Effect of posture and artificial tears on corneal power measurements with a handheld automated keratometer

Effect of posture and artificial tears on corneal power measurements with a handheld automated keratometer Andrew K.C. Lam, PhD, FAAO, Rufina Chan, MOptom, FAAO, Roger Chiu, HCert(Optom) Purpose: To study the effect of posture and artificial tears on handheld automated keratometry. Setting: Department of Optometry and Radiography, Hong Kong Polytechnic University, Hong Kong, China. Methods: Thirty-five subjects were recruited to have corneal curvature measurement in 1 eye (randomly selected) by a Medmont topographic keratometer and a Nidek handheld keratometer. In handheld keratometry, the measurements were taken with the subject in a sitting and lying posture (both with and without the use of artificial tears). The sequence of measurements was randomly assigned, but the application of artificial tears was always the last. The steepest and flattest corneal curvatures were compared between the 4 conditions. The corneal power was converted to orthogonal power vector components and rectangular Fourier form (M, J0, J45) for another comparison. Results: There was a significant difference in the steepest and flattest meridians between the 4 conditions (P⬍.01). However, the mean difference between the handheld keratometer and the topographic keratometer was less than 0.50 diopter, and the intraclass correlation coefficient (ICC) was very high (0.96), indicating good clinical reliability. When analyzed in rectangular Fourier form, the difference was also significant but the ICCs were lower (0.97, 0.89, and 0.64 for M, J0, and J45, respectively). The greatest difference was when subjects were in the lying posture and had received artificial tears. Conclusions: Handheld keratometry provided different results from topographic keratometry. The difference was greatest with the use of artificial tears. Cataract surgeons should take this into consideration, especially when using the handheld keratometry in the operating theater in patients under general anesthesia. Results show that the power vector method is best for studying corneal shape. J Cataract Refract Surg 2004; 30:645–652  2004 ASCRS and ESCRS

C

orneal refractive power and axial length are 2 important parameters in calculating intraocular lens (IOL) power during cataract surgery.1 With the popular-

Accepted for publication May 3, 2003. From the Department of Optometry and Radiography, Hong Kong Polytechnic University, Hong Kong, China. None of the authors has a financial or proprietary interest in any material or method mentioned. Reprint requests to Dr. Andrew Lam, Department of Optometry and Radiography, Hong Kong Polytechnic University, Hong Kong, China.  2004 ASCRS and ESCRS Published by Elsevier Inc.

ity of refractive surgery, the original corneal power may not be known to the cataract surgeon; hence, error in measuring the true corneal power affects the determination of IOL power.2–4 Corneal power can be measured by a standard keratometer or by the more sophisticated topographic keratometer. Standard keratometry estimates the central corneal curvature using 4 points along a thin annulus of the cornea but can still provide accurate IOL power calculations.5 Handheld automated keratometers provide flexibility in measuring the cornea with the subject in different 0886-3350/04/$–see front matter doi:10.1016/S0886-3350(03)00554-6

EFFECT OF POSTURE AND ARTIFICIAL TEARS ON CORNEAL POWER

Table 1. The mean corneal power (⫾SD) in rectangular Fourier form (M, J0, J45) with 4 measuring conditions. Medmont Topographic Keratometer

Parameter M (D) J0 (D) J45 (D)

Nidek Handheld Keratometer Sitting

Lying and with Tears

Lying 42.95 ⫾ 1.31

Significance*

ICC

42.96 ⫾ 1.34

F ⫽ 8.91, P⬍.01

0.97

F ⫽ 8.88, P⬍.01

0.89

F ⫽ 2.73, P ⫽ .048

0.64

43.21 ⫾ 1.34

42.97 ⫾ 1.29

⫺0.64 ⫾ 0.44

⫺0.67 ⫾ 0.45

⫺0.71 ⫾ 0.44

⫺0.81 ⫾ 0.47

⫺4.1 ⫻ 10⫺5 ⫾ 0.28

⫺0.10 ⫾ 0.22

⫺0.044 ⫾ 0.27

⫺0.088 ⫾ 0.32

ICC ⫽ intraclass correlation coefficient *Repeated-measures ANOVA

postures. Previous investigation of handheld automated keratometers suggest that measurements should be made with the instrument held vertically (6 o’clock/ 12 o’clock) for accurate axis determination.6,7 Whether manual keratometry or handheld automated keratometry provides better repeatability in calculating IOL power is controversial.8,9 However, when patients with congenital cataract or mental retardation require cataract surgery, using a manual keratometer may not be possible and corneal measurement must be done with a handheld automated keratometer in the operating theater with the patient under general anesthesia.10 In this case, the handheld automated keratometer cannot be operated along a vertical plane while the patient is in the sitting posture. The axis compensator, a feature of handheld keratometers designed to compensate for the tilting effect,6 may not work properly and may provide inaccurate keratometric results. This study compared corneal power measurements with a handheld automated keratometer, held vertically and facing downward, with those with a topographic keratometer. The effect of artificial tears on corneal measurement was also studied.

Subjects and Methods Thirty-five subjects (21 men) were recruited from the Optometry Clinic, Hong Kong Polytechnic University. Their

mean age was 20.8 years (range 20 to 25 years). All subjects had a general eye examination at the Optometry Clinic and were free of corneal pathology. The Human Subjects Ethics Sub-committee of Hong Kong Polytechnic University approved the study, and all subjects gave informed consent before data collection began. One eye of each subject was randomly chosen for measurement. The sequence of measurements was randomly assigned, but the application of artificial tears was always the last. Corneal power was measured with a Medmont E300 topographic keratometer. This keratometer took repeated captures and provided a rating of the accuracy of focus and centration. Each subject had 2 readings taken by 1 practitioner, and only measurements with a rating greater than 90 were used.11 The simulated K-reading was used for analysis. Another practitioner did measurements with a Nidek KM500 handheld keratometer. Two consecutive readings were taken, and the mean was used for analysis. The measurements were taken with the subject in the sitting posture, and the keratometer was held vertically. The subject was also asked to lie on a bench; the handheld keratometer faced downward, and another 2 readings were taken. Keratometric measurements were also taken with the patient in a lying posture with the application of 1 drop of artificial tears (Opti-Tears威). This was to simulate the situation in the operating theater in which the practitioner uses artificial tears before keratometry. Measurements were obtained right after the artificial tears were instilled. All the practitioners were masked to the keratometric results, and all the measurements were taken within 15 minutes. Measurements using the handheld keratometer were aligned with the vertical axis of the eye without the instrument tilted. The practitioner then aligned the mire ring to the subject’s cornea as described in the instruction manual.

Table 2. The mean corneal power (⫾SD) in the steepest and flattest meridians with 4 measuring conditions. Nidek Handheld Keratometer

Meridian

Medmont Topographic Keratometer

Sitting

Lying

Lying and with Tears

Significance*

ICC

Steepest (D)

43.90 ⫾ 1.52

43.67 ⫾ 1.47

43.70 ⫾ 1.49

43.83 ⫾ 1.54

F ⫽ 4.41, P⬍.01

0.96

Flattest (D)

42.51 ⫾ 1.30

42.26 ⫾ 1.26

42.20 ⫾ 1.28

42.09 ⫾ 1.29

F ⫽ 16.2, P⬍.01

0.96

ICC ⫽ intraclass correlation coefficient *Repeated-measures ANOVA

646

J CATARACT REFRACT SURG—VOL 30, MARCH 2004

EFFECT OF POSTURE AND ARTIFICIAL TEARS ON CORNEAL POWER

Table 3. The mean difference (95% limits of agreement; ie, 1.96 ⫻ standard deviation of the difference) in rectangular Fourier form (M, J0, J45) for the steepest and flattest meridians.

95% limits of agreement (1.96 ⫻ the standard deviation of the difference) were determined as suggested by Bland and Altman.14

Medmont Versus Nidek Sitting

Lying

Lying and with Tears

M (D)

0.24 (⫾ 0.78)

0.26 (⫾0.85)

0.25 (⫾0.98)

J0 (D)

0.03 (⫾0.35)

0.07 (⫾0.35)

0.17 (⫾0.61)

J45 (D)

0.10 (⫾0.38)

0.04 (⫾0.36)

0.09 (⫾0.63)

Steepest (D)

0.23 (⫾0.80)

0.20 (⫾0.83)

0.07 (⫾1.25)

Flattest (D)

0.25 (⫾0.85)

0.32 (⫾0.97)

0.42 (⫾1.01)

Parameter

There were 4 sets of measurements: 1 with the topographic keratometer and 3 with the handheld keratometer. The 2 readings from each set were averaged, and the mean was used for analysis. In averaging the 2 readings, the corneal power was analyzed as though it were a spherocylindrical lens using orthogonal power vector methods developed by Thibos and coauthors.12 The corneal power of the flattest meridian was used as the spherical component, the toricity was designated as the positive cylinder, and the axis was specified as the axis of the flattest meridian; that is, Corneal power ⫽ Fs ⫹ Fc ⫻ ␤ where Fs is the power of flattest meridian, Fc is the toricity, and ␤ is the axis of the flattest meridian. Formulas used to convert from a positive cylinder to a polar Fourier description (mean spherical equivalent) were M ⫽ Fs ⫹ Fc/2 ␣ ⫽ ␤ ⫺ 90 The conversion of polar Fourier description to rectangular Fourier form (M, J0, J45) used the following equations: Jackson cross-cylinder at axis 0⬚ with power, J0 ⫽ (Fc/2)(cos 2␣)

Results The mean refractive error of the subjects was ⫺2.41 diopters (D) ⫾ 2.29 (SD) in the spherical component and ⫺0.67 ⫾ 0.90 D in the cylindrical component, where the mean spherical equivalent was ⫺2.74 ⫾ 2.44 D. By considering the topographic results as the standard, the mean corneal astigmatism was 1.38 ⫾ 0.88 D. Table 1 shows the corneal power results in rectangular Fourier form under 4 measuring conditions. A repeated-measures ANOVA demonstrated a significant difference for each parameter under 4 measuring conditions (Table 1). Student-Newman-Keuls post-hoc tests revealed that the Medmont results contributed to a significant difference in Medmont and the Nidek results with artificial tears contributed to the significant difference in J0. When the corneal power was converted back to steepest and flattest curvatures from the rectangular Fourier form, the differences were still significant (Table 2). The ICC results were nearly 1.00. (The closer the ICC value is to 1.00, the stronger the reliability.) The mean difference in curvature between the Medmont and Nidek models ranged from 0.07 to 0.42 D (Table 3). They were all less than 0.50 D. The difference obtained with the Medmont topographic keratometer was plotted against that from the Nidek keratometer. Figures 1 to 3 show the mean differences with the 95% limits of agreement for parameters M, J0, and J45 respectively.

Jackson cross-cylinder at axis 45⬚ with power, J45 ⫽ (Fc/2)(sin 2␣)

Discussion

The corneal power was analyzed in rectangular Fourier form (M, J0, J45). The results were first tested for normality (P⬎.05, Kolmogorov-Smirnov test), and parametric tests were used in the analysis. Repeated-measures analysis of variance (ANOVA) was used to compare results between the topographic keratometer and the handheld keratometer under 3 modes of measurements. The intraclass correlation coefficient (ICC), a reliability coefficient calculated from variance estimates obtained through an analysis of variance,13 was also calculated for each parameter by considering the topographic results to be the standard. Plots of the difference between the topographic keratometer and handheld keratometer against their means and the

Many studies report that handheld automated keratometers have good reliability in measuring corneal curvature.6–8,15,16 We compared a handheld keratometer and topographic keratometer. We used the Medmont model because it was found to be the most repeatable among other topographic keratometers.11 The main source of error in handheld keratometry derives from whether the keratometer can be held vertically.6–8 However, the main benefit of a handheld device is its portability; thus, it can be used in different practices and with the subject in different postures. Edwards

J CATARACT REFRACT SURG—VOL 30, MARCH 2004

647

EFFECT OF POSTURE AND ARTIFICIAL TEARS ON CORNEAL POWER

Figure 1. (Lam) The mean and difference in parameter M between the Medmont and Nidek with the subject in sitting posture (A ), lying posture (B ), and lying posture with artificial tears (C ). The solid line indicates the mean of difference. The dotted lines indicate the 95% limits of agreement.

and Cho15 report good repeatability with the Nidek handheld keratometer with the subject in a sitting posture. When considering the steepest and flattest meridians, we found a mean difference between the Medmont and Nidek (sitting) results of just 0.25 D (Table 3). Thus, the 2 instruments can be considered clinically comparable. This was reflected in the ICC results. The 95% limits of agreement were similar for the steepest and flattest meridians. The performance of the Nidek keratometer deteriorated when the patient was in a 648

lying posture (95% limits of agreement ⫾0.83 D and ⫾0.97 D for steepest and flattest meridians, respectively). Its performance was even worse with the application of artificial tears with the subject in a lying posture (95% limits of agreement ⫾1.25 D and ⫾1.01 D for steepest and flattest meridians, respectively). Although there was a significant difference in the steepest and flattest curvatures (Table 2), the ICC results were very good. Portney and Watkins13 suggest that, in general, ICC values above 0.75 indicate good reliability but that

J CATARACT REFRACT SURG—VOL 30, MARCH 2004

EFFECT OF POSTURE AND ARTIFICIAL TEARS ON CORNEAL POWER

Figure 2. (Lam) The mean and difference in parameter J0 between the Medmont and Nidek with the subject in sitting posture (A ), lying posture (B ), and lying posture with artificial tears (C ). The solid line indicates the mean of difference. The dotted lines indicate the 95% limits of agreement.

for most clinical measurements, the ICC should be at least 0.9 to ensure reasonable validity. When considering just the steepest and flattest meridians in measuring corneal curvature, the influence of the axis is disregarded. The power vector method of Thibos and coauthors12 overcomes this. This method considers a spherocylinder lens as the sum of a spherical lens and 2 cross-cylinders, 1 at axis 0 degrees and the other at axis 45 degrees. These 3 parameters (M, J0, and J45) can be interpreted as coordinates (x, y, z) of a

vector representation of the power profile.12 It makes problems involving lens combinations or comparisons easy. The vector components were significantly different between the 4 measuring conditions (Table 1). The ICC results also indicated doubtful validity with the Nidek keratometer (ICC ⫽ 0.64 in the J45 vector component). When the Nidek keratometer was held vertically, the difference compared with the Medmont keratometer was small (95% limits of agreement

J CATARACT REFRACT SURG—VOL 30, MARCH 2004

649

EFFECT OF POSTURE AND ARTIFICIAL TEARS ON CORNEAL POWER

Figure 3. (Lam) The mean and difference in parameter J45 between the Medmont and Nidek with the subject in sitting posture (A ), lying posture (B ), and lying posture with artificial tears (C ). The solid line indicates the mean of difference. The dotted lines indicate the 95% limits of agreement.

⫾0.78 D, ⫾0.35 D, and ⫾0.38 D for M, J0, and J45, respectively). Its performance was poor with the application of artificial tears with the subject in a lying posture (95% limits of agreement ⫾0.98D, ⫾0.61D, and ⫾0.63D for M, J0, and J45, respectively). Previous studies report the effect of artificial tears on corneal topography.17–19 The effect is greater with high-viscosity preparations18,19 and is time dependent.19 These studies measured the cornea with the subject in a sitting posture, in which the tear film thickness may 650

be more affected by gravity. We applied artificial tears with the subjects in a lying posture to spread the tears more evenly on the corneal surface. With the handheld keratometer, which presented findings just along the principal meridians, without the use of other indicators such as surface irregularity and surface asymmetry,17 the effect from artificial tears on cornea was demonstrated already. As the Nidek keratometer is portable, it can be used in different settings such as in the operating theater. It is not uncommon to use artificial tears for

J CATARACT REFRACT SURG—VOL 30, MARCH 2004

EFFECT OF POSTURE AND ARTIFICIAL TEARS ON CORNEAL POWER

keratometry in the operating theater, especially in patients under general anesthesia. Therefore, cataract surgeons should be cautious when using a handheld keratometer for preoperative keratometry with the patient in a lying posture and has had an application of artificial tears. Although 2 investigators were assigned for individual instruments, interobserver variability should be minimal as previous studies demonstrate good interobserver repeatability from Nidek keratometer7 and Medmont topographer.11 One reason for the poor reliability with the handheld keratometer could be the alignment method used in our study. We did not ask the subjects to look at the internal fixation light inside the handheld keratometer, which was different from using the topographic keratometer. The investigator tried her best to align the mire ring to the subject’s cornea as described in the instruction manual. We adopted this alignment method to simulate the situation of general anesthesia. In a real situation, the practitioner would align the handheld keratometer with the patient’s cornea instead of with the patient fixating on the internal target. This alignment method could lead to disagreement with the topographic keratometer. The use of a handheld keratometer with the subject in a lying posture (with and without the use of artificial tears) was done with alignment to the vertical axis of the eye. Tilting the instrument could introduce great variation.6 Cataract surgeons should take this into consideration as well. In the current study, considering only the steepest and flattest meridians (with good ICC results) did not reveal whether the Nidek keratometer has questionable validity. The conversion of corneal shape from the steepest and flattest curvatures to the rectangular Fourier form is therefore a better method studying corneal shape. Also, the use of a handheld keratometer with the subject in a lying posture with artificial tears will further influence the reliability corneal curvature measurement. However, for cataract surgeons interested in the accuracy of spherical equivalent from a handheld keratometer, its deviation from the topographer could be minimal and no cause for concern.

References 1. Holladay JT. Standardizing constants for ultrasonic biometry, keratometry, and intraocular lens power calculations. J Cataract Refract Surg 1997; 23:1356–1370

2. Seitz B, Langenbucher A. Intraocular lens power calculation in eyes after corneal refractive surgery. J Refract Surg 2000; 16:349–361 3. Feiz V, Mannis MJ, Garcia-Ferrer F, et al. Intraocular lens power calculation after laser in situ keratomileusis for myopia and hyperopia; a standardized approach. Cornea 2001; 20:792–797 4. Odenthal MTP, Eggink CA, Melles G, et al. Clinical and theoretical results of intraocular lens power calculation for cataract surgery after photorefractive keratectomy for myopia. Arch Ophthalmol 2002; 120:431–438 5. Husain SE, Kohnen T, Maturi R, et al. Computerized videokeratography and keratometry in determining intraocular lens calculations. J Cataract Refract Surg 1996; 22:362–366 6. Lam AKC. A hand-held keratometer. Ophthalmic Physiol Opt 1995; 15:227–230 7. Noonan CP, Rao GP, Kaye SB, et al. Validation of a handheld automated keratometer in adults. J Cataract Refract Surg 1998; 24:411–414 8. Leyland M, Benjamin L. Clinical assessment of a handheld automated keratometer in cataract surgery. Eye 1997; 11:854–857 9. Manning CA, Kloess PM. Comparison of portable automated keratometry and manual keratometry for IOL calculation. J Cataract Refract Surg 1997; 23:1213–1216 10. Mittelviefhaus H, Gentner C. Messungenauigkeiten der keratometrie bei der Intraokularlinsenberechnung fu¨r Sa¨uglinge. Ophthalmologe 2000; 97:186–188 11. Cho P, Lam AKC, Mountford J, Ng L. The performance of four different corneal topographers on normal human corneas and its impact on orthokeratology lens fitting. Optom Vis Sci 2002; 79:175–183; errata, 462 12. Thibos LN, Wheeler W, Horner D. Power vectors: an application of Fourier analysis to the description and statistical analysis of refractive error. Optom Vis Sci 1997; 74:367–375 13. Portney LG, Watkins MP. Foundations of Clinical Research; Applications to Practice. Norwalk, CT, Appleton & Lange, 1993; 505–528 14. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1:307–310 15. Edwards MH, Cho P. A new, hand-held keratometer: comparison of the Nidek KM-500 auto keratometer with the B&L keratometer and the Topcon RK-3000A keratometer. J Br Contact Lens Assoc 1996; 19:45–48 16. Shin JA, Manny RE, Kleinstein RN, et al. Short-term repeatability of hand-held keratometry measurements. Optom Vis Sci 1999; 76:247–253 17. Pavlopoulos GP, Horn J, Feldman ST. The effect of artificial tears on computer-assisted corneal topography in normal eyes and after penetrating keratoplasty. Am J Ophthalmol 1995; 119:712–722

J CATARACT REFRACT SURG—VOL 30, MARCH 2004

651

EFFECT OF POSTURE AND ARTIFICIAL TEARS ON CORNEAL POWER

18. Emde CWD, Brewitt H, Winter R. Vera¨nderungen der Hornhaut-Topographien nach Anwendung von Tra¨enersatzmitteln. Klin Monatsbl Augenheilkd 1996; 208: 431–437

652

19. Novak KD, Kohnen T, Chang-Godinich A, et al. Changes in computerized videokeratography induced by artificial tears. J Cataract Refract Surg 1997; 23:1023– 1028

J CATARACT REFRACT SURG—VOL 30, MARCH 2004