tuesday Wednesday Thursday Friday
Pat
ien
ts p
er D
ay
Results – Patient Volumes
The model assumed a continuation of current policy, which meant that observation patients would remain in the pre/post surgical suite until discharged
or the end of the day. At the end of the day, all remaining patients were transferred to an inpatient unit, which results in longer stays and increased costs.
Observation Patients to Floor per Day
• With current policies, there would be fewer than two patients per day needing placement at the end of the day.
• As a result, additional inpatient beds dedicated to observation patients would not be needed.
Note: The variability by day of the week was due to the surgeon specialty mix.
Excel analysis resulted in 1.3 beds per day, without insight into daily variation or downstream effects.
SIMUL8 Corporation | SIMUL8.com | [email protected]
Results – Unexpected Findings
However, by using a simulation, we were able to capture additional performance metrics, which suggested that there may be other potential issues.
FY 2013 FY 2018
Maximum Schedule
Annual Patient Volume
14,000 15,000 16,000
Days with Delayed
Surgeries 67% 77% 82%
Number of Delayed
Surgeries 6 daily 9 daily 10 daily
Number of Observation
Patients to Floor 1.3 daily 1.5 daily 2.1 daily
Additional Performance Metrics
• The simulation queues showed that many patients were seeing delayed surgery starts.
• With current state processes and policies, this would happen on over 65% of days.
• When delays did occur, it would affect on average 6 patients per day.
In addition, the frequency and duration of delays will increase if the growth target is reached.
SIMUL8 Corporation | SIMUL8.com | [email protected]
Results – Operational Implications
Delayed surgeries are caused by a bed shortage, which prevents patients from being prepped for their procedure on time. This directly affect profitability, either in foregone revenue or increased staffing costs.
0
5
10
15
20
25
30
35
40
45
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Nu
mb
er
in U
se
Hour of the Day
Note: The second surge in O.R. volumes depicts delayed patients finally getting through pre-op into surgery.
Observation patients remain in Pre/Post Unit
Not enough bed capacity for arriving patients
Delayed prep causes delayed surgery start times
Patients are cancelled or staff must work overtime
Example Day – Effect of Bed Shortage
Pre/Post Beds
O.R. Rooms
Maximum Bed Capacity
SIMUL8 Corporation | SIMUL8.com | [email protected]
Recommendations – Alternatives
Given this information, the natural question is – how do we fix it?
There were three alternative solutions that were simulated, in order to measure the real impact that implementation would have. 1. Pre-Admission Testing Rooms – Repurpose the four pre-admission testing rooms that were
adjacent to the pre/post suite. These could be retrofitted before construction was complete as recovery spaces.
2. Family Waiting Policy – The plan for the new unit was to allow patients’ families to remain in their patient’s prep room during the surgery, and return the patient to the same location for recovery.
3. Observation Patient Policy – Modify the policy to indicate that observation patients should be moved to an inpatient unit if they will be staying for longer than a pre-determined threshold.
SIMUL8 Corporation | SIMUL8.com | [email protected]
Recommendations – Voting Results
Please Vote – Which alternative was the most effective?
A. Reclaim 4 pre-admission testing rooms.
B. Ask families to move to the waiting room during surgery.
C. Move observation patients to inpatient beds after surgery.
SIMUL8 Corporation | SIMUL8.com | [email protected]
Recommendations – Best Technical
Modifying the family waiting policy was the most effective at balancing the needs of the inpatient units and operating rooms.
Family Remains in Pre/Post
Room during Surgery
Family Moves to Another
Location during Surgery
% Days with Delays 77% 45%
# of Patients Delayed 10 daily / 2,647 annual 1 daily / 287 annual
# Observation Patients to
Floor 2 daily / 417 annual 0 annual
• The change in policy would minimize the number of delayed cases and eliminate the need for inpatient beds to house observation patients, releasing bed capacity for other uses.
• Additional improvement could be made by modifying the O.R. block schedule to distribute observation patients more evenly throughout the week.
SIMUL8 Corporation | SIMUL8.com | [email protected]
Recommendations – Voting Results
Due to other factors, this alternative was not implemented.
Please Vote – Which was the primary barrier?
A. The solution was too technically complex to implement.
B. We did not have the right executives in the room to be able to make the policy decision.
C. There were other programs being implemented that were perceived to be in conflict.
D. Political divisions created barriers to buy-in.
SIMUL8 Corporation | SIMUL8.com | [email protected]
Recommendations – Trade-Offs
Ultimately, it was institutional concern about Value Based Purchasing (which rewards hospitals for patient satisfaction scores) that drove the
decision to modify the observation patient policy instead.
1.5
3.4
4.3
5.4
6.7
9.3
10.3 77%
45%
32%
20%
13%
5% 2%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
0.0
2.0
4.0
6.0
8.0
10.0
12.0
No Limit 42 hours 40 hours 38 hours 36 hours 30 hours 24 hours
% o
f D
ays w
ith
Dela
ys
Nu
mb
er
of
Pati
en
ts t
o F
loo
r
Policy Cut-Off Point
Daily Obs to Floor % Days with Shortage
The Ultimate Trade-Off
• The trade-off was a decision for the executive team.
• As more observation patients were moved to inpatient units, the number of delays dropped dramatically.
• Ultimately, the policy was modified so that every observation patient was moved to an inpatient unit after surgery.
The other factor to consider is the impact
on E.R. throughput.
SIMUL8 Corporation | SIMUL8.com | [email protected]
Lessons Learned
OVERALL PROJECT • Unexpected findings – On several occasions, the analysis results did not turn
out as expected. Eventually, we discovered that the simulation was operating correctly – but the process was not operating as it had been described.
• Scope creep – The scope of the project grew several times, as we uncovered additional questions that needed to be answered.
• Stakeholder buy-in – Changing policy presents challenges, depending on the stakeholders and their entrenched beliefs. The best technical solution will not always be implemented.
RELATED TO DESIGN • Rules of Thumb – Architecture and construction teams often rely on industry
standards when designing physical spaces, such as “four beds per OR”. But every situation is unique and this approach results in over/under-built spaces.
• Earlier is Better – Simulation is helpful at any stage of the process, but to reduce costs, earlier is always better. If we had completed this analysis a few months earlier, we would not have needed to redo several rounds of architectural plans, which prevented us from considering several alternatives.
A few last thoughts…
SIMUL8 Corporation | SIMUL8.com | [email protected]
Discussion and Questions
“ Great ideas need landing gear as well as wings.” – C. D. Jackson
SIMUL8 Corporation | SIMUL8.com | [email protected]
Next Steps
If you enjoyed today’s discussion, please join us in
September for the next workshop!
Are you facing complex processes and an overwhelming
amount of work to do? Suggest a future topic!
Join the simulation community by connecting with us on
LinkedIn, Twitter, or on our website at SIMUL8.com!
SIMUL8 Corporation | SIMUL8.com | [email protected]
Appendix – Additional Analysis
SIMUL8 Corporation | SIMUL8.com | [email protected]
Alternative 1 Impacts
77%
70% 65%
58%
49%
43%
35%
29%
22% 17%
12% 7% 6% 4% 3% 2% 0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55
% D
ays
wit
h S
ho
rta
ge
Number of Pre/Post Beds
SIMUL8 Corporation | SIMUL8.com | [email protected]
Patient Delay Durations
4
18
5 8
15
26
44
30 32 33
0
5
10
15
20
25
30
35
40
45
50
0
5
10
15
20
25
Monday Tuesday Wednesday Thursday Friday
Ave
rag
e T
ime
in
Qu
eu
e (
min
)
Pa
tie
nts
th
at
Wa
ite
d f
or
Pre
p
SIMUL8 Corporation | SIMUL8.com | [email protected]
Block Time Utilization
66% 70%
73% 75% 76% 80%
87%
-
1.00
2.00
3.00
4.00
5.00
6.00
7.00
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
CVS Other Uro Gyn Gen NOS ENT
Ra
tio
Blo
ck
vs
. A
ve
rag
e D
ura
tio
n
Blo
ck
Tim
e U
tili
za
tio
n
SIMUL8 Corporation | SIMUL8.com | [email protected]
Block Time Utilization
77% 77% 74% 69%
76%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
Uti
lizati
on
Min
Avg
Max
66% 70%
73% 75% 76% 80%
87%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
CVS Other Uro Gyn Gen NOS ENT
Uti
lizati
on
Min
Avg
Max