1. Patient Safety: A Human
Factors Approach
Sept 4, 2015
Paul Barach, BSc, MD, MPH, Maj ( ret.)
Clinical Professor
Wayne State University School of Medicine
4. 4
No
system
beyond
this
point
10-2 10-3 10-4 10-5 10-6
Civil Aviation
Nuclear Industry
Railways (France)
Chartered Flight
Road Safety
Chemical Industry (total)
Fatal
risk
ED/ Medical risk
(total)
Anesthesiology
ASA1
Pedi Cardiac Surgery
Patient ASA 3-5
Fatal Iatrogenic
adverse events
Very
unsafe
Ultra
safe
Average
rate
per
exposure
of
catastrophes
and
associated
deaths
in
various
industries
and
human
acAviAes
Unsafe
Safe
Hymalaya
mountaineering
Microlight spreading
activity
NICU
5. Does the day of surgery matter for outcomes ?
operations performed on Fridays were associated with a higher 30-day
mortality rate than those performed on Mondays through Wednesdays:
2.94% vs. 2.18%;
Odds ratio, 1.36; 95% CI, 1.24–1.49)
6. March
27,
1977:
KLM
747-‐200
and
Pan
Am
747-‐100;
Tenerife,
Canary
Islands:
578
dead
7.
8. Collision
KLM
747-‐200
and
Pan
Am
747-‐100;
1977,
Tenerife,
Canary
Islands:
578
dead
contribu8ng
factors:
• bomb
threat
Las
Palmas
• poor
visibility
(mist)
• runway
ligh8ng
out
of
order
• airport
extremely
crowded
• (many
planes
parked
on
the
taxiways)
• impa8ence
/
hurry
/
irrita8on
(we’ve
waited
too
long….)
• ambiguous
communica8on
“you
are
‘cleared’ “
-‐-‐-‐
for
what?
“is
he
not
clear
then…?”
• Steep
hierarchy
gradient
• emergency
pa8ent
arrives
in
ER
-‐-‐>
OR
• anesthesia
understaffed
• OR
overbooked
• anesthesia
induc8on
takes
very
long
(we’ve
waited
too
long….
get
on
with
it)
• instruments
not
ready
• ambiguous
communica8on
I
thought
you
said:
‘give
protamine’.
….
• Steep
hierarchy
gradient
?
Recognize
this
?
9. Introduction to Human Factors
l ‘To say accidents are due to human failing is
like saying falls are due to gravity. It is true
but it does not help us prevent them’ Trevor
Kletz
l Human factors engineering is about
designing the workplace and the equipment
in it to accommodate for limitations of human
performance
11. Role of Human Factors
l User-Centered Design
l Systems designed to fit people (not vice-versa).
l Reduces training time.
l Minimizes human error.
l Improves comfort, safety, and productivity.
16. 16
FATIGUE MANAGEMENT
Anesthesia
and
fatigue
Australian
Incident
MonitotingStudy,
1987-‐1997
MORRIS
&
Morris,
Anaesth.Intensive Care
2000
Nature of incidents
Relative percentage of
advense events
ONo fatigue
OFatigue
5 10 15 20 25 30%
Fluid
error
Drug
error
Dose
error
Obstructions
17. Approaches to Problem-Solving
l Equipment Design – change physical equipment
l Task Design – change how task is accomplished
l Environmental Design – change features of the work
environment such as temperature, lighting, sound
l Training – change worker behavior by providing skills
and teaching procedures
l Selection – recognizes individual differences in ability to
accomplish work
18. “If an error is possible, someone will
make it. The designer must assume that
all possible errors will occur and design
so as to minimize the chance of the error
in the first place, or its effects once it
gets made”
Norman, The Design of Everyday
Things, 2001
19.
20. Congenital Heart Surgery and Human Factors
• Bristol Infirmary Inquiry report (2000): 30% of
children undergoing heart surgery were given
less than adequate care characterized by a lack
of communication, leadership, and teamwork
• Manitoba Pediatric Cardiac Inquest (2001) linked
human factors to less than adequate care
• Duke, heart-lung ABO incompatible transplant,
US
• Radboud Medical Centre, Nimegen, Netherlands
21. Congenital HD discharge mortality, 2011
l Ventricular septal defect (VSD) repair -- 0.6% (range, 0% to
5.1%),
l Tetralogy of Fallot (TOF) repair --1.1% (range, 0% to 16.7%),
l Complete atrioventricular canal repair (AVC)-- 2.2% (range, 0% to
20%),
l Arterial switch operation (ASO)-- 2.9% (range, 0% to 50%),
l ASO --VSD-- 7.0% (range, 0% to 100%),
l Fontan operation --1.3% (range, 0% to 9.1%),
l Truncus arteriosus repair-- 10.9% (0% to 100%),
l Norwood procedure-- 19.3% (range, 0% to 100%).
l Mortality rates between centers for the Norwood procedure, for
which the Bayesian-estimated range (95% probability interval)
after risk-adjustment was 7.0% (3.7% to 10.3%) to 41.6% (30.6%
to 57.2%).
Jacobs et al Ann Thorac Surg 2011;92:2184–92.
22. Pediatric Cardiac Surgery
A highly complex, low error-tolerant
l Highly dependent upon a sophisticated
organizational structure, coordinated efforts of
team members, and high levels of cognitive and
technical performance
l High-risk populations such as neonates in
particular, exhibit a fragile physiology
l Human factors, institution and surgeon-specific
volumes, complexity of cases, and systems
failures have been linked to variable outcomes
-deLeval 2000; Walsh 2001
23.
24. Research questions
l How do teams learn and recover so well?
l How do adverse conditions, mediated by team and
task processes, lead to negative outcomes (non-
routine events and negative team outcomes)?
l Can we reduce the negative outcomes by means of
an intervention focused at the team level (non-
technical skills) or through the conditions adjustment
loop?
30. Teamwork in the Cardiac
Operating Theatre
S
1A
SN
P
ACR
Perfusion
HLM
Anaesthetic
Workstation
2A
AR
Pumps
& Drips
Coding for
TEAMS:
S1=Primary
Surgeon,
S2=Assisting
Surgeon1
S3=Assisting
Surgeon2
A1=Anesthetist
A2=Anesthetic
Nurse
P1=Perfusionist
P2=Perfusionist
N1= Assisting
Nurse
N2=Circulating
Nurse
31. Observation Method
• 2 HF trained PHD observers
• Handwritten notes
• Scoring case complexity (1-25)
• Coding case outcome at discharge (1-4)
• Technical and non-technical skills
• High interrater reliability/kappy >0.7
Schraagen, JM, et al, 2010, 2011
32.
33. Observation Data
l 102 cases-Boston Children’s; U of Chicago and U of Miami
l 9/1/05 - 12/30/07
l 102 cases
l ~ 700 hours of observations
l @1300 annotated events
l ~ 70%: < 1 year old
l Mean case complexity - 11.7 (range 3.5-24.5)
l 42 cases, Netherlands
l 10/08-3/10
l 200 hours of observations
l Mean case complecity, 10.7
l 400 events
Galvan C, Bache E, Mohr J, Barach P. Progress Pediatric Cardiology,
2005;20:13-20.; Schrageen J, Barach P. 2009
34. My
‘Idiot’s
Guide’
to
Human
factors:
l ‘Hard
Stuff’:
l people
interacAng
with
machines
l People
interacAng
with
computers
l People
interacAng
with
automaAon
l ‘So_
Stuff’:
l People
working
with
people:
l Team
performance
l handovers
l Culture
35. Safety/learning at the “Coal Face”
l Initiation of bypass without sufficient heparin is catastrophic
l Hospital A
l Surgeon: Heparin please
l Anaesthetist: Okay, heparin
l Anaesthetist: Heparin going in
l Surgeon: Are we ready to go on bypass?
l Anaesthetist: Yes, ready
l Perfusionist: Yes, I’m ready
l Hospital B:
l Surgeon: Okay?
l Anaesthetist: Yes
l Surgeon: Alright then
“It’s fine if you know how we do it here.”
“About 6 months ago when we had
a bit of an incident with someone
new, but they weren’t here long.”
No recent heparin incidents
Catchpole K, 2011, in press
36. Process Mapping
l Ovals are beginnings and ends
l Boxes are steps or activities
l Diamonds are decision points
l Questions with yes/no answers
l Arrow indicates direction and sequence
37. 37Draft 4-2-04
Pediatric Cardiovascular Surgical Care
Our aim is to improve the process of cardiovascular surgical care, starting with
the child's referral for surgery and ending with the child's first post-discharge follow-up visit.
Cardiologist
Presents Case at
Cardiac Cath
Conference
Does Child
Need
Surgery?
Cardiologist
Notifies Child/
Family About
Surgery
Child Arrives for
Surgical Clinic
Visit
Child Arrives for
Pre-Op Hospital
Visit
Child Arrives for
Surgery (day of,
unless from NICU
or PICU)
(T, W, TH)
(H&P, pre-op teaching,
schedule surgery,
reserve room for
surgery )
Child and Family
Wait in Pre-op
Holding Room
(M400)
Transport child
to OR
Family to Surgical
Waiting Room
PICU Receives
Patient
Information From
Surgery, Via NP
PICU Receives
Multiple Updates
From Surgery,
Via NP
Report (what
happened in OR,
what lines, etc.)
OR team
transports child
to PICU
Child arrives in
PICU and is
stabilized
Discharged
Home (from
PICU,
Intermediate, or
Floor)
No
Surgery
Child has
Appointment with
Cardiologist
Cardiologist
Follows-Up with
Child/Family
Nurse Sets up
PICU
First Follow-Up in Clinic
(1-2 weeks post discharge)
Cardiologist
Makes Referral
for Surgery
NP Calls Family
to Answer
Questions and
Schedule Clinic
Visit
Yes
Diagnostic
Evaluation
Complete?
Completed while
Child on Table
Yes
No
Discharge
Planning Begins -
Case Managers
Pull Census
Report
Page 2
Page 3
Pre-op events
and initial
sedation
CHD detected
prenatally, in NICU,
by pediatrician, or
other modes of
presentation
RESULTS
Barach P. Anesthesia and Analgesia, 2007
38. Technical Aspects
l CTA based observational tool
l Checklist with narrative
Schraagen JM, et al, 2009.
39. Risk Mapping and Risk analysis
Main Prospective methods
l Work Domain Analysis
l Preliminary hazard analysis (PHA)
l Failure mode and effect analysis (FMEA)
l failure mode effect and criticality analysis
(FMECA)
l Hazard and operability study (HAZOP)
l Hazard analysis and critical control point
(HACCP)
l Probabilistic risk assessment (PRA)
39
Pascal
Bonnabry,
forum
Romand,
Lausanne
19.4.2005
40. Systems errors
l Adverse outcomes
l rarely have a single cause
l are the result of multiple system errors that
“line up” eventually to create a system failure
l Correction of system errors must focus on
the system processes, not the individuals
l A human factors engineering approach is
needed
l Improvement mediated thru the
microsystem
Carthey J, et al 2001; Catchpole K, et al 2007; Galvin C et al, 2005;
Barach P, et al 2008, Schraagen J, et al, 2010, 2011
41. Anesthesiologist meets with patient in surgical holding area
Pre-op events and premedication
Patient transported to OR
Patient enters OR
Insertion of lines and induction of anesthesia
Patient prepared for surgery
Incision
Dissection
Cannulation
Go on cardiopulmonary bypass (CPB)
Identification of structures
Surgical repair
Off CPB
Heparin reversed
Hemostasis
Chest closed
Prepare for move and update ICU
Team leaves with patient for ICU
Arrive at ICU
ICU nurses take over
Anesthesiologist or surgeon gives ICU attending report
Transport to OR
Pre-Surgery/
Anes. Induction
Surgery/Pre-
Bypass
Surgery/Bypass
Surgery/Post-
Bypass
Transport to ICU
Handoff
Process Flow Domain Major Events
2%
21%
12%
15%
45%
5%
0%
42. Major Team Failures
Paediatric Cardiac
l Swab causes compression of right coronary artery
l Ex-sanguination during post-bypass heamofiltering
l Omission of key surgical step
l Premature separation from bypass due to breakdown in teamwork
l Aortic homograft ruptured during sternotomy
l Incorrectly labeled homograft
l Difficult management of activated clotting time
Orthopaedics
l Multiple uncertainty leads to teamwork and task breakdown.
Examples of minor failures implicated in major failure sequences:
Communication/co-ordination failures in 5 out of 8 major failures
Absences in 4 out of 8 major failures
Equipment failures in 4 out of 8 major failures
Vigilance/awareness failures in 3 out of 8 major failures
43. Outcom
e N
Average
case
complexit
y
(Aristotle
score)
Average
length of
surgery
Average
No
of major
events/
case
Average No
of minor
events /case
1 50 10.5 200.7 1.06 15.3
2 7 14.3 190.3 1.23 17
3 9 13.6 174.9 1.00 13.6
4 4 18.7 330.1 2.25 11.5
Outcome scale: 1- excellent; 2-moderate ill; 3-severely ill; 4-death
Outcomes Related to Complexity and Number of
Events
.Bognar A, Bacha E, Nevo I, Ahmad A, Barach P. Society of Cardiovascular Anesthesia,
May 2005.
44. Fig. 4 The distribution of types of major events
0
5
10
15
C
ardiovascular
Ventilation
BleedingLine
Placem
ent
SurgicalTechn...
C
ardiopulm
onar...
Blood
P
roduct
C
om
m
unication...
C
ognitiveInstrum
ent
M
edication
Echo
SterilityM
onitoring
Transport
Type of the event
Numberofevents
Fig. 5 The distribution of types of minor events
0
100
200
300
C
om
m
unication...
Instrum
ent
Line
Placem
ent
SterilityC
ardiopulm
onar...
Transport
M
onitoring
C
ardiovascular
Ventilation
SurgicalTechn...
C
ognitiveM
edication
Blood
P
roduct
BleedingEcho
Type of the event
Numberofevent
Figure 4. 44% of
major events were
cardiovascular,
ventilation and
bleeding problems
(patient related
problems)
Figure 5. 44 % of
all minor events
communication/
coordination and
instrumentation
problems were
detected (not
patient related
problems)
Distribution of Major and Minor Events
45. Identifying non-technical skills
Current approach:
l Mini STAR, e.g.
l How well did you sleep last night?
l Are you well-prepared?
l Do you have any concerns about equipment, people,
process?
l Safety Culture Assessment (U. Chicago)
l Patient Safety statements
l Workload, staffing and supervision
l Communication in the OR
l Detailed process checklist paediatric cardiac
surgery
l Non-technical skills checklist (NOTECHS)
48. NOTECHS Tool – Part 2
2 dimensions (total 4)
Schraagen, JM, et al 2009, 2010
49. Conceptual model based on Reason’s model showing the role of the environment as a latent condition or
barrier to adverse events in health care settings.
Sources: Dickerman and Barach (2008); Joseph et al 2008; Patti and Barach (2011); Cassin and Barach
(2012); Sanchez and Barach (2012)
Socio-technical approach to safety and quality
50. Process Organisation
– Task Allocation
– Task sequence
– Discipline and composure
Teamwork
– Leadership
– Involvement
– Briefing
Threat and Error Management
– Checklists
– Predicting and Planning
– Situation Awareness
Lessons from Nuclear Power
and Aviation
Technology
Training Regimes
54. High Reliability Organizations
l Environment rich with potential for errors
l Unforgiving social and political environment
l Learning through experimentation difficult
l Complex processes
l Complex technology
Weick, KE and Sutcliffe, KM, 1999
55. Mindfulness and Safety in HRO’s
1. Preoccupation with failure
Regarding small, inconsequential errors as a
symptom that something is wrong; finding the
half-event
2. Sensitivity to operations
Paying attention to what’s happening on the front
line at the shop floor
3. Reluctance to simplify
Encouraging diversity in experience, perspective,
and opinion
4. Commitment to resilience
Developing capabilities to detect, contain, and
bounce-back from events that do occur
5. Deference to expertise
Pushing decision making down to the
person with the most related knowledge and
expertise
58. Human Factors Contributing to Mishaps
l Normalization of deviance
l Poor communication
l Production pressure
l Fatigue and stress
l Emergency operations
l Inadequate provider experience
l Inadequate familiarity with equipment, device, surgical procedure,
anesthetic technique
l Lack of skilled assistance or supervision
l Afferent overload (excess stimuli or noise)
l Normalcy bias (assuming alarms are ‘false alarms’
l Faulty or absent policy and procedures
Prielipp R, Anesthesia & Analgesia. 2010;110(5):1499-1502.
59. Apply human factors thinking to
your work environment
1. Human behaviour can be predicted with
reasonable accuracy
2. Avoid reliance on memory
3. Make things visible
4. Review and simplify processes
5. Standardize common processes and procedures
6. Routinely use checklists
7. Decrease the reliance on vigilance
60. “No matter how well equipment is
designed, no matter how sensible
regulations are, no matter how much
humans can excel in their
performance, they can never be
better than the system that bounds
them.”
Captain Daniel Maurino, Human Factors Coordinator
International Civil Aviation Organization