Relationship Between Climate, Disease Severity, and Causative Organism for Contact Lens–Associated Microbial Keratitis in Australia

●

sc●

●

watpagcmdtmc●

ttiiwwqcadi●

drtc(E

A

aL(

aBW

6

Relationship Between Climate, Disease Severity, andCausative Organism for Contact Lens–Associated

Microbial Keratitis in Australia

FIONA STAPLETON, LISA J. KEAY, PAUL G. SANFILIPPO, SUCHI KATIYAR, KATIE P. EDWARDS,

Mimscwaafbcb6A

mkqsteaetkw

Aewcvomcsa

ccf

PURPOSE: To evaluate associations between diseaseeverity, causative organism, and climatic variation inontact lens-related microbial keratitis in Australia.DESIGN: Prospective, observational case series.METHODS: Contact lens wearing patients (n � 236)ith presumed microbial keratitis presenting to private

nd hospital ophthalmologists in Australia between Oc-ober 1, 2003 and September 30, 2004 were identifiedrospectively. Clinical details, management information,nd microbiology data were collected and cases wereraded for severity based on lesion size and locationriteria. Causative organisms were assigned to “environ-ental” or “endogenous” groups. Climate zone andaytime temperature and humidity were determined forhe geographic location of each event. The main outcomeeasures were disease severity, causative organism, and

limate zone.RESULTS: Severe contact lens-related microbial kerati-

is was more likely to occur in warmer, humid regions ofhe country (P < .001), compared with smaller, increas-ngly peripheral corneal lesions that were more commonn cooler conditions (P < .001). Culture-proven keratitisas predominantly caused by environmental organismsith Pseudomonas aeruginosa being recovered most fre-uently. Environmental organisms were isolated moreommonly from tropical regions of the country and alsoccounted for nearly all cases of vision loss that occurreduring the study period. Humidity did not have a signif-cant effect on causative organism.

CONCLUSIONS: Climatic conditions play a role inisease severity and causative organism in contact lens-elated microbial keratitis and therefore have implica-ions for practitioners involved in contact lens care andontact lens wearers who live in or travel to the tropics.Am J Ophthalmol 2007;144:690–698. © 2007 bylsevier Inc. All rights reserved.)

ccepted for publication Jun 26, 2007.From the Institute for Eye Research (F.S., T.N.); School of Optometry

nd Vision Science, University of New South Wales, Australia (F.S.,.J.K., P.G.S., S.K., K.P.E.); and Vision Cooperative Research CentreF.S., L.J.K., S.K., K.P.E.), Sydney, Australia.

Inquiries to Fiona Stapleton, School of Optometry and Vision Sciencend Vision CRC, Level 3, North Wing, Rupert Myers Building Gate 14,

Aarker Street, University of New South Wales, Sydney, 2052, New Southales, Australia; e-mail: [email protected]

© 2007 BY ELSEVIER INC. A90

ICROBIAL KERATITIS IS A SERIOUS OCULAR IN-

fectious disease that can lead to significantvision loss and ophthalmic morbidity. It is rare

n the absence of predisposing factors,1,2 and was onceost commonly associated with ocular trauma or ocular

urface disease. However, the increasingly prevalent use ofontact lenses in the general community has resulted in aell-established association between microbial keratitisnd contact lens wear in the literature.3–6 The pathogenssociated with contact lens-related microbial keratitis isrequently bacterial, although fungi and protozoa haveeen reported.7,8 In several large series of culture-provenontact lens-related cases, Pseudomonas aeruginosa haseen cited as the primary causative organism (23% to3%), followed by gram-positive bacteria and fungi andcanthamoeba.7,9,10

Individual regional studies carried out in different cli-ate zones have suggested that, for all causes of microbial

eratitis, gram-positive bacterial species are more fre-uently recovered in temperate zones11 and gram-negativepecies and fungi in tropical climates.12–15 Similar climaterends have been observed for contact lens-related dis-ase.8,10,11 The relationship between climatic conditionsnd the observed microbial spectrum is important as manyye units or independent practitioners may not have accesso microscopy or culture facilities. In these instances,nowledge of the “regional” etiology of keratitis is helpfulhen instigating empirical therapy.A retrospective case series from a regional center inustralia examining all causes of microbial keratitis has

stablished that visual outcome was strongly associatedith causative organism.16 We therefore hypothesized thatontact lens-related microbial keratitis resulting from en-ironmental pathogens was more likely to result in a severeutcome, that disease from environmental pathogens wasore common in tropical climates, and that climatic

onditions that favor growth of environmental organisms,pecifically higher temperature and humidity would bessociated with more severe disease.

We report a 12-month prospective study to evaluateausative organism, disease severity, and environmentalonditions (climate zone, temperature, and humidity)or presumed contact lens-related microbial keratitis in

LL RIGHTS RESERVED. 0002-9394/07/$32.00doi:10.1016/j.ajo.2007.06.037

●

Zucopppt

oaca

kOucw

V

METHODS

CASE ASCERTAINMENT: The Royal Australian and Newealand College of Ophthalmologists membership register wassed to identify practicing ophthalmologists in Australia, in-luding those in training and in research positions. A totalf 661 practitioners were contacted initially and invited toarticipate in the study via a mailed information pack. Theack included a letter of invitation quoting a uniqueractitioner identification code and password for use on

TABLE 1. Severity Classification for Presumed C

Grade

Severe keratitis with vision loss Visual loss/surgery

Severe keratitis without vision loss Culture positive

Culture negative/not cultur

Mild keratitis Mild

*Report from practitioner or lab; 3� colonies on 1� solid mediu

morphology to organisms on smear. This grade allows for some vari

of the test result. Generally, if the organism was only found on one

negative.†Dimension in the longest meridian.

he study website (www.mk.study.org). In addition to e

MICROBIAL KERATITIS IN AUSTRAOL. 144, NO. 5

nline case reporting, the website enabled practitioners toccess information on the approval status of ethics appli-ations at their relevant local health service, eligibility,nd diagnostic criteria and patient consent forms.

New cases of presumed contact lens-related microbialeratitis presenting to ophthalmic practitioners betweenctober 1, 2003 and September 30, 2004, were identified

sing either the internet or postal/fax report. Initial dataollected included contact lens type, mode of wear,hether the lesion was cultured, and the organisms recov-

FIGURE. Climate zones basedon temperature and humidityderived from Australian Govern-ment Bureau of Meteorologydata: http://www.bom.gov.au/lam/climate/levelthree/cpeople/travelmap.htm.

ct Lens-Related Microbial Keratitis in Australia

Criteria

Vision loss of 2 or more lines of best-corrected visual acuity or

surgical intervention

Identification of a causative organism from the corneal scrape*

Any part of the lesion within central 4 mm

Outside central 4 mm with a hypopyon

Outside central 4 mm, �2 mm in diameter†

Lesion outside central 4 mm, �2 mm in diameter

� on one solid and growth in broth; growth in 2 broth � similar

between laboratory and practitioner criteria and their interpretation

ium or after long periods of incubation, the result was considered

onta

ed

m; 3

ability

med

red. Subsequent detailed information was obtained re-

LIA AND RELATED FACTORS 691

gPtftma1citeoi

lmtmsfirttw1ecois

●

ceotct

ltl

aoavbaavc“bcatc

●

6

arding clinical details and management information.ractitioners were actively surveyed at the end of eachwo-month response period either via email, telephone,ax, or mail to prompt responses, irrespective of whetherhey had seen an eligible case. Case detection was aug-ented by retrospective inpatient and casualty (where

vailable) records audits at major ophthalmic centers (n �1) with catchment populations of 200,000 or more. Dataapture techniques ranged from manual prospective casedentification by hospital staff to retrospective identifica-ion from diagnostic codes or pathology results. A visit toach of the 11 hospitals was scheduled for the conclusionf the study to collect detailed, de-identified clinicalnformation from the records.

In Australia, optometrists fit the majority of contactenses, but are not usually involved in the treatment oficrobial keratitis, with the exception of a small group of

herapeutically certified optometrists in Victoria and Tas-ania. For this reason, supplementary case reporting was

ought through surveillance of 3,288 optometrists identi-ed through Australian state and territory optometricegistration boards. Optometrists were similarly con-acted at the start of the study and were provided withhe information pack. Active follow-up was conductedith the therapeutically certified optometrists (n �73) and those members of regional contact lens soci-ties (n � 444). Regular active follow-up was notonducted with the remainder of the optometrists. Theptometrists’ case reporting provided supplementarynformation, including pre-event visual acuity and vi-ual outcome data post-event.

MICROBIOLOGY: For the purposes of the study, eligibleases were individuals presenting with a lesion which wasither culture-proven or culture negative/not scraped withne or more of the following signs/symptoms; any part ofhe lesion within the central 4 mm of the cornea, anteriorhamber response, or significant pain (either reported byhe practitioner or the patient).

Eligible cases were current contact lens wearers usingenses for low refractive error correction and ages between 15o 64 years old at the date of event (cases were excluded when

TABLE 2. Proportion of Cases in Each CatZones (%) for Presumed Contact Len

Disease Severity Tropi

Severe keratitis with vision loss 11 (1

Severe keratitis without vision loss 37 (6

Mild keratitis 8 (1

Total 56 (1

NS � not significant.

ens wear was intended for medical or therapeutic reasons). r

AMERICAN JOURNAL OF92

The criteria used for grading the severity of eligible casesre shown in Table 1. To enable meaningful comparisonsf disease severity between climatic conditions, cases weressigned to three descriptive groups. “Severe keratitis withision loss” was defined as a loss of more than 2 lines ofest-corrected spectacle acuity compared with pre-eventcuity (where available), fellow eye acuity (where avail-ble) or compared with 20/20. “Severe keratitis withoutision loss” represents the sum of culture positive andulture negative/not cultured cases that did not fit themild” criteria. The “mild” group was retained separatelyecause there remains uncertainty in the ophthalmicommunity as to whether small, more peripheral events aren acute inflammatory process or truly infective in na-ure17; however, this issue is beyond the scope of theurrent study.

CLIMATE ZONE: Regional variance analysis of organism

y of Disease Severity by Grouped Climatelated Microbial Keratitis in Australia

Temperate Total P value

22 (12.2) 33 (14.0) NS

79 (43.9) 116 (49.2) .018

79 (43.9) 87 (36.8) �.001

180 (100) 236 (100)

TABLE 3. Causative Organisms in Culture-Proven Casesof Contact Lens-Related Microbial Keratitis (n � 63)

Gram-negative bacteria

Pseudomonas aeruginosa or spp. 35*

Serratia spp. 6

Klebsiella oxygenae 1

Other gram-negative bacteria 1

Subtotal 43 (68.2%)

Gram-positive bacteria

Staphylococcus aureus 4

Coagulase-negative staphylococcus 4

Nocardia sp. 2

Streptococcus pneumoniae 1

Corynebacterium sp. 1

Other gram-positive bacteria 4

Subtotal 16 (25.4%)

Acanthamoeba 2 (3.2%)

Fungus/yeast 2 (3.2%)

Total 63 (100%)

*Includes five polymicrobial Pseudomonas aeruginosa cul-

tures and four Pseudomonas spp.

egors-Re

cal

9.6)

6.1)

4.3)

00)

ecovered and case severity across Australia is based on

OPHTHALMOLOGY NOVEMBER 2007

cmct

wBpMa

●

misat

tcNp

t(nbrs

●

ritp1bsdo

V

limate zone data available from the Australian Govern-ent Bureau of Meteorology (www.bom.gov.au). Australia

an be divided into six main climatic regions or zones onhe basis of air temperature and humidity (Figure).

Eligible cases’ latitudes and longitudes were matchedith the Australian Weather Stations (AWS) of theureau of Meteorology using the AWS coordinates. Tem-erature and humidity data collected by the Bureau ofeteorology at the AWS sites were matched with the day

nd location of event initiation.

DATA ANALYSIS: A proportions test was used to deter-ine associations between climate zone and disease sever-

ty, culture result and between lesion locations, diseaseeverity, and organism type. A one-way analysis of vari-nce with multiple comparisons test was used to determinehe effect of environmental conditions on culture result.

Micro-organisms have been classified into environmen-al and endogenous groups. Environmental organisms in-lude gram-negative bacteria, Acanthamoeba, fungi, andocardia, whereas endogenous organisms comprise gram-

TABLE 4. Causative Organisms of Contact Lens-Relat

Organism

Gram-negative bacteria

Pseudomonas aeruginosa

P. aeruginosa � Bacillus

P. aeruginosa � Candida parapsil

P. aeruginosa � Klebsiella

P. aeruginosa � S. aureus

P. aeruginosa � Serratia spp. � Xylosoxidans spp. � Stenotroph

Pseudomonas spp.

Serratia liquefaciens

Serratia marcescens

Serratia spp.

Klebsiella oxgenae

Other gram-negative bacteria

Subtotal gram-negative

Gram-positive bacteria

Coagulase-negative staphylococcus

Staphylococcus aureus

Nocardia spp.

Strep. pneumoniae

Corynebacterium spp.

Staphylococcus epidermidis

Staphylococcus spp.

Other gram-positive bacteria

Subtotal gram-positive

Acanthamoeba

Fungi filamentous

Fungi spp.

Subtotal culture-proven

Culture-negative

Unknown

Total

ositive bacteria (not including Nocardia). This classifica- g

MICROBIAL KERATITIS IN AUSTRAOL. 144, NO. 5

ion system has been used to differentiate commensalmost commonly gram-positive) flora from more virulentoncommensal (environmental) microorganisms that maye associated with more severe disease outcome in terms ofate of vision loss, cost of disease, and duration ofymptoms.18

RESULTS

CASE ASCERTAINMENT: The overall ophthalmologistesponse rate was 96.2% (636/661), with 1.8% (12) refus-ng to participate. The minimum response rate for any ofhe six reporting periods was 88.2%. A total of 345 cases ofresumed microbial keratitis were reported during the2-month study period. Of these, 16 cases were excludedecause the onset of disease was outside the period ofurveillance. A further seven cases were excluded as theyid not meet the inclusion criteria; five were keratoconic,ne aphakic, and one wore bandage contact lenses.In 46 cases, there was insufficient information to

icrobial Keratitis in Australia by Climate Zones (n [%])

Tropical Temperate Total

16 (59.3) 10 (27.8) 26 (41.3)

1 (3.7) 0 1 (1.6)

0 1 (2.8) 1 (1.6)

1 (3.7) 0 1 (1.6)

1 (3.7) 0 1 (1.6)

as sp. 1 (3.7) 0 1 (1.6)

2 (7.4) 2 (5.5) 4 (6.3)

0 1 (2.8) 1 (1.6)

1 (3.7) 3 (8.3) 4 (6.3)

0 1 (2.8) 1 (1.6)

0 1 (2.8) 1 (1.6)

0 1 (2.8) 1 (1.6)

23 (85.2) 20 (55.5) 43 (68.2)

1 (3.7) 2 (5.5) 3 (4.8)

0 4 (11.1) 4 (6.3)

0 2 (5.5) 2 (3.2)

0 1 (2.8) 1 (1.6)

0 1 (2.8) 1 (1.6)

0 1 (2.8) 1 (1.6)

1 (3.7) 0 1 (1.6)

1 (3.7) 2 (5.5) 3 (4.8)

3 (11.1) 13 (36.1) 16 (25.4)

1 (3.7) 1 (2.8) 2 (3.2)

0 1 (2.8) 1 (1.6)

0 1 (2.8) 1 (1.6)

27 (100) 36 (100) 63 (100)

7 72 79

1 5 6

35 113 148

ed M

omon

rade the severity. For the purpose of the incidence

LIA AND RELATED FACTORS 693

adgoapmmtscta

●

pw

zi13caph

tvlW1wg

ant b

6

nalysis, these cases were apportioned according to theistribution of eligible/noneligible cases. The reviewersraded the remaining 276 cases and were in agreementn the initial severity grading in 247 (89.5%). Finalgreement was reached for the remaining 10.5% withanel discussion. After panel review, 31 (11.2%) did noteet the diagnostic criteria for contact lens-relatedicrobial keratitis (predictive value positive 88.8%). Of

hese 31, 24 were cases of corneal inflammation orterile/marginal keratitis, three were adenoviral kerato-onjunctivitis, one had herpes simplex keratitis, andwo had ocular trauma without evidence of an associ-ted infection.

CLIMATE ZONE: Of the 245 eligible cases, six did notrovide an address associated with event initiation and

TABLE 5. Comparison of Organisms IsolaLens-Related Micro

Organism Tropical (%

Environmental organisms

Pseudomonas aeruginosa* 74.1 (n

Serratia spp. 3.7 (n

Nocardia spp. 0

Acanthamoeba 3.7 (n

Fungi spp. 0

Other 7.4 (n

Subtotal 88.9 (n

Endogenous organisms

Staphylococcus aureus 0

Coagulase negative

Staphylococcus

3.7 (n

Other 7.4 (n

Subtotal 11.1 (n

NS � not significant.

*Includes polymicrobial cultures in which Pse

organism grown.

TABLE 6. Temperature and Humidity (DOrganisms for Presumed Contact Le

Climate Condition

Environmental O

(n � 49

Mean maximum temperature (C) 25.6 � 4

Mean minimum temperature (C) 15.6 � 6

Mean humidity 9 a.m. (%) 69.9 � 1

Mean humidity 3 p.m. (%) 56.0 � 1

ANOVA � analysis of variance; NS � not sig

*Differences indicated are statistically signific

ere excluded from the analysis. When grouped by climate k

AMERICAN JOURNAL OF94

one, 11 cases of microbial keratitis occurred in zone 1, 45n zone 2, one in zone 3, two in zone 4, 168 in zone 5, and2 in zone 6. Australia has a low population density in zonesand 4 (mostly desert) and as n � 3 for these two zones

ombined, they have been excluded from the subsequentnalysis. Zones 1 and 2 (tropical) and zones 5 and 6 (tem-erate) are climatically similar in terms of temperature andumidity and so have been grouped together for analysis.Of the remaining 236 cases with sufficient information

o grade disease severity, there were 33 (14.0%) that hadisual acuity reduction of at least 2 lines on the Snellenetter chart or required surgical intervention (Table 2).

ithin this group, 11 (33.3%) were culture negative and0 of these occurred in the temperate climate zone. Thereere 116 cases in the “severe keratitis without vision loss”roup, 41 (17.4%) of which were culture-proven. Mild

rom Grouped Climate Zones for Contacteratitis in Australia

27 Temperate (%) n � 36 P value

) 30.6 (n � 11) .002

11.1 (n � 4) NS

5.5 (n � 2) NS

2.8 (n � 1) NS

5.5 (n � 2) NS

13.9 (n � 5)

) 69.5 (n � 25) NS

11.1 (n � 4) NS

8.3 (n � 3) NS

11.1 (n � 4)

30.5 (n � 11) NS

onas aeruginosa was the predominant micro-

f Event Initiation) for Various Causativelated Microbial Keratitis in Australia

Organism Type

s Endogenous Organisms

(n � 14)

Culture Negative

(n � 75)

P value

(ANOVA)

25.5 � 8.8 21.1 � 8.9* .011

12.3 � 7.2 9.9 � 6.3* �.001

76.6 � 15.2 73.4 � 20.6 NS

48.5 � 20.1 54.4 � 24.2 NS

nt.

y multiple comparisons testing.

ted fbial K

) n �

� 20

� 1)

� 1)

� 2)

� 24

� 1)

� 2)

� 3)

udom

ay ons-Re

rganism

)

.8*

.1*

4.1

5.6

nifica

eratitis occurred in 87 (36.8%) cases.

OPHTHALMOLOGY NOVEMBER 2007

●

esocmb(c4

gpAp.q(pwtha

●

D

chwi

tatrirmis

twmHo

tCapw.cgcnlcc

ant b

25

V

MICROBIOLOGIC CHARACTERISTICS: At the initialxamination, 63% (148/236) of corneal lesions werecraped for microbiologic identification of the causativerganism. Positive cultures were obtained in 42.6% (63) ofases, with gram-negative bacteria accounting for theajority of isolates (68.2%), followed by gram-positive

acteria (25.4%) and Acanthamoeba (3.2%) and fungi3.2%) (Table 3). Pseudomonas aeruginosa was the mostommonly recovered microorganism, being observed in9.2% of culture-proven cases.The differences in microbial spectrum based on the two

rouped climate zones are shown in Table 4. Culture-roven isolates occurred more often in the tropical zones ofustralia with an observed rate of 77.1% (27/35) com-

ared with 31.9% (36/113) in temperate conditions (P �001). Similarly, gram-negative bacteria were most fre-uently recovered in tropical regions, occurring in 85.2%23/27) cases compared with 55.6% (20/36) in the tem-erate zone (P � .05). In contrast, gram-positive bacteriaere more commonly isolated in 36.1% (13/36) of cases in

emperate conditions (P � .05). Interestingly, there was aigh culture negative rate of 63.7% (72/113) in coolerreas compared with tropical areas (P � .01).

COMPARISONS BETWEEN ENVIRONMENTAL AND EN-

OGENOUS MICROORGANISMS: There was a strong asso-iation between Pseudomonas aeruginosa and warmer,umid areas of Australia (P � .005) (Table 5). Serratia spp.ere the second most commonly observed organisms and

TABLE 7. Corneal Location and Causative OKeratitis

Corneal Location Environmental Organisms

Central 12 (26.1)

Midperipheral 28 (60.8)

Peripheral 6 (13.1)*

Total 46 (100)

NS � not significant.

*Differences indicated are statistically signific

TABLE 8. Disease Severity and CausativeKeratitis

Disease Severity

Environmen

Tropical Te

Keratitis with vision loss 9

Keratitis without vision loss 15

Total cases 24

n contrast to Pseudomonas were more common in the c

MICROBIAL KERATITIS IN AUSTRAOL. 144, NO. 5

emperate region of the country (13.9%). Staphylococcusureus (11.1%) and Nocardia spp. (5.5%) were exclusive toemperate conditions. One case of Acanthamoeba wasecovered from each grouped climate region and bothnstances of fungi were recovered in the temperate envi-onment. Although environmental organisms occurredore frequently in the tropics and endogenous organisms

n temperate areas, this difference was not statisticallyignificant.

Average temperature and humidity data for the period ofhe study is shown in Table 6. Environmental organismsere associated with a higher average minimum andaximum temperature than were endogenous microbes.umidity did not have a significant effect on causative

rganism.Corneal infiltrate location data was available for 60 of

he culture-proven and 75 of the culture negative cases.ulture growth and type of causative organism were

ssociated with corneal infiltrate location, but only foreripheral lesions (Table 7). Culture negative infiltratesere more likely to occur in the peripheral cornea (P �

05). Conversely, however, there was no significant asso-iation between central corneal location and culturerowth, with 52% of culture-proven lesions occurringentrally. Within the culture positive subgroup, endoge-ous organisms were more likely to cause a peripheral

esion (P � .05). Although environmental organisms moreommonly caused central disease than endogenous mi-robes, no significant relationship was observed between

ism (%) for Contact Lens-Related Microbialustralia

Organism Type

enous Organisms Culture Negative P value

1 (7.1) 12 (16.0) NS

6 (42.9) 30 (40.0) NS

7 (50)* 33 (44.0) .01

14 (100) 75 (100)

y a test of proportions.

nism for Contact Lens-Related Microbialustralia

anisms Endogenous Organisms

e Total Tropical Temperate Total

21 1 0 1

28 2 11 13

49 3 11 14

rganin A

Endog

Orgain A

tal Org

mperat

12

13

entral corneal location and organism type.

LIA AND RELATED FACTORS 695

((miAoc

T

sodTofdpr

esicwCebwcl

icc(rtwmsmwpieoPplic

atgcr(cadpo

dcampasdmoiphtabiNllwicscrqcpAfdtcriTcc

hha

6

Of the culture-proven cases, vision loss occurred in 42.9%21/49) of cases caused by environmental organisms and 7.1%1/14) of cases caused by endogenous organisms. Environ-ental organisms were the most common cause for vision loss

n contact lens wearers who develop microbial keratitis inustralia, accounting for 95.5% of cases (Table 8). No cases

f vision loss were caused by endogenous microbes in cooleronditions.

DISCUSSION

HIS STUDY REPORTS A PROSPECTIVE SURVEILLANCE

tudy of presumed contact lens-related microbial keratitisccurring over a one-year period in Australia, stratified byisease severity, causative organism, and climate zone.his is the first study to perform a contemporaneous studyf disease across climate zones. The high response raterom practitioners provides some confidence that the fullisease spectrum has been captured over the 12-montheriod and the large dataset enables full exploration of theelevant factors.

Of the 236 eligible cases, 14% (33) lost vision from thevent and 63% (149) had “severe” disease based on lesionize and location. Severe keratitis was more likely to occurn warmer, humid regions of the country (P � .001),ompared with smaller, increasingly peripheral lesions thatere more common in cooler conditions (P � .001).ulture-proven keratitis was predominantly caused by

nvironmental organisms, with Pseudomonas aeruginosaeing recovered most frequently. Environmental organismsere isolated more commonly from tropical regions of theountry and also accounted for nearly all cases of visionoss that occurred during the study period.

In this series, “severe keratitis without vision loss”ncluded culture-proven, central, and larger peripheralorneal lesions. Mild disease occurred in 36.8% (87) of allases with a significant propensity for temperate regions90.8%). Not withstanding differences in diagnostic crite-ia, mild disease rates from other groups were determinedo be 63.3%3 in temperate climatic conditions comparedith 36% in a tropical environment.19 The majority oficrobes recovered from culture-proven scrapings in our

tudy were gram-negative (68.2%) and these occurredost frequently in both climate zones. A regional effectas noted, however, with a higher proportion of culture-roven cases being observed in tropical conditions; this isn agreement with other studies conducted in a similarnvironment.7,10 The most common species identifiedverall, and within each of the grouped climate zones, wasseudomonas aeruginosa (49.2%). There was a significantropensity for tropical Australia as the preferred climaticandscape for this microorganism (P � .005). The situations quite different, though, when a temperate climate is

AMERICAN JOURNAL OF96

ssociation with gram-positive bacteria in such condi-ions,8,11 and this was also reflected in our results. In thisram-positive cohort, Staphylococcus aureus was the mostommon organism isolated (11.1%), and this is compa-able to Bennett and associates (16.7%)11 and Schein26%)8. Interestingly, the culture negative rate wasonsiderably higher in temperate conditions and thisppears to correlate with the increased rate of “mild”isease observed. In this case, if a practitioner diderform a corneal scrape, they were more likely tobtain a negative result.Aside from Pseudomonas and Staphylococcus, the pre-

ominant remaining organisms in the microbial spectrumomprised of Serratia spp., Nocardia spp., Acanthamoeba,nd fungal species. Serratia spp., and, in particular, Serratiaarcescens has emerged in recent years as an importantathogen in microbial keratitis.20,21 As with Pseudomonaseruginosa, Serratia marcescens has the potential to causeevere disease and is commonly resistant to contact lensisinfecting solutions21 and antimicrobial agents. Serratiaarcescens accounted for 6.3% of culture-proven cases inur series and was the second most common organismsolated after Pseudomonas aeruginosa, with a higher pro-ortion occurring in the temperate zone. Other studiesave found rates of 4.7%7 and 7.3%5 in subtropical andemperate conditions, respectively. Serratia marcescens wasn equally prevalent cause of contact lens-related micro-ial keratitis (18%), together with Pseudomonas aeruginosan another study reporting infections in a tropical region.22

ocardia spp. have been infrequently reported in theiterature in both general microbial keratitis and contactens-related microbial keratitis populations. Nocardia spp.as recovered from two cases (1.6%) where both occurred

n the temperate environment; this is in contrast to onease (4.7%) noted in subtropical Hong Kong.19 A similarlymall number of cases of Acanthamoeba, one from eachlimate zone, were observed in our study. This is compa-able to the results of Schein and associates,8 althoughuite disparate to a Scottish Study in which 70% ofontact lens-related microbial keratitis was caused by thisrotozoa.4 Our findings did not reflect an increased rate ofcanthamoeba in cooler conditions; therefore, nonclimatic

actors such as patient hygiene, mode of contact lensisinfection, and water supply may have played a role inhe Scottish Study. Fungal infection is a relatively rareomplication of contact lens wear,23 and this is alsoeflected in our results. In a sample of contact lens wearersn South India,24 fungal keratitis occurred in 3.6% of cases.his is in contrast to a general series of microbial keratitisases in a similar region of the country where 44% wereaused by ocular trauma.25

We would hypothesize that microbial growth is en-anced in warmer, humid climates, because we noted aigher rate of culture-proven cases from these areas. Welso anticipated that tropical conditions may favor growth

OPHTHALMOLOGY NOVEMBER 2007

ciwpptiIoPgo

srlbi

hmromidtemtlaemcv

TSCAcHhiaa

Tac

V

ase only when temperature was considered. Higher min-mum and maximum daytime temperatures were associatedith the recovery of environmental organisms when com-ared with culture negative cases. Certainly, the relativerevalence of fungal keratitis increases toward tropical lati-udes25 and Pseudomonas spp., the species most frequentlysolated in our study, grows best at warmer temperatures.26

nterestingly, humidity had no significant effect on causativerganism in this study and was an unexpected finding.seudomonas aeruginosa, for example, has a predilection forrowth in moist environments, which is probably a reflectionf its natural existence in soil and water.

In conclusion, this study supports the high degree ofuspicion for Pseudomonas aeruginosa in contact lens-elated keratitis. Severe microbial keratitis and visualoss in contact lens wearers is more likely to be causedy environmental pathogens and is more likely to occur

contact lens-related keratitis. Cornea 1989;8:281–285.

1

1

1

1

1

1

1

MICROBIAL KERATITIS IN AUSTRAOL. 144, NO. 5

umidity was associated with the recovery of environ-ental organisms. We believe this is the first conclusive

eport that climatic conditions impact both causativerganism and disease severity in contact lens-relatedicrobial keratitis, which has implications for those

nvolved in the management of contact lens-relatedisease. Contact lens wearers who live in or travel to theropics are more likely to have disease caused bynvironmental organisms and are at a higher risk of aore severe outcome and vision loss. These cases need

o be managed cautiously with scraping and microbio-ogical analysis and on the assumption that Pseudomonaseruginosa is implicated, commenced on appropriatempiric antimicrobial therapy. Patients for their partay be educated on the importance of hygiene and

ompliance issues and to seek appropriate care earlyia an ophthalmic practitioner if symptoms should

HIS STUDY WAS SUPPORTED BY THE INSTITUTE FOR EYE RESEARCH, SYDNEY, NEW SOUTH WALES, AUSTRALIA (DRStapleton and Naduvilath), the University of New South Wales (Drs Stapleton, Sanfilippo, Keay, and Edwards), the Vision Cooperative Researchentre (Drs Edwards, Keay, and Katiyar), CIBA Vision USA, Atlanta, Georgia and the National Health and Medical Research Council, Canberra,CT, Australia (Dr Keay). The authors indicate no financial conflict of interest, Keays. Involved in design and conduct of study (F.S., L.K., K.E., S.K.);

ollection, management, analysis, and interpretation of the data (L.K., K.E., P.S., S.K., T.N., F.S.); and preparation of the manuscript (P.S., F.S., L.K.).uman research ethics approval was obtained from the University of New South Wales Human Research Ethics Committee and from 63 regional areaealth services in Australia. The study was carried out in accordance with the Declaration of Helsinki and with federal and state ethics and privacy laws

n Australia. Informed consent was obtained from cases to allow collection of clinical data and microbiological information. The authors gratefullycknowledge the contribution of RANZCO in providing access to members contact details and for publishing study updates in their regular newsletter,lso to Optometrists Association Australia and the state registration boards for providing access to lists of practitioners.

We acknowledge with gratitude the clinical and medical records staff (C. Petsoglou, J. Males, K. McLellan, M. Morris, J. Stretton, G. Snibson, H.aylor, A. Poon, K. Lingcoln, P. Versace, I. Francis, N. Morlet, I. Sim, Mei-Ling Tay-Kearney, A. Apel, G. Wharton, P. Mitchell, J. Wolford, G. Lee)t participating hospitals for facilitating the data audit, also importantly all ophthalmic practitioners in Australia. We would like to thank the steeringommittee (O. Schein, D. Cavanagh, J. Dart, B. Holden, F. Stapleton, H. Taylor, D. Sweeney, D. Fonn) for protocol and data review.

REFERENCES

1. Miedziak AI, Miller MR, Rapuano CJ, et al. Risk factors inmicrobial keratitis leading to penetrating keratoplasty. Oph-thalmology 1999;106:1166–1170.

2. Musch DC, Sugar A, Meyer RF. Demographic and predis-posing factors in corneal ulceration. Arch Ophthalmol 1983;101:1545–1548.

3. Dart JK, Stapleton F, Minassian DC. Contact lenses andother risk factors in microbial keratitis. Lancet 1991;338:650–653.

4. Seal DV, Kirkness CM, Bennett HG, Peterson M, KeratitisStudy Group. Population based cohort study of microbialkeratitis in Scotland: incidence and features. Cont LensAnterior Eye 1999;22:49–57.

5. Bourcier T, Thomas F, Borderie V, et al. Bacterial keratitis:predisposing factors, clinical and microbiological review of300 cases. Br J Ophthalmol 2003;87:834–838.

6. Keay L, Edwards K, Naduvilath T, et al. Microbial keratitis:predisposing factors and morbidity. Ophthalmology 2006;113:109–116.

7. Alfonso E, Mandelbaum S, Fox MJ, Forster RK. Ulcerativekeratitis associated with contact lens wear. Am J Ophthal-mol 1986;101:429–433.

8. Schein OD, Ormerod LD, Barraquer E, et al. Microbiology of

9. Galentine PG, Cohen EJ, Laibson PR, et al. Corneal ulcersassociated with contact lens wear. Arch Ophthalmol 1984;102:891–894.

0. Houang E, Lam D, Fan D, Seal D. Microbial keratitis in HongKong: relationship to climate, environment and contact-lensdisinfection. Trans R Soc Trop Med Hyg 2001;95:361–367.

1. Bennett HG, Hay J, Devonshire P, et al. Antimicrobialmanagement of presumed microbial keratitis: guidelines fortreatment of central and peripheral ulcers. Br J Ophthalmol1998;82:137–145.

2. Hagan M, Newman M, Wright E, et al. Causes of suppurativekeratitis in Ghana. Br J Ophthalmol 1995;79:1024–1028.

3. Dunlop AS, Wright ED, Howlader SA, et al. Suppurativecorneal ulceration in Bangladesh. A study of 142 casesexamining the microbiological diagnosis, clinical and epide-miological features of bacterial and fungal keratitis. Aust NZOphthalmol 1994;22:105–110.

4. Ormerod LD. Causation and management of microbial keratitisin subtropical Africa. Ophthalmology 1987;94:1662–1668.

5. The Hong Kong Microbial Keratitis Study Group. Aetiologi-cal agents and risk factors of microbial keratitis in HongKong—preliminary report on first 50 cases. Hong Kong JOphthalmol 1997;1:99–103.

6. Keay L, Edwards K, Naduvilath T, et al. Microbial keratitis:predisposing factors and morbidity. Ophthalmology 2006;

LIA AND RELATED FACTORS 697

1

1

1

2

2

2

2

2

2

2

6

7. Holden BA, Reddy MK, Sankaridurg PR, et al. Contact-lensinduced peripheral ulcers with extended wear of disposablehydrogel lenses: histopathologic observations on the natureand type of corneal infiltrate. Cornea 1999;18:538–543.

8. Keay L, Edwards K, Naduvilath T, et al. Factors affecting themorbidity of contact lens-related microbial keratitis: a pop-ulation study. Invest Ophthalmol Vis Sci 2006;47:4302–4308.

9. Lam DS, Houang E, Fan DS, et al. Incidence and risk factorsfor microbial keratitis in Hong Kong: comparison withEurope and North America. Eye 2002;16:608–618.

0. Varaprasathan G, Miller K, Leitman T, et al. Trends in theetiology of infectious corneal ulcers at the F. I. ProctorFoundation. Cornea 2004;23:360–364.

1. Parment PA. The role of Serratia marcescens in soft contactlens-associated ocular infections: a review. Acta Ophthalmol

AMERICAN JOURNAL OF98

2. Alexandrakis G, Alfonso E, Miller D. Shifting trends inbacterial keratitis in South Florida and emerging resis-tance to fluoroquinolones. Ophthalmology 2000;107:1497–1502.

3. Wilhelmus KR, Robinson NM, Font RA, et al. Fungalkeratitis in contact lens wearers. Am J Ophthalmol 1988;106:708–714.

4. Sharma S, Gopalakrishnan S, Aasuri MK, et al. Trends incontact lens-associated microbial keratitis in Southern India.Ophthalmology 2003;110:138–143.

5. Leck AK, Thomas PA, Hagan M, et al. Aetiology ofsuppurative corneal ulcers in Ghana and South India, andepidemiology of fungal keratitis. Br J Ophthalmol 2002;86:1211–1215.

6. Lennette E, Balows A, Hausler W, et al. editors In: Manualof clinical microbiology, 4th edn. Washington, DC: Ameri-

OPHTHALMOLOGY NOVEMBER 2007

FSmde

V

Biosketch

iona Stapleton, MSc, PhD, McOptom, DCLP, FAAO, FBCLA, is based at the University of New South Wales inydney. Her research areas include the epidemiology and pathogenesis of contact lens-related disease, particularlyicrobial keratitis and corneal inflammation. Dr Stapleton was a principal investigator in a national surveillance study to

etermine the incidence of contact lens related-corneal infection in Australia and in a parallel case control study tostablish determinants of disease in the United Kingdom.

MICROBIAL KERATITIS IN AUSTRALIA AND RELATED FACTORSOL. 144, NO. 5 698.e1