C

UD

a

KPPC

1

a[dtpabtopbp

coa

Eh

1d

Contact Lens & Anterior Eye 35 (2012) 39– 45

Contents lists available at ScienceDirect

Contact Lens & Anterior Eye

jou rn al h om epa ge: www.elsev ier .com/ locate /c lae

orrelation of central and peripheral corneal thickness in healthy corneas

sama Fares, Ahmad Muneer Otri, Mouhamed Ali Al-Aqaba, Harminder S. Dua ∗

ivision of Ophthalmology and Visual Sciences, University of Nottingham, United Kingdom

r t i c l e i n f o

eywords:achymetryentacamorneal thickness

a b s t r a c t

Purpose: To study the thickness profile of the normal cornea in order to establish any correlation betweencentral and peripheral points.Methods: Sixty-seven eyes of 40 patients were subjected to central corneal thickness measurement (CCT)with an ultrasound pachymeter (UP) and corneal thickness mapping with the Oculus Pentacam. Thecorneal apex thickness (CAT), pupil centre thickness (recorded as CCT and corresponded to CCT of UP) andthickness at the thinnest location (CTL) were obtained and compared with each other. Corneal thicknessdata at 3 mm and 7 mm temporally, nasally, superiorly and inferiorly from the corneal apex were obtained.The mean corneal thickness values along the 2, 4, 6, 8 and 10 mm diameter concentric circles, with theCTL as the centre, were also obtained. The above data at different points were statistically correlated.Results: There was no significant difference between CCT readings measured by UP and Pentacam(P = 0.721). There was high positive correlation between the CAT values and the thickness at 3 mm

(R ≥ 0.845, P < 0.001) and at 7 mm points (R ≥ 0.654, P < 0.001). A gradual increase in thickness was notedfrom the centre to the periphery with a high positive correlation between the CTL values and the meanthickness at the circles of 2, 4, 6, 8 and 10 mm (R ≥ 0.635, P < 0.001).Conclusion: The results suggest that central corneal thickness can serve as a good guide for predictingperipheral thickness. For surgical procedures specifically undertaken at mid-peripheral and peripheralzones, the actual measurements at the site of surgery may confer some advantage.

Britis

© 2011

. Introduction

Morphometric data on central corneal thickness (CCT) hasssumed considerable clinical importance in relation to refractive1] and non refractive corneal surgery and glaucoma [2,3]. Theepth of incisions and of ablations in refractive surgery and thehickness of the cut in lamellar corneal surgery, all require fairlyrecise measurement of corneal thickness. Some incisions are madet the mid-periphery (arcuate incisions) [4], at the periphery (lim-al relaxing incisions) [5] or from the edge of the optical zone tohe periphery as in radial keratotomy [6]. In manual preparationf donor material for the DSEK procedure the depth of the initialeripheral incision in the donor sclerocorneal disc can be estimatedy the CCT. By implication, the CCT data is used as a surrogate foreripheral corneal thickness (PCT) in some instances.

Ultrasound pachymetry has been reported to give very accurate,

linically reliable and reproducible data [7,8], but determinationf the actual points to apply the probe is operator dependentnd can lead to inaccuracies when mapping the corneal thickness.

∗ Corresponding author at: Division of Ophthalmology and Visual Sciences, B floor,ye Ear Nose Throat Centre, University Hospital, Queens Medical Centre, Notting-am, NG7 2UH, United Kingdom.

E-mail address: harminder.dua@nottingham.ac.uk (H.S. Dua).

367-0484/$ – see front matter © 2011 British Contact Lens Association. Published by Elsoi:10.1016/j.clae.2011.07.004

h Contact Lens Association. Published by Elsevier Ltd. All rights reserved.

The advent of devices such as the Scheimpflug rotating camera(Pentacam; Oculus, Inc., Wetzlar, Germany), slit-scanning opticalpachymetry (Orbscan; Bausch & Lomb, Rochester, New York, USA)and high speed optical coherence tomography (OCT; Carl ZeissMeditec, Inc., CA, USA) has allowed accurate mapping of cornealthickness at multiple points on the cornea [9]. The Pentacam forexample collects information from up to 25,000 data points [10].Using such data we have been able to correlate CCT to the PCT andestablish a comprehensive thickness profile of the normal cornea.This information is reported in this paper.

2. Methods

This prospective clinical study enrolled 67 eyes of 40 patients.Inclusion criteria were healthy corneas with a best corrected visualacuity of 6/6 or better. The following exclusion criteria wereapplied: amblyopia, previous history of ocular disease or trauma,previous ocular surgery including refractive surgery and contactlens wear. Thirteen eyes of 13 patients were thus excluded. The

study was performed according to the Tenets of the Declarationof Helsinki. We certify that all applicable institutional and govern-mental regulations concerning the ethical use of human volunteerswere followed during this study.

evier Ltd. All rights reserved.

40 U. Fares et al. / Contact Lens & Anterior Eye 35 (2012) 39– 45

F ss (CCs by Pena

2

sN(

ammi

cjusatmastmctsd

(nwsav1t

j30apm

ig. 1. (A) Scatter plot showing linear regression between central corneal thicknehowing linear regression between corneal apex thickness (CAT in �m) measured

nd (B) represents the 45◦ line for perfect agreement.

.1. Measurements techniques

All eyes were subjected to central corneal thickness mea-urement with an ultrasound pachymeter (UP) (SP 3000, Tomey,agoya, Japan) and corneal thickness mapping with the Pentacam

Oculus, Inc., Wetzlar, Germany).All measurements were performed between 10 A.M and 1 P.M.

fter subjects had been awake for more than 1 h [11,12]. Eachodality was performed by expert operators blinded to the otherodality results. Pentacam measurements were done first to elim-

nate any corneal change which might be caused by the UP probe.Pentacam images were obtained with the subject seated with

hin on a chinrest and forehead against a forehead strap. The sub-ect was asked to fixate ahead on a target. The Oculus Pentacamses a rotating Scheimpflug camera and a monochromatic slit lightource (blue light-emitting diode, 475 nm) which rotate togetherround the optical axis of the eye. The operator visualised a real-ime image of the patient’s eye on a computer screen, with the

achine marking the pupil centre, edge and corneal apex. The oper-tor could manually, with a guidance of arrows displayed on thecreen; align and focus the image in the anteroposterior, horizon-al and vertical axes. The automatic release mode was activated to

inimise observer-dependent variables. Thus, Pentacam’s rotatingamera started capturing when correct focus and alignment withhe corneal apex had been achieved. Fifty slit images of the anterioregment were obtained within 2 s. The calibration of Pentacam wasone by the manufacturer.

The corneal apex thickness (CAT), pupil centre thicknessrecorded as CCT and corresponded to CCT of UP) and thick-ess at the thinnest location (CTL) were obtained and comparedith each other. PCT data at 3 mm and 7 mm temporally, nasally,

uperiorly and inferiorly from the corneal apex was obtaineds single-data points. Furthermore, the mean corneal thicknessalues were obtained along concentric circles of 2, 4, 6, 8 and0 mm in diameter with the thinnest corneal location as the cen-re.

Ultrasound pachymetry (UP) was then performed with the sub-ects seated upright using a handheld ultrasound pachymeter (SP000, Tomey, Nagoya, Japan). After instillation of oxybuprocaine

.4% eye drops, the handheld probe was placed gently on the corneat the point corresponding to the pupil centre and was aligned per-endicularly. Ten CCT readings were taken and averaged. The UPachine was calibrated by the manufacturer.

T in �m) as measured by Pentacam and ultrasound pachymetry. (B) Scatter plottacam and CCT (�m) measured by ultrasound pachymetry. The hashed line in (A)

2.2. Statistical analysis

MedCalc version 11 and SPSS version 16.0 (SPSS, Chicago) wereused for the statistical analysis. All data were checked for normalityusing Kolmogorov–Smirnov test. The following comparisons andcorrelations were made.

Comparison of CCT measurements with UP and Pentacam wasmade by using independent-sample t-test. “Effect size statisticsprovide an indication of the magnitude of the difference betweengroups (not just whether the difference could have occurredby chance). The following formula was used to calculate theeffect size (strength of the difference between our groups): Etasquared = t2/(t2 + (N1 + N2 − 2)); where t is the t test value obtainedfrom the SPSS output, N1 is the number of samples in the first groupand N2 is the number of samples in the second group [13].”

Correlation of CCT and CAT with the 3 and 7 mm PCT in thevertical (90–270◦) and horizontal (0–180◦) meridians was madeusing the Pearson product-moment correlation coefficient. Simi-larly correlation of CTL with mean thickness along the 2, 4, 6, 8 and10 mm circles was also made using the Pearson product-momentcorrelation coefficient.

3. Results

Of the 40 patients studied 24 were males (60%) and 16 werefemales (40%). The mean age was 38.65 years ± 14.58 (range 19–76years). Keratometry readings (K) measured in dioptres (D) wereas following: flat meridian (43.12 ± 1.43 D, range 40.3–47.4 D),steep meridian (44.18 ± 1.54 D, range 41.3–49.3 D) and average K(43.65 ± 1.44 D, range 41.1–48.4 D).

3.1. Comparison and correlation of ultrasound pachymetry withPentacam pachymetry

The mean pupil centre thickness (recorded as CCT) measured byUP was 548.10 ± 34.21 �m (range 442–628 �m), which was slightlythinner than mean CCT (also the pupil centre thickness) measuredby Pentacam (mean 550.21 ± 33.83 �m; range 444–631 �m). Therewas no significant difference between CCT readings measured by

UP and Pentacam (P = 0.721). The magnitude of the difference in themeans (mean difference = 2.10, 95% CI: −9.52 to 13.73) was verysmall (eta squared = 0.001). There was high positive correlationbetween the CCT values by US and Pentacam (R = 0.975, P < 0.001;

U. Fares et al. / Contact Lens & Anterior Eye 35 (2012) 39– 45 41

Fn

PiU

atpU

3

3

(ms0

8iew

Table 1Corneal thickness obtained by the Pentacam at the three central points.

Cornealthickness (�m)

Pupil centre Thinnest location Cornea apex

Mean ± SD 550.21 ± 33.83 547.30 ± 34.64 551.22 ± 33.58Range 444–631 432–627 450–632

Table 2Corneal thickness at 3 and 7 mm points from the corneal apex.

Parameters Mean (�m) ± SD Range Increased thicknesscompared to CAT (%)

Temporal3 mm 562.73 ± 35.98 540–633 2.37 mm 639.15 ± 34.59 553–730 16

Nasal3 mm 577.94 ± 31.48 496–645 4.87 mm 676.78 ± 42.62 582–761 22.8

Superior3 mm 581.57 ± 33.41 506–645 5.57 mm 671.22 ± 44.28 571–766 21.8

Inferior3 mm 576.97 ± 35.26 452–633 4.77 mm 664.21 ± 41.80 586–763 20.5

Fb

ig. 2. Bland and Altman plots showing agreement between central corneal thick-ess (CCT) measured by Pentacam and ultrasound pachymetry.

earson’s correlation coefficient). Furthermore, there was high pos-tive correlation between the CAT (547.30 ± 34.64 �m) and CCT byS (R = 0.977, P < 0.001; Pearson’s correlation coefficient) (Fig. 1).

When the difference of the measurements were comparedgainst their means (Bland Altman plots) the Pentacam appearedo overestimate the CCT by 2.1 �m compared with UP. Ninety-fiveercent of the differences between CCT readings of Pentacam andP were between 17.09 �m and −12.89 �m (Fig. 2).

.2. Analysis of Pentacam pachymetry data

.2.1. Three central pointsThere was no significant difference between mean CCT and CTL

P = 0.624), between mean CTL and CAT (P = 0.507) and betweenean CCT and CAT (P = 0.862). The data of CCT, CAT and CTL are

hown in Table 1. The pupil centre was located within a circle of.60 mm diameter from the corneal apex.

The position of the CTL was in the inferotemporal quadrant in

5.7% (57 eyes; 78.79% for right eyes and 91.18% for left eyes) and

n the superotemporal quadrant in 2.99% (2 eyes). In 6 eyes it wasxactly on the temporal horizontal semi-meridian and in 1 eye itas exactly on the inferior vertical semi-meridian (Fig. 3).

ig. 3. Percentage distribution of the quadrant wise location of the thinnest point of thelack dots represent number of eyes where the thinnest point was located exactly on the

CAT = corneal apex thickness.

For right eyes all the CTL points were within a circle of radius0.515 mm from the corneal apex (1.03 mm diameter). For left eyesall the CTL points were within a circle of radius 0.580 mm from thecorneal apex (1.16 mm diameter).

The corneal thickness at 3 mm from the apex of cornea was2.3–5.5% thicker than the CAT. At 7 mm from the apex of cornea, thenasal cornea demonstrated the greatest thickness values, followedby the superior cornea, the inferior cornea and the temporal cornea.These were between 16% and 23% thicker than the CAT (Table 2).

3.2.2. Correlation between CAT and corneal thickness at 3 mmand 7 mm points

There was high positive correlation between the CAT valuesand the thickness at 3 mm points (R ≥ 0.845, P < 0.001) (Table 3and Fig. 4). Furthermore, there was also good positive correla-

tion between the CAT and the thickness at 7 mm points (R ≥ 0.654,P < 0.001) (Table 4 and Fig. 5) but this was comparatively less thanat the 3 mm points.

cornea in right and left eyes (T = temporal, N = nasal, S = superior, I = inferior). The semi-meridians (n = number).

42 U. Fares et al. / Contact Lens & Anterior Eye 35 (2012) 39– 45

Fig. 4. Scatter plots showing linear regression between corneal apex thickness (CAT intemporally and nasally. The plots show good correlation.

Table 3Correlation-coefficient between corneal apex thickness (CAT) and corneal thicknessat 3 mm points from the corneal apex.

3 mm points Pearson correlation (R) Variance (R2) P-Value

Superior 0.840 0.706 <0.001Inferior 0.955 0.912 <0.001

3w

etT

TCa

ment of glaucoma [14] and monitoring certain corneal diseasessuch as keratoconus [15,16] and Fuchs endothelial dystrophy [17].For all of the above indications, reliance is placed on measure-ment of the central corneal thickness by manual placement of the

Temporal 0.853 0.728 <0.001Nasal 0.870 0.757 <0.001

.2.3. Correlation between CTL and corneal thickness at circlesith a diameter of 2, 4, 6, 8 and 10 mm

The mean of corneal thickness values along circles with a diam-

ter of 2 mm, 4 mm, 6 mm, 8 mm and 10 mm, with the centre ofhe circles being the thinnest location are demonstrated in Table 5.here was also a high positive correlation between the CTL values

able 4orrelation-coefficient between corneal apex thickness (CAT) and corneal thicknesst 7 mm points from the corneal apex.

7 mm points Pearson correlation (R) Variance (R2) P-Value

Superior 0.656 0.430 <0.001Inferior 0.683 0.466 <0.001Temporal 0.654 0.428 <0.001Nasal 0.673 0.453 <0.001

�m) and corneal thickness (�m) at the distance of 3 mm superiorly, inferiorly,

and the mean thickness at the circles of 2 mm, 4 mm, 6 mm, 8 mmand 10 mm (R ≥ 0.635, P < 0.001). A gradual increase in thicknesswas noted from the centre to the periphery (Table 5 and Fig. 6).

4. Discussion

Corneal thickness measurements are essential in the work upof patients for laser or incisional refractive surgery [1], assess-

Table 5The mean of corneal thickness values along circles with a diameter of 2, 4, 6, 8 and10 mm and the percentage of increased thickness in relation to the thickness at thecorneal thinnest location (CTL) and to the central corneal thickness (CCT).

Parameter Mean (�m) ± SD Increased thicknesscompared to CTL (%)

Increased thicknesscompared to CCT (%)

2 mm 557.96 ± 31.09 1.9 1.44 mm 585.06 ± 31.67 6.9 6.36 mm 633.22 ± 32.55 15.7 15.18 mm 700.88 ± 39.25 28.1 27.410 mm 784.81 ± 47.73 43.4 42.6

CTL = corneal thickness at thinnest location; CCT = central corneal thickness.

U. Fares et al. / Contact Lens & Anterior Eye 35 (2012) 39– 45 43

F AT int

ursm[otwc

dMmmlocsaCwmca

ig. 5. Scatter plots showing linear regression between corneal apex thickness (Cemporally and nasally.

ltrasound probe. Clinically this has proven to be reasonable andeliable. Relatively newer instruments have enabled us to mea-ure corneal thickness at several points on the cornea giving aore accurate estimate of thickness at any given point of interest

18,19]. We undertook this study to explore the thickness profilef the whole cornea and compare and correlate it with the cen-ral corneal thickness measurements to ascertain whether thereas any theoretical advantage or disadvantage of relying solely on

entral corneal thickness measurements.Recent studies have reported that the repeatability and repro-

ucibility of CCT measurements with Pentacam was high [10,18].iranda et al. concluded that the variability in the PCT measure-ents obtained by Pentacam were clinically acceptable [20]. Theajority of pachymetry studies with Pentacam in the published

iterature have focused only on the CCT and compared it with CCTbtained by UP. Although we too compared CCT between Penta-am and UP our main aim was to correlate CCT with PCT. Sometudies report that UP overestimates CCT compared to Pentacamnd others report the reverse with the Pentacam overestimatingCT when compared to UP. Our CCT values measured by Pentacam

ere slightly but not statistically significantly higher than thoseeasured by UP. The underestimation of thickness with ultrasound

ould be related to the flattening of the cornea caused by UP probend the displacement of the tear-film [21,22]. Our results agreed

�m) and corneal thickness (�m) at the distance of 7 mm superiorly, inferiorly,

with Ucakhan et al. [16] who reported Pentacam CCT values 2.7 �mgreater than UP, in healthy corneas. Lackner et al. reported that CCTvalues using Pentacam were 2.4% thinner than UP however theytook the thinnest corneal location (CTL) as CCT [18]. O’Donnell andMaldonado-Codina also reported lower measurements for the CCTwith Pentacam than with UP by 0.01%, but they did not specify thepoint of the cornea they took to represent CCT [10]. This variationin comparing UP CCT with Pentacam CCT, CTL or CAT could explainwhy some observers recorded lower values with Pentacam com-pared to UP. It has also been suggested that local anaesthesia usedfor UP measurements could cause corneal swelling [22].

Bland and Altman reported that the best way to analyse theagreement between two measurements is to plot the differencebetween those measurements against their mean [23]. The meandifference between CCT by Pentacam and UP in our study wasnot significant (2.1 �m) and agrees with previous reported studies[10,18,20] indicating that variations in measurement with thesetwo modalities is not clinically relevant, however, when record-ing thickness and planning surgery which may involve workingclose to the margins of safety, it may be prudent to use both meth-

ods if available. However, the close concordance of both methodssuggests that either one of them can be used with good accuracy.To date, UP remains the most common technique for measuringCCT but its main disadvantage is not having a fixation target. Thus,

44 U. Fares et al. / Contact Lens & Anterior Eye 35 (2012) 39– 45

Fig. 6. (A)–(E) Scatter plots showing linear regression between corneal thickness at thinnest location (CTL in �m) and corneal thickness (�m) along circles with a diametero thickni

aPbatd

f 2 mm, 4 mm, 6 mm, 8 mm and 10 mm. (F) Graph showing progressive increase ins depicted).

pplication of its probe may vary in repeated measurements [24].entacam measures corneal thickness as a function of the difference

etween the anterior and the posterior corneal surfaces and has thedvantage of overlapping up to 50 Scheimpflug images centred onhe cornea. Moreover it also provides simultaneous topographicalata [25].

ess from CTL towards the periphery (mean thickness along circle or each diameter

The Pentacam determines three different central points on thecornea namely the CCT, CAT and CTL. However, in practical terms

all three points are located within a circle of 1.2 mm and 0.60 mmdiameter for CTL and CCT respectively using the corneal apex asthe centre. Differences in the thickness measurements at thesepoints were not statistically significant. Hence, though important as

& An

rn

tCtwclchppnCef

1iiw

mr[irqZtei[

cIstbmgssts

C

e

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[29] Klein SR, Epstein RJ, Randleman JB, Stulting RD. Corneal ectasia after laser in situ

U. Fares et al. / Contact Lens

eference points for research and analysis, their location and thick-ess is not clinically relevant in planning surgery.

The distance of the peripheral thickness points as depicted onhe thickness map generated by the Pentacam are related to theAT. Similarly, in this study the peripheral thickness measured athe 3 mm and 7 mm distances in each of the four semi-meridiansere from the CAT. Our Pentacam measurements at 3 mm from

orneal apex showed the superior cornea to be the thickest, fol-owed by the nasal cornea, the inferior cornea and the temporalornea. This was agreeable with Khoramnia et al. and Zheng et al.,owever, their measurements were obtained at 3 mm from theupil centre (not corneal apex) [26,27]. Interestingly at the 7 mmoints, the nasal cornea was the thickest. At least 71% of the thick-ess at 3 mm from the corneal apex could be predicted by knowingAT values. This percentage decreased to 43% at 7 mm points. Zhengt al. too found a high correlation of CAT with the PCT at 3 mm pointsrom the pupil [27].

On the other hand, the average thickness along the 2, 4, 6, 8 and0 mm circles were measured from the CTL. There was a gradual

ncrease in thickness from the CTL to the periphery, which occurredn a fairly predictable manner. Each peripheral point correlated well

ith the CTL and with the adjacent peripheral point (Fig. 6).We identified that the thinnest point of the cornea (CTL) as

easured from the corneal apex was located predominantly infe-otemporal. This agreed with results reported by Khoramnia et al.26], who found that the thinnest location was 92% and 8% in thenferotemporal and superotemporal respectively. Ashwin et al. tooeported the thinnest point to be located in the inferotemporaluadrant in more than 96% in right eyes and 81% in left eyes [28].heng et al. also noted that the thinnest point located in the infero-emporal quadrant in about 78% of tested eyes [27]. This fact couldxplain the positioning of post-LASIK ectasia inferotemporally ornferiorly [29] and the location of the cone apex in keratoconus30].

The experience gathered over the years in planning and exe-uting refractive surgery based on CCT measurements is vast.ntuitively one would believe that this is an efficient and safe mea-urement for the purpose. The data generated in this study provideshe evidence in support of that belief. The good correlation observedetween central and peripheral measurements suggests that forost purposes the central corneal thickness can serve as a good

uide for predicting peripheral thickness. For surgical procedurespecifically undertaken at mid-peripheral and peripheral zones,uch as insertion of intracorneal rings and limbal relaxing incisions,he actual measurements at or along the site of surgery may conferome advantage.

onflict of interest

None of the authors have any financial or other conflict of inter-st.

eferences

[1] Javaloy J, Vidal MT, Villada JR, Artola A, Alio JL. Comparison of fourcorneal pachymetry techniques in corneal refractive surgery. J Refract Surg

2004;20:29–34.

[2] Argus WA. Ocular hypertension and central corneal thickness. Ophthalmology1995;102:1810–2.

[3] Gordon MO, Beiser JA, Brandt JD, Heuer DK, Higginbotham EJ, Johnson CA, et al.The ocular hypertension treatment study: baseline factors that predict the

[

terior Eye 35 (2012) 39– 45 45

onset of primary open-angle glaucoma. Arch Ophthalmol 2002;120:714–20[discussion 829–30].

[4] Wilkins MR, Mehta JS, Larkin DF. Standardized arcuate keratotomy for postk-eratoplasty astigmatism. J Cataract Refract Surg 2005;31:297–301.

[5] Muller-Jensen K, Fischer P, Tan M. [Para-limbic relaxing incisions for reductionof astigmatism within the scope of catarct surgery]. Klin Monbl Augenheilkd2000;217:257–62.

[6] Pak KH, Kim JH. Radial keratotomy for the purpose of reducing glasses powerin high myopia. Korean J Ophthalmol 1992;6:83–90.

[7] Marsich MW, Bullimore MA. The repeatability of corneal thickness measures.Cornea 2000;19:792–5.

[8] Miglior S, Albe E, Guareschi M, Mandelli G, Gomarasca S, Orzalesi N. Intraob-server and interobserver reproducibility in the evaluation of ultrasonicpachymetry measurements of central corneal thickness. Br J Ophthalmol2004;88:174–7.

[9] Konstantopoulos A, Hossain P, Anderson DF. Recent advances in ophthalmicanterior segment imaging: a new era for ophthalmic diagnosis? Br J Ophthalmol2007;91:551–7.

10] O’Donnell C, Maldonado-Codina C. Agreement and repeatability of centralthickness measurement in normal corneas using ultrasound pachymetry andthe OCULUS Pentacam. Cornea 2005;24:920–4.

11] Read SA, Collins MJ. Diurnal variation of corneal shape and thickness. OptomVis Sci 2009;86:170–80.

12] Harper CL, Boulton ME, Bennett D, Marcyniuk B, Jarvis-Evans JH, Tullo AB. Diur-nal variations in human corneal thickness. Br J Ophthalmol 1996;80:1068–72.

13] Pallant J. Statistical techniques to compare groups, in SPSS, Survival Manual.Open University Press; 2008. pp. 232–42.

14] Doughty MJ, Zaman ML. Human corneal thickness and its impact on intraocularpressure measures: a review and meta-analysis approach. Surv Ophthalmol2000;44:367–408.

15] Pflugfelder SC, Liu Z, Feuer W, Verm A. Corneal thickness indices discriminatebetween keratoconus and contact lens-induced corneal thinning. Ophthalmol-ogy 2002;109:2336–41.

16] Ucakhan OO, Ozkan M, Kanpolat A. Corneal thickness measurements in normaland keratoconic eyes: Pentacam comprehensive eye scanner versus noncon-tact specular microscopy and ultrasound pachymetry. J Cataract Refract Surg2006;32:970–7.

17] Borboli S, Colby K. Mechanisms of disease: Fuchs’ endothelial dystrophy. Oph-thalmol Clin North Am 2002;15:17–25.

18] Lackner B, Schmidinger G, Pieh S, Funovics MA, Skorpik C. Repeatability andreproducibility of central corneal thickness measurement with Pentacam, Orb-scan, and ultrasound. Optom Vis Sci 2005;82:892–9.

19] Prospero Ponce CM, Rocha KM, Smith SD, Krueger RR. Central and peripheralcorneal thickness measured with optical coherence tomography, Scheimpflugimaging, and ultrasound pachymetry in normal, keratoconus-suspect, andpost-laser in situ keratomileusis eyes. J Cataract Refract Surg 2009;35:1055–62.

20] Miranda MA, Radhakrishnan H, O’Donnell C. Repeatability of corneal thicknessmeasured using an Oculus Pentacam. Optom Vis Sci 2009;86:266–72.

21] Ho T, Cheng AC, Rao SK, Lau S, Leung CK, Lam DS. Central corneal thick-ness measurements using Orbscan II, Visante, ultrasound, and Pentacampachymetry after laser in situ keratomileusis for myopia. J Cataract RefractSurg 2007;33:1177–82.

22] Li EY, Mohamed S, Leung CK, Rao SK, Cheng AC, Cheung CY, et al. Agree-ment among 3 methods to measure corneal thickness: ultrasound pachymetry,Orbscan II, and Visante anterior segment optical coherence tomography. Oph-thalmology 2007;114:1842–7.

23] Bland JM, Altman DG. Statistical methods for assessing agreement between twomethods of clinical measurement. Lancet 1986;1:307–10.

24] Gordon A, Boggess EA, Molinari JF. Variability of ultrasonic pachometry. OptomVis Sci 1990;67:162–5.

25] Ambrosio Jr R, Alonso RS, Luz A, Coca Velarde LG. Corneal-thickness spatialprofile and corneal-volume distribution: tomographic indices to detect kera-toconus. J Cataract Refract Surg 2006;32:1851–9.

26] Khoramnia R, Rabsilber TM, Auffarth GU. Central and peripheral pachymetrymeasurements according to age using the Pentacam rotating Scheimpflug cam-era. J Cataract Refract Surg 2007;33:830–6.

27] Zheng Y, Huang G, Huang W, He M. Distribution of central and peripheralcorneal thickness in Chinese children and adults: the Guangzhou twin eyestudy. Cornea 2008;27:776–81.

28] Ashwin PT, Shah S, Pushpoth S, Wehbeh L, Ilango B. The relationship of centralcorneal thickness (CCT) to thinnest central cornea (TCC) in healthy adults. ContLens Anterior Eye 2009;32:64–7.

keratomileusis in patients without apparent preoperative risk factors. Cornea2006;25:388–403.

30] Demirbas NH, Pflugfelder SC. Topographic pattern and apex location of kerato-conus on elevation topography maps. Cornea 1998;17:476–84.