Clinical and Experimental Ophthalmology
TQ\wHJ 2006;
.H(}[eG_ 34
$:u7Dv}\ : 880–885
gc2|V6( doi:10.1111/j.1442-9071.2006.01342.x
OU.9 #|q U © 2006 Royal Australian and New Zealand College of Ophthalmologists
J'&B:PZObB K!-
&Zv Correspondence:
MY?O/,6 Dr Garry Brian, 5 Hazelmere Parade, Sherwood, Qld 4075, Australia. Email:
grbrian@tpg.com.au 3
!> L? Received 11 April 2006; accepted 19 June 2006.
dJQ }{,+6 Original Article
|Y8Mk2,s Cataract and its surgery in Papua New Guinea
LF9aw4:>Ou Jambi N Garap
'x45E.wYw MMed(Ophthal)
ndKvJH 4 ,
C?/r}ly<\ 1,2
fD\^M{5f Sethu Sheeladevi
CFkW@\] MHM
TRySl5jx@ ,
p;n3`aVh 3
-VreBKn Garry Brian
liugaRO8J FRANZCO
>p]WCb'PH ,
VNr 2,4
W>q*.9}Y" BR Shamanna
<@U. MD
AT%*
~tr ,
je- ,S>U 3
F=
_uNq Praveen K Nirmalan
VW{aUgajO MPH
JK(`6qB>(6 3
0okO+QU,a and Carmel Williams
#uD)0zdw MA
8m/FKO (r 4
TdNuD V 1
Pp1zW3+Q The Fred Hollows Foundation – Papua New Guinea Eye Care Program,
xUiWiOihr6 2
#L@} .Giz Department of Ophthalmology, School of Medicine and Health
~8H&m,{j Sciences, University of Papua New Guinea, Port Moresby, Papua New Guinea;
p|FlWR'mA 3
2y9:'c| International Center for Advancement of Rural Eye Care,
xQNw&'|UU L.V. Prasad Eye Institute, Hyderabad, India; and
msA' 5> 4
X[GIOPDx The Fred Hollows Foundation (New Zealand), Auckland, New Zealand
u"`5 Key words:
'h
!h! blindness
%|: ;Ti ,
XPHQAo[(s cataract
XysFwi ,
NqiB8hZ~ Papua New Guinea
i&A{L}eCr: ,
a~8:rW^ surgery
gIV3n#-{L ,
HBt|}uZ?6i vision impairment
,bB( 24LD .
%iK%$ I
kVV\*"9y NTRODUCTION
2-B6IPeI Just north of Australia, tropical Papua New Guinea (PNG)
7MKD_`g has more than five million people spread across several major
T&o,I and hundreds of other smaller islands. Almost 50% of the
sFbfFUd
land area is mountainous, and 85% of inhabitants are rural
Wr]O
dwellers. Forty per cent of the population is age 14 years or
*D\0.K,o younger, and 9% is 50 years or older.
pwa.q 1
wY' "ab Papua New Guinea was administered by Australia until
@''&nRC1 1975, when independence was granted. Since that time, governance,
+n@f'a"> particularly budgetary, economic performance, law
g
E\A9L~b and justice, and development and management of basic
;l$9gD>R health and other services have declined. Today, 37% of the
BmJkt3j." population is said to live below the poverty line, personal
v2IEJ
and property security are problematic, and health is poor.
K|`+C1! There are significant and growing economic, health and education
_GE=kw;: disparities between urban and rural inhabitants.
u@AI&[Z Papua New Guinea has one referral hospital, in Port
\s,Iz[0Vfz Moresby. This has an eye clinic with one part-time and two
Kp8fh-4_ full-time consultant ophthalmologists, and several ophthalmology
3 t/ R 2
M training registrars. There are also two private ophthalmologists
\^Ep>Pq`] in the city. Elsewhere, four provincial hospitals
)X7e$<SU* have eye clinics, each with one consultant ophthalmologist.
E3y6c)< One of these, supported by Christian Blind Mission and
5(423"(y based at Goroka, provides an extensive outreach service.
k}X[u8A Visiting Australian and New Zealand ophthalmology teams
I"B8_ and an outreach team from Port Moresby General Hospital
}6<)yW}U provide some 6 weeks of provincial service per year.
Dh2#$[/@1 Cataract and its surgery account for a significant proportion
oBb?"2 ~9 of ophthalmic resource allocation and services delivered
g`kY]lu in PNG. Although the National Department of Health keeps
AU/#b(mI some service-related statistics, and cataract has been considered
Jff 79)f in three PNG publications of limited value (two district
;xhOj<: service reports
kQy&I3 2,3
d`85P+Qen| and a community assessment
U{"f.Z:Ydo 4
AfN&n= d K ), there has
21ViHV been no systematic assessment of cataract or its surgery.
8qN"3 Et A
k5wi' BSTRACT
eI0F!Yon Purpose:
kymn)Ea To determine the prevalence of visually significant
3^XVQS*** cataract, unoperated blinding cataract, and cataract surgery
l~rb]6E for those aged 50 years and over in Papua New Guinea.
vGMJ ^q Also, to determine the characteristics, rate, coverage and
7u::5 W-q outcome of cataract surgery, and barriers to its uptake.
7lwTZ*rnY Methods:
|:N>8%@6c Using the World Health Organization Rapid
9qr UM`z$g Assessment of Cataract Surgical Services protocol, a population-
gJh}CrU- based cross-sectional survey was conducted in
J,(7.+`~# 2005. By two-stage cluster random sampling, 39 clusters of
ck$M(^)l 30 people were selected. Each eye with a presenting visual
8 aZ$5^z acuity worse than 6/18 and/or a history of cataract surgery
'y.Jc
S!| was examined.
{OCJ(^8i Results:
)&,K94
Of the 1191 people enumerated, 98.6% were
vOe0}cR examined. The 50 years and older age-gender-adjusted
#]y5zi prevalence of cataract-induced vision impairment (presenting
=JW[pRI5a acuity less than 6/18 in the better eye) was 7.4% (95%
f`?0WJ(M confidence interval [CI]: 6.4, 10.2, design effect [deff]
oNIFx5*Z =
`eC+% O
1.3).
jgq{pZ#E That for cataract-caused functional blindness (presenting
I+"
lrU acuity less than 6/60 in the better eye) was 6.4% (95% CI:
VoM6 5.1, 7.3, deff
CW-A e =
#ib^Kg 1.1). The latter was not associated with
{cR=N~_EO gender (
+c) TDH P
Kt6>L5:94 =
41yOXy ;~l 0.6). For the sample, Cataract Surgical Coverage
X;VQEDMPU at 6/60 was 34.5% for Eyes and 45.3% for Persons. The
cDeZMsV Cataract Surgical Rate for Papua New Guinea was less than
h e&V# # 500 per million population per year. The age-genderadjusted
Qqq
<e prevalence of those having had cataract surgery
L/i(KF{ was 8.3% (95% CI: 6.6, 9.8, deff
*_"u)<J =
6I>^Pf'ND 1.3). Vision outcomes of
/oL8;:m surgery did not meet World Health Organization guidelines.
Js{=i>D Lack of awareness was the most common reason for not
;#/0b{XFj seeking and undergoing surgery.
78]gtJ Conclusion:
B4Lx{uno Increasing the quantity and quality of cataract
8\#
^k#X surgery need to be priorities for Papua New Guinea eye
X+l'bp]Ry care services.
Fo}7hab Cataract and its surgery in Papua New Guinea 881
%|+E48 © 2006 Royal Australian and New Zealand College of Ophthalmologists
W)4QOS& This paper reports the cataract-related aspects of a population-
O3qM1-k}S based cross-sectional rapid assessment survey of
X3B{8qx_> those 50 years and older in PNG.
d8OL!Rk M
DDeE(E ETHODS
e:-8k_0| The National Ethical Clearance Committee of The Medical
i9m*g*"2 Research Advisory Committee granted ethics approval to
s|-FH X survey aspects of eye health and care in Papua New Guinea
GEe`ZhG,
(MRAC No. 05/13). This study was performed between
RU&_j*U December 2004 and March 2005, and used the validated
*>."V5{;S World Health Organization (WHO) Rapid Assessment of
*b)Q5dw@1 Cataract Surgical Services
Zyy
e%Ly 5,6
\#,t O%D protocol. Characterization of
YQ]\uT>}& cataract and its surgery in the 50 years and over age group
a07=tD was part of that study.
-]S.<8<$ As reported elsewhere,
R_kQPP
7
s;1h-Oq( the sample size required, using a
tIuoD+A
W prevalence of bilateral cataract functional blindness (presenting
<@uOCRbV visual acuity worse than 6/60 in both eyes) of 5% in the
XL5Es:"+?S target population, precision of
g,kzQ}_ ±
K@:m/Z}|4 20%, with 95% confidence
%0-wpuHc(] intervals (CI), and a design effect (deff) of 1.3 (for a cluster
Hs(D/&6% size of 30 persons), was estimated as 1169 persons. The
{nMAm/kyj sample frame used for the survey, based on logistics and
3,eIB( security considerations, included Koki wanigela settlement
}e]f in the Port Moresby area (an urban population), and Rigo
& >JDPB?5 coastal district (a rural population, effectively isolated from
r)T[(D'Tm- Port Moresby despite being only 2–4 h away by road). From
F%/h* this sample frame, 39 clusters (with probability proportionate
=
+Xc4a to population size) were chosen, using a systematic random
Ufid%T' sampling strategy.
M>H=z#C>/A Within each cluster, the supervisor chose households
5OP`c< using a random process. Residency was defined as living in
(ywo
a that cluster household for 6 months or more over the past
q/&y*)&'O year, and sharing meals from a common kitchen with other
8cxai8 members of the household. Eligible resident subjects aged
p@Q5b}xCG_ 50 years and older were then enumerated by trained volunteers
K
k[`dR; from the Port Moresby St John Ambulance Services.
qr(`&hB-L This continued until 30 subjects were enrolled. If the
`,Vv["^ PB required number of subjects was not obtained from a particular
n(;:*<Rh cluster, the fieldworkers completed enrolment in the
-uR72f nearest adjacent cluster. Verbal informed consent was
RDQ^dui obtained prior to all data collection and examinations.
UW
Xm?v2j A standardized survey record was completed for each
hfM;/ participant. The volunteers solicited demographic and general
E;(Rm>lB information, and any history of cataract surgery. They
]$iN#d|ZU also measured visual acuity. During a methodology pilot in
cPF<D$B the Morata settlement area of Port Moresby, the kappa statistic
C2F0tr| for agreement between the four volunteers designated
{padD p to perform visual acuity estimations was over 0.85.
zY9H% The widely accepted and used ‘presenting distance visual
zA+@FR? acuity’ (with correction if the subject was using any), a measure
r.3KPiYK of ocular condition and access to and uptake of eye care
g:)DNy services, was determined for each eye separately. This was
DPHQ,dkp done in daylight, using Snellen illiterate E optotypes, with
MXh
"Y*} four correct consecutive or six of eight showings of the
t-e5ld~a smallest discernible optotype giving the level. For any eye
#;tT8[Ewuw with presenting visual acuity worse than 6/18, pinhole acuity
H#d:kil Ny was also measured.
9}tG\0tL* An ophthalmologist examined all eyes with a history of
qoU3"8 cataract surgery and/or reduced presenting vision. Assessment
SAv<& of the anterior segment was made using a torch and
wsfd8T4 loupe magnification. In a dimly lit room, through an undilated
2&(sa0*y pupil, the status of the visually important central lens
83ipf"]* was determined with a direct ophthalmoscope. An intact red
1&)_(|p[C reflex was considered indicative of a ‘normal’ clear central
C*70;:b lens. The presence of obvious red reflex dark shading, but
bha?eN transparent vitreous, was recorded as lens opacity. Where
p='-\M74K
present, aphakia and pseudophakia with and without posterior
b!>\2DlyJ capsule opacification were noted. The lens was determined
s4= "kT] to be not visible if there were dense corneal opacities
keAcKhj or other ocular pathologies, such as phthisis bulbi, precluding
'Pltn{iq[ any view of the lens. The posterior segment was examined
gQWd&)'muf with a direct ophthalmoscope, also through an
g=)J~1&p undilated pupil.
i, n D5@# A cause of vision loss was determined for each eye with
VDEv>u4 a presenting visual acuity worse than 6/18. In the absence of
^(x^6d any other findings, uncorrected refractive error was considered
V,7%1TZ: to be that cause if the acuity then improved to better
S5=Udd" than 6/18 with pinhole. Other causes, including corneal
n]CbDbNw7) opacity, cataract and diabetic retinopathy, required clinical
=d;Vk findings of sufficient magnitude to explain the level of vision
D vkxI<Xa loss. Although any eye may have more than one condition
':
F}3At contributing to vision reduction, for the purposes of this
8Z#j7)G
study, a single cause of vision loss was determined for each
j0-McLc eye. The attributed cause was the condition most easily
0(>rG{u treated if each of the contributing conditions was individually
J QKdW treatable to a vision of 6/18 or better. Thus, for example,
!|wzf+
V when uncorrected refractive error and lens opacity coexisted,
|?m` xO refractive error, with its easier and less expensive treatment,
WPuz]Ty was nominated as the cause. Where treatment of a condition
9|r* pK[ present would not result in 6/18 or better acuity, it was
j#A%q"]8 determined to be the cause rather than any coincident or
%uESrc-; associated conditions amenable to treatment. Thus, for
>
dZ3+f example, coincident retinal detachment and cataract would
DtCEm(b0 be categorized as ‘posterior segment pathology’.
Xg<[fwW Participants who were functionally blind (less than 6/60
`d]Z)*9 in the better eye) because of unoperated cataract were interrogated
*U7%|wd about the reasons for not having surgery. The
Q'NmSX)0 responses were closed ended and respondents had the option
[`fI:ao| of volunteering more than one barrier, all of which were
'JmBh@A recorded in a piloted proforma. The first four reasons offered
RNJFSD. were considered for analysis of the barriers to cataract
q%3<Juq~$ surgery.
:>3=gex@^0 Those eyes previously operated for cataract were examined
qc-,+sn( to characterize that surgery and the vision outcome. A
!|{IVm/J detailed history of the surgery was taken. This included the
*MM#Z?mP age at surgery, place of surgery, cost and the use of spectacles
2OBfHO~D afterward, including reasons for not wearing them if that was
mX# "+X| the case.
:J_oj:0r"f The Rapid Assessment of Cataract Surgical Services data
)$M,Ul entry and analysis software package was used. The prevalences
AHh#Fx+K of visually significant cataract, unoperated blinding
n2-0.
Er cataract and cataract surgery were determined. Where prevalence
D@54QJ< estimates were age and gender adjusted for the population
%CH6lY=lI of PNG, the estimated population structure for the
Q\_{d0
0 882 Garap
I*)VZW et al.
fib#CY © 2006 Royal Australian and New Zealand College of Ophthalmologists
? 5|/
C year 2000
[I+9dSM1t 1
gT52G?- was used, and 95% CI were derived around these
7k8n@39? point estimates. Additional analysis for potential associations
|"7Pv
skT of cataract, its surgery and surgical outcomes employed the
#a~"K|'G STATA (version 8.0; StataCorp LP, TX, USA). Fisher’s exact
F9o6V|v test and the chi-square test for bivariate analysis and a multiple
d,#.E@Po logistic regression model for multivariate analysis were
&w\I<J`T used. Odds ratios (OR) and 95% CI were estimated. A
:hqZPajE P
z/|BH^Vw -
RI.2F*| value of
s;.=5wcvi? <
dk5|@?pe 0.05 was taken as significant for this analysis.
.[pUuVq] The Cataract Surgical Coverage (Eyes) (CSC(Eyes)) was
Hcv u7uD calculated. This is a surgical service impact indicator. It measures
Gp2!xKgm the proportion of cataract that has been operated on
&"S/Lt in a defined population at a particular point in time, being
iD_TP the eyes having had cataract surgery as a percentage of the
&9 B_/m3 combined total of all of those eyes operated with those
3oBC
currently blind (less than 6/60) from cataract (CSC(Eyes) at
Y`li> .\ 6/60
,;i
A2 =
@EZ>f5IO+ 100
$G";2(-k a
rw u3Nb /(
Wc4vCVw a
tbG8MXX +
Mtp%co )f b
6Z c)0I' ), where
[MmM 9J[" a
$WaZ_kt =
/bcY6b=: pseudophakic
AzjMv6N +
[m#NfA:h, aphakic eyes,
xqWj|jA and
"$@,n7k b
n7iE8SK|k =
^, wnp@ eyes with worse than 6/60 vision caused by cataract).
In(NF# 8
0xP:9rm The Cataract Surgical Coverage (Persons) (CSC(Persons))
OmNn,PCl8 was determined. This considers people with operated
(#u{ U= cataract (either or both eyes) as a proportion of those having
p7 [(z
operable cataract. (CSC(Persons) at 6/60
!sK#zAR2
=
H m8y]>$ 100(
dq1TRFu x
%UG/ak%z +
f\'G`4e
y
/Sj_y*x1e )/
qw0~*0
} (
T{HfP x
,\>g +
1^sb T[%R y
yk#rd~2Z0 +
Hdna{@~ z
E7LbSZ ), in which
hXnw..0" x
pXNtN5@FQ =
~Jxlj(" 0( persons with unilateral pseudophakia
=W|vOfy or unilateral aphakia and worse than 6/60 vision
4w9F+*- caused by cataract in the other eye,
QpJIDM/ y
i6A$1(:h =
ZmP1C`> persons with bilateral
{[lx!QF 8& previously operated cataract, and
R05T5Q1]A z
vcV!K^M- =
GJ>ypEWo persons with bilateral
2X^iV09 cataract causing vision worse than 6/60 in each).
%cd
]xQpCp 8
Pgx+\;w" The Cataract Surgical Rate, being the number of cataract
`43vxcMg operations per year per million of population, was also
% dYI5U89 estimated.
^X$
I= ro R
~6MMErSj ESULTS
XWc|[>iO Of the 1191 people enumerated, 5 subjects were not available
C~:b* X during the survey and 12 refused participation. Data
t3K>\ : from these 17 were not considered in the analysis. Of the
L#[HnsLp_ remaining 1174 (98.6%), 606 (51.6%) were female, and 914
Vj29L?3 (77.9%) were domiciled in rural Rigo.
;+r0
O0;9 Cataract caused 35.2% of vision impairment (presenting
Re7{[*Q4 vision less than 6/18) and 62.8% of functional blindness
Hc>([?P%t (presenting vision less than 6/60) in the 2348 eyes sampled
rT o%=0P (Table 1). It was second to refractive error (45.7%)
)!BB/'DRQ 7
*P&OxVz in the
uAC hu] former, and the leading cause of the latter.
HH+TjX/b For the 1174 subjects, cataract was the most prevalent
LZbRQ"!!o cause of vision impairment (46.7%) and functional blindness
Dy98[cL (75.0%) (Table 1). On bivariate analysis, increasing age
p0"BO4({{ (
:m)c[q8 P
$4ka +nfU <
l)|z2H 0.001), illiteracy (
<Wp`[S]r P
y>?k<