Sepsis at the Hospital System Level
One Sepsis Journey
Alan Whippy, MD
Medical Director of Quality and Safety
The Permanente Medical Group
October 26, 2012
Kaiser Permanente Northern
California
KP’s Integrated Model
Kaiser Foundation
Health Plan
Kaiser Foundation
Hospitals
Permanente
Medical Group
Membership
Facilities
3.4 million
21 hospitals
Employees
64,000
Physicians
7,169
10,000 Foot View
At the Heart of Hospital Mortality
Condition
2011
AMI
Sepsis
Mortality
# Admissions
Rate
5,254
18,746
3.5%
10%
# Deaths
183
1,926
4
Why Sepsis?
We miss it
We underestimate it
We under treat it
We cause it
5
Understanding the Landscape
Malignant
intravascular
inflammation
Causing cytopathic
tissue hypoxia
 Shock
 Quiet Shock
 ? Pending Shock
Why Sepsis?
42%
47%
No Sepsis on
Admission
138,319
*Principle or Secondary Dx of Sepsis, Severe Sepsis or Septic Shock POA, 2011
Paradigm Shift
We need a systems
approach to sepsis
like
AMI
Stroke
Trauma
Goals of KP Sepsis Program
1. Find and Name it
2. Stratify it
2. Treat
Stratify
Risk
3.
it early
4. Prevent it
Save Lives
Construction
Mortality
Diagnostic
Spring 08
Sepsis
Design
Team
Pilot
Sepsis
Summit
Nov 08
Implement
Learn
and
Improve
Ja
nM 06
ay
-0
Se 6
p0
Ja 6
nM 07
ay
-0
Se 7
p0
Ja 7
nM 08
ay
-0
Se 8
p0
Ja 8
nM 09
ay
-0
Se 9
p0
Ja 9
nM 10
ay
-1
Se 0
p1
Ja 0
nM 11
ay
Se 11
p1
Ja 1
nM 12
ay
-1
2
35%
5%
25% Mortality Rate
30%
25%
20%
15%
14%
10%
3.5% Admission Rate
8.7%
0%
Observed to Expected
Sepsis Mortality
Med Care. 2008 Mar;46(3):232-9. Risk-adjusting hospital inpatient mortality using automated inpatient,
outpatient, and laboratory databases. Escobar GJ, Greene JD, Scheirer P, Gardner MN, Draper D, Kipnis
Observed to Expected
Sepsis LOS
Med Care. 2008 Mar;46(3):232-9. Risk-adjusting hospital inpatient mortality using automated inpatient,
outpatient, and laboratory databases. Escobar GJ, Greene JD, Scheirer P, Gardner MN, Draper D, Kipnis P.
Crude Hospital Mortality
2007
931 fewer deaths/yr
330 fewer pts/day
2009 2010
2011 2012
What Every Hospital Was Asked to Do
Plan Handoffs
Develop Teams
Measure
& PDSA
500 to 1000 ml Fluid
boluses q 30 min
8-12
MAP ≥ 65?
< 65
Norepinephrine
D
I
R
E
C
T
E
D
Triage to
ED Room
ED MD
to HBS
ED RN to
ICU RN
HBS and
Intensivist
≥ 65
ScvO2 ≥70?
<70
≥ 70
If Hct low,
transfuse to 30
<70
Dobutamine
Lower
Lactate
Abx in 1 hr
ABX
ABX
Within
Within22hrs:
hrs:
Central
CentralLine
LinePlaced
Placed
<8-12
CL in 2 hrs
Within
Within11hr:
hr: Start
Start
EARLY GOAL
CVP 8-12?
Lactates on ED
Blood Cultures
Team
Aggressive
Aggressivefluid
fluid
resuscitation
resuscitation
55%




































































DENOM
47%
T
H
E
R
A
P
Y
44%
84%
37%
69%
91%
29%
85%
32%
77%
37%
Kaiser Foundation Hospital - San Jose
DRAFT
EMERGENCY DEPARTMENT
ED NURSES FLOW SHEET
Tools and Equipment
ADDRESSOGRAPH
Room / Bed # ____________
EARLY GOAL DIRECTED THERAPY (EGDT)
ED AND ICU ORDERSET
Date _________________________________
Time
Pain*
P
BP
R
Addressograph
Pg__ of __
T
GCS* MAP*
(EMV)
SCVO2
Score Character
Page 1 of 4
FiO2/
SpO2
CVP
RASS*
Intervention
mm/Hg
Score
Note actions taken in response to the findings on
the left.
Train
99%
80%
36%
63%
10
X Check box to activate an order
10
Mark chart: Allergic to ___________________________________________
No Known Allergies
Weight _______(kg)
Height______
Check if applies: Pregnant  Lactating
66%
10
Sepsis Implementation
Training
Diagnosis
Diagnosisof
of
Severe
SevereSepsis
Sepsisor
or
Septic
SepticShock
Shock
Presentation
Adopt Algorithms
Lower Lactate
Sepsis Care
Lactates
10
ED ORDERS
10
ICU ORDERS
10
 Administer oxygen titrate FiO2 to maintain SpO2 between 90-94%.
Notify physician if patient requires over 50% FiO2 via face mask or
more than 10% increase in 1 hour to achieve needed goal
 ___________________, RN ______________(Date/Time)
10
 Continue
 Discontinue
 Continue
 Discontinue
10
58%




56%
100%
0%
0%
0%
10
 Insert Foley catheter
 ___________________, RN ______________(Date/Time)
VITAL SIGNS
 Measure intake and output hourly
 ___________________, RN ______________(Date/Time)
 Record vital signs (heart rate, blood pressure, respiratory rate,
SpO2) per unit standards and as needed
 ___________________, RN ______________(Date/Time)
10
10
10
 Continue
 Discontinue
 Continue
 Discontinue
 Continue
 Discontinue
 Continue
 Discontinue
10
10
10
10
10
 Measure central venous pressure (CVP) every 30 minutes until
goal has been achieved and for at least 2 hours after goal has been
reached, then monitor per unit standards and as needed
 ___________________, RN ______________(Date/Time)
OUTPUT
Time
TYPE
INTAKE
AMOUNT
Time
TYPE
PARENTERAL FLUIDS
AMOUNT
Time
#
SITE
GAUGE
TYPE
STARTED
ABSORBED
 Monitor mixed venous oxygen saturation (ScvO2) continuously
 ___________________, RN ______________(Date/Time)
TOTALS
Initals Signature/Title
Collaborate!
Initials
Signature/Title
Initials
Signature/Title
Collaborate!
Collaborate!
1. Frontline
Engagement
6. Clear
Communication
2. Scripted
Processes
Model for a Learning
Organization
3. Reliable
Execution
5. Scientific
Change Process
4. Organizational
Learning
Terry P. Clemmer, MD
Source: T Clemmer
Hospital
LDSLDS
Hospital
1.1.Frontline
Frontline
Engagement
Engagement
Leadership Alignment
Teams
Deliberate
Practice
MD Champion, Improvement
Advisor, Elephant
World View
ER
18 November 30, 2012
ICU
Everything
Else
2.2.Scripted
Scripted
Processes
Processes
All ED patients with infection are screened for
Getting Lactates
Sep
% of Blood Cultures in ED with Lactate Test
Lactates
100%
90%
Lactates on ED
Blood Cultures
55%
DENOM
97%
80%
98%
91%
99%
83%
70%
95%
96%
99%
97%
60%
97%
82%
98%
94%
50%
92%
88%
20
09
20 -01
09
20 -05
09
20 -09
10
20 -01
10
20 -05
10
20 -09
11
20 -01
11
20 -05
11
20 -09
12
20 -01
12
20 -05
12
-0
9
95%
94%
100%
Identify at triage if suspected infection and 2 SIRS criteria
To < 96.8 (36.0) or > 100.4 (38.0)
HR > 90
SUSPECTED SEPSIS
RR > 20
DOCUMENT SIRS
WBC > 12K or < 4K or > 10% bands
-OR- Altered LOC
E
A
R
L
Y
CBC, Lactate, BC
Consider IV fluids and ABX
SBP ≤ 90?
yes
SBP >90
20 ml/kg fluid
R
st
bolus in 1 hr
E
C
SBP ≤ 90
O
Document Septic
Septic
G Document
Shock (Time
(Time Zero)
Zero)
N Shock
I
T
I
O
N
The Golden Hours
Document
Document Sepsis
Sepsis
<2
Lactate high?
no
Sepsis
Resuscitation
≥4
2-3.9
Aggressive IV fluid resuscitation
Early ABX
Repeat lactate in 3-6 hrs
Document Severe
Sepsis
EGDT Goals from Time Zero
1.
2.
3.
4.
5.
6.
Start Antibiotic in 1 hr
First CVP or ScvO2 within 2 hrs
CVP ≥ 8-12 within 6 hrs
MAP ≥ 65 within 6 hrs
ScvO2 ≥ 70 within 6 hrs
Repeat lactate is lower than
initial lactate w/in 3-12 hrs
Document
Document Severe
Severe
Sepsis
Sepsis (Time
(Time Zero)
Zero)
EGDT
Aggressive fluid
resuscitation
CVP ≥ 8-12
<8
≥ 8-12
Within 1 hr:
MAP ≥ 65?
Start ABX
≥ 65
Within 2 hrs:
Measure CVP or
ScvO2
ScvO2 ≥70?
≥ 70
< 65
<70
EARLY GOAL D
I
500 – 1000 ml Fluid
R
boluses q 30 min
E
C
T
Norepinephrine
E
D
If Hct low,
transfuse to 30
<70
Dobutamine
Repeat lactate
3-12 hrs
T
H
E
R
A
P
Y
3.3.Reliable
Reliable
Execution
Execution
Diagnosis
Central
Line
Placed
MAP
CVP
ScvO2
at Goal
Lactate
Clearance
ABX started
IV Fluid
Time
Zero
1 hour
2
hours
6
hours
6-12
hours
3.3.Reliable
Reliable
Execution
Execution
EGDT Bundle July 2009 to Dec 2011
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
Abx in 1 hr
Central Line
in 2 hr
CVP ≥ 8-12
MAP ≥ 65
21 Hospital Success rates
ScvO2 ≥ 70
Lower
Repeat
Lactate
Web-based EGDT Tracking System
•Semi-automated.
•Identifies and allows review
of yesterday’s ED cases
Collaboration and Learning
• Co-locate BC
and lactate
• Sepsis Alerts
• Velcro Clock
• Critical lab
calls
• RRT lactate FU
• RRT lactate
screens
• EGDT in OR
This Works
28%
In the last 12 months
we had 45% fewer
deaths than in the first
year for this population
KPNC EGDT Bundle vs. Mortality
70%
Mortality: 27%
26%
60%
24%
50%
22%
20%
40%
18%
30%
16%
Mortality: 18%
14%
12%
20%
10%
Bundle
10%
0%
Q3 09
Q4 09
Q1 10
Q2 10
Q3 10
EGDT Mortality
Q4 10
Q1 11
Q2 11
EGDT Bundle
Q3 11
Q4 11
We’re Not Alone
Pre-Intervention:
Average Mortality 44.8%
Post intervention:
Average Mortality 24.5%
60
50
40
30
20
10
Standard
EGDT
Otero RM, Nguyen HB, et al Chest, 130:5 Nov 2006 (1579-2093)
s
au
l
t.
P
B
JH
H
H
S
d
Sa
m
nU
G
oo
H
an
N
N
E
U
P
N
JC
D
C
U
M
B
ID
M
C
B
M
FS
B
ir
H
rt
LL
U
0
2009
Trzeciak Chest 2006, 129:225-235
Shapiro Crit Care Med 2006,34;1025-1032
Jones, Chest 2007, 132:425-432
Micek, Crit Care Med 2006,34:2702-2713
2.0 liters (1.2- 3.4 liters)
5.0 liters (3.8- 7.2 liters)
285 subjects
The Tale of a Thousand Lines
We’ve reviewed over 4000 EGDT cases
4 Pneumothoraces
0 BSI
Lives lost due to central line
0
Mortality 2718%
Lives saved by EGDT
1. Frontline
Engagement
6. Clear
Communication
2. Scripted
Processes
Model for a Learning
Organization
3. Reliable
Execution
5. Scientific
Change Process
4. Organizational
Learning
Terry P. Clemmer, MD
Source: T Clemmer
Hospital
LDSLDS
Hospital
4.4.Organizational
Organizational
Learning
Learning
Early Line Placement Matters
MAP Goal Met
CVP Goal Met
ScvO2 Goal Met
CL after 2 hrs
75%
53%
30%
CL before 2 hrs
86%
87%
60%
Data from Q3 2009, KPNC
4.4.Organizational
Organizational
Learning
Learning
Bundling
Outcome by Bundle Completion, 2010-2012YTD
25
21.6
20
15.9
15
10
5
2332 pts
0
1519 pts
Patients
webundle
do
incomplete bundle do better if
complete
better
What if the BP is Normal?
Lactate > 4 mMol / L
70
“Quiet Shock” Mortality
60
50
40
30
20
10
0
MAP > 100
EGDT
All Patients
Control
Donnino et al. Chest 2003 124: 90S
4.4.Organizational
Organizational “Quiet” Shock
Learning
Learning
EGDT Survival Trends
30
25
Low WNL
Low
20
15
Low
WNL
WNL
10
5
0
270 132
725 360
1220 788
Those with
normal 2011BP2012 YTD
2010
benefit
even more
low BP nl BP
Late 2009
4.4.Organizational
Organizational
Learning
Learning
Seeking Lactate
Clearance
Mortality and Intermediate Lactate Clearance
(Q2'12, 1495 Admissions)
15%
9%
Overall Mortality
8%
Mortality
Without Lactate
Clearance
Mortality With
Lacate
Clearance
4.4.Organizational
Organizational
Learning
Learning
Improved Survival
Intermediate lactate Raw and O/E Mortality Trends
25%
1.8
1.6
20%
1.4
15%
1
0.8
10%
0.6
0.4
5%
3-3.99 2-2.99
3-3.99 2-2.99
3-3.99 2-2.99
3-3.99 2-2.99
0%
0.2
0
2009
3-3.99 Raw
2010
2-2.99 Raw
2011
3-3.99 O/E
2012
2-2.99 O/E
O/E Mortality
Raw Mortality
1.2
Work in Progress
•
•
•
•
EGDT in OR
Pediatric Sepsis pilots
Research projects
Intermediate lactate
bundle??
• Electronic Early Warning
System development
• Delirium prevention; long
term functional and
cognitive outcomes
• Reducing HA sepsis:
CDI, HAP, SSI programs
Coordinated by Design