Intern
ati
o
n
a
l
Jo
urn
a
l
o
f
P
u
b
lic Hea
l
th Science (IJ
P
HS)
V
o
l.3
,
No
.2
,
Jun
e
2
014
, pp
. 12
9
~
13
6
I
S
SN
: 225
2-8
8
0
6
1
29
Jo
urn
a
l
h
o
me
pa
ge
: h
ttp
://iaesjo
u
r
na
l.com/
o
n
lin
e/ind
e
x.ph
p
/
IJPHS
Measuri
n
g Effi
ci
ency Of Hospit
al
s By DE
A: An E
m
pirical
Evidence F
r
om
Pakist
an
Syed
Az
iz
Ras
ool
1
,
Ab
dul S
a
boor
2
, Mu
hammad
Raashi
d
2
1
School of
Econ
omic Scien
ces, Federa
l Urdu University
of
Arts,
Sciences
& Tech
nolog
y
,
Islamab
a
d, Pak
i
stan
2
Department of
Economics & A
g
ri-E
cono
imcs,
PMAS-Arid Agricultur
e
Un
iv
ersity
, R
a
walpind
i
,
Pakistan
Article Info
A
B
STRAC
T
Article histo
r
y:
Received Apr 11, 2014
Rev
i
sed
May
3, 201
4
Accepted
May 27, 2014
There h
a
s
been
incre
a
s
i
ng focus
on effi
cien
c
y
m
eas
urem
ent
in h
eal
th s
ecto
r
around the world. This empirical stud
y
ai
ms at measuring efficiency
lev
e
l of
non-profit priv
ate organ
i
zation b
y
usin
g
Data En
velopment Analy
s
is (DEA)
in th
e health
sector of Pak
i
stan
.
DEA is non-par
a
metric lin
ear
pr
ogramming
based approach for homog
eneous decisio
n
making units. Lay
t
on
Rehmatullah Benevolen
t Trust (non-profit
private organization) is the subject
m
a
tter for inv
e
stigat
ional
ana
l
y
s
is
constitut
i
ng over 16 sub-units spreadin
g
across the countr
y
. Secondar
y
data of num
ber of patient beds, specialists and
nurses in all th
e 16 br
anch
es
of LRBT hospitals has b
een
used app
l
y
i
ng
quantit
ativ
e spe
c
i
fic
a
tion
tool
, bo
th sca
l
e
and
tech
nica
l eff
i
c
i
enc
y
l
e
vels
in a
n
environm
ent wh
ere m
u
ltip
le o
f
i
nputs and outpu
t
s
are in
place, has been used
for final
evalu
a
t
i
on. Th
e outco
m
e
s so expecte
d
will help po
li
c
y
m
a
k
e
rs to
form
ulate eff
e
c
t
i
v
e plans
and pro
g
ram
s
in order to enhanc
e the
ef
fici
enc
y
of
health measures
conducted b
y
no
n-
profit priv
ate
organizations.
Keyword:
DEA
IRS
CRS
BHU
RHU
Copyright ©
201
4 Institut
e
o
f
Ad
vanced
Engin
eer
ing and S
c
i
e
nce.
All rights re
se
rve
d
.
Co
rresp
ond
i
ng
Autho
r
:
Sy
ed Azi
z
R
a
s
ool
,
PhD
Scho
lar
,
Sch
o
o
l
o
f
Economic Scien
ces,
Fed
e
ral Urdu
Un
i
v
ersity of
Arts, Scien
c
e
& Tech
no
log
y
,
Zer
o
po
in
t, Secto
r
, G-
7,
I
s
lamab
a
d,
4
400
0, Pak
i
stan
.
Em
a
il: syed
.raso
o
l
0
@
g
m
ail.c
o
m
1.
INTRODUCTION
There
i
s
a c
o
n
t
i
nual
di
scussi
on
i
n
eco
n
o
m
i
cs re
gar
d
i
n
g
p
r
o
p
er
ext
e
nt
a
n
d
pa
rt
of
t
h
e
st
at
e persi
s
t
s
si
nce t
h
e
devel
opm
ent
of t
h
e
m
odern ec
on
o
m
i
c
s. C
u
rre
nt
scenari
o
of the
world em
phas
i
ze the im
porta
nce of
th
is b
a
sic no
ti
o
n
of th
e eco
n
o
m
ics, th
at is
“efficien
t
use of
sca
r
ce
re
so
urces”
,
as po
p
u
l
a
t
i
on of
t
h
e wo
rl
d
growi
n
g in pac
e
along
with other com
p
lexities. There
f
or
e
,
increasi
ng em
phasis has be
e
n
placed on efficiency
m
easurem
ent
o
f
bot
h
pu
bl
i
c
a
n
d
p
r
i
v
at
e sect
or
w
h
et
he
r t
h
e
y
are
pr
ofi
t
or
no
n
-
p
r
o
f
i
t
o
r
ga
ni
zat
i
ons
.
Org
a
n
i
zatio
n
a
l
stru
cture o
f
a
m
o
d
e
rn
state co
n
s
titu
tes bo
th p
r
o
f
it an
d
n
on-profit o
r
g
a
n
i
zatio
n
s
. Th
e
need for non-profit
orga
nizations
is gr
owing in pace
with
the inc
r
ease in
natural disasters and
poverty.
At the
sam
e
ti
me, th
e
requ
irem
en
t o
f
no
n-pro
f
it o
r
g
a
n
i
zation
is to
b
e
ab
le to
p
r
esen
t th
e resu
lts o
f
so
m
e
k
i
n
d
wit
h
t
h
e am
ount
o
f
m
oney
i
n
v
o
l
v
ed.
The
peo
p
l
e
w
h
o i
n
v
o
l
v
e
i
n
c
o
nt
ri
b
u
t
i
n
g t
o
wa
rd
s t
h
e
s
e n
o
n
-
p
r
o
fi
t
p
r
i
v
at
e
orga
nizations a
r
e m
o
stly long
away the
n
the
actual proj
ect
places s
o
this t
h
ing m
a
kes efficiency
m
easurem
ent
m
o
re im
port
a
nt
fo
r t
h
ose
peo
p
l
e
. T
h
e
di
ffi
c
u
l
t
y
l
i
e
s i
n
un
de
rst
a
n
d
i
n
g t
h
e ef
fi
ci
ency
of
no
n-
pr
o
f
i
t
or
ga
ni
zat
i
ons
a
n
d
t
h
e am
ou
nt
of
co
nt
ri
b
u
t
i
o
n
s
t
h
at
t
h
ey
pr
o
v
i
d
e t
o
t
h
e
s
o
ci
et
y
.
As m
e
nt
i
oned
abo
v
e t
h
at
t
h
e wo
rl
d i
s
fac
i
ng t
h
e e
v
i
l
s
of
po
ve
rt
y
,
hu
nge
r an
d sca
r
c
e
reso
urce
s,
sp
ecially th
e po
pu
lated
co
un
t
r
ies lik
e Pak
i
st
an
is facing
the
s
e kinds of iss
u
es m
o
re seve
rely as co
m
p
are
to the
d
e
v
e
l
o
p
e
d
coun
tries.
So,
k
e
ep
ing
in v
i
ew t
h
ese
g
litch
e
s
it is realistic th
at wh
y sno
w
b
a
l
lin
g
weigh
t
is
b
e
ing
placed on m
e
asuri
ng the e
f
ficiency
of non-profit orga
ni
zations. T
h
ere
f
or
e, the foc
u
s in this study is on
efficiency m
easurem
ent of
no
n-
pr
o
f
i
t
h
o
spi
t
al
as n
o
n
-
p
r
o
fi
t
p
r
i
v
at
e
or
gani
zat
i
on.
Bo
d
y
o
f
literat
u
re, confirm
s
t
h
e ro
le
o
f
allo
cativ
e efficien
cy an
d
op
ti
m
a
li
t
y
. Th
eir
wo
rk
ran
g
e
s fro
m
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
252
-88
06
IJP
H
S
V
o
l
.
3,
No
. 2,
Ju
ne
2
0
1
4
:
12
9 – 1
3
6
13
0
cor
p
orat
i
o
n t
o
i
ndi
vi
dual
s
,
cove
ri
n
g
pu
bl
i
c
and
p
r
i
v
at
e
bo
di
es i
n
cl
u
d
i
n
g n
o
n
-
p
r
o
fi
t
or
gani
zat
i
o
n
s
. The
secto
r
al
d
i
str
i
bu
tio
n of
n
on-
pr
of
it or
g
a
n
i
zatio
n
s
is p
a
r
tially n
o
n
-
pr
of
it an
d fu
lly n
on-
pr
of
it or
g
a
n
i
zatio
ns.
Health
sector
o
f
Pak
i
stan
wh
ich
con
s
titu
tes abo
u
t
R
u
p
e
es 42
b
illio
n
th
at is 0
.
2
3
percen
t
o
f
GDP also
co
nstitu
tes fu
lly fun
d
e
d
no
n-profit bod
ies sp
ecially ho
sp
itals. Th
erefore th
e scen
ar
i
o
requ
ires
h
a
v
i
ng
a
d
e
tailed
inv
e
stig
atio
n of effici
en
cy m
easu
r
emen
t o
f
non
-p
ro
fit
p
u
b
lic organ
i
zatio
n
s
.
Paki
st
an i
s
t
h
e
one
of t
h
e hi
ghe
r p
o
pul
o
u
s
cou
n
t
r
y
i
n
t
h
e So
ut
h
Asi
a
n
regi
on
wi
t
h
1
7
0
.
6 m
i
l
l
i
o
n
resi
de
nt
s. Pa
ki
st
an ha
s
pract
i
ced
basi
c he
al
t
h
ser
v
i
ces
del
i
v
ery
sy
st
em
cont
ai
ni
ng
o
f
B
HUs
, R
u
ral
H
eal
t
h
Cen
t
ers (R
HUs), Matern
ity an
d
C
h
ild
Health
Cen
t
ers (MC
H
s), Civ
il Ho
sp
itals an
d
Dispen
saries as wel
l
. Th
is
acceptable institutional m
e
c
h
anism
for
public health
services delive
r
y stand by
968
hospitals, 4,813
d
i
sp
en
sar
i
es,
5,34
5 b
a
sic h
e
alth
un
its,
5
72
r
u
r
a
l health
ce
nters, 908 m
a
ternity and c
h
il
d
health ce
nters and
29
3 TB
cent
e
rs
.
Th
e Basic
Health
Un
it (BHU) is term
ed
as a h
ealth
uni
t
t
h
at
pr
ovi
des a
n
ecessary
m
e
di
cal
cove
r t
o
50
0
0
t
o
1
0
0
0
0
l
i
v
i
n
g
wi
t
h
i
n
t
h
e
prem
i
s
es of
u
n
i
o
n c
o
unci
l
co
veri
ng
a
n
a
r
ea
of
1
5
t
o
2
5
s
q
uare
m
i
l
e
s. T
h
e
Basic Health
Un
it (BHU) aim
s
to
p
r
ov
id
e
treat
m
e
n
t
an
d
med
i
cal facilit
i
e
s in
rural commu
n
ity, b
a
sic h
ealth
educat
i
o
n,
pre
gna
ncy
cases and
basi
c heal
t
h
of chi
l
d
ren a
n
d w
o
m
e
n, im
muni
zat
i
o
n, an
d
execut
i
o
n o
f
di
sease
eradication program
s
such as
TB, poli
o
, etc. BHU ca
n se
rve as th
e su
ppo
rt b
y
m
a
k
i
n
g
h
e
alth
serv
ices with
in
th
e reach
for i
n
d
i
v
i
du
als, fam
i
lies an
d co
mmu
n
ities [1
].
Keepi
ng i
n
vi
ew t
h
e im
port
a
n
ce of p
u
b
l
i
c
sect
or ser
v
i
ces i
n
t
h
e heal
t
h
sect
or
, preci
sel
y
of
no
n-
pr
ofi
t
pu
bl
i
c
or
ga
ni
zat
i
on as
m
o
st
of t
h
e p
o
pul
at
i
on i
n
t
h
e c
o
u
n
t
ry
i
s
bel
o
w t
h
e po
vert
y
l
i
n
e,
t
h
i
s
st
udy
has
been
conducted for
efficiency anal
ysis
of t
h
e La
y
t
on R
a
hm
at
ul
l
a
B
e
nevol
e
n
t
Tru
s
t
(LR
B
T
)
hos
pi
t
a
l
s
, Pa
ki
st
an.
Th
e serv
ices of th
e
ho
sp
ital
are exte
nded i
n
all four
pr
ov
inces of th
e cou
n
try. So,
the
efficiency m
easure
m
ent
of
t
h
e
n
o
n
-
p
r
of
i
t
or
gani
zat
i
o
n
i
n
t
h
e
f
o
rm
of
LR
B
T
h
o
s
p
i
t
a
l
s
has
bee
n
t
a
ke
n i
n
t
o
acc
ou
nt
.
The efficiency
analysis pre
p
a
r
ed
a
nd
de
vel
o
ped
by
t
h
e res
earche
r
s o
f
ten
arise th
e relativ
e d
i
fficu
lty
enco
u
n
t
e
re
d i
n
eval
uat
i
n
g
t
h
e pe
rf
orm
a
nce o
f
a
fi
rm
or
an
or
ga
ni
zat
i
on.
These
or
ga
ni
zat
i
ons
o
f
t
e
n
ha
v
e
m
u
lt
i
p
l
e
i
nput
s and o
u
t
p
ut
s
t
o
be anal
y
zed
. So, f
o
r t
h
i
s
p
u
r
p
ose we are
usi
ng
Dat
a
Envel
opm
ent
Anal
y
s
i
s
(DE
A
) m
e
t
hod
ol
o
g
y
t
o
m
easure t
h
e e
ffi
cien
cy o
f
n
on-
pro
f
it pu
b
lic or
gan
i
zatio
n
i.e.
no
n-p
r
o
f
it ho
sp
itals in
our case
.
Th
is is th
e area wh
ich
h
a
s b
e
en
ign
o
red
v
e
ry
m
u
ch
in
Pak
i
stan
on
ly v
e
ry
litt
le wo
rk
h
a
s
b
e
ing
don
e
by the researc
h
ers es
pecially in
case of no
n
-
p
r
o
f
i
t
pri
v
at
e or
ga
ni
zat
i
on i
n
heal
t
h
sect
or.
Or
ga
ni
zat
i
ona
l
beha
vi
o
r
i
s
al
way
s
i
m
port
a
n
t
whet
her t
h
at
i
s
pu
bl
i
c
o
r
pri
v
at
e su
b
di
vi
si
on t
h
ere
f
ore;
t
h
e t
o
ol
s f
o
r ef
f
i
ci
ency
measurem
ent are c
ontinuous
lyim
perative.
In
m
o
st o
f
th
e
real wo
rl
d
scen
ari
o
s situ
ation
s
are
q
u
ite co
m
p
lex
an
d to
m
eet th
ese co
m
p
lex
ities th
e m
e
th
o
d
o
l
og
y shou
ld
be sim
p
le an
d
co
m
p
reh
e
n
s
iv
e
at th
e sam
e
ti
me. In
Paki
st
an
, t
h
e
r
e
exi
s
t
a fe
w
n
o
n
-
p
r
o
fi
t
pri
v
at
e or
ga
ni
zat
i
ons
i
n
heal
t
h
sect
o
r
,
but
t
h
ei
r c
o
n
t
ri
but
i
o
n
nee
d
s
t
o
be
analyzed,
whet
her they are
pe
rf
orm
i
ng ef
fi
ci
ent
l
y
or
n
o
t
?
T
h
e
ma
j
o
r obj
ec
tiv
e
s
of
th
e s
t
u
d
y
ar
e
:
a)
Eval
uat
e
a
n
d a
n
al
y
ze t
h
e e
ffi
ci
ency
o
f
pri
v
a
t
e or
ga
ni
zat
i
o
n
w
hose
ai
m
i
s
not
t
o
see
k
fo
r
pr
ofi
t
.
b)
Use o
f
no
n
-
pa
r
a
m
e
t
r
i
c
m
e
t
hod, DE
A m
e
t
hodol
ogy
f
o
r ef
fi
ci
ency
m
easure si
nce i
t
i
s
com
p
ati
b
l
e
i
n
usi
n
g
m
u
l
tip
le in
pu
ts and
o
u
t
p
u
t
s.
1.
1.
Sys
t
ematic Re
view.
The fi
nal
g
o
al
of t
h
e
n
o
n
-
p
r
o
f
i
t
pu
bl
i
c
or
ga
ni
zat
i
ons i
s
n
o
t
fi
nanci
a
l
.
T
h
eref
ore
,
we ca
nn
ot
use t
h
e
classical fin
a
ncial co
n
t
ro
l in n
on-profit o
r
g
a
n
i
zation
s
.
Sin
ce th
is is somewh
at less fo
cused
area
we find
it
im
port
a
nt
t
o
l
o
ok at
t
h
e m
e
t
h
od
ol
o
g
i
e
s t
h
at
has
been
use
d
t
o
m
easure eff
i
ci
ency
i
n
o
r
ga
ni
zat
i
ons
w
h
i
c
h a
r
e
not
dri
v
en
by
eco
nom
i
c
resul
t
s
. St
ei
n
g
r
u
b
e
r
et al.
(
2
0
0
6
)
di
sc
usse
d i
n
hi
s w
o
r
k
t
h
at
t
h
e nee
d
f
o
r ai
d
orga
nizations i
s
growi
ng in
pace with
incre
a
se in the natural disaster and
pove
rty [2].
At the sam
e
time th
e
need
f
o
r
no
n
-
p
r
o
f
i
t
o
r
ga
ni
zat
i
ons
t
o
be a
b
l
e
t
o
prese
n
t
t
h
e
r
e
sul
t
s
t
h
at
a
r
e
bei
n
g e
x
pect
ed
by
t
h
e
pe
opl
e
w
h
o
h
a
s
b
e
en
invo
lv
ed in
t
h
e
fu
lfi
l
l
m
en
t o
f
fi
n
a
ncial n
eed
s
of t
h
ese
orga
nizations
.
Henser
(2001) a
n
alyzed
that in
recent years, much
of
the anal
ytical and intellectual effort conce
r
ne
d
with the econom
i
cs
of
he
alth system
s
in
d
e
v
e
l
o
p
i
n
g
cou
n
t
ries h
a
s tend
ed
to
fo
cu
s on
asp
ects of wh
at will b
e
d
e
fin
e
d
as “allo
cativ
e” efficien
cy [3
].
Hsu
(20
10) d
e
d
u
c
ed
th
at th
e relativ
e ro
les o
f
th
e pub
lic an
d
p
r
iv
ate sect
o
r
s in
h
ealt
h
care facility h
a
v
e
an
d
cont
i
n
ue t
o
e
v
ol
ve
o
v
er t
i
m
e [4]
.
Im
pro
v
e
m
ent
s
st
em
b
ack
to 19
th
cen
t
ur
y n
e
oclass
ical econom
i
c
s
that
mark
et so
l
u
tion
s
m
a
in
to
m
o
re efficien
t allo
catio
n of
m
e
a
n
s,
do
ub
ts ov
er
how
the
healthcare m
a
rket woul
d
respon
d
t
o
th
ese in
stitu
tio
n
s
an
d, m
o
re recen
tly, to
n
e
w
p
u
b
lic m
a
n
a
g
e
men
t
th
eo
ries
an
d
t
h
e influ
e
n
ce
of
m
u
l
tilaterals. Ch
arn
e
s
et al
.
(1
9
7
8
)
di
sc
uss
e
d fi
rst
t
i
m
e
abo
u
t
D
E
A m
odel
s
an
d c
o
ncl
ude
d t
h
at
use
of t
e
rm
s
lik
e D
M
U
’
s (decisio
n
m
a
k
i
ng
u
n
it) and pro
g
ram
s
w
ill help
to em
p
h
a
size th
at your interest is cen
t
ered
o
n
d
ecision
m
a
k
i
n
g
b
y
no
t-for pro
f
it en
tities rath
er th
an
th
e mo
re cu
sto
m
ary
firm
s an
d
in
dustries [5
]. Keng
il
et
al
.
(
2
01
0
)
d
e
fi
ned
t
h
at
Dat
a
en
vel
o
pm
ent
anal
y
s
i
s
(
D
E
A
)
i
s
a
n
e
m
pi
ri
call
y
based m
e
t
hod
ol
o
g
y
t
h
at
excludes
the
need for s
o
me of t
h
e as
sum
p
tions
an
d l
i
m
i
t
a
t
i
ons
of
c
u
st
om
ar
y effic
i
ency m
easurement
approaches
[6]
.
Jacobs (200
0) d
i
scu
s
sed
th
at efficien
cy esti
m
a
tes
in
DEA
m
a
y
b
e
su
bj
ect to
th
e sa
m
e
b
i
as if
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
PH
S
I
S
SN
:
225
2-8
8
0
6
Mea
su
ring
Efficien
cy o
f
Ho
spita
ls b
y
DEA:
An
Emp
i
rica
l
Evid
en
ce from
Pa
kistan
(S
yed Aziz Ra
so
o
l
)
13
1
in
efficien
t un
its u
s
ing
low lev
e
ls of th
e en
do
g
e
no
us re
s
o
urce are set toughe
r efficiency
targets tha
n
e
qually
i
n
effi
ci
e
n
t
uni
t
s
usi
n
g m
o
re of t
h
e
reso
u
r
c
e
[7]
.
K
o
nt
o
d
i
m
opoul
o
s
an
d
Ni
akas (
2
00
6
)
revi
e
w
e
d
t
h
a
t
DEA
receive
d wide
acceptance
by researc
h
ers and practitione
rs in m
a
ny public and
pri
v
ate sectors [8].
Ko
nt
o
d
i
m
opo
u
l
os an
d Ni
a
k
as
[8]
anal
y
zed t
h
e technical and scale efficienci
es of pub
lic d
i
strict h
o
sp
it
als in
Tamil Nadu,
India.
Em
pirical analysis us
ed t
h
e
data
o
f
29
di
st
ri
ct
s
hos
pi
t
a
l
s
f
o
r t
h
e y
ear
20
0
4
-
2
0
0
5
,
Technical and
scale efficienci
es were
ass
e
ss
ed by
usi
n
g D
E
A.
Out
put
va
ri
abl
e
s i
n
cl
u
d
e
num
ber o
f
i
n
p
a
t
i
e
nt
s,
num
ber o
f
out
pat
i
e
nt
vi
si
t
s
, n
u
m
b
er of chi
l
d
del
i
v
eri
e
s, n
u
m
ber of sur
g
er
i
e
s, and n
u
m
b
er of em
ergency
cases
at
t
e
nded
,
w
h
i
l
e
num
ber o
f
st
af
f an
d n
u
m
b
er o
f
be
ds
were
us
ed as i
n
put
var
i
abl
e
s. Deat
h r
a
t
e
un
der
fi
ve
y
ears
,
l
i
f
e exp
ect
ancy
at
bi
rt
h
an
d
go
od
heal
t
h
we
re
use
d
as
o
u
t
p
ut
vari
a
b
l
e
s a
nd
num
ber
of
p
h
y
s
i
c
i
a
ns an
d
be
d
s
pe
r
10
0
0
occ
upa
nt
s and am
ount
of t
o
t
a
l
ex
pen
d
i
t
u
re
on
heal
t
h
wer
e
use
d
as i
nput
v
a
ri
abl
e
s. Zere
et al
. (
200
6)
concl
ude
d t
h
at
t
h
e t
echni
cal
and scal
e efficiencies of dist
rict hos
pitals
of Nam
i
bia [9
]
.
DEA was
u
s
ed to
assess technica
l efficiency. T
o
tal rec
u
rrent
expe
nditure
,
nu
m
b
er
of nurs
es
and num
b
er
of beds
were use
d
as
in
pu
t v
a
riab
les and
t
o
tal num
b
e
r o
f
ou
tp
atien
t
v
i
sits
an
d i
n
pat
i
e
nt
da
y
s
were
u
s
ed
as o
u
t
p
ut
vari
abl
e
s.
A
l
v
a
r
a
do
(
2006
)
ex
am
in
ed
the p
r
od
u
c
tiv
e eff
i
cien
cy of
th
e p
r
im
ar
y h
ealth
car
e cen
t
ers
o
f
C
h
ile [
1
0
]
.
Stu
d
y
u
s
ed
th
e cross sectio
n
a
l d
a
ta o
f
2
4
Ch
ilean
m
u
n
i
cip
a
lities
lo
cated
in
th
e
u
r
b
a
n
area fo
r
th
e year 2
001
. DEA
was em
ployed to evaluate techni
cal and scale efficiencies.
Osei
et al.
(2
0
0
5
)
co
ncl
u
de
d t
h
at
a st
udy
rel
y
i
n
g
o
n
a
relativ
e small sa
m
p
le o
f
17
ho
sp
itals in
Gh
an
a i
n
20
0
0
f
o
un
d t
h
at
t
h
ey
suf
f
ere
d
f
r
om
seri
ou
s
pu
re
technical ine
ffi
ciency together with
obvi
ous s
cale inefficienc
y
[11].
2.
R
E
SEARC
H M
ETHOD
The
data for this em
pirical work
o
f
a
n
o
n
-
p
r
o
fi
t
pri
v
at
e or
ga
ni
zat
i
o
n
i
s
t
a
ken
f
r
o
m
t
h
e an
nua
l
rep
o
rt
s
,
a
n
n
u
al
ne
wsl
e
t
t
e
rs a
n
d
pers
o
n
al
rec
o
r
d
s
of
LR
B
T
hos
pi
t
a
l
s
an
d i
n
f
o
rm
at
i
on o
f
heal
t
h
dep
a
rt
m
e
nt
o
f
t
h
e p
r
ovi
nci
a
l
go
ve
rnm
e
nt
ha
s al
so
bee
n
use
d
.
I
n
p
u
t
s
a
n
d
o
u
t
p
ut
s
dat
a
i
s
a
n
al
y
ze f
o
r t
h
e
y
ear 2
0
1
0
.
2.
1.
Sample
Selection
Th
e sam
p
le u
n
d
e
r st
u
d
y
will in
clu
d
e
16
ho
sp
itals o
f
Layton
Rah
m
atu
lla
Ben
e
vo
len
t
Tru
s
t (LRBT)
Pak
i
stan. Th
ese 1
6
ho
sp
itals
in
clu
d
e
2
tertiary h
o
sp
itals
i.e. K
a
r
a
ch
i and
Laho
r
e
an
d
14 seco
nd
ar
y ho
sp
itals.
Th
e
p
r
ov
in
ce-wid
e
d
i
stribu
ti
o
n
of t
h
ese
h
o
sp
itals is as fo
ll
o
w
s.
Si
nd
h h
a
s f
o
u
r
hos
pi
t
a
l
s
, Pu
n
j
ab
has si
x
,
K
h
y
b
er Pa
kht
un
k
h
wa
has t
h
ree
and B
a
l
u
c
h
i
s
t
a
n has t
w
o
of
t
h
em
. It
i
s
assu
m
e
d t
h
at
al
l
h
o
s
pi
t
a
l
s
ha
ve
ho
m
ogeno
us
deci
si
on
m
a
ki
ng
u
n
i
t
s
.
2.
2.
Appr
oaches
to Study E
ffici
ency.
Di
ffe
re
nt
m
e
t
hods t
o
exam
i
n
e t
h
e effi
ci
enc
y
are no
rm
all
y
t
a
ken i
n
t
o
ac
cou
n
t
ei
t
h
er p
a
ram
e
t
r
i
c
or
no
n
-
pa
ram
e
t
r
i
c
,
w
h
e
r
e param
e
t
r
i
c
m
e
t
hod
s are base
d on
th
e assu
m
p
tio
n th
at a
p
a
rticular fu
nctio
n
a
l
form
whe
r
eas non-param
etric
meth
ods
do not.
An alternative term can be used
as
that
m
e
thod can be statistical or
no
n
-
st
at
i
s
t
i
cal
[7]
.
St
at
i
s
t
i
cal
m
e
t
hods assu
m
e
t
h
at
t
h
e dat
a
i
s
of t
h
e st
ochast
i
c
nat
u
r
e
. Non
-
para
m
e
t
r
i
c
m
e
t
hods s
u
c
h
as DE
A t
e
n
d
t
o
be
det
e
rm
i
n
i
s
t
i
c
, whe
r
eas s
t
at
i
s
t
i
c
al
m
e
t
h
ods
d
u
e t
o
t
h
ei
r nat
u
re a
nd
ba
sed
o
n
fr
ont
i
e
r re
g
r
es
si
on m
odel
s
u
s
ed t
o
be
para
m
e
t
r
i
c
and st
o
c
hast
i
c
. Us
ual
l
y
t
h
e fro
nt
i
e
r m
odel
s
m
a
ke
speci
fi
c
assu
m
p
tio
n
s
ab
ou
t th
e i
n
eff
i
cien
cy ter
m
in
th
e m
o
d
e
ls w
h
ich
ar
e v
e
ry r
e
str
i
ctiv
e (W
agstaf
f, 1989
)
[
1
2
]
.
St
ocha
st
i
c
C
o
s
t
Fr
ont
i
e
r
(SC
F
)
us
ual
l
y
co
n
s
t
r
uct
s
a
v
e
ry
sm
oot
h pa
ram
e
t
r
i
c
fr
o
n
t
i
e
r.
DEA
has
o
n
e
of
t
h
e
ad
v
a
n
t
ag
es that it can
m
a
n
a
g
e
m
i
scella
n
e
ou
s
p
r
od
u
c
t
i
o
n
fun
c
tion
s
with
m
u
ltip
l
e
in
pu
t and
o
u
t
p
u
t
technologies.
2.
3.
Da
ta
En
velop
ment
An
alysis
(D
ea
)
Mo
del
The
DEA
,
a
m
a
t
h
em
at
i
c
al
, l
i
n
ear p
r
o
g
r
am
m
i
ng t
ool
,
wa
s de
vel
o
ped
b
y
C
h
arne
s et
al
. (1
9
7
8
)
t
o
make efficienc
y
m
easurem
ent precise a
nd ac
curate
discip
lin
e esp
ecially fo
r th
e
situ
atio
ns in
wh
ich
th
ere are
com
p
lex
situations rega
rding to
the
m
u
ltiplicity of differe
n
t inputs a
n
d out
puts [5]. It
ha
s bee
n
use
d
all
ove
r
the gl
obe
es
pe
cially in the Greece [13].
Data Env
e
lopmen
t An
alysis
(DEA), as m
e
n
tio
n
e
d in
th
e start is a lin
ear
prog
ramm
in
g
meth
od
wh
ich
helps in the m
easurem
ent of e
fficien
cy. It ba
sically co
m
e
s f
r
om
the theo
ret
i
cal
l
y
based co
ncept
of p
r
od
u
c
t
i
o
n
efficiency.
DE
A a
n
alyses the
collaboration
betwee
n inpu
t
s
t
o
a pr
od
uct
i
o
n pr
ocess (f
or
exam
pl
e,
reso
urce
s
use
d
i
n
a
hos
pi
t
a
l
)
an
d t
h
e
o
u
t
put
s
(
f
o
r
e
x
am
pl
e,
num
ber
of
pat
i
e
nt
s t
r
eat
e
d
)
of t
h
at
p
r
oces
s [
7
]
.
DEA ha
s all three
ki
nds
of
m
odels that are, acco
rding t
o
scale.
The
r
e
are
norm
ally two
scale
assum
p
t
i
ons e
m
pl
oy
ed:
cons
t
a
nt
ret
u
r
n
s t
o
scal
e (C
R
S
)
m
odel
and
va
r
i
abl
e
ret
u
r
n
s t
o
scal
e
(
V
R
S
)
m
odel
.
Th
e CRS assum
p
t
i
o
n
is su
itab
l
e wh
en
all DM
U’s a
r
e
working at a
n
optimal scale.
In
p
u
t
-
o
r
i
e
nt
at
i
o
n
m
e
t
hod c
o
n
cerns
w
h
e
n
t
h
ere a
r
e
decreas
i
ng i
n
put
s i
n
d
a
t
a
t
o
have
m
o
re e
ffi
ci
ent
Decision Maki
ng
units (DMU’s
),
while
output
-orientation m
e
thod c
o
nce
r
ns
of inc
r
easi
ng
quality of s
e
rvices
p
r
ov
id
ed
b
y
a
h
ealth
un
it with
ou
t ch
ang
i
ng
th
e qu
ality o
f
th
e inp
u
t
s
u
s
ed
in
th
at p
r
o
cess. A v
a
riab
le ret
u
rn
to
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
252
-88
06
IJP
H
S
V
o
l
.
3,
No
. 2,
Ju
ne
2
0
1
4
:
12
9 – 1
3
6
13
2
scal
e (VR
S
)
m
odel
poi
se
d
of
t
h
ree
t
y
pes;
fi
rst
,
co
nst
a
nt
ret
u
rn
t
o
scal
e (C
R
S
)
,
whi
c
h
doe
s
not
ha
ve t
o
i
m
p
r
ov
e in
pu
ts o
r
ou
tpu
t
s, seco
nd
, in
creasing
return
to
scal
e (IRS) m
o
d
e
l th
at n
eed
s to
in
fl
u
e
n
ce th
e
qu
ality
o
f
p
r
ov
id
ed
serv
ices. Th
ird
,
d
ecreasi
n
g
ret
u
rn
to
scale (DRS) m
o
d
e
l
will allev
i
ate t
h
e inp
u
t
v
o
l
ume. Th
e
soft
ware
Data
DEA
P
2.
1
vers
ion is
use
d
a
s
t
h
e
DE
A s
o
ft
w
a
re in
this
rese
arch
.
DEA
ca
n
be
defi
ned
as t
h
e
fract
i
o
n
o
f
w
e
i
ght
ed
s
u
m
of
out
put
s t
o
i
t
s
wei
ght
e
d
s
u
m
of i
n
p
u
t
s
[1
1]
,[
1
2
]
.
S
u
p
p
o
se we ha
ve Gi
ven
n
ou
tpu
t
s an
d
m
inputs
,
efficiency (
h
0)
fo
r ho
sp
ital 0
is d
e
fin
e
d
as
fo
llo
ws:
The
weights are specific to ea
ch unit so t
h
at 0
≤
w
≤
1 a
nd
a val
u
e
of
u
n
i
t
y
im
pl
y
co
m
p
let
e
t
echni
cal
efficien
cy.
DEA co
m
p
u
t
es all p
o
ssib
l
e sets
o
f
weigh
t
s wh
i
c
h
satisfy all c
o
n
s
t
r
ain
t
s.
Ho
sp
ital u
n
it th
at i
s
th
e
h
i
gh
est ef
f
i
cien
cy w
i
t
h
th
e
giv
e
n d
a
ta
of
inp
u
t
s and
ou
tputs f
o
r
t
h
e sp
eci
f
i
c p
e
r
i
od
of
time in
our
case it is
year
201
0.
Outp
ut Variables
1.
Ou
t
p
atien
t
v
i
sits, th
e
nu
m
b
er o
f
ou
tp
atien
t
v
i
sits with
in
a year.
2.
Inpatient cases
case-m
i
x adjusted the
annual num
b
er of
hos
p
italized cases.
3.
To
tal nu
m
b
er
o
f
surgery, t
h
e
to
tal
num
ber
o
f
s
u
r
g
eri
e
s
wi
t
h
i
n
a
y
ear.
Inpu
t Vari
abl
e
s
1.
Bed
s
, t
h
e
n
u
m
b
e
r of ex
istin
g
p
a
tien
t
bed
s
wi
th
in
th
e ho
sp
ital read
y
for
u
s
e.
2.
Sp
ecialists, th
e to
tal nu
m
b
er of sp
ecialist
m
e
d
i
cal do
ctors em
p
l
o
y
ed
in
th
e ho
sp
itals.
3.
Nurses, Th
e to
t
a
l n
u
m
b
e
r
o
f
nu
rses em
p
l
o
y
ed
in th
e
ho
sp
ital.
3.
RESULTS
A
N
D
DI
SC
US
S
I
ON
Th
is sectio
n
will p
r
esen
t th
e resu
lts o
b
t
ain
e
d
fr
o
m
th
e e
m
p
i
rical st
u
d
y
of non
-profit p
u
b
lic
or
ga
ni
zat
i
on
L
R
B
T
’s
16
h
o
s
p
i
t
a
l
uni
t
s
i
n
t
h
e
f
o
u
r
pr
o
v
i
n
ces
o
f
t
h
e
c
o
u
n
t
r
y
by
u
s
i
n
g
DE
A
m
e
t
hod
ol
o
g
y
.
DE
A
diffe
re
ntiate the technical e
ffi
cient
and i
n
efficient units across the c
o
unt
ry
usi
n
g s
p
eci
fi
c i
n
p
u
t
s
a
n
d
out
p
u
t
s
.
3.
1.
Descripti
v
e s
t
atis
tics of inputs
and
outputs.
Tabl
e 1
s
h
o
w
s
t
h
e descri
pt
i
v
e
st
at
i
s
t
i
c
s
of t
h
e
i
n
put
s use
d
. A
v
era
g
e nu
m
b
er
of d
o
ct
o
r
an
d ot
he
r
in
pu
ts shows t
h
at th
ere is
on
avera
g
e a
v
ailable num
b
er of
doct
o
rs
i
n
each
ho
sp
ital u
n
it
for th
e su
rg
eries and
ot
he
r m
e
di
cal
servi
ces
o
f
t
h
e
pat
i
e
nt
s a
nd
si
m
i
l
a
rl
y
t
h
e st
anda
r
d
de
vi
at
i
o
n s
h
o
w
s t
h
at
o
n
a
v
era
g
e t
h
e
r
e can
b
e
vari
at
i
o
n o
r
de
vi
at
i
on ca
n e
x
i
s
t
bet
w
ee
n i
n
p
u
t
s
i
n
th
e
ho
spitals u
n
its o
f
LRBT. Th
e m
a
x
i
m
u
m
an
d
m
i
n
i
m
u
m
v
a
lu
es
o
f
t
h
e i
n
pu
ts sh
ow larg
est and
lowest v
a
lu
es
of a
v
ailable inputs and
we m
a
y s
a
y that these are the
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
PH
S
I
S
SN
:
225
2-8
8
0
6
Mea
su
ring
Efficien
cy o
f
Ho
spita
ls b
y
DEA:
An
Emp
i
rica
l
Evid
en
ce from
Pa
kistan
(S
yed Aziz Ra
so
o
l
)
13
3
larg
est or lo
west n
u
m
b
e
r o
f
i
n
pu
ts av
ailab
l
e fo
r
p
a
tien
t
s treated
in
a h
o
sp
ital wh
eth
e
r th
ey are ou
t –
p
atien
t
vi
si
t
s
, m
a
jor s
u
rge
r
i
e
s
or a
n
nu
al
num
ber
of
h
o
spi
t
a
l
i
zed
pe
r
s
on
s
du
ri
n
g
t
h
e
y
ear
20
1
0
.
Tab
l
e 2
sh
ows th
e d
e
scri
p
tive statistics
o
f
th
e ou
tp
uts use
d
. Ave
r
age
num
b
er of treated pe
rsons a
nd
explains that
there are on
avera
g
e
11
2
2
37
pe
rso
n
s
vi
si
t
s
t
h
e h
o
spi
t
al
s as out
-pat
i
e
nt
s, 6
4
33
p
e
rso
n
s
h
o
s
p
italized
in a year and
11
086
p
e
rso
n
s
h
a
s
b
een
ad
mitted
in
h
o
sp
it
als for m
a
j
o
r
su
rg
eries. Stand
a
rd
devi
at
i
o
n s
h
ow
s t
h
at
on
a
v
era
g
e t
h
ere
can
b
e
va
ri
at
i
o
n
o
r
devi
at
i
o
n ca
n e
x
i
s
t
bet
w
een
o
u
t
p
ut
s i
n
t
h
e
h
o
spi
t
a
l
uni
t
s
of LR
B
T
.
The m
a
xim
u
m
and m
i
nim
u
m
val
u
es o
f
t
h
e out
put
s s
h
ow
l
a
rgest
an
d l
o
west
val
u
es o
f
out
put
s
and
we m
a
y sa
y that these are the largest or lowest nu
m
b
er of
pat
i
e
nt
s re
spect
i
v
el
y that are treated during a
year in
a ho
sp
ital u
n
it.
Table
1.
Descr
i
ptive s
t
atistic
s
of In
pu
t
vari
a
bles
Input Variables
Mean
S.D
Max
Min
Beds,
the nu
m
b
er
of existing patient
beds within the hospital ready
for
use
.
43.
18
22.
39
95
10
Specialists, the tot
a
l nu
m
b
e
r
of sp
eci
alist
m
e
dica
l doctors e
m
plo
y
ed
in the hospitals.
9.
25
7 30
3
Nur
s
es,
T
h
e total n
u
m
ber
of nur
se
s e
m
plo
y
ed in the hospital.
23.
43
6.
44
40
15
T
a
bl
e 2. Descr
i
pti
v
e
s
t
ati
s
ti
c
s
o
f
out
p
u
t
va
ri
abl
e
s
3.
2.
Technical Efficiency
(TE) Scores
by DE
A
Table 3
shows
the res
u
lts of
tec
hnical efficiency obtaine
d
by us
i
ng
open-access Data E
nvel
opm
ent
An
alysis Prog
ra
mm
e (DEAP
2
.
1
)
[12
]
. In
Tab
l
e 4
.
3
th
e resu
lts are calcu
l
a
ted
with
CCR co
n
s
tan
t
retu
rn
to
scale
m
o
d
e
l.
A ho
sp
ital un
it is said
to
b
e
tech
n
i
cal ef
fi
cient if its efficiency sc
ore i
s
‘1
’ a
n
d
ine
f
ficient
ot
he
rwi
s
e [
13]
.
The degree of
inefficie
n
cy
then
de
pends
upon the e
ffici
ency scor
es
o
f
h
o
sp
ital un
its o
f
LRBT.
There
f
ore, acc
or
di
n
g
t
o
t
h
e r
e
sul
t
s
sh
ow
n i
n
t
h
e t
a
bl
e 4.
3,
5 (3
1.
25
perce
n
t
)
h
o
s
p
i
t
a
l
uni
t
s
out
of
16
ho
spi
t
a
l
s
are technical e
fficient
with t
h
e effici
en
cy sco
r
es of
1
.
00
0 an
d 11(
68
.75
p
e
rcen
t) h
o
s
p
ital u
n
its
are
i
n
efficien
t
by
usi
ng a
v
ai
l
a
bl
e dat
a
f
o
r s
p
eci
fi
c i
n
put
s
and
o
u
t
p
ut
s t
h
at
are bei
n
g
us
ed i
n
ou
r a
n
al
y
s
i
s
for t
h
e y
e
ar 2
0
1
0
.
The M
ean
TE
appea
r
e
d
as 0
.
70
3
or
70
.3
pe
rcent
f
o
r al
l
16
uni
t
s
of LR
B
T
h
o
spi
t
a
l
s
t
a
k
e
n i
n
t
h
i
s
st
ud
y
for t
h
e
give
n data
of c
onsi
d
ere
d
inputs and
out
puts
.
Am
ong
those 11
tech
nical inefficient units 5(31.25 pe
rcent
)
have
efficiency sc
ores ranges
fr
o
m
0
.
5
0
t
o
0.70
or
w
e
m
a
y say t
h
ey ar
e
50 pe
r
cent
t
o
70 perc
en
t technical efficient
and 2 (0
.1
2
5
perce
n
t
)
ha
vi
n
g
s
o
res
m
o
re t
h
an 0.
7
0
and rem
a
in
in
g
4
(0.25
p
e
rcen
t) h
o
sp
ital un
its h
a
ve
efficiency sc
ore 0.50
or bel
o
w.
3.3. The
Vari
able Re
turn To Scal
e DE
A Model And
Effi
ciency
Sc
ores
Tab
l
e 4
sho
w
s th
e resu
lts ob
tain
ed
fro
m
t
h
e ex
ten
d
e
d
DEA m
o
d
e
l i.e. v
a
riab
le ret
u
rn
to
scale
(VRS), Ban
k
e
r, Ch
arn
e
s an
d
Co
op
er
(
BCC)
m
o
d
e
l th
e o
b
j
e
ctiv
e o
f
th
e i
n
clu
s
ion
of th
is m
o
d
e
l in
th
e an
alysis
is that it stretches the efficiency scores to the s
cale efficiency scores
of eac
h unit along with the
VR
S
technical effici
ency score
s
. T
h
e TE scores obtained
fr
om
a
CRS DEA into two c
o
m
ponents, one due to scale
i
n
effi
ci
e
n
cy
an
d t
h
e
ot
her
o
n
e
i
s
d
u
e t
o
“
p
ur
e” t
ech
ni
cal
i
n
effi
ci
ency
we
can cal
c
u
l
a
t
e
t
h
ese t
w
o
com
pone
nt
s
by conducting
bot
h a CRS and a VRS DE
A upon the same da
ta [22]. If technical e
ffici
ency scores for each
h
o
s
p
ital u
n
it in
th
e two
efficien
cy sco
r
es
are differe
n
t then t
h
is shows
th
at the hospi
t
al units havi
ng scale
inefficie
n
cies and t
h
ese scale inefficie
n
cies can be
calc
u
lated from
the difference
betwee
n the VR
S technica
l
efficiency sc
ore and CRS te
c
hnical e
fficiency score
s
.
Vari
a
b
l
e
ret
u
r
n
t
o
scal
e (V
R
S
) n
o
rm
al
l
y
fo
rm
s a conve
x h
u
l
l
of t
w
o
i
n
t
e
rsect
i
ng
p
l
ans w
h
i
c
h
envel
op t
h
e d
a
t
a
poi
nt
s m
o
r
e
t
i
ght
l
y
as co
m
p
are t
o
t
h
e C
R
S
DEA
w
h
i
c
h m
a
kes a coni
cal
h
u
l
l
an
d t
h
us
provide techni
cal efficiency
score
s
w
h
i
c
h a
r
e g
r
eat
er t
h
a
n
or e
q
ual
t
o
t
h
ose
whi
c
h are
obt
ai
ne
d
usi
n
g
C
R
S
m
odel [13]
.
O
u
t
p
ut
Variables
Mean S.
D Max
Min
Annual nu
m
b
er
of
Out-
patient visits
1122
37.
5
8153
7.
54
3285
00
5840
0
I
npatient cases cas
e-m
i
x adjusted the
annual nu
m
b
er
of
hospitalized cases.
6433.
1
5
4304.
0
6
1825
0
730
Total nu
m
b
er of surgery, the tota
l nu
m
b
er of surgeries
within a
year
.
1108
6.
875
6240.
8
8
2920
0
1825
0
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
252
-88
06
IJP
H
S
V
o
l
.
3,
No
. 2,
Ju
ne
2
0
1
4
:
12
9 – 1
3
6
13
4
In
tab
l
e 4 th
e l
a
st co
lu
m
n
sho
w
s th
e
n
a
ture o
f
th
e
scales
whet
her t
h
at is
increasi
n
g ret
u
rn to scale
(IRS
)
or
dec
r
e
a
sing
retu
r
n
s t
o
scale
(DR
S
)
.
T
h
e
nature
of scale ine
ffici
encies (d
ue t
o
IR
S
or
DR
S)
f
o
r
a
p
a
rticu
l
ar ho
spital u
n
it can
be d
e
term
in
ed
by o
b
s
erv
i
n
g
whethe
r the
non- inc
r
easing
re
turn to scale
(NIRS)
TE sco
r
e is equ
a
l to
t
h
e
VRE TE sco
r
es, if th
ey ar
e equ
a
l
D
R
S ap
p
l
y
o
t
her
w
ise
I
R
S apply.
Table 3.
Technical
Effi
ciency
sc
ores CRS DEA
S. No
Hospital uni
t
Technical Ef
ficie
n
cy (TE)
1
Kor
a
ngi Hospital,
(
K
ar
achi)
Sindh
1.
000
2 T
a
ndoBago,
Sindh
0.
500
3 Rashidabad,
Sindh
0.
750
4
Gam
b
at Hospital,
Sindh
0.
698
5
Akor
aKhattak Hospital,
KPK
0.
591
6
Kalakalay Hospital,
(
S
wat)
KPK
0.
695
7
M
a
nsher
a
Hospital KPK
1.
000
8 L
a
r
Hospital,
(
M
ultan)
Punjab
0.
750
9
Shahpur
Saddar
,
Hospital (
n
ear
Sar
godha)
Punjab
0.
677
10
M
a
ndr
a
Hospital,
Punjab
1.
000
11
Khanewal
m
i
ni Ho
spital, Punjab
1.
000
12
Pasr
ur
Distr
i
ct Sialkot,
Punjab
0.
467
13
Arifwala Hospital,
(near
Sahiwal)
Pu
njab
0.
493
14
L
a
hor
e Hospital,
Punjab
1.
000
15
Sibi
m
i
ni Hospital, Baluchistan
0.
096
16
Quetta Hospital, B
a
luchistan
0.529
Mean 0.
703
Ta
ble
4
.
Co
mpa
r
iso
n
of VRS TE &
CRS TE
S. No
Hospital uni
t
CRSTE
VRSTE
Scale
1
Kor
a
ngi Hospital,
(
K
ar
achi)
Sindh
1.
000
1.
000
1.
000
-
2 T
a
ndoBago,
Sindh
0.
500
0.
600
0.
834
IRS
3 Rashidabad,
Sindh
0.
750
0.
786
0.
954
IRS
4
Gam
b
at Hospital,
Sindh
0.
698
0.
715
0.
977
IRS
5
Akor
aKhattak Hospital,
KPK
0.
591
0.
652
0.
970
IRS
6
Kalakalay Hospital,
(
S
wat)
KPK
0.
695
0.
697
0.
996
IRS
7
M
a
nsher
a
Hospital KPK
1.
000
1.
000
1.
000
-
8 L
a
r
Hospital,
(
M
ultan)
Punjab
0.
750
1.
000
0.
750
IRS
9
Shahpur
Saddar
,
Hospital (
n
ear
Sar
godha)
Punjab
0.
677
0.
679
0.
988
DRS
10
M
a
ndr
a
Hospital,
Punjab
1.
000
1.
000
1.
000
-
11
Khanewal
m
i
ni Ho
spital, Punjab
1.
000
1.
000
1.
000
-
12
Pasr
ur
Distr
i
ct Sialkot,
Punjab
0.
467
0.
500
0.
933
IRS
13
Ar
ifwala Hospital, (
n
ear
Sahiwal) Pu
njab
0.
493
0.
600
0.
822
IRS
14
L
a
hor
e Hospital,
Punjab
1.
000
1.
000
1.
000
IRS
15
Sibi
m
i
ni Hospital, Baluchistan
0.
096
0.
750
0.
128
IRS
16
Quetta Hospital,
B
a
luchistan
0.
529
0.
600
0.
882
IRS
Mean
0.
703
0.
786
0.
866
CRSTE = technical efficie
n
c
y
from
CRS DEA
VRSTE
= technical efficie
n
c
y
from
VRS DEA
Scale = scale e
fficiency = CR
STE/VRST
E
4.
CO
NCL
USI
O
N
The t
e
c
hni
cal
effi
ci
ency
of
1
6
hos
pi
t
a
l
u
n
i
t
s
o
f
LR
B
T
, a
no
n
-
p
r
o
f
i
t
p
r
i
v
at
e or
ga
ni
zat
i
on,
has
bee
n
taken i
n
to ac
count. T
h
e te
chnical e
fficiency show
s that a
m
o
n
g
th
ese 16
u
n
its
of LRBT
h
o
s
pitals 5
(31
.
25
p
e
rcen
t
)
h
o
s
p
ital un
its are th
e efficien
t, with
t
h
e g
i
v
e
n d
a
ta
o
f
i
n
pu
ts
an
d ou
tpu
t
s
u
s
ed
in th
is stud
y
and
th
e d
a
ta is taken
fro
m
th
e an
nu
al repo
rts
an
d
p
e
rson
al reco
rd
s
o
f
LR
BT h
o
s
p
itals.
Th
e
h
o
s
p
ital un
its o
f
Ko
ra
ngi
Karac
h
i
,
La
ho
re,
K
h
anewal
m
i
ni
, M
a
nd
ra a
n
d
Manshe
ra
a
r
e said
to be
t
h
e efficient one
s and
are
th
ese are th
e un
its wh
ich are
p
r
ov
id
in
g
t
h
eir serv
ices
with
min
i
m
u
m
lev
e
l
of
inpu
ts. Th
ese un
its ar
e peer
s
f
o
r
t
h
e ot
he
r 11 (
6
9
.
7
5
pe
rce
n
t
)
hos
pi
t
a
l
uni
t
s
whi
c
h are
no
t
on effi
ci
ent
fr
ont
i
e
r a
nd sa
i
d
t
o
be i
n
ef
fi
ci
ent
.
Efficien
t un
its are u
s
ing
less in
pu
ts and
larger o
u
t
pu
t h
a
s been produced
by them as
com
p
are to their
counte
r
p
a
r
t
n
e
r p
e
er
s
,
in
e
f
f
i
c
i
en
t un
its
.
T
h
e av
e
r
age
T
E
of
all
h
o
sp
itals is
0
.
7
0
3
with
t
h
e CR
S DEA m
o
d
e
l u
s
i
ng
m
u
lt
i
-
st
age appr
oac
h
. The
n
DEA
wi
t
h
VR
S has al
so bee
n
use
d
in order to calculate th
e scale efficiencies of
all h
o
s
p
ital u
n
i
ts an
d
also
to
an
alyze th
e retu
rn
to
scal
es of all DMU’s.
th
e resu
lts ob
t
a
in
ed
fro
m
th
e VRS
DEA
are
n
o
rm
al
l
y
hi
gh
beca
use
of t
h
e i
n
t
e
r
s
ect
i
on
of t
h
e planes which form
a
conv
e
x
hul
l
i
n
st
ea
d of coni
cal
hul
l
w
h
i
c
h i
s
f
o
rm
ul
at
ed i
n
C
R
S
m
odel
,
t
h
e
con
v
e
x
h
u
l
l
i
n
case of
VR
S e
nvel
op t
h
e dat
a
poi
nt
s m
o
re t
i
ght
l
y
so t
h
e
res
u
l
t
s
wi
t
h
t
h
i
s
m
ode
l
no
rm
all
y
shows hi
gh
res
u
l
t
s
. Th
us r
e
sul
t
s
gene
rat
e
d
by
t
h
e D
E
A m
odel
usi
n
g
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
PH
S
I
S
SN
:
225
2-8
8
0
6
Mea
su
ring
Efficien
cy o
f
Ho
spita
ls b
y
DEA:
An
Emp
i
rica
l
Evid
en
ce from
Pa
kistan
(S
yed Aziz Ra
so
o
l
)
13
5
DEAP in a
VR
S m
odel shows
on a
v
era
g
e te
c
hnical e
fficiency
score
s
as
0.
78
6.
In c
o
ncl
u
si
on
,
t
h
e st
u
d
y
ha
s c
o
n
f
i
r
m
e
d t
h
at
DEA
n
o
t
onl
y
hel
p
s
heal
t
h
p
o
l
i
c
ym
akers an
d m
a
nagers
t
o
answe
r
t
h
e
quest
i
o
n “H
o
w
wel
l
are t
h
e hos
pi
t
a
l
s
perf
orm
i
ng
?
”
bu
t
al
so “B
y ho
w m
u
ch coul
d
t
h
ei
r
per
f
o
r
m
a
nce b
e
im
pro
v
e
d
?
”
We i
n
d
o
rs
e f
u
r
t
her a
n
al
y
s
es
o
f
t
h
e
h
o
s
p
i
t
a
l
s
whi
c
h a
r
e
per
f
o
rm
i
ng
best
ha
ve t
o
mak
e
in
o
r
d
e
r t
o
en
h
a
n
ce
h
eal
th
facilities.
POLIC
Y
AN
D RECO
M
M
E
NDETIO
N
S
C
once
r
ni
ng
ho
spi
t
a
l
s
uni
t
s
w
i
t
h
out
p
u
t
s
fal
l
i
ng s
h
o
r
t
of t
h
e DEA t
a
r
g
et
s,
LR
B
T
m
a
nagem
e
nt
and
go
ve
rnm
e
nt
sect
or p
o
l
i
c
y
m
a
kers c
oul
d hel
p
t
h
e n
o
n
-
p
r
ofi
t
pri
v
at
e or
ga
n
i
zat
i
on t
o
im
prove t
h
e e
ffi
ci
ency
by
im
proving access of the
funds from
governm
e
nt sector to
these kinds
of
non-profit orga
nizations
. Paki
stan is
a devel
o
pi
n
g
cou
n
t
r
y
,
a
n
d a
s
al
l
devel
o
pi
ng c
o
unt
ri
es, s
o
t
h
e s
o
ci
et
y
i
s faci
n
g
m
a
jor
i
ssues i
n
heal
t
h
an
d
ed
u
cation
secto
r
s th
erefo
r
e,
g
o
v
e
rn
m
e
n
t
institu
tio
n
s
are
always
in
n
e
ed
o
f
h
e
l
p
ing
h
a
nd
s o
f
th
ese
non
-p
rofit
or
ga
ni
zat
i
ons i
n
or
de
r t
o
im
pr
o
v
e and e
n
h
a
nce t
h
e l
i
v
i
n
g st
anda
r
d
s o
f
t
h
e resi
dent
s
of t
h
e co
u
n
t
r
y
.
The
u
tilizatio
n
o
f
h
ealth
resou
r
ces also
n
e
ed
s
to
b
e
im
p
r
o
v
e
d
and
th
e m
a
n
a
g
e
m
e
n
t
o
f
t
h
e LRBT shou
ld
also
increase t
h
eir
perform
a
nce in
order to i
n
crea
se th
e efficien
cy o
f
all ho
sp
ital u
n
its.
Heal
t
h
pr
om
ot
i
on i
s
a key
i
s
sue f
o
r any
co
unt
ry
t
h
at
i
s
f
aci
ng p
r
o
b
l
e
m
s
bot
h at
t
h
e basi
c and a
t
ad
v
a
n
ce lev
e
l o
f
h
ealth
facilit
atio
n
,
lik
e
Pak
i
stan
’s scen
ario th
is situ
atio
n
p
u
t
s i
n
creasi
n
g p
r
essure on
the n
on-
p
r
o
f
it sect
o
r
t
o
p
e
rfo
r
m
co
m
p
eten
tly. Th
e
presen
t stud
y
ratio
n
a
lly d
e
m
o
nstrated
th
e ap
plicab
ility o
f
DEA i
n
h
ealth
sector to
id
en
tify th
e least
and m
o
st efficient healt
h
units in
LR
B
T
,
Paki
st
an as
m
e
nt
i
oned abo
v
e
efficiency i
n
health services a
t
prim
ary as well as
at adv
a
n
c
e lev
e
l can
co
ver th
e way
for
d
e
stin
ation
REFERE
NC
ES
[1]
Abbas et al. An
aly
z
ing th
e eff
i
ciency
d
i
ffer
e
n
ces among basic h
ealth units
in Sargodha District.2011.
Munich
Pe
rsona
l Re
PE
c Arc
h
ive
,
Onlin
e at http://mpra.ub.
uni-muench
e
n.de/32244
/MPRA
Paper No. 32244, posted 14
.
July
2011
/ 17:5
0
.
[2]
Halena Steing
ru
ber.2006. Ef
f
i
ciency
in non-profit organizations- How do
aid organizations estimate external and
intern
al eff
i
c
i
en
c
y
and how is
it com
m
unicat
e
d
?
B
ach
elor
th
e
s
is in business adm
i
nistrati
on f
i
nanc
ial
control
autumn term 20
06 at School of
Busine
ss, Econo
mics and
Law,
Gothenbuerg.
[3]
Hensher M., “Financing Health
Sy
stems through Effi
ciency
Gains” Working paper for the commission o
n
Macroeconomics and Health
, 20
01.
[4]
Hsu J., “The relative Efficiency
o
f
Public and
Priv
ate Service D
e
liver
y
”,
World Health
Report, pp
.
25-30, 2010
.
[5]
Charnes, A
.
, W. W. Coop
er
and
B.Golan
y
. 1
985. Foundatio
n
s of data
env
e
lopment
anal
y
s
is for Pareto-
Koopmans efficient
empirical pr
oduction
functio
n
s. Journal of
Econometrics., 30(
1): 91-107.
[6]
Kengil B. C.
, N. Gokm
en, H. T
o
zan, “
E
ffi
ci
enc
y
M
eas
ur
es
in
t
h
e Health S
e
rv
i
ces
with DEA-An Overview”,
Journal of Naval Science and
En
gineering
, vol/is
sue: 1(6), pp
. 1-
14, 2010
.
[7]
Rowena Jacobs. 2000. Alternative Methods to Examine
Hospital Efficiency
: D
a
ta
Env
e
lopmen
t Analy
s
is and
Stochastic Frontier Analy
s
is. Dis
c
u
ssion paper
.
The University
of
Yo
rk. Centr
e
fo
r
Health
Economics.
[8]
Kontodim
opoulos N et al. 2006
. Bal
a
ncing
effi
c
i
enc
y
of h
eal
th
services and
equ
i
t
y
of a
ccess in
rem
o
te are
a
s in
Greece
. h
ttp:
//w
ww.ncbi.nlm
.n
ih
.gov/pubm
ed/15
927299.
[9]
Akazili. J
., M
.
Adjuik and
E.
Zere
. 2008
. Using data env
e
lop
m
ent analy
s
is
to
measure th
e
ex
te
nt o
f
technical
effic
i
enc
y
of
pu
blic
hea
lth
cen
tr
es in
Ghan
a. BM
C In Health
Hum Rights 8:11.
[10]
Os
ei D., S
.
George, M
.
Kirigia
,
“
T
echn
i
ca
l Effi
ci
enc
y
and cos
t
ef
fect
ivenes
s
of Heal
th centr
es
in Ghana A pilot
study
”
,
Cos
t
E
ffe
cti
venes
s
and
Re
s
our
ce Al
locat
i
o
n
, vol/issue: 3(9)
, pp
. 123-145
, 2
005.
[11]
M.Linna and U
.
Hakkinen.
2000
.Determinants of
Cost efficien
cy
of Fi
nish Hospitals: A Comparis
on of DEA and
SFA: 1-35.
[12]
Hollingsworth,
B., P. J. Dawson a
nd N. Manid
a
kis. 1999.
Effici
ency
measurem
ent of Hea
lth Care : A review of
non-parametric methods
and app
lic
ations
., Health Care Mana
g
e
ment Scien
c
e., 2
(
3): 161-172
.
[13]
Ozcan Y. A., R. D. Luke, “
H
ealt
h
Care Deliver
y
Restructur
ing an
d Productivit
y
C
h
ange: Assessing the Veterans
Integrated Serv
ice Network (VISN)
Using the
Malmquist Approach”,
M
e
dica
l Care Resear
ch
and Review
,
vol/issue: 11(1), pp.
1-16
,
2010
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
252
-88
06
IJP
H
S
V
o
l
.
3,
No
. 2,
Ju
ne
2
0
1
4
:
12
9 – 1
3
6
13
6
BIOGRAP
HI
ES OF
AUTH
ORS
S
y
ed
Aziz Raso
ol, PhD Scholar,
School of
Econo
mic Scien
ces
, Federal Urdu Univ
ersity
of
Arts, Sciences &
Techno
log
y
, Islamabad, 44000
, P
a
kistan
.
Em
ail:
s
y
ed.r
as
ool0@gm
ail.
com
Cell No.
+92-33
4-5488010
Dr. Abdul Saboo
r, Professor,
Department o
f
Economics & Agr
i
-Econo
imcs,
PMAS-Arid Agricultur
e
Univ
ersi
t
y
, R
a
walpind
i
,
Pakistan.
Em
ail:dr
a
bdul.saboor@uaar.edu.pk
Phone #:
+92-51
-9290022
Cell No.
+92-33
1-6501706
Muhammad Raashid,PhD Scholar,
Department o
f
Economic
s &Agri-Econoimcs,
PMAS-Arid Agricultur
e
Univ
ersi
t
y
, R
a
walpind
i
,
Pakistan.
Email: mraashid
505@gmail.com
.
Cell No.
+92-34
5-5398879
Evaluation Warning : The document was created with Spire.PDF for Python.