Você está na página 1de 51

OIT 262: Operations

Class 7: Process Improvement in Health Care


Yonatan Gur
Graduate School of Business
Stanford University

Plan for Today


Reminder: Littlefield begins Wed. 4/29, 1:00pm
- Register your teams (username, password)
- Plan on meeting before 4/29 to strategize
- Inventory Management review session on 4/29

I.

Process Flow at Mass General

II. Inventory Build-up at Bottleneck


III. Waiting Times at Non-Bottleneck Steps
IV. Recommendations
V. MGH Guest: Cecilia Zenteno
Stanford GSB

OIT 262 Prof. Gur Class 7

Goals for the MGH Case


Capstone Case in
Process Analysis Segment
How can operations help deliver
health care/services more
efficiently?
How can process and queueing
analysis help inform managerial
decisions and strategies?

Stanford GSB

OIT 262 Prof. Gur Class 7

Massachusetts General Hospital


Mass General Hospital
Founded 1811
950 beds, 4.6 million square feet
#2 hospital in US (2014-2015, US News)
consistently 1st or 2nd in recent years
Anesthesia Department
Birthplace of anesthesia
278 physicians, 198 nurses
Supports patients before, during, and after
surgery
Pre Admission Testing Area (PATA)
Responsible for outpatients (43% of surgeries)
Evaluate safety of anesthesia before surgery
Inform patient
Obtain legal acknowledgement and consent
Stanford GSB

OIT 262 Prof. Gur Class 7

PATA Problems
Dr. Peter Slavin
President, Massachusetts General Hospital
Dear Dr. Slavin,

The
very nice and
apologetic,
butinofVermont,
the almost
Last week I brought
myproviders
mother intowere
the Pre-Admission
Testing
Area. We live
almost 3
hours away, and4had
to make
specialintrip
forclinic,
this appointment,
which her oncologist, Dr. Paul
hours
we aspent
the
only 1 hours
Schneider, said was
necessary
to ensureface
a safe
and with
successful
surgery.
of that
was actually
time
anyone!
When we arrived at the clinic, the waiting room was so full, it was five minutes before my mother and I
Average patient visit length: 3h15 min
could get two seats together. We sat there for a full half-hour before they sent us back to get her blood
Face
timewaited
with back
providers:
1h20room
min for another 45 minutes before being moved
pressure reading.
We then
in the waiting
Provider
frustrated
& came
providers
to an exam room.
It was 20OT,
minutes
before a patients
nurse finally
in and she mostly just asked questions I
had already answered on a form provided by the front desk. After the nurse left, it was almost another
half-hour before the doctor finally came in and he also asked many of the same questions. The
providers were very nice and apologetic, but of the almost 4 hours we spent in the clinic, only 1 hours
of that was actually face time with anyone!

two families in the waiting room said their relatives never even
Even more aggravating,had
whiletomy
mother
was in appointment.
surgery this morning,
families
thesame
waiting room said their
have
a PATA
Onetwo
even
had inthe
relatives never even hadcondition
to have a PATA
appointment.
One
even
had
the
same
condition
as my mother so Im not sure
as my mother
why our PATA visit was even necessary.
I brought my mom from PATA
out-of-state
because
were of
toldsurgical
that Massout-patients
General provides the best care in
can only
seewe65%
all of New England, maybe even the country, but thats not at all what we experienced. I sincerely hope
or visit
delayed
surgeries
that we can expect moreCancelled
from our next
to MGH.
~ Claire Bradley
Rutland, VT

Stanford GSB

OIT 262 Prof. Gur Class 7

PATA Process Flow

Stanford GSB

OIT 262 Prof. Gur Class 7

PATA Floorplan
PATA Floor Plan

Exam Room 7

Office of the
Nursing
Director

Exam
Room
8

Exam
Room 6

Exam
Room
9

(Charge Nurse Station)


Exam
Room 5

Supply Closet

Exam
Room 4

Bathroom

Chart Room

Closet
Exam
Room 3
Exam
Room 2
Exam
Room 1

Kitchen
Bathroom
Lab
(Blood
work)

Exam
Room
11
Exam
Room
12

Lab (Vitals + EKG)

EKG Bed
Cabinets/Counter
Top
Computer Station

Front Desk

Stanford GSB

Exam
Room
10

OIT 262 Prof. Gur Class 7

Sign In at Front Desk


PATA Floor Plan

Exam Room 7

Office of the
Nursing
Director

Exam
Room
8

Exam
Room 6

Exam
Room
9

(Charge Nurse Station)


Exam
Room 5

Supply Closet

Exam
Room 4

Bathroom

Chart Room

Closet
Exam
Room 3
Exam
Room 2
Exam
Room 1

Kitchen
Bathroom
Lab
(Blood
work)

Exam
Room
11
Exam
Room
12

Lab (Vitals + EKG)

EKG Bed
Cabinets/Counter
Top
Computer Station

Front Desk

Stanford GSB

Exam
Room
10

OIT 262 Prof. Gur Class 7

Lab for Vitals and EKG


PATA Floor Plan

Exam Room 7

Office of the
Nursing
Director

Exam
Room
8

Exam
Room 6

Exam
Room
9

(Charge Nurse Station)


Exam
Room 5

Supply Closet

Exam
Room 4

Bathroom

Chart Room

Closet
Exam
Room 3
Exam
Room 2
Exam
Room 1

Kitchen
Bathroom
Lab
(Blood
work)

Exam
Room
11
Exam
Room
12

Lab (Vitals + EKG)

EKG Bed
Cabinets/Counter
Top
Computer Station

Front Desk

Stanford GSB

Exam
Room
10

OIT 262 Prof. Gur Class 7

Charge Nurse Station


PATA Floor Plan

Exam Room 7

Office of the
Nursing
Director

Exam
Room
8

Exam
Room 6

Exam
Room
9

(Charge Nurse Station)


Exam
Room 5

Supply Closet

Exam
Room 4

Bathroom

Chart Room

Closet
Exam
Room 3
Exam
Room 2
Exam
Room 1

Kitchen
Bathroom
Lab
(Blood
work)

Exam
Room
11
Exam
Room
12

Lab (Vitals + EKG)

EKG Bed
Cabinets/Counter
Top
Computer Station

Front Desk

Stanford GSB

Exam
Room
10

OIT 262 Prof. Gur Class 7

10

Exam Room
PATA Floor Plan

Exam Room 7

Office of the
Nursing
Director

Exam
Room
8

Exam
Room 6

Exam
Room
9

(Charge Nurse Station)


Exam
Room 5

Supply Closet

Exam
Room 4

Bathroom

Chart Room

Closet
Exam
Room 3
Exam
Room 2
Exam
Room 1

Kitchen
Bathroom
Lab
(Blood
work)

Exam
Room
11
Exam
Room
12

Lab (Vitals + EKG)

EKG Bed
Cabinets/Counter
Top
Computer Station

Front Desk

Stanford GSB

Exam
Room
10

OIT 262 Prof. Gur Class 7

11

Chart Room
PATA Floor Plan

Exam Room 7

Office of the
Nursing
Director

Exam
Room
8

Exam
Room 6

Exam
Room
9

(Charge Nurse Station)


Exam
Room 5

Supply Closet

Exam
Room 4

Bathroom

Chart Room

Closet
Exam
Room 3
Exam
Room 2
Exam
Room 1

Kitchen
Bathroom
Lab
(Blood
work)

Exam
Room
11
Exam
Room
12

Lab (Vitals + EKG)

EKG Bed
Cabinets/Counter
Top
Computer Station

Front Desk

Stanford GSB

Exam
Room
10

OIT 262 Prof. Gur Class 7

12

Lab (Blood Work)


PATA Floor Plan

Exam Room 7

Office of the
Nursing
Director

Exam
Room
8

Exam
Room 6

Exam
Room
9

(Charge Nurse Station)


Exam
Room 5

Supply Closet

Exam
Room 4

Bathroom

Chart Room

Closet
Exam
Room 3
Exam
Room 2
Exam
Room 1

Kitchen
Bathroom
Lab
(Blood
work)

Exam
Room
11
Exam
Room
12

Lab (Vitals + EKG)

EKG Bed
Cabinets/Counter
Top
Computer Station

Front Desk

Stanford GSB

Exam
Room
10

OIT 262 Prof. Gur Class 7

13

PATA Process Flow Diagram and Capacity


Arrival rate = 8 pts/hr
Arrival rate = 4 pts/hr

7am-12 and 2-3pm


12-2pm (lunch)
Queue 1

Patients
arrive

Check-in

Queue 2

Wait

Capacity = 30 pts/hr

(Waiting Room)

Vitals + EKG
in Lab

Wait

Capacity =12 pt/hr


(lunch: 6 pt/hr)

(Waiting Room
+ Exam Room)

Capacity = 7 pt/hr
(lunch: 2.8 pt/hr)
Provider path

RN Chart
Review
5 min/pt

RN Visit

Wait

RN Chart
Write-Up

(Exam Room)

10 min/pt

RN Visit
27 min/pt

Queue 3

MD Chart
Review

MP Visit

Provider path

11 min/pt

MD Visit

Capacity = 7.5 pt/hr


(lunch: 3.75 pt/hr)

37 min/pt
Queue 4

MD Chart
Write-Up
17 min/pt

Stanford GSB

Blood Work
in Lab

Wait
(Waiting Room)

Capacity = 30 pt/hr
(lunch: 20 pt/hr)

OIT 262 Prof. Gur Class 7

Check-out

Patients
Leave

Capacity = 60 pt/hr

14

Calculating PATA Process Capacities


Check-in
Service time = 2 min/pt
Service rate = 30 pt/h
m = 1 attendant
Capacity = 30 pts/hr

MD Visit
Service time = 64 min/pt

Vitals + EKG in Lab


Service time =10 min/pt
Service rate = 6 pt/h
m = 2 technicians
Capacity = 12 pt/hr
(lunch: 6 pt/hr)
RN Visit
Service time =43 min/pt

Stanford GSB

Chart review = 5 min/pt


Visit with patient = 27 min/pt
Chart write-up = 11 min/pt

Service rate = 1.4 pt/h


m = 5 nurses
Capacity = 7 pt/hr
(lunch: 2.8 pt/hr)

Chart review = 10 min/pt


Visit with patient = 37 min/pt
Chart write-up = 17 min/pt

Service rate = 0.94 pt/h


m = 8 MDs
Capacity = 7.5 pt/hr
(lunch: 3.75 pt/hr)

Blood Work in Lab


Service time = 6 min/pt
Service rate = 10 pt/h
m = 3 technicians
Capacity = 30 pt/hr
(lunch: 20 pt/hr)
Check-out
Service time = 1 min/pt
Service rate = 60 pt/h
m=1 attendant
Capacity = 60 pt/hr

OIT 262 Prof. Gur Class 7

15

PATA Process Capacities

Non-Lunch
Step
Check-in
Vitals + EKG in Lab
RN Visit
MD Visit
Blood Work in Lab
Check-out

Service Time
(min/pt)
2
10
43
64
6
1

Service Rate
(pts/hr)
30
6
1.40
0.94
10
60

Lunch

# of
Capacity
# of
Capacity
Employees (pts/hr) Employees (pts/hr)
1
30
1
30
2
12
1
6
5
7
2
2.8
8
7.5
4
3.75
3
30
2
20
1
60
1
60

The RNs are the bottleneck

Stanford GSB

OIT 262 Prof. Gur Class 7

16

PATA Process Flow Diagram and Capacity


Arrival rate = 8 pts/hr
Arrival rate = 4 pts/hr

7am-12 and 2-3pm


12-2pm (lunch)
Queue 1

Patients
arrive

Check-in

Queue 2

Wait

Capacity = 30 pts/hr

(Waiting Room)

Vitals + EKG
in Lab

Wait

Capacity =12 pt/hr


(lunch: 6 pt/hr)

(Waiting Room
+ Exam Room)

Bottleneck
Provider path

RN Chart
Review
5 min/pt

RN Visit

Wait

RN Chart
Write-Up

(Exam Room)

10 min/pt

RN Visit
27 min/pt

Queue 3

MD Chart
Review

MP Visit

Capacity = 7 pt/hr
(lunch: 2.8 pt/hr)

Provider path

11 min/pt

MD Visit

Capacity = 7.5 pt/hr


(lunch: 3.75 pt/hr)

37 min/pt
Queue 4

MD Chart
Write-Up
17 min/pt

Stanford GSB

Blood Work
in Lab

Wait
(Waiting Room)

Capacity = 30 pt/hr
(lunch: 20 pt/hr)

OIT 262 Prof. Gur Class 7

Check-out

Patients
Leave

Capacity = 60 pt/hr

17

What Does Patient Flow Look Like


Before the RN step
Patients can flow through at the arrival rate until the waiting
room is full (like in Cranberry)
No info on waiting room capacity lets just assume its large
and never fills up (those who cant find a seat just stand)
Then flow rate at steps before the RN is 8 pts/hr in non-lunch
times, 4 pts/hr during lunch
At the RN step and afterwards
RN capacity limits flow
So flow rate = 7 pts/hr in non-lunch times, 2.8 pts/hr during
lunch
Stanford GSB

OIT 262 Prof. Gur Class 7

18

Utilization Analysis

Non-Lunch
Step
Check-in
Vitals + EKG in Lab
RN Visit
MD Visit
Blood Work in Lab
Check-out

Stanford GSB

Flow Rate
(pts/hr)
8
8
7
7
7
7

# of
Employees
1
2
5
8
3
1

Lunch
Capacity
(pts/hr)
30
12
7
7.5
30
60

Util.
0.27
0.67
1.00
0.93
0.23
0.12

OIT 262 Prof. Gur Class 7

Flow Rate
(pts/hr)
4
4
2.8
2.8
2.8
2.8

Capacity
# of Employees (pts/hr)
1
30
1
6
2
2.8
4
3.75
2
20
1
60

Util.
0.13
0.67
1.00
0.74
0.14
0.05

19

Time Out: Whats Going on in This Process


RNs are overloaded
Like in Cranberry, they build up a backlog of work
Will work through backlog after patients stop arriving
Can analyze using inventory build-up diagrams
Other stations have utilizations < 1
Waiting times in front of other stations will be driven by
randomness in arrivals/processing
Can analyze using queueing tools
PATA has waiting times driven both by too little capacity (at RN
station) and randomness (at other stations)
Stanford GSB

OIT 262 Prof. Gur Class 7

20

Analyzing Inventory Buildup at the RN Station

Capacity
7 am 12 pm: 7 pts/hr
12 pm 2 pm: 2.8 pts/hr
2 pm end of day: 7 pts/hr
Arrivals
7 am 12 pm: 8 pts/hr
12 pm 2 pm: 4 pts/hr
2 pm 3 pm: 8 pts/hr
After 3 pm: 0 pts/hr

Stanford GSB

OIT 262 Prof. Gur Class 7

Backlog Accumulates at
7 am 12 pm: 1 pt/hr
12 pm 2 pm: 1.2 pts/hr
2 pm 3 pm: 1 pt/hr
3 pm+: -7 pts/hr

21

Inventory Buildup at RN Station


Peak: 8.4 patients at 3 pm

4:12 pm

Stanford GSB

OIT 262 Prof. Gur Class 7

22

Average Inventory at RN Station


Time
Length (hours) Start Inv End Inv Avg Inv
7 am - 12 pm
5
0
5
2.5
12 pm - 2 pm
2
5
7.4
6.2
2 pm - 3 pm
1
7.4
8.4
7.9
3 pm - 4:12 pm
1.2
8.4
0
4.2
Grand Average
4.11

Avg Patients Waiting = 4.11


Comment: Another way is by:
Avg Patients Waiting = AUC/b
If was triangle, we would get
0.5*h*b/b = 0.5*h = 4.2
(pretty close!)

Stanford GSB

OIT 262 Prof. Gur Class 7

23

Average Patient Waiting Time


Average flow rate (out of RN queue)
7 patients/hr from 7 am to 12 pm (5 hours)
2.8 patients/hr from 12 pm to 2 pm (2 hours)
7 patients/hr from 2 pm to 4:12 pm (2.2 hours)
Weighted average = 6.1 patients/hr
Average patient waiting time
Littles Law: Inv = Flow Rate x Flow Time
Flow Time = Inv/Flow Rate
Waiting Time = (4.11 pts) / (6.1 pts/hr) = 0.67 hrs = 40 min
Average patient waits 40 minutes at RN station
Comment: another way is by AUC/(number of patients)
If triangle we get 0.5 x 9.2 x 8.2 / 56 = 0.664 hrs (very close!)
Stanford GSB

OIT 262 Prof. Gur Class 7

24

What Have We Done?


RNs are overloaded
Like in Cranberry, they build up a backlog of work
Will work through backlog after patients stop arriving
Can analyze using inventory build-up diagrams
40 minutes of waiting time at RN
Other stations have utilizations < 1
Waiting times in front of other stations will be driven by
randomness in arrivals/processing
Can analyze using queueing tools
PATA has waiting times driven both by too little capacity (at RN
station) and randomness (at other station)
Stanford GSB

OIT 262 Prof. Gur Class 7

25

Whats Next?
RNs are overloaded
Like in Cranberry, they build up a backlog of work
Will work through backlog after patients stop arriving
Can analyze using inventory build-up diagrams
40 minutes of waiting time at RN
Other stations have utilizations < 1
Waiting times in front of other stations will be driven by
randomness in arrivals/processing
Can analyze using queueing tools
PATA has waiting times driven both by too little capacity (at RN
station) and randomness (at other station)
Stanford GSB

OIT 262 Prof. Gur Class 7

26

PATA Process Flow Diagram and Capacity


Arrival rate = 8 pts/hr
Arrival rate = 4 pts/hr

7am-12 and 2-3pm


12-2pm (lunch)

Queue 2
Avg Wait Time = 40 min

Queue 1

Patients
arrive

Check-in

Wait

Capacity = 30 pts/hr

(Waiting Room)

Vitals + EKG
in Lab

Wait

Capacity =12 pt/hr


(lunch: 6 pt/hr)

(Waiting Room
+ Exam Room)

Capacity = 7 pt/hr
(lunch: 2.8 pt/hr)
Provider path

5 min/pt

RN Visit

RN Visit
27 min/pt

Wait

RN Chart
Write-Up

(Exam Room)

10 min/pt

MP Visit

RN Chart
Review

Queue 3

MD Chart
Review

Provider path

11 min/pt

MD Visit

Capacity = 7.5 pt/hr


(lunch: 3.75 pt/hr)

37 min/pt
Queue 4

MD Chart
Write-Up
17 min/pt

Stanford GSB

Blood Work
in Lab

Wait
(Waiting Room)

Capacity = 30 pt/hr
(lunch: 20 pt/hr)

OIT 262 Prof. Gur Class 7

Check-out

Patients
Leave

Capacity = 60 pt/hr

27

Waiting at Vitals + EKG (Queue 1, 9 am to 12 pm only)


Arrivals
Arrival rate = 1/a = 8 patients/hr
Average Interarrival time = a = 60/8 = 7.5 minutes
Std Dev of Interarrival Times from 9 am to 12 pm (Fig 2a) = 8.9 min
CVa = Std Dev/Mean = 8.9/7.5 = 1.2
Service
Average Processing Time = 10 min (case p. 10)
Std Dev of Processing Time = 3.5 min (case p. 10 footnote 9)
CVp = Std Dev/Mean = 3.5/10 = 0.35
Number of Technicians/Stations = m = 2
u = p/ma = 10/(2x7.5) = 0.667
Tq = 6.4 minutes
Stanford GSB

OIT 262 Prof. Gur Class 7

28

Waiting Time Analysis at MD & Blood Work


Vitals + EKG

MD

Blood Work

a [min]

7.5

8.6

8.6

Std Dev a [min]

8.9

1.7

3.4

CVa

1.2

0.2

0.4

p [min]

10

64

Std Dev p [min]

3.5

29

CVp
m

0.35
2

0.45
8

0.33
3

0.67

0.93

0.23

Tq [min]

6.38

11.77

0.02

Queue 1

Queue 3

Queue 4

Arrival rate (1/a) after RN equals the capacity at RN = 7/60 = 1/8.6


Total waiting time from queueing effects = about 18 minutes
CVa and CVp data for MD and Blood Work on page 14 of case.
Stanford GSB

OIT 262 Prof. Gur Class 7

29

PATA Process Flow Diagram and Total Flow Times


Queue 1
Avg Wait Time = 6.4 min

Patients
arrive

Check-in

Vitals + EKG
in Lab

Wait

Time = 2 min

Queue 3
Avg Wait Time = 12 min

MD Chart
Review

5 min/pt

RN Visit

RN Visit
27 min/pt

Wait

RN Chart
Write-Up

(Exam Room)

10 min/pt

11 min/pt

MD Visit

Total Time = 47 min

37 min/pt
MD Chart
Write-Up
17 min/pt

Total Wait Time = 58 minutes


Stanford GSB

(Waiting Room
+ Exam Room)

Total Time = 32 min

Provider path
RN Chart
Review

Wait

Time = 10 min

(Waiting Room)

Provider path

MP Visit

Queue 2
Avg Wait Time = 40 min

Queue 4
Average Wait Time = 0.02 min

Wait

Blood Work
in Lab

Check-out

(Waiting Room)

Time = 6 min

Time = 1 min

Total Service Time = 98 minutes


OIT 262 Prof. Gur Class 7

Patients
Leave

Avg Flow Time = 156 minutes


30

Summary of Initial Analysis


1 hour (about 1/3 of total time in system) is non-value added
waiting
Even more waiting (from patient perspective) during NP/MD writeup times (value is added during these times but the patient is not
face-to-face with a provider)
The non-value added waiting may be controllable via the design of
the system
Processing capacity/staff
Physical capacity/exam rooms
Scheduling
Now that we know whats driving the non-value added waiting
times, we can see how proposals impact that time
Stanford GSB

OIT 262 Prof. Gur Class 7

31

Analysis of Taskforce Recommendations


We asked you to analyze three specific recommendations:

Extend hours to 6:30 pm, increase time between appointments to


45 minutes

Add an RN

Add an MD

Lets look at each of these


Stanford GSB

OIT 262 Prof. Gur Class 7

32

Extend Hours to 6:30 pm, Schedule Appointments Every 45 Minutes


Current System
Non-lunch: 4 arrivals every 30
minutes 8 patients/hr
Lunch: 2 arrivals every 30 minutes
4 patients/hr
Proposal
Non-lunch: 4 arrivals every 45
minutes 5.3 patients/hr
Lunch: 2 arrivals every 45 minutes
2.67 patients/hr
To maintain same number of total arrivals
(56) in 1 day, you actually need to extend
scheduled arrivals from 3 pm to 6:30 pm
(not 6 pm as in the case)
Does this help?
Stanford GSB

OIT 262 Prof. Gur Class 7

33

Slow Down Arrivals: Analysis


Eliminates build-up at RN only queueing times remain
- waiting time at RN is not zero!
Reduces queueing waiting times at other steps (lower
arrival rate)
Total average waiting time is down from about 1 hour to
less than 15 minutes
Vitals + EKG
a
11.3
Std Dev a
8.9
CVa
0.8
p
10
Std Dev p
3.5
CVp
0.35
m
2
u
0.44
Tq
1.01

NP
11.3
11.3
1.0
43
21
0.49
5
0.76
11.26

MD
11.3
2.3
0.2
64
29
0.45
8
0.71
1.09

Blood Work
11.3
4.5
0.4
6
2
0.33
3
0.18
0.01

What are the main drawbacks?


Stanford GSB

OIT 262 Prof. Gur Class 7

34

Add a Registered Nurse (RN)


Current RN utilization is 100%, resulting
in inventory buildup, current average
wait time of 40 minutes
What happens if we add 1 RN?
Assume extra RN at both lunch and
non-lunch
RN capacity goes from 7 pts/hr to
8.4 pts/hr during non-lunch
RN capacity goes from 2.8 pts/hr to
4.2 pts/hr during lunch
Both are more than demand, so 1
extra RN eliminates inventory buildup due to insufficient capacity
Does this help?
Stanford GSB

OIT 262 Prof. Gur Class 7

35

Add an RN: Analysis

No more inventory build-up at RN, but now MD is bottleneck


MD capacity is 7.5 patients/hr, only slightly more than RN
Can expect inventory buildup at MD step
Even without deterministic inventory buildup at RN, will still have
queueing delays (utilization still high around 95%)

Does it worth the cost?


Stanford GSB

OIT 262 Prof. Gur Class 7

36

Add an Anesthesiologist (MD)


Current MD utilization is 93%, resulting
in queueing delays due to randomness,
current average wait time of almost 12
minutes
What happens if we add 1 MD?
Focus on non-lunch (9 am 12 pm)
time used in queueing analysis
above
MD staff goes from m = 8 to m = 9
Utilization at MD step goes from
93% to 83%
Does this help?
Stanford GSB

OIT 262 Prof. Gur Class 7

37

Add an MD: Analysis


Waiting time drops to 2.68 minutes (reduction of less than 10
minutes - at most it could be 12 minutes)
Anesthesiologists are expensive and this is a minimal decrease in
waiting time

Does it worth the cost as a standalone measure?


Stanford GSB

OIT 262 Prof. Gur Class 7

38

Your Recommendations: How to Fix PATA

Stanford GSB

OIT 262 Prof. Gur Class 7

39

MGH Wrap-up
Recall our framing thoughts:
How can operations help deliver health care
more efficiently?
How can process and queueing analysis
help inform managerial decisions and
strategies?
How process analysis helps?
Complicated system to analyze
Tools allow for solid first order
assessment of the problem and impact
of potential actions
Further analysis and simulation may be
useful to predict impact more accurately
And this is only the tip of the iceberg!
Queueing analysis has been improving
the design of health care systems in
many aspects to reduce patient waiting
times at minimal costs
Stanford GSB

OIT 262 Prof. Gur Class 7

40

Process Improvement in Health Care


MGH PATA Redesign

Stanford GSB
April 24, 2015

PATA Follow-Up
Immediate Actions
1. RN and MD assigned to patient at the same time.

RN Visit

Gave MD 20 25min to do chart review


Eliminated wait time between providers

Wait

2. RNs given a 5-minute patient history rule


3. No batching of intake forms at front desk

MD Visit

4. Only charge nurse allowed to modify the appointment


schedule/tracking sheet
5. Redesigned card RNs filled out in visit
Reduced patient wait time ~20min
Helped with staff morale and set up clinic for long term strategies

42

PATA Follow-Up
Long-Term Actions
Kelsey McCarty hired as Sr. Project Manager
Business plan to redesign entire process
Presented to CEO and funded at
$2.6M over 3 years
1.

See 100% of patients


Ensure positive patient experience
Generate high quality evaluations

Phone Program for low acuity, low complexity patients


RNs collect basic info over phone + 2min assessment on day of surgery
Triage grid for surgeons: phone vs clinic + lab and EKG orders

2.

Single-provider model in clinic with Nurse Practitioners (NPs)


NPs can do both RN Assessment and anesthesia evaluation

3.

Predefined schedule: Patients assigned in advance 90min visit

4.

Web-based intake form for patients to complete in advance

5.

RFID tracking of all patient and providers in PATA

2h appt = 30min for labs + 90min appt with NPs


43

MGH-MIT Collaboration: Work to date


Implemented & Results

Ready In
toImplementation
start implementation

OR block reallocation
(elective cases)

Intra-day surgical scheduling


Post-surgery Recovery Area patient flow

OR Open Blocks
(non-elective cases)

Cancer Center

Inventory management of surgical


supplies - Part I

Primary Care Redesign Rx


Management

New
In design
projects
Hospital bed capacity
management
(predictions &
decision support)
Primary Care Redesign
Provider Scheduling

Critical Care
Non-Oncology Infusion
Surgical Observation
Patient Flow
Optimization

44

OR Scheduling of Non-Elective Cases: Symptoms


T. Carnes, B. Daily, P. Dunn, R. Levi, S. Moss, D. Price, C. Zenteno

Max Prescribed Wait Time


Non Urgent

24 hours

Urgent

4 hours

Emergent

45 minutes

~30% patients waited more than


recommended wait time

Non-elective cases = wait list cases


Data source: MGH OR case data. Time frame: July 2009 vs May 2010
Surgeries performed on weekdays, excluding holidays.
45

OR Scheduling of Non-Elective Cases:


Centralized Open Blocks
Objective: Increase timely access for waitlist cases

Before project
~2.4 Reserved ORs owned
by particular services on
weekdays

Centralization
Jan Sept 2012

Two extra blocks


became available
October 2012

Simulation Model
5 Open Blocks

Block = {Operating Room, Weekday},


7am 5pm

~30% patients < 3% pts. wait more than


recommended wait time
Assumes OR is the only limiting factor
46

OR Scheduling of Non-Elective Cases: Results


Even with 9% increase in overall case volume, the average wait time decreased by 25%
for all non-elective cases.

100%
80%

% Patients had procedure within designated


wait time
87.7%
79.3%

86.2%
75.3%

70.0%
65.3%

60%

OR Type

Prime Time
Utilization

Elective

87.2%

86.2%

Non-electives

86.1%

80%

40%
20%
0%

Non-Urgent

Urgent
2011

Emergent

OR Type

Avg # Running
ORs/day

Elective

45.8

47.3

Non-electives

2.4

2.7

2012

Wait time = Booking time Into OR

Data source: MGH OR case data


Time frame: June Dec 2011 vs June Dec 2012
Surgeries performed on weekdays, excluding holidays.
47

Success Drivers and Lessons Learned


Work only on top priority burning problems
Make sure data is available to identify the problem
Understand mutual cultures, organization and community
Leadership engagement to support implementation
Analyzing impact of implementation and monitoring results
is a big challenge Very dynamic environment!

48

Q&A
azentenolangle@mgh.harvard.edu

49

OIT 262 Roadmap: 18 Sessions of Operations Greatest Hits


Class

Date

Topic

4/3

Introduction, basic concepts, Littles Law

4/6

Process Analysis

4/10

Process Improvement

4/13

Randomness and Waiting Times

4/17

Optional Review Session

4/20

Batching and the Product-Process Matrix

4/24

Process Improvement in Health Care

4/27

Supply Chain Inventory Management

5/1

Quality and Process Improvement in Health Care

10

5/4

Gurs Research Lecture

11

5/8

Toyota Production System

12

5/11

Measuring and Quantifying Quality

13

5/15

Quality Management in Luxury Services

14

5/18

Supply Chain Simulation

15

5/22

Managing Product Variety in Supply Chains

16

5/27

Forecasting Game, Contracts & Incentives

17

5/29

Leadtime Reduction and Quick Response

18

6/1

Student Presentations & Course Summary

Stanford GSB

Overview
Process Analysis
Types of processes
Dimensions of process capability
Process flow diagrams
Capacity, utilization, inventory
Uncertainty (randomness) in processes
Bottlenecks

Process Improvement and Quality Management


Process in service industries
Quality control, Total Quality Management
Just-in-time, Lean Manufacturing
Six Sigma, Statistical Process Control
Supply Chain Management
Order fulfillment and distribution
Postponement
Quick response/dual response
Information sharing, contracts, and incentives in supply chains
Fast fashion production

OIT 262 Prof. Gur Class 7

50

Upcoming Classes
Next Class: Tools: supply chain and inventory management
Preparation: book reading + problems (do not turn in,
will be solved in class / review session)
Wed. 4/29: Review session, Littlefield begins 1:00pm
Fri. 5/1: Quality and process in healthcare (guest speaker)
Preparation: case reading
Mon. 5/4: Gurs research lecture (Quality, lookahead and
browsing path optimization in online services);
Littlefield ends
Fri. 5/8: Quality and lean operations in TPS
Toyota case (first priority)
Textbook reading (expands on the ideas of TPS that
are discussed in the case)
NY Times article on standardization of teaching
Stanford GSB

OIT 262 Prof. Gur Class 7

51

Você também pode gostar