background image

Resuscitation

 

83 (2012) 63–

 

69

Contents

 

lists

 

available

 

at

 

SciVerse

 

ScienceDirect

Resuscitation

j o

 

u

 

r n

 

a l

 

h o m

 

e p a g e

 

:

 

w w w . e l s e v i e r . c o m / l o c a t e / r e s u s c i t a t i o n

Clinical

 

Paper

Impact

 

of

 

resuscitation

 

system

 

errors

 

on

 

survival

 

from

 

in-hospital

 

cardiac

 

arrest

Joseph

 

P.

 

Ornato

a

,

, Mary

 

Ann

 

Peberdy

a

,

b

,

 

Renee

 

D.

 

Reid

a

,

 

V.

 

Ramana

 

Feeser

a

,

 

Harinder

 

S.

 

Dhindsa

a

,

 

for

the

 

NRCPR

 

Investigators

c

a

Department

 

of

 

Emergency

 

Medicine,

 

Virginia

 

Commonwealth

 

University,

 

Richmond,

 

VA,

 

United

 

States

b

Department

 

of

 

Internal

 

Medicine

 

&

 

Emergency

 

Medicine,

 

Virginia

 

Commonwealth

 

University,

 

Richmond,

 

VA,

 

United

 

States

a

 

r

 

t

 

i

 

c

 

l

 

e

 

i

 

n

 

f

 

o

Article

 

history:

Received

 

14

 

May

 

2011

Received

 

in

 

revised

 

form

 

9

 

September

 

2011

Accepted

 

11

 

September

 

2011

Keywords:
Inhospital
Resuscitation
Quality
Outcomes
Errors

a

 

b

 

s

 

t

 

r

 

a

 

c

 

t

Background:

 

An

 

estimated

 

350,000–750,000

 

adult,

 

in-hospital

 

cardiac

 

arrest

 

(IHCA)

 

events

 

occur

 

annually

in

 

the

 

United

 

States.

 

The

 

impact

 

of

 

resuscitation

 

system

 

errors

 

on

 

survival

 

during

 

IHCA

 

resuscitation

 

has

not

 

been

 

evaluated.

 

The

 

purpose

 

of

 

this

 

paper

 

was

 

to

 

evaluate

 

the

 

impact

 

of

 

resuscitation

 

system

 

errors

on

 

survival

 

to

 

hospital

 

discharge

 

after

 

IHCA.

Methods

 

and

 

results:

 

We

 

evaluated

 

subjective

 

and

 

objective

 

errors

 

in

 

118,387

 

consecutive,

 

adult,

 

index

IHCA

 

cases

 

entered

 

into

 

the

 

Get

 

with

 

the

 

Guidelines

 

National

 

Registry

 

of

 

Cardiopulmonary

 

Resuscitation

database

 

from

 

January

 

1,

 

2000

 

through

 

August

 

26,

 

2008.

 

Cox

 

regression

 

analysis

 

was

 

used

 

to

 

determine

the

 

relationship

 

between

 

reported

 

resuscitation

 

system

 

errors

 

and

 

other

 

important

 

clinical

 

variables

 

and

the

 

hazard

 

ratio

 

for

 

death

 

prior

 

to

 

hospital

 

discharge.

 

Of

 

the

 

108,636

 

patients

 

whose

 

initial

 

IHCA

 

rhythm

was

 

recorded,

 

resuscitation

 

system

 

errors

 

were

 

committed

 

in

 

9,894/24,467

 

(40.4%)

 

of

 

those

 

with

 

an

 

ini-

tial

 

rhythm

 

of

 

ventricular

 

fibrillation

 

or

 

pulseless

 

ventricular

 

tachycardia

 

(VF/pVT)

 

and

 

in

 

22,599/84,169

(26.8%)

 

of

 

those

 

with

 

non-VF/pVT.

 

The

 

most

 

frequent

 

system

 

errors

 

related

 

to

 

delay

 

in

 

medication

 

admin-

istration

 

(>5

 

min

 

time

 

from

 

event

 

recognition

 

to

 

first

 

dose

 

of

 

a

 

vasoconstrictor),

 

defibrillation,

 

airway

management,

 

and

 

chest

 

compression

 

performance

 

errors.

 

The

 

presence

 

of

 

documented

 

resuscitation

system

 

errors

 

on

 

an

 

IHCA

 

event

 

was

 

associated

 

with

 

decreased

 

rates

 

of

 

return

 

of

 

spontaneous

 

circu-

lation,

 

survival

 

to

 

24

 

h,

 

and

 

survival

 

to

 

hospital

 

discharge.

 

The

 

relative

 

risk

 

of

 

death

 

prior

 

to

 

hospital

discharge

 

based

 

on

 

hazard

 

ratio

 

analysis

 

was

 

9.9%

 

(95%

 

CI

 

7.8,

 

12.0)

 

more

 

likely

 

for

 

patients

 

whose

 

ini-

tial

 

documented

 

rhythm

 

was

 

non-VF/pVT

 

when

 

resuscitation

 

system

 

errors

 

were

 

reported

 

compared

 

to

when

 

no

 

errors

 

were

 

reported.

 

It

 

was

 

34.2%

 

(95%

 

CI

 

29.5,

 

39.1)

 

more

 

likely

 

for

 

those

 

with

 

VF/pVT.

Conclusions:

 

The

 

presence

 

of

 

resuscitation

 

system

 

errors

 

that

 

are

 

evident

 

from

 

review

 

of

 

the

 

resuscitation

record

 

is

 

associated

 

with

 

decreased

 

survival

 

from

 

IHCA

 

in

 

adults.

 

Hospitals

 

should

 

target

 

the

 

training

 

of

first

 

responders

 

and

 

code

 

team

 

personnel

 

to

 

emphasize

 

the

 

importance

 

of

 

early

 

defibrillation,

 

early

 

use

of

 

vasoconstrictor

 

medication,

 

and

 

compliance

 

with

 

ACLS

 

protocols.

© 2011 Elsevier Ireland Ltd. All rights reserved.

1.

 

Introduction

The

 

Institutes

 

of

 

Medicine

 

(IOM)

 

landmark

 

publication

 

(“To

 

Err

is

 

Human”)

 

estimated

 

that

 

at

 

least

 

44,000

 

and

 

perhaps

 

as

 

many

 

as

98,000

 

Americans

 

die

 

in

 

hospitals

 

each

 

year

 

as

 

a

 

result

 

of

 

med-

ical

 

errors.

1

Although

 

the

 

magnitude

 

of

 

the

 

problem

 

has

 

been

夽 A

 

Spanish

 

translated

 

version

 

of

 

the

 

abstract

 

of

 

this

 

article

 

appears

 

as

 

Appendix

in

 

the

 

final

 

online

 

version

 

at

 

doi:10.1016/j.resuscitation.2011.09.009

.

∗ Corresponding

 

author

 

at:

 

Department

 

of

 

Emergency

 

Medicine,

 

Virginia

 

Com-

monwealth

 

University,

 

1250

 

East

 

Marshall

 

Street

 

 

Main

 

Hospital

 

2nd

 

Floor,

 

Suite

500,

 

Richmond,

 

VA

 

23298-0401,

 

United

 

States.

 

Tel.:

 

+1

 

804

 

828

 

5250;

 

fax:

 

+1

 

804

828

 

8590.

E-mail

 

address:

 

ornato@aol.com

 

(J.P.

 

Ornato).

c

For

 

the

 

American

 

Heart

 

Association’s

 

Get

 

With

 

the

 

Guidelines

 

 

Resuscitation

(National

 

Registry

 

of

 

Cardiopulmonary

 

Resuscitation)

 

Investigators,

 

see

Appendix

 

A

.

questioned,

2

the

 

Canadian

 

Adverse

 

Events

 

(AE)

 

Study

 

confirms

an

 

alarming

 

frequency

 

of

 

in-hospital

 

AEs

 

(7.5

 

per

 

100

 

hospital

admissions;

 

95%

 

confidence

 

interval

 

[CI],

 

5.7–9.3),

 

36.9%

 

(95%

 

CI,

32.0–41.8%)

 

of

 

which

 

are

 

potentially

 

preventable.

 

Death

 

occurred

in

 

20.8%

 

(95%

 

CI,

 

7.8–33.8%)

 

of

 

cases.

The

 

American

 

Heart

 

Association

 

(AHA)

 

Get

 

with

 

the

 

Guidelines

National

 

Registry

 

of

 

Cardiopulmonary

 

Resuscitation

 

(NRCPR)

 

col-

lects

 

data

 

on

 

adult

 

and

 

pediatric

 

in-hospital

 

cardiac

 

arrest

 

(IHCA)

events

 

from

 

approximately

 

10%

 

of

 

hospitals

 

in

 

the

 

United

 

States.

3

From

 

this

 

registry,

 

NRCPR

 

investigators

 

have

 

documented

 

lower

survival

 

from

 

adult

 

in-hospital

 

cardiac

 

arrest

 

(1)

 

on

 

nights

 

and

weekends

 

likely

 

due,

 

at

 

least

 

in

 

part,

 

to

 

system

 

factors,

3

(2)

 

when

defibrillation

 

is

 

delayed

 

greater

 

than

 

2

 

min

 

in

 

patients

 

whose

 

ini-

tial

 

IHCA

 

rhythm

 

is

 

ventricular

 

fibrillation

 

or

 

pulseless

 

ventricular

tachycardia

 

(VF/pVT),

4

and

 

(3)

 

in

 

certain

 

hospital

 

locations.

5

The

 

purpose

 

of

 

this

 

paper

 

was

 

to

 

determine

 

whether

 

the

presence

 

of

 

resuscitation

 

system

 

errors

 

reported

 

to

 

NRCPR

 

are

0300-9572/$

 

 

see

 

front

 

matter ©

 

 2011 Elsevier Ireland Ltd. All rights reserved.

doi:

10.1016/j.resuscitation.2011.09.009

background image

64

J.P.

 

Ornato

 

et

 

al.

 

/

 

Resuscitation

 

83 (2012) 63–

 

69

associated

 

with

 

lower

 

likelihood

 

of

 

survival

 

in

 

adult

 

patients

 

who

experience

 

an

 

IHCA.

2.

 

Methods

2.1.

 

Data

 

collection

 

and

 

integrity

NRCPR

 

is

 

a

 

prospective,

 

observational,

 

multi-center

 

perfor-

mance

 

improvement

 

registry

 

of

 

IHCA

 

events.

 

Hospitals

 

join

voluntarily

 

and

 

pay

 

an

 

annual

 

fee

 

for

 

data

 

support

 

and

 

report

 

gen-

eration.

Hospital

 

medical

 

records

 

on

 

sequential

 

IHCA

 

events

 

are

abstracted

 

by

 

trained,

 

NRCPR-certified,

 

performance

 

improvement

personnel

 

at

 

each

 

participating

 

institution.

 

All

 

data

 

elements

 

have

standardized

 

definitions

 

allowing

 

aggregate

 

data

 

analysis

 

from

multiple

 

sites,

 

and

 

all

 

data

 

transfer

 

is

 

in

 

compliance

 

with

 

the

 

Health

Insurance

 

Portability

 

and

 

Accountability

 

Act.

 

Oversight

 

for

 

opera-

tions

 

is

 

provided

 

by

 

the

 

American

 

Heart

 

Association,

 

a

 

scientific

advisory

 

board,

 

and

 

an

 

executive

 

database

 

steering

 

committee.

Multiple

 

efforts

 

are

 

taken

 

to

 

assure

 

data

 

integrity,

 

including

 

data

abstractor

 

certification

 

prior

 

to

 

allowing

 

data

 

entry,

 

over

 

300

 

soft-

ware

 

checks

 

and

 

smart

 

skips

 

to

 

assist

 

with

 

accurate

 

data

 

entry,

 

and

ongoing

 

abstractor

 

training

 

with

 

monthly

 

user’s

 

group

 

calls

 

and

 

an

annual

 

user’s

 

group

 

conference.

 

Even

 

though

 

the

 

most

 

challeng-

ing

 

data

 

points

 

to

 

collect

 

during

 

resuscitation

 

are

 

event

 

times

 

and

intervals,

 

these

 

are

 

documented

 

in

 

a

 

high

 

percentage

 

of

 

cases

 

in

NRCPR

 

hospitals.

 

For

 

example,

 

in

 

this

 

analysis

 

the

 

time

 

intervals

from

 

IHCA

 

onset

 

to

 

start

 

of

 

CPR

 

and

 

first

 

vasopressor

 

administra-

tion

 

were

 

captured

 

in

 

89%

 

and

 

76%,

 

respectively.

 

Further

 

details

 

of

the

 

NRCPR

 

database

 

and

 

data

 

integrity

 

can

 

be

 

found

 

in

 

previous

publications.

6,7

2.2.

 

Study

 

outcomes

The

 

primary

 

study

 

outcome

 

was

 

survival

 

to

 

hospital

 

discharge.

Secondary

 

outcomes

 

were

 

return

 

of

 

spontaneous

 

circulation

(ROSC)

 

and

 

survival

 

for

 

24

 

h

 

after

 

IHCA.

2.3.

 

Inclusion/exclusion

 

criteria

The

 

current

 

analysis

 

includes

 

all

 

consecutive,

 

adult

 

(age

≥18

 

years),

 

initial,

 

pulseless

 

IHCA

 

events

 

entered

 

from

 

549

 

dif-

ferent

 

hospitals

 

from

 

January

 

1,

 

2000

 

through

 

August

 

26,

 

2008.

All

 

adults

 

(

≥18

 

years

 

of

 

age)

 

who

 

experienced

 

an

 

in-hospital

resuscitation

 

event

 

and

 

who

 

had

 

documentation

 

of

 

initial

 

heart

rhythm

 

were

 

eligible

 

for

 

inclusion.

 

An

 

event

 

is

 

defined

 

as:

 

(1)

cardiopulmonary

 

arrest

 

requiring

 

chest

 

compressions

 

and/or

 

defib-

rillation,

 

or

 

(2)

 

acute

 

respiratory

 

compromise

 

requiring

 

emergency

assisted

 

ventilation

 

leading

 

to

 

cardiopulmonary

 

arrest

 

requiring

chest

 

compressions

 

and/or

 

defibrillation.

 

All

 

events

 

must

 

also

 

elicit

a

 

resuscitation

 

response

 

by

 

facility

 

personnel

 

and

 

have

 

a

 

resusci-

tation

 

record

 

completed.

 

Events

 

are

 

excluded

 

if

 

the

 

arrest

 

begins

outside

 

of

 

the

 

hospital,

 

is

 

limited

 

to

 

a

 

shock

 

delivered

 

by

 

an

implanted

 

cardioverter-defibrillator

 

(ICD),

 

or

 

occurs

 

on

 

a

 

patient

with

 

a

 

pre-existing

 

do

 

not

 

attempt

 

resuscitation

 

order.

 

For

 

patients

having

 

multiple

 

IHCA

 

events

 

during

 

the

 

same

 

hospitalization,

 

only

the

 

first

 

event

 

was

 

analyzed.

2.4.

 

Categorization

 

and

 

counting

 

of

 

resuscitation

 

system

 

error

types

The

 

NRCPR

 

database

 

contains

 

a

 

section

 

for

 

the

 

data

 

abstractor

at

 

each

 

hospital

 

site

 

to

 

capture

 

“self-reported”

 

system

 

errors

 

that

were

 

noted

 

during

 

or

 

following

 

the

 

resuscitation

 

effort

 

on

 

the

 

code

record

 

and/or

 

hospital

 

chart

 

or

 

to

 

document

 

errors

 

reported

 

by

resuscitation

 

members.

 

Types

 

of

 

error

 

categories

 

included

 

alert-

ing

 

hospital-wide

 

resuscitation

 

response,

 

airway

 

management,

vascular

 

access,

 

chest

 

compression,

 

defibrillation,

 

medications,

leadership,

 

protocol

 

deviation,

 

and

 

equipment

 

function

 

issues,

 

with

specific

 

subcategories

 

listed

 

in

 

Table

 

1

.

 

Each

 

type

 

of

 

individual

 

error

had

 

a

 

checkbox

 

that

 

was

 

to

 

be

 

filled

 

in

 

by

 

the

 

data

 

abstractor

 

if

supported

 

by

 

the

 

code

 

record

 

and/or

 

hospital

 

chart

 

documenta-

tion,

 

or

 

conveyed

 

by

 

personnel

 

on

 

scene.

 

For

 

analysis

 

of

 

this

 

data,

one

 

point

 

was

 

assigned

 

for

 

each

 

individual

 

error

 

box

 

checked

 

on

 

a

given

 

IHCA

 

event.

 

We

 

totaled

 

the

 

number

 

of

 

cases

 

with

 

individual

system

 

errors

 

and

 

created

 

subtotals

 

for

 

the

 

various

 

types

 

of

 

errors

to

 

calculate

 

the

 

number

 

of

 

cases

 

in

 

which

 

any

 

error

 

occurred

 

and

to

 

catalogue

 

and

 

quantify

 

the

 

specific

 

types

 

of

 

errors

 

that

 

occurred.

We

 

counted

 

and

 

added

 

one

 

point

 

on

 

each

 

case

 

for

 

each

 

NRCPR

“process

 

of

 

care

 

exception”

 

error

 

defined

 

as:

 

(1)

 

delay

 

of

 

>5

 

min

from

 

IHCA

 

event

 

recognition

 

to

 

the

 

first

 

dose

 

of

 

a

 

vasoconstric-

tor

 

(epinephrine

 

or

 

vasopressin)

 

on

 

events

 

that

 

had

 

a

 

duration

 

of

>5

 

min;

 

or

 

(2)

 

delay

 

of

 

>2

 

min

 

from

 

IHCA

 

event

 

recognition

 

to

 

the

first

 

defibrillation

 

shock

 

in

 

patients

 

whose

 

initial

 

documented

 

IHCA

rhythm

 

was

 

VF/pVT.

 

“Resuscitation

 

system

 

errors”

 

were

 

defined

as

 

the

 

total

 

number

 

of

 

reported

 

system

 

errors

 

plus

 

the

 

number

 

of

“process

 

of

 

care

 

exceptions”

 

in

 

each

 

case.

 

When

 

the

 

process

 

of

 

care

exception

 

matched

 

a

 

self

 

reported

 

error

 

(e.g.,

 

delay

 

in

 

time

 

to

 

first

shock,

 

time

 

to

 

first

 

vasoconstrictor

 

administration),

 

only

 

one

 

point

was

 

assigned

 

for

 

the

 

error

 

to

 

avoid

 

double

 

counting.

2.5.

 

Statistical

 

analysis

All

 

data

 

analyses

 

were

 

performed

 

using

 

PASW

 

Statistics

 

ver-

sion

 

17.0.2

 

(SPSS,

 

Inc.,

 

Chicago,

 

IL).

 

Chi

 

square

 

and

 

95%

 

confidence

intervals

 

were

 

used

 

for

 

comparison

 

of

 

descriptive

 

variables.

 

ANOVA

using

 

Scheffe’s

 

test

 

for

 

multiple

 

comparisons

 

was

 

used

 

to

 

analyze

the

 

relationship

 

between

 

0,

 

1,

 

2,

 

3

 

or

 

more

 

errors

 

and

 

outcome

variables.

 

Cox

 

regression

 

analysis

 

was

 

used

 

to

 

determine

 

the

 

rela-

tionship

 

between

 

reported

 

resuscitation

 

system

 

errors

 

as

 

well

 

as

other

 

important

 

clinical

 

variables

 

and

 

the

 

hazard

 

ratio

 

for

 

death

prior

 

to

 

hospital

 

discharge.

 

The

 

status

 

of

 

the

 

primary

 

outcome

(i.e.,

 

life

 

or

 

death)

 

was

 

ascertained

 

at

 

three

 

time

 

points

 

follow-

ing

 

resuscitation:

 

(1)

 

whether

 

ROSC

 

occurred;

 

(2)

 

24

 

h

 

following

the

 

IHCA

 

event;

 

and

 

(3)

 

at

 

hospital

 

discharge.

 

Cox

 

regression

 

anal-

ysis

 

was

 

also

 

used

 

to

 

determine

 

the

 

hazard

 

ratio

 

for

 

the

 

time

 

of

day

 

(day/evening

 

or

 

7a–11p

 

vs.

 

night

 

or

 

11p–7a)

 

and

 

day

 

of

 

week

(weekday

 

or

 

M–F,

 

weekend

 

or

 

S–Su)

 

on

 

the

 

reporting

 

of

 

resuscita-

tion

 

system

 

errors

 

during

 

a

 

IHCA

 

event.

3.

 

Results

A

 

total

 

of

 

118,387

 

in-hospital,

 

adult,

 

index

 

IHCA

 

cases

 

were

entered

 

into

 

the

 

NRCPR

 

database

 

from

 

January

 

1,

 

2000

 

through

August

 

26,

 

2008.

 

Of

 

these,

 

84,440

 

(71.3%)

 

had

 

no

 

system

 

errors

recorded

 

and

 

33,947

 

(28.7%)

 

had

 

one

 

or

 

more

 

system

 

errors

recorded.

 

Of

 

the

 

cases

 

with

 

system

 

errors,

 

26,919

 

(22.7%)

 

had

1

 

error,

 

5614

 

(4.7%)

 

had

 

2

 

system

 

errors,

 

and

 

1414

 

(1.2%)

had

 

3

 

or

 

more

 

system

 

errors.

 

Of

 

all

 

cases

 

in

 

which

 

the

 

initial

rhythm

 

was

 

recorded,

 

84,169/108,636

 

(77.5%)

 

had

 

non

 

VF/pVT

and

 

24,467/108,636

 

(22.5%)

 

had

 

VF/pVT.

 

Errors

 

were

 

committed

 

in

22,599/84,169

 

(26.8%)

 

of

 

non

 

VF/pVT,

 

and

 

in

 

9894/24,467

 

(40.4%)

of

 

those

 

with

 

VF/pVT

 

as

 

the

 

first

 

documented

 

rhythm.

The

 

distribution

 

of

 

system

 

errors

 

for

 

all

 

IHCA

 

patients

 

by

 

cate-

gory

 

is

 

noted

 

in

 

Table

 

1

.

 

The

 

most

 

frequent

 

system

 

errors

 

related

to

 

delay

 

in

 

medication

 

administration

 

(>5

 

min

 

time

 

from

 

event

recognition

 

to

 

first

 

dose

 

of

 

a

 

vasoconstrictor),

 

defibrillation,

 

air-

way

 

management,

 

and

 

chest

 

compression

 

performance

 

errors.

 

The

5.4%

 

of

 

cases

 

having

 

a

 

delay

 

in

 

defibrillation

 

of

 

>2

 

min

 

repre-

sents

 

the

 

percentage

 

of

 

patients

 

with

 

defibrillation

 

delays

 

using

 

all

background image

J.P.

 

Ornato

 

et

 

al.

 

/

 

Resuscitation

 

83 (2012) 63–

 

69

65

Table

 

1

Distribution

 

of

 

resuscitation

 

system

 

errors.

Error

 

category

N

%

 

of

 

cases

%

 

of

 

reported

 

resuscitation

 

system

 

errors

Alerting

Delay

 

in

 

alerting

 

code

 

team

232

 

0.2

1.1

Pager

 

issue

 

66

 

0.1

Other

 

266

 

0.2

Airway

Aspiration

 

related

 

to

 

provision

 

of

 

airway

 

225

 

0.2

17.0

Airway

 

insertion

 

delay

2672

2.3

Delayed

 

recognition

 

of

 

airway

 

misplacement

 

328

 

0.3

Intubation

 

attempted,

 

not

 

achieved

319

 

0.3

Multiple

 

intubation

 

attempts

 

3314

 

2.8

Other

 

1939

 

1.6

Vascular

 

access

Delay

 

in

 

obtaining

 

access

846

0.7

3.4

Inadvertent

 

arterial

 

cannulation

37

<0.1

Infiltration

 

or

 

inadvertent

 

disconnection

 

of

 

IV

 

line

415

 

0.4

Other

 

480

 

0.4

Chest

 

compression

Compression

 

rate

 

of

 

∼100/min

 

not

 

maintained

 

2530

2.1

7.2

Interruption

 

of

 

compressions

 

>10

 

s

 

at

 

any

 

time

 

177

 

0.1

Delay

 

in

 

starting

 

chest

 

compressions

 

155

 

0.1

No

 

board

 

used

 

beneath

 

patient

 

during

 

compressions

 

758

 

0.6

Other

 

120

 

0.1

Defibrillation

Defibrillation

 

delay

 

>2

 

min

 

from

 

event

 

recognition

 

in

patients

 

with

 

an

 

initial

 

CA

 

rhythm

 

of

 

VF/pVT

 

due

 

to

insufficient

 

trained

 

personnel

 

or

 

defibrillator

 

not

immediately

 

available

6364

 

5.4

19.0

Energy

 

level

 

too

 

high/low

 

based

 

on

 

AHA

 

guidelines

 

485

 

0.4

Incorrect

 

defibrillator

 

paddle/pad

 

placement

 

32

 

<0.1

Defibrillator

 

malfunction

131

0.1

Shock

 

given,

 

not

 

indicated

 

957

 

0.8

Shock

 

indicated,

 

not

 

given

607

 

0.5

Other

 

1068

 

0.9

Medications

First

 

vasopressor

 

(epinephrine

 

or

 

vasopressin)

 

delay

>5

 

min

 

from

 

event

 

recognition

 

in

 

cases

 

with

 

event

duration

 

>5

 

min

20,035

16.9

42.5

Wrong

 

route

 

of

 

administration

 

202

 

0.2

Wrong

 

dosage

 

575

 

0.5

Wrong

 

medication

 

selection

 

1190

 

1.0

Code

 

team

 

leadership

Delay

 

in

 

identifying

 

team

 

leader

276

0.2

2.8

Knowledge

 

of

 

equipment

 

107

 

0.1

Knowledge

 

of

 

medications/protocols

539

 

0.5

Knowledge

 

of

 

team

 

member

 

roles

 

223

 

0.2

Code

 

team

 

oversight

 

196

 

0.2

Too

 

many

 

individuals

 

present

 

in

 

room

 

105

 

0.1

Protocol

 

deviation

 

from

 

AHA

 

Guidelines

Deviation

 

from

 

AHA

 

ACLS

 

guideline

 

recommendations

 

1521

 

1.3

3.1

Other

 

84

 

0.1

Equipment

 

issues

Availability

 

1100

 

0.9

3.9

Malfunction

 

540

 

0.5

Other

 

387

 

0.3

patients

 

as

 

the

 

denominator

 

including

 

those

 

without

 

initial

 

shock-

able

 

rhythms.

 

There

 

were

 

6364

 

patients

 

with

 

defibrillation

 

delays

of

 

>2

 

min

 

out

 

of

 

20,125

 

patients

 

with

 

an

 

initial

 

shock-able

 

rhythm

documented,

 

yielding

 

a

 

delay

 

in

 

defibrillation

 

in

 

31.6%

 

of

 

patients

in

 

whom

 

defibrillation

 

was

 

indicated.

Table

 

2

 

displays

 

the

 

demographic

 

characteristics

 

of

 

IHCA

 

events

with

 

and

 

without

 

reported

 

resuscitation

 

system

 

errors.

 

Cases

 

in

which

 

system

 

errors

 

were

 

documented

 

were

 

more

 

likely

 

to

 

be

 

male,

not

 

witnessed/monitored

 

at

 

the

 

time

 

of

 

arrest,

 

during

 

the

 

night

(11p–7a),

 

on

 

weekends

 

(S–Su),

 

initially

 

in

 

VF/pVT,

 

medical

 

(vs.

 

sur-

gical

 

or

 

cardiac)

 

type

 

admissions,

 

or

 

non-ICU

 

patients

 

(vs.

 

ICU,

 

ED,

or

 

OR/PACU

 

patients)

 

than

 

cases

 

in

 

which

 

no

 

system

 

errors

 

were

documented.

Resuscitation

 

system

 

errors

 

occurred

 

in

 

the

 

highest

 

percentage

of

 

IHCA

 

events

 

in

 

non-ICU

 

inpatient

 

areas

 

(9838/24,378,

 

40.4%);

they

 

were

 

least

 

frequently

 

noted

 

on

 

IHCA

 

events

 

occurring

 

in

ICU/stepdown

 

units

 

(17,692/69,643,

 

25.8%),

 

EDs

 

(3491/12,830,

27.2%),

 

or

 

Operating

 

Room/Post-Anesthesia

 

Care

 

Units

 

(685/2913,

23.5%)

 

(p

 

=

 

.0001).

The

 

presence

 

of

 

documented

 

resuscitation

 

system

 

errors

 

on

 

an

IHCA

 

event

 

was

 

associated

 

with

 

decreased

 

rates

 

of

 

ROSC,

 

survival

to

 

24

 

h,

 

and

 

survival

 

to

 

hospital

 

discharge

 

(

Fig.

 

1

)

 

in

 

all

 

patients

 

as

 

a

group

 

and

 

in

 

those

 

whose

 

initial

 

documented

 

IHCA

 

rhythm

 

was

separated

 

into

 

VF/pVT

 

or

 

non-VF/pVT

 

categories.

 

Characteristics

that

 

increased

 

the

 

likelihood

 

of

 

death

 

prior

 

to

 

hospital

 

discharge

in

 

patients

 

whose

 

initial

 

rhythm

 

was

 

non-VF/pVT

 

included:

 

male

gender,

 

when

 

the

 

event

 

occurred

 

at

 

night

 

or

 

on

 

a

 

weekend,

 

or

 

when

there

 

were

 

documented

 

resuscitation

 

system

 

errors

 

(

Fig.

 

2

).

 

There

was

 

a

 

lower

 

likelihood

 

of

 

death

 

prior

 

to

 

hospital

 

discharge

 

if

 

the

event

 

was

 

witnessed

 

or

 

monitored,

 

if

 

the

 

type

 

of

 

patient

 

was

 

car-

diac

 

or

 

surgical

 

(as

 

opposed

 

to

 

general

 

medical),

 

and

 

if

 

the

 

patient

location

 

was

 

OR/PACU.

background image

66

J.P.

 

Ornato

 

et

 

al.

 

/

 

Resuscitation

 

83 (2012) 63–

 

69

Table

 

2

Demographic

 

characteristics

 

of

 

in-hospital

 

CA

 

events

 

with

 

and

 

without

 

resuscitation

 

system

 

errors.

No

 

resuscitation

 

system

 

errors

 

Resuscitation

 

system

 

errors

 

p

N

 

(total

 

=

 

118,387)

 

84,440

 

(71.3%)

 

33,947

 

(28.7%)

 

Age

 

[mean,

 

95%

 

CI]

66.07

 

[65.96,

 

66.18]

66.36

 

[66.20,

 

66.53]

 

.008

Male

48,428/84,440

 

(57.4%)

 

19,951/33,947

 

(58.8%)

 

.0001

Witnessed

 

or

 

monitored

 

arrest

 

69,361/70,152

 

(98.9%)

 

25,729/26,636

 

(96.6%)

 

.0001

Initial

 

documented

 

CA

 

rhythm

 

VF/pVT

 

14,573/76,143

 

(19.1%)

 

9894/32,493

 

(30.4%)

 

.0001

Patient

 

type

Cardiac

 

29,329/81,750

 

(35.9%)

 

11,487/33,757

 

(34.0%)

.0001

Medical

34,777/81,750

 

(42.5%)

15,483/33,757

 

(45.9%)

Surgical

17,644/81,750

 

(21.6%)

6787/33,757

 

(20.1%)

CA

 

event

 

location

ICU

 

or

 

stepdown/telemetry

 

area

 

50,951/77,057

 

(66.1%)

 

17,692/31,707

 

(55.8%)

.0001

Non-ICU

 

inpatient

 

area

 

14,539/77,057

 

(18.9%)

 

9838/31,707

 

(31.0%)

Emergency

 

department

 

9339/77,057

 

(12.1%)

 

3491/31,707

 

(11.0%)

Operating

 

room

 

or

 

PACU

2228/77,057

 

(2.9%)

685/31,707

 

(2.2%)

Time

 

of

 

day

Day-evening

 

(7a–11p)

 

54,888/80,480

 

(68.2%)

 

22,387/33,777

 

(66.3%)

.0001

Night

 

(11p–7a)

 

25,592/80,480

 

(31.8%)

 

11,390/33,777

 

(33.7%)

Day

 

of

 

week

Weekday

 

(M–F)

 

56,783/81,970

 

(69.3%)

 

23,236/10.711

 

(68.4%)

.003

Weekend

 

(S–Su)

 

25,187/81,970

 

(30.7%)

 

10,711/33,947

 

(31.6%)

Fig.

 

1.

 

Effect

 

of

 

any

 

resuscitation

 

system

 

errors

 

on

 

an

 

IHCA

 

event

 

and

 

the

 

rate

 

of

 

ROSC,

 

survival

 

for

 

24

 

h,

 

and

 

survival

 

to

 

hospital

 

discharge

 

for

 

all

 

patients

 

and

 

those

 

with

 

an

initial

 

documented

 

IHCA

 

rhythm

 

of

 

non-VF/pVT

 

and

 

VF/pVT.

Fig.

 

2.

 

Association

 

between

 

the

 

number

 

of

 

resuscitation

 

system

 

errors

 

during

 

an

 

IHCA

 

event

 

and

 

a

 

significantly

 

increased

 

hazard

 

ratio

 

for

 

death

 

prior

 

to

 

hospital

 

discharge

stratified

 

by

 

patients

 

with

 

an

 

initial

 

IHCA

 

rhythm

 

of

 

non-VF/pVT

 

or

 

VF/pVT.

background image

J.P.

 

Ornato

 

et

 

al.

 

/

 

Resuscitation

 

83 (2012) 63–

 

69

67

Characteristics

 

that

 

increased

 

the

 

likelihood

 

of

 

death

 

prior

 

to

hospital

 

discharge

 

in

 

patients

 

whose

 

initial

 

rhythm

 

was

 

VF/pVT

included:

 

when

 

the

 

event

 

occurred

 

at

 

night

 

or

 

on

 

a

 

weekend,

 

or

when

 

there

 

were

 

documented

 

resuscitation

 

system

 

errors

 

(

Fig.

 

2

).

There

 

was

 

a

 

lower

 

likelihood

 

of

 

death

 

prior

 

to

 

hospital

 

discharge

if

 

the

 

type

 

of

 

patient

 

was

 

cardiac

 

or

 

surgical

 

(as

 

opposed

 

to

 

gen-

eral

 

medical)

 

or

 

if

 

the

 

patient

 

location

 

was

 

ED.

 

The

 

relative

 

risk

 

of

death

 

prior

 

to

 

hospital

 

discharge

 

based

 

on

 

hazard

 

ratio

 

analysis

 

was

9.9%

 

(95%

 

CI

 

7.8,

 

12.0)

 

more

 

likely

 

for

 

patients

 

whose

 

initial

 

docu-

mented

 

rhythm

 

was

 

non-VF/pVT

 

when

 

resuscitation

 

system

 

errors

were

 

reported

 

compared

 

to

 

when

 

no

 

errors

 

were

 

reported.

 

It

 

was

34.2%

 

(95%

 

CI

 

29.5,

 

39.1)

 

more

 

likely

 

for

 

those

 

with

 

VF/pVT.

Only

 

certain

 

types

 

of

 

resuscitation

 

system

 

errors

 

were

 

asso-

ciated

 

with

 

an

 

increased

 

hazard

 

ratio

 

for

 

death

 

prior

 

to

 

hospital

discharge.

 

For

 

patients

 

whose

 

initial

 

rhythm

 

was

 

non-VF/pVT,

delays

 

in

 

obtaining

 

vascular

 

access

 

(i.e.,

 

an

 

intravenous

 

line)

 

and

medication

 

errors

 

increased

 

the

 

hazard

 

ratio

 

for

 

death

 

prior

 

to

hospital

 

discharge

 

(

Fig.

 

3

).

 

For

 

patients

 

whose

 

initial

 

rhythm

 

was

VF/pVT,

 

defibrillation

 

problems

 

and

 

medication

 

errors

 

increased

the

 

hazard

 

ratio

 

for

 

death

 

prior

 

to

 

hospital

 

discharge.

4.

 

Discussion

The

 

principal

 

finding

 

in

 

this

 

study

 

is

 

that

 

the

 

presence

 

of

 

resus-

citation

 

system

 

errors

 

is

 

associated

 

with

 

decreased

 

survival

 

from

IHCA

 

in

 

adults.

 

More

 

errors

 

were

 

noted

 

in

 

patients

 

whose

 

initial

documented

 

IHCA

 

rhythm

 

was

 

VF/pVT

 

as

 

opposed

 

to

 

those

 

with

non-shock-able

 

rhythms.

 

This

 

finding

 

is

 

particularly

 

relevant

 

clin-

ically,

 

given

 

that

 

the

 

majority

 

of

 

survivors

 

of

 

IHCA

 

are

 

those

 

with

initial

 

VF/pVT.

3

Our

 

findings,

 

although

 

much

 

broader,

 

support

 

those

 

reported

by

 

Chan

 

et

 

al.

4

who

 

evaluated

 

6789

 

sequential

 

patients

 

with

 

VF

as

 

the

 

first

 

documented

 

rhythm

 

in

 

the

 

NRCPR

 

database

 

and

 

found

that

 

30.1%

 

of

 

this

 

cohort

 

underwent

 

defibrillation

 

more

 

than

 

2

 

min

after

 

initial

 

recognition

 

of

 

their

 

IHCA.

 

Patients

 

with

 

delayed

 

defib-

rillation

 

had

 

a

 

significantly

 

lower

 

likelihood

 

of

 

ROSC

 

(adjusted

 

odds

ratio,

 

0.55;

 

95%

 

CI,

 

0.49–0.62;

 

p

 

<

 

0.001)

 

and

 

survival

 

at

 

24

 

h

 

after

the

 

cardiac

 

arrest

 

(adjusted

 

odds

 

ratio,

 

0.52;

 

95%

 

CI,

 

0.46–0.58;

p

 

<

 

0.001).

 

The

 

Chan

 

study

 

was

 

the

 

first

 

large

 

scale

 

analysis

 

of

 

data

in

 

IHCA

 

patients

 

identifying

 

a

 

specific

 

defibrillation

 

delay

 

time

 

cut-

off

 

(2

 

min

 

or

 

less

 

after

 

event

 

recognition)

 

that

 

negatively

 

impacts

survival.

 

Our

 

analysis,

 

in

 

a

 

larger

 

sample

 

size

 

from

 

the

 

same

 

reg-

istry,

 

confirms

 

that

 

defibrillation

 

system

 

errors,

 

including

 

a

 

>2

 

min

delay

 

from

 

IHCA

 

recognition

 

to

 

initial

 

defibrillation

 

accounts

 

for

higher

 

mortality

 

in

 

the

 

initial

 

VF/pVT

 

group,

 

but

 

it

 

also

 

found

 

an

association

 

between

 

medication

 

errors

 

and

 

a

 

lower

 

likelihood

 

of

survival.

Peberdy

 

et

 

al.

3

found

 

that

 

survival

 

to

 

discharge

 

following

 

in-

hospital

 

cardiac

 

arrest

 

is

 

lower

 

during

 

nights

 

(14.7%

 

[95%

 

CI,

14.3–15.1%]

 

vs.

 

19.8%

 

[95%

 

CI,

 

19.5–20.1%])

 

or

 

weekends

 

(20.6%

[95%

 

CI,

 

20.3–21%]

 

vs.

 

17.4%

 

[95%

 

CI,

 

16.8–18%])

 

compared

 

with

day/evening

 

or

 

weekdays,

 

even

 

after

 

accounting

 

for

 

many

 

poten-

tially

 

confounding

 

IHCA

 

event

 

and

 

hospital

 

factors.

 

Our

 

current

study

 

confirms

 

that

 

nights

 

and

 

weekends

 

are

 

associated

 

with

 

an

increased

 

hazard

 

ratio

 

for

 

death

 

prior

 

to

 

hospital

 

discharge

 

in

patients

 

regardless

 

of

 

the

 

initial

 

documented

 

IHCA

 

rhythm

 

and

 

goes

a

 

step

 

beyond

 

the

 

previously

 

reported

 

data

 

in

 

demonstrating

 

an

increased

 

hazard

 

ratio

 

for

 

death

 

during

 

those

 

times

 

in

 

which

 

there

is

 

an

 

increase

 

in

 

resuscitation

 

system

 

errors,

 

thus

 

suggesting

 

a

 

link

between

 

increased

 

error

 

on

 

nights

 

and

 

weekends

 

and

 

decreased

survival

 

during

 

those

 

times.

Kayser

 

et

 

al.,

5

demonstrated

 

that

 

ED

 

location

 

was

 

an

 

inde-

pendent

 

predictor

 

of

 

improved

 

survival,

 

speculating

 

that

 

this

 

was

due

 

to

 

both

 

the

 

requirement

 

for

 

ED

 

staff

 

to

 

receive

 

basic

 

and

the

 

advanced

 

cardiac

 

life

 

support

 

training,

 

and

 

their

 

frequent

experience

 

in

 

performing

 

resuscitation

 

compared

 

to

 

clinicians

working

 

on

 

general

 

hospital

 

floors.

 

Our

 

paper

 

supports

 

this

 

hypoth-

esis

 

by

 

confirming

 

that

 

IHCA

 

events

 

occurring

 

in

 

the

 

ED

 

have

 

a

relatively

 

low

 

percentage

 

of

 

cases

 

with

 

resuscitation

 

system

 

errors.

Not

 

all

 

types

 

of

 

errors

 

were

 

associated

 

with

 

an

 

increased

 

hazard

ratio

 

for

 

death

 

prior

 

to

 

hospital

 

discharge

 

in

 

our

 

study.

 

Olasveen-

gen

 

et

 

al.,

17

randomized

 

out-of-hospital

 

cardiac

 

arrest

 

patients

 

to

receive

 

ACLS

 

treatment

 

with

 

and

 

without

 

intravenous

 

drug

 

admin-

istration

 

and

 

determined

 

that

 

survival

 

to

 

discharge

 

was

 

the

 

same

 

in

both

 

groups.

 

However,

 

they

 

were

 

not

 

able

 

to

 

identify

 

the

 

timeframe

in

 

which

 

the

 

drug

 

therapy

 

was

 

given

 

in

 

the

 

group

 

that

 

received

an

 

intravenous

 

line

 

and

 

medication.

 

In

 

our

 

IHCA

 

population,

 

we

were

 

able

 

to

 

demonstrate

 

decreased

 

survival

 

when

 

the

 

first

 

vaso-

constrictor

 

was

 

administered

 

>5

 

min

 

after

 

IHCA

 

onset

 

in

 

patients

whose

 

arrest

 

lasted

 

for

 

at

 

least

 

5

 

min.

 

The

 

difference

 

in

 

our

 

find-

ings

 

is

 

likely

 

due

 

to

 

the

 

fact

 

that

 

the

 

time

 

from

 

event

 

onset

 

to

 

first

drug

 

administration

 

is

 

usually

 

much

 

shorter

 

in-

 

vs.

 

out-of-hospital.

In

 

our

 

study,

 

the

 

mean

 

time

 

from

 

event

 

onset

 

to

 

first

 

epinephrine

was

 

2.4

 

(95%

 

CI,

 

2.3,

 

2.4)

 

min.

 

In

 

contrast,

 

the

 

time

 

interval

 

from

collapse

 

to

 

first

 

epinephrine

 

in

 

standard

 

vs.

 

high

 

dose

 

epinephrine

pre-hospital

 

trials

 

is

 

approximately

 

20

 

min.

18

Much

 

of

 

the

 

focus

 

on

 

patient

 

safety

 

and

 

error

 

prevention

 

in

 

hos-

pitals

 

focuses

 

on

 

interventions

 

that

 

relate

 

to

 

patient

 

interactions

that

 

are

 

of

 

higher

 

volume

 

and

 

lower

 

acuity

 

compared

 

to

 

resus-

citation.

 

In

 

our

 

study,

 

the

 

impact

 

of

 

resuscitation

 

team

 

errors

 

on

survival

 

varied

 

widely

 

by

 

both

 

error

 

type

 

and

 

the

 

initial

 

docu-

mented

 

heart

 

rhythm,

 

with

 

some

 

errors

 

causing

 

little

 

or

 

no

 

impact

and

 

others

 

being

 

associated

 

with

 

a

 

significantly

 

lower

 

survival,

 

par-

ticularly

 

in

 

the

 

group

 

of

 

patients

 

with

 

initial

 

VF/pVT.

 

Given

 

that

resuscitation

 

practices

 

need

 

to

 

occur

 

consistently

 

well

 

throughout

all

 

areas

 

of

 

the

 

hospital

 

and

 

that

 

responders

 

will

 

always

 

have

 

vary-

ing

 

degrees

 

of

 

expertise

 

and

 

experience,

 

our

 

findings

 

suggest

 

that

resuscitation

 

training

 

should

 

be

 

targeted

 

to

 

emphasize

 

avoiding

 

the

types

 

of

 

errors

 

having

 

the

 

greatest

 

impact

 

on

 

survival

 

(e.g.,

 

delays

 

in

initial

 

defibrillation

 

and

 

medication

 

administration

 

and

 

adherence

to

 

ACLS

 

protocols).

An

 

increasing

 

body

 

of

 

evidence

 

indicates

 

that

 

effective

 

lead-

ership

 

and

 

team

 

work

 

rather

 

than

 

just

 

individual

 

knowledge,

skills,

 

and

 

attitudes

 

are

 

required

 

to

 

optimize

 

outcomes

 

and

 

min-

imize

 

errors

 

in

 

a

 

variety

 

of

 

medical

 

emergencies.

8–15,16

Specific

behaviors

 

have

 

been

 

identified

 

that

 

contribute

 

to

 

effective

 

leader-

ship

 

including

 

providing

 

orienting

 

remarks,

 

inviting

 

team

 

member

contributions,

 

promoting

 

exchange

 

of

 

information

 

and

 

clear

 

com-

munication,

 

and

 

avoiding

 

performing

 

physical

 

tasks

 

that

 

can

 

be

assigned

 

to

 

others

 

during

 

the

 

emergency.

8

In

 

addition,

 

there

 

are

four

 

teamwork

 

behavior

 

principles

 

that

 

can

 

help

 

to

 

avoid

 

medi-

cal

 

errors:

 

(1)

 

the

 

leader

 

should

 

voice

 

specific

 

findings

 

rather

 

than

diagnosing

 

the

 

problem

 

prematurely;

 

(2)

 

all

 

members

 

of

 

the

 

team

should

 

“think

 

out

 

loud”

 

and

 

“talk

 

to

 

the

 

room”

 

as

 

the

 

case

 

unfolds;

(3)

 

the

 

leader

 

should

 

direct

 

period

 

reviews

 

of

 

quantitative

 

infor-

mation

 

(e.g.,

 

drug

 

dose,

 

time,

 

response);

 

and

 

(4)

 

all

 

members

 

of

 

the

team

 

should

 

double-check

 

crucial

 

data.

8

Unfortunately,

 

the

 

majority

 

of

 

these

 

leadership

 

and

 

team

 

behav-

iors

 

could

 

not

 

be

 

measured

 

in

 

this

 

study.

 

This

 

is

 

the

 

likely

explanation

 

for

 

why

 

we

 

were

 

unable

 

to

 

demonstrate

 

an

 

association

with

 

reported

 

“poor

 

team

 

leadership”

 

and

 

survival

 

in

 

our

 

study.

 

The

definition

 

of

 

“good

 

team

 

leadership”

 

is

 

subjective

 

and

 

may

 

not

 

be

reported

 

consistently

 

among

 

institutions

 

or

 

that

 

that

 

magnitude

 

of

the

 

effect

 

on

 

survival

 

of

 

other

 

resuscitation

 

errors

 

drowns

 

out

 

the

impact

 

of

 

poor

 

team

 

leadership.

Our

 

observations

 

confirm

 

the

 

association

 

between

 

the

 

presence

of

 

resuscitation

 

system

 

errors

 

that

 

are

 

evident

 

from

 

review

 

of

 

the

resuscitation

 

record

 

and

 

decreased

 

survival

 

from

 

IHCA

 

in

 

adults.

However,

 

they

 

do

 

not

 

point

 

to

 

a

 

specific

 

solution

 

to

 

the

 

problem.

A

 

number

 

of

 

recent

 

simulator-based

 

studies

 

have

 

identified

qualitatively

 

and

 

quantitatively

 

similar

 

problems

 

to

 

those

 

noted

background image

68

J.P.

 

Ornato

 

et

 

al.

 

/

 

Resuscitation

 

83 (2012) 63–

 

69

Fig.

 

3.

 

Association

 

between

 

various

 

types

 

of

 

resuscitation

 

system

 

errors

 

and

 

the

 

hazard

 

ratio

 

for

 

death

 

prior

 

to

 

hospital

 

discharge

 

in

 

patients

 

whose

 

initial

 

IHCA

 

rhythm

 

was

non-VF/pVT

 

and

 

VF/pVT.

during

 

clinical

 

resuscitation.

4,19–24

Team

 

training

 

can

 

improve

 

per-

formance

 

during

 

simulation

 

of

 

medical

 

emergencies,

25–28

but

 

not

all

 

training

 

is

 

equal

 

or

 

effective.

12–14

Siassakos

 

et

 

al.,

12–14

have

 

iden-

tified

 

specific

 

elements

 

required

 

to

 

improve

 

outcome,

 

including

multi-professional

 

training

 

of

 

all

 

healthcare

 

providers

 

who

 

man-

age

 

an

 

emergency

 

in

 

a

 

realistic

 

simulation

 

setting.

 

These

 

elements

need

 

to

 

be

 

incorporated

 

into

 

team

 

and

 

leadership

 

training,

 

which

are

 

now

 

recommended

 

in

 

the

 

2010

 

American

 

Heart

 

Association

Guidelines

 

for

 

Adult

 

and

 

Pediatric

 

Advanced

 

Cardiovascular

 

Life

Support.

29,30

4.1.

 

Limitations

General

 

limitations

 

of

 

NRCPR

 

include:

 

(1)

 

registry

 

hospitals

 

may

not

 

be

 

representative

 

of

 

all

 

hospitals;

 

(2)

 

there

 

is

 

no

 

on-site

 

valida-

tion

 

of

 

data

 

collection;

 

(3)

 

and

 

there

 

is

 

no

 

follow-up

 

after

 

hospital

discharge.

 

In

 

addition,

 

although

 

medication

 

use

 

is

 

tracked,

 

NRCPR

does

 

not

 

attempt

 

to

 

assess

 

clinical

 

eligibility

 

for

 

each

 

medication.

These

 

limitations

 

are

 

similar

 

to

 

those

 

of

 

other

 

contemporary

 

in-

hospital

 

registries.

The

 

specific

 

limitation

 

pertinent

 

to

 

this

 

analysis

 

is

 

that

 

NRCPR

data

 

is

 

self-reported

 

by

 

having

 

a

 

trained

 

abstractor

 

review

 

hospi-

tal

 

charts

 

and

 

code

 

records.

 

The

 

Get

 

with

 

the

 

Guidelines

 

NRCPR

 

is

a

 

voluntary

 

data

 

collection/analysis

 

quality

 

improvement

 

project.

Participating

 

hospitals

 

pay

 

a

 

fee

 

to

 

the

 

American

 

Heart

 

Associa-

tion

 

to

 

have

 

the

 

data

 

analyzed,

 

benchmarked,

 

and

 

reported

 

back

to

 

them

 

quarterly.

 

With

 

hundreds

 

of

 

hospitals

 

involved

 

through-

out

 

the

 

entire

 

United

 

States

 

and

 

very

 

minimal

 

hospital

 

user

 

fees,

 

it

has

 

not

 

been

 

feasible

 

to

 

provide

 

independent

 

data

 

validation

 

at

 

the

hospital

 

level.

The

 

abstractors

 

work

 

with

 

uniform

 

definitions

 

and

 

uniform

methods

 

of

 

data

 

acquisition.

 

The

 

abstractors

 

review

 

the

 

physi-

cian

 

and

 

nurse

 

narrative

 

notes

 

in

 

the

 

hospital

 

chart

 

and

 

the

 

“code

record”

 

on

 

each

 

event.

 

Thus,

 

all

 

of

 

the

 

errors

 

identified

 

came

 

from

 

a

review

 

of

 

documentation

 

in

 

the

 

patient

 

charts.

 

The

 

abstractors

 

do

not

 

capture

 

whether

 

the

 

errors

 

were

 

“recognized”

 

in

 

the

 

narrative

notes

 

by

 

the

 

doctors

 

and

 

nurses

 

running

 

the

 

code

 

or

 

are

 

evident

from

 

the

 

documentation

 

of

 

events

 

themselves

 

(which

 

is

 

the

 

case

for

 

the

 

majority

 

of

 

errors

 

that

 

we

 

looked

 

at).

There

 

is

 

no

 

way

 

for

 

the

 

abstractor

 

to

 

verify

 

whether

 

errors

 

may

have

 

occurred

 

but

 

were

 

not

 

documented.

 

This

 

may

 

have

 

resulted

in

 

an

 

under-reporting

 

of

 

resuscitation

 

team

 

errors.

 

If

 

this

 

occurred,

if

 

anything,

 

it

 

would

 

increase

 

the

 

potential

 

importance

 

of

 

our

findings.

 

Finally,

 

we

 

cannot

 

exclude

 

the

 

possibility

 

that

 

individu-

als

 

completing

 

the

 

code

 

sheets

 

could

 

have

 

checked

 

off

 

more

 

“error

boxes”

 

in

 

patients

 

who

 

did

 

not

 

achieve

 

ROSC

 

than

 

the

 

boxes

 

they

would

 

have

 

checked

 

had

 

the

 

patient

 

been

 

resuscitated.

4.2.

 

Public

 

health

 

importance

The

 

public

 

health

 

importance

 

of

 

these

 

findings

 

is

 

con-

siderable.

 

Eisenberg

 

and

 

Mengert,

31

estimated

 

that

 

there

 

are

350,000–750,000

 

adult,

 

IHCA

 

events

 

per

 

year

 

in

 

the

 

United

 

States.

Eliminating

 

resuscitation

 

system

 

errors

 

has

 

the

 

potential

 

to

 

save

21,000–44,000

 

additional

 

lives

 

per

 

year

 

in

 

the

 

United

 

States

 

from

IHCA.

 

This

 

figure

 

is

 

ten

 

times

 

larger

 

than

 

the

 

estimated

 

2000–4000

additional

 

lives

 

saved

 

per

 

year

 

from

 

out-of-hospital

 

cardiac

 

arrest

that

 

drove

 

widespread

 

deployment

 

of

 

public

 

access

 

defibrillation

programs

 

throughout

 

the

 

country.

13

5.

 

Conclusions

We

 

conclude

 

that

 

the

 

presence

 

of

 

resuscitation

 

system

 

errors

that

 

are

 

evident

 

from

 

review

 

of

 

the

 

resuscitation

 

record

 

is

 

associ-

ated

 

with

 

decreased

 

survival

 

from

 

IHCA

 

in

 

adults.

 

Hospitals

 

should

target

 

their

 

training

 

of

 

first

 

responders

 

and

 

code

 

team

 

personnel

to

 

emphasize

 

the

 

importance

 

of

 

early

 

defibrillation

 

when

 

indi-

cated,

 

early

 

use

 

of

 

vasoconstrictor

 

medication,

 

and

 

compliance

 

with

established

 

AHA

 

ACLS

 

resuscitation

 

protocols.

Conflict

 

of

 

interest

 

statement

None

 

of

 

the

 

authors

 

have

 

any

 

relevant

 

conflicts.

Funding

 

sources

None.

Acknowledgement

None.

Appendix

 

A.

Get

 

with

 

the

 

Guidelines

 

– National

 

Registry

 

of

 

Cardiopulmonary

Resuscitation

 

(NRCPR)

 

investigators:

background image

J.P.

 

Ornato

 

et

 

al.

 

/

 

Resuscitation

 

83 (2012) 63–

 

69

69

Mary

 

E.

 

Mancini,

 

Robert

 

A.

 

Berg,

 

Emilie

 

Allen,

 

Elizabeth

 

A.

 

Hunt,

Vinay

 

M.

 

Nadkarni,

 

Scott

 

Braithwaite,

 

Graham

 

Nichol,

 

Kathy

 

Dun-

can,

 

Tanya

 

Lane

 

Truitt,

 

Melinda

 

Smyth,

 

Brian

 

Eigel,

 

Paul

 

S.

 

Chan,

Tim

 

Mader,

 

Karl

 

B.

 

Kern,

 

Sam

 

Warren,

 

Thomas

 

Noel,

 

Romergryko

Geocadin,

 

Dana

 

Edelson,

 

Vince

 

Mosesso

 

and

 

Comilla

 

Sasson.

References

1. Institute

 

of

 

Medicine.

 

To

 

err

 

is

 

human:

 

building

 

a

 

safer

 

health

 

system.

 

Washing-

ton,

 

DC:

 

National

 

Academy

 

Press;

 

2000.

2. University

 

study

 

identifies

 

problems

 

with

 

IOM

 

report.

 

Hosp

 

Case

 

Manag

2000;8(Suppl.

 

3–4):146.

3. Peberdy

 

MA,

 

Ornato

 

JP,

 

Larkin

 

GL,

 

et

 

al.

 

Survival

 

from

 

in-hospital

 

cardiac

 

arrest

during

 

nights

 

and

 

weekends.

 

JAMA

 

2008;299:785–92.

4. Chan

 

PS,

 

Krumholz

 

HM,

 

Nichol

 

G,

 

Nallamothu

 

BK.

 

Delayed

 

time

 

to

 

defibrillation

after

 

in-hospital

 

cardiac

 

arrest.

 

N

 

Engl

 

J

 

Med

 

2008;358:9–17.

5.

 

Kayser

 

RG,

 

Ornato

 

JP,

 

Peberdy

 

MA.

 

Cardiac

 

arrest

 

in

 

the

 

emergency

 

depart-

ment:

 

a

 

report

 

from

 

the

 

National

 

Registry

 

of

 

Cardiopulmonary

 

Resuscitation.

Resuscitation

 

2008;78:151–60.

6. Nadkarni

 

VM,

 

Larkin

 

GL,

 

Peberdy

 

MA,

 

et

 

al.

 

First

 

documented

 

rhythm

 

and

 

clin-

ical

 

outcome

 

from

 

in-hospital

 

cardiac

 

arrest

 

among

 

children

 

and

 

adults.

 

JAMA

2006;295:50–7.

7.

 

Peberdy

 

MA,

 

Kaye

 

W,

 

Ornato

 

JP,

 

et

 

al.

 

Cardiopulmonary

 

resuscitation

 

of

 

adults

in

 

the

 

hospital:

 

a

 

report

 

of

 

14720

 

cardiac

 

arrests

 

from

 

the

 

National

 

Registry

 

of

Cardiopulmonary

 

Resuscitation.

 

Resuscitation

 

2003;58:297–308.

8.

 

Hunziker

 

S,

 

Johansson

 

AC,

 

Tschan

 

F,

 

et

 

al.

 

Teamwork

 

and

 

leadership

 

in

 

cardiopul-

monary

 

resuscitation.

 

J

 

Am

 

Coll

 

Cardiol

 

2011;57:2381–8.

9.

 

Schaefer

 

HG,

 

Helmreich

 

RL.

 

The

 

importance

 

of

 

human

 

factors

 

in

 

the

 

operating

room.

 

Anesthesiology

 

1994;80:479.

10.

 

Schaefer

 

HG,

 

Helmreich

 

RL,

 

Scheidegger

 

D.

 

Human

 

factors

 

and

 

safety

 

in

 

emer-

gency

 

medicine.

 

Resuscitation

 

1994;28:221–5.

11.

 

Schaefer

 

HG,

 

Helmreich

 

RL,

 

Scheidegger

 

D.

 

Safety

 

in

 

the

 

operating

 

theatre

 

part

 

1:

 

interpersonal

 

relationships

 

and

 

team

 

performance.

 

Curr

 

Anaesth

 

Crit

Care

 

1995;6:48–53.

12.

 

Siassakos

 

D,

 

Bristowe

 

K,

 

Draycott

 

TJ,

 

et

 

al.

 

Clinical

 

efficiency

 

in

 

a

 

simulated

 

emer-

gency

 

and

 

relationship

 

to

 

team

 

behaviours:

 

a

 

multisite

 

cross-sectional

 

study.

BJOG

 

2011;118:596–607.

13.

 

Siassakos

 

D,

 

Draycott

 

TJ,

 

Crofts

 

JF,

 

Hunt

 

LP,

 

Winter

 

C,

 

Fox

 

R.

 

More

 

to

 

teamwork

than

 

knowledge,

 

skill

 

and

 

attitude.

 

BJOG

 

2010;117:1262–9.

14.

 

Siassakos

 

D,

 

Fox

 

R,

 

Crofts

 

JF,

 

Hunt

 

LP,

 

Winter

 

C,

 

Draycott

 

TJ.

 

The

 

management

of

 

a

 

simulated

 

emergency:

 

better

 

teamwork,

 

better

 

performance.

 

Resuscitation

2011;82:203–6.

15.

 

Tourgeman-Bashkin

 

O,

 

Shinar

 

D,

 

Zmora

 

E.

 

Causes

 

of

 

near

 

misses

 

in

 

critical

 

care

of

 

neonates

 

and

 

children.

 

Acta

 

Paediatr

 

2008;97:299–303.

16.

 

Manser

 

T.

 

Teamwork

 

and

 

patient

 

safety

 

in

 

dynamic

 

domains

 

of

 

health-

care:

 

a

 

review

 

of

 

the

 

literature.

 

Acta

 

Anaesthesiol

 

Scand

 

2009;53:

143–51.

17.

 

Olasveengen

 

TM,

 

Sunde

 

K,

 

Brunborg

 

C,

 

Thowsen

 

J,

 

Steen

 

PA,

 

Wik

 

L.

 

Intravenous

drug

 

administration

 

during

 

out-of-hospital

 

cardiac

 

arrest:

 

a

 

randomized

 

trial.

JAMA

 

2009;302:2222–9.

18.

 

Brown

 

CG,

 

Martin

 

DR,

 

Pepe

 

PE,

 

et

 

al.

 

A

 

comparison

 

of

 

standard-dose

 

and

 

high-

dose

 

epinephrine

 

in

 

cardiac

 

arrest

 

outside

 

the

 

hospital.

 

The

 

Multicenter

 

High-

Dose

 

Epinephrine

 

Study

 

Group.

 

N

 

Engl

 

J

 

Med

 

1992;327:1051–5.

19. Abella

 

BS,

 

Alvarado

 

JP,

 

Myklebust

 

H,

 

et

 

al.

 

Quality

 

of

 

cardiopulmonary

 

resusci-

tation

 

during

 

in-hospital

 

cardiac

 

arrest.

 

JAMA

 

2005;293:305–10.

20. Abella

 

BS,

 

Sandbo

 

N,

 

Vassilatos

 

P,

 

et

 

al.

 

Chest

 

compression

 

rates

 

during

cardiopulmonary

 

resuscitation

 

are

 

suboptimal:

 

a

 

prospective

 

study

 

during

 

in-

hospital

 

cardiac

 

arrest.

 

Circulation

 

2005;111:428–34.

21.

 

Kramer-Johansen

 

J,

 

Myklebust

 

H,

 

Wik

 

L,

 

et

 

al.

 

Quality

 

of

 

out-of-hospital

 

car-

diopulmonary

 

resuscitation

 

with

 

real

 

time

 

automated

 

feedback:

 

a

 

prospective

interventional

 

study.

 

Resuscitation

 

2006;71:283–92.

22. Chan

 

PS,

 

Nichol

 

G,

 

Krumholz

 

HM,

 

Spertus

 

JA,

 

Nallamothu

 

BK.

 

Hospital

 

varia-

tion

 

in

 

time

 

to

 

defibrillation

 

after

 

in-hospital

 

cardiac

 

arrest.

 

Arch

 

Intern

 

Med

2009;169:1265–73.

23. Hunziker

 

S,

 

Tschan

 

F,

 

Semmer

 

NK,

 

Howell

 

MD,

 

Marsch

 

S.

 

Human

 

factors

 

in

resuscitation:

 

lessons

 

learned

 

from

 

simulator

 

studies.

 

J

 

Emerg

 

Trauma

 

Shock

2010;3:389–94.

24.

 

Hunziker

 

S,

 

Tschan

 

F,

 

Semmer

 

NK,

 

et

 

al.

 

Hands-on

 

time

 

during

 

cardiopulmonary

resuscitation

 

is

 

affected

 

by

 

the

 

process

 

of

 

teambuilding:

 

a

 

prospective

 

ran-

domised

 

simulator-based

 

trial.

 

BMC

 

Emerg

 

Med

 

2009;9:3.

25.

 

Thomas

 

EJ,

 

Taggart

 

B,

 

Crandell

 

S,

 

et

 

al.

 

Teaching

 

teamwork

 

during

 

the

Neonatal

 

Resuscitation

 

Program:

 

a

 

randomized

 

trial.

 

J

 

Perinatol

 

2007;27:

409–14.

26.

 

Thomas

 

EJ,

 

Williams

 

AL,

 

Reichman

 

EF,

 

Lasky

 

RE,

 

Crandell

 

S,

 

Taggart

 

WR.

 

Team

training

 

in

 

the

 

neonatal

 

resuscitation

 

program

 

for

 

interns:

 

teamwork

 

and

 

quality

of

 

resuscitations.

 

Pediatrics

 

2010;125:539–46.

27.

 

Cooper

 

S.

 

Developing

 

leaders

 

for

 

advanced

 

life

 

support:

 

evaluation

 

of

 

a

 

training

programme.

 

Resuscitation

 

2001;49:33–8.

28.

 

Hunziker

 

S,

 

Buhlmann

 

C,

 

Tschan

 

F,

 

et

 

al.

 

Brief

 

leadership

 

instructions

 

improve

cardiopulmonary

 

resuscitation

 

in

 

a

 

high-fidelity

 

simulation:

 

a

 

randomized

 

con-

trolled

 

trial.

 

Crit

 

Care

 

Med

 

2010;38:1086–91.

29.

 

Bhanji

 

F,

 

Mancini

 

ME,

 

Sinz

 

E,

 

et

 

al.

 

Part

 

16:

 

education,

 

implementation,

and

 

teams:

 

2010

 

American

 

Heart

 

Association

 

Guidelines

 

for

 

Cardiopul-

monary

 

Resuscitation

 

and

 

Emergency

 

Cardiovascular

 

Care.

 

Circulation

2010;122:S920–33.

30.

 

Mancini

 

ME,

 

Soar

 

J,

 

Bhanji

 

F,

 

et

 

al.

 

Part

 

12:

 

education,

 

implementation,

 

and

teams:

 

2010

 

International

 

Consensus

 

on

 

Cardiopulmonary

 

Resuscitation

 

and

Emergency

 

Cardiovascular

 

Care

 

Science

 

with

 

Treatment

 

Recommendations.

 

Cir-

culation

 

2010;122:S539–81.

31.

 

Eisenberg

 

MS,

 

Mengert

 

TJ.

 

Cardiac

 

resuscitation.

 

N

 

Engl

 

J

 

Med

2001;344:1304–13.


Document Outline