and
good
people.
They
would
not
intentionally
harm
babies.”
Therefore,
the
fault
must
rest
with
the
computer
system.
The
estimated
cost
for
the
new
computer
system
was
a
hefty
$5
million.
Nonetheless,
they
felt
certain
they
knew
the
root
cause
and
this
was
the
necessary
solution,
and
they
were
angry
at
the
existing
IT
vendor
for
doing
this
to
them.
At
that
point,
expecting
the
group
to
ask
the
five
whys
would
have
quickly
devolved
into
the
five
whos,
getting
them
even
angrier
as
they
identified
more
and
more
people
to
blame.
Edward
Blackman,
who
headed
up
continuous
improvement
for
IT,
had
been
asked
to
the
meeting.
He
was
pretty
certain
he
was
invited
just
to
rubber-stamp
the
team’s
conclusion,
but he
could
not
stand
by
idly
and
let
the
hospital
spend
millions
of
dollars
on
computer
technology
that
might
not
even
solve
the
problem.
He
had
to
find
a
way
to
get
this
powerful
group
to
reopen
the
investigation
of
the
causes
of
the
prob-
lem,
and
he
had
always
learned
important
things
by
going
to
the
gemba.
What
he
did
not
want
to
do
was
to
go
off
on
his
own
and investigate and
then
have
to
report
back
to
the
team
members
that
they
were
wrong—which
would
almost
certainly
lead
to
defensive
reactions.
Instead,
he
respectfully
asked
them
if
they
would
be
willing
to
go
to
the
gemba
to
investigate
the
problem
further.
The
actual
clinic
was
just
a
walk
upstairs,
so
they
agreed
as
long
as
it
was
quick.
They
had
plenty
of
time
since
the
meeting
had
been
scheduled
for
four
hours,
and
they
had
spent
less
than
an
hour
get-
ting
to
what
they
were
sure
was
the
root
cause
and
solution.
Before
beginning
to
study
the
gemba,
Blackman
gathered
key
stakeholders,
including
the
director
of
the
wing, nursing
managers,
nurses,
IT
analysts,
quality coor-
dinators,
technicians,
and
administrative
assistants,
and
spent
10
minutes
in
a
confer-
ence
room
discussing
what
was
really
going
on.
The
driving
question
was
“What
is
actually
happening?,”
not
just
what
should
be
happening
according
to
procedure.
It
was
clear
that
nobody
really
knew,
so
they
hit
the
gemba
and
proceeded
to
investigate
for
two
hours.
They
interviewed
people
responsible
for
the
process,
they
videotaped
and
timed
responses,
they
played
the
roles
of
patients
and
responding
clinicians,
and
they
value-stream
mapped
the
current-state
process.
A
picture
started
to
emerge,
and
it
became
clear
to
all
that
the
problem
had
little
to
do
with
a
bad
computer
system.
It
was
an
issue
of
how
the
existing
software
was
set
up
and
how
people,
including
patients’
families,
were
trained.
The
map
of
the
floor
in
Figure
12.2
gives
a
picture
of
what
was
happening.
On
one
side
of
the
oval-shaped
floor
are
the
patient
rooms
for
babies.
In
the
center
of
the
oval,
out
of
the
line
of
sight
and
sound
of
the
patient
rooms,
is
the
registration
desk,
where
the
alert
messages
were
received.
Patient
-.
Registration
Desk
3101
B
WA
S
N
171
Bi
(O
SEVON
S
{0
d
1)
N
When
the
group
went
to
talk
to
the
administrative
assistant
at
the
desk,
Edward
asked
to
do
the
questioning,
as
he
did
not
want
a
bunch
of
high-powered
people
intim-
idating
her
and
casting
blame.
He
asked
what
happened
when
she
received
an
alert.
The
answer:
“I
turn
it
off.”
The
crowd
was
getting
restless.
He
calmly
asked
why.
Answer:
“Because
they
are
always
false
alarms.”
The
murmuring
of
the
crowd
grew
louder.
Edward
asked
how
she
knew
that.
Answer:
“When
there
is
an
alert,
I
am
sup-
posed
to
get
a
voice
follow-up.
If
there
is
no
voice
follow-up,
I
assume
it
is
a
false
alarm.
The
parents
can
signal
for
help
or
I
can
get
an
automated
signal
from
sensors
on
the
babies
detecting
a
breathing
problem.
The
babies
roll
around
all
the
time
and
trigger
the
sensors
stuck
onto
them,
causing
mainly
false
alarms”
(see
the
process
flow
in
Figure
12.3).
The
crowd
started
questioning
whether
they
really
understood
what
was
going
on
and
were
now
more
open
to
investigation.
