Relation between corneal thickness and intraocular pressure measurement by noncontact and applanation tonometry

Relation between corneal thickness and intraocular pressure measurement by noncontact and applanation tonometry ¨ mer F. Recep, MD, Hikmet Hasıripi, MD, Nurullah C O ¸ ag˘ıl, MD, Hikmet Sarıkatipog˘lu, MD ABSTRACT Purpose: To assess whether there is a significant difference in intraocular pressure (IOP) measurements between noncontact and applanation methods and if so, whether the difference is correlated with corneal thickness. Setting: Is¸ık Eye Clinic, Ankara, Turkey. Methods: This prospective study comprised 120 eyes of 60 randomly selected patients. Noncontact tonometry, applanation tonometry, and pachymetry were performed in each eye. The results were compared and an analysis was performed to determine whether there was a correlation between the differences in IOP measurements and corneal thickness in the entire group and in subgroups composed of the halves and quarters of the group. Results: The mean IOP in the entire group was 20.38 mm Hg ⫾ 4.97 (SD) with the noncontact tonometer and 18.84 ⫾ 4.47 mm Hg with the Goldmann applanation tonometer. The difference between the measurements with the noncontact method and those with the applanation method was statistically significant (P ⬍ .01) except in cases having a corneal thickness between 513 ␮m and 539 ␮m (P ⬎ .01). There was a positive correlation between corneal thickness and the difference in measurements by noncontact and applanation tonometry, and the magnitude of correlation was greater in cases with thicker corneas. Conclusion: The reliability of tonometers decreased with increasing corneal thickness, in which case higher readings were found, especially with the noncontact tonometer. J Cataract Refract Surg 2001; 27:1787–1791 © 2001 ASCRS and ESCRS

T

onometry remains the best method to diagnose and follow cases of glaucoma. Several tonometers were introduced to measure intraocular pressure (IOP) without significantly disturbing the patient. Today, the most commonly used methods are noncontact and applanation tonometry. Accepted for publication February 28, 2001. ¨ mer F. Recep, MD, Tepebas¸ı Mh. Foc¸a Sk. 70/8, Reprint requests to O Kec¸io¨ren, Ankara, Turkey. © 2001 ASCRS and ESCRS Published by Elsevier Science Inc.

Most studies comparing Goldmann applanation and noncontact tonometry found the latter to be reliable within the normal IOP range; however, reliability decreases at the higher pressure ranges and is limited by an abnormal cornea or poor fixation.1 There are several sources of error in tonometry, and the factors influencing the results of Goldmann applanation tonometry also affect those of noncontact tonometry. There may be other sources of error with both methods. 0886-3350/01/$–see front matter PII S0886-3350(01)00900-2

RELATION OF CORNEAL THICKNESS TO IOP

Corneal variables are 1 source of error in tonometry. In this study, we measured IOP taken with a noncontact and a Goldmann applanation tonometer and compared the results statistically. We also looked for a correlation between corneal thickness and the difference in IOP measurements between the noncontact and applanation techniques.

Patients and Methods One hundred twenty eyes of 60 patients were included in this prospective study. The patients were randomly selected from our clinical practice regardless of the IOP range. The mean age of the 28 women (46.67%) and 32 men (53.33%) was 51.12 years ⫾ 17.64 (SD) (range 14 to 76 years). The inclusion criteria were the absence of corneal disorders, previously diagnosed glaucoma, and current ocular drug use. The examination was performed in the following order: autorefractometry, noncontact tonometry (Topcon CT 60 computerized tonometer), refraction, biomicroscopy, applanation tonometry, pachymetry, and ophthalmoscopy. Noncontact and applanation tonometry were performed approximately 10 to 15 minutes apart, which is a safe interval for repeated tonometry.2 All IOP measurements were performed by the same ¨ .F.R.). The IOP was measured consecuphysician (O tively 3 times with the noncontact tonometer without any interval, and the mean value was used for statistical analysis. One measurement was taken with the applanation method. With each method, both eyes were measured, 1 immediately after the other. Corneal thickness was measured 10 times with an ultrasonic pachymeter (Teknar, Ophthasonic A-scan/Pachymeter III), and the mean value was calculated. Descriptive statistics, a paired t test, and linear regression analysis were used for statistical analysis. Both eyes of each patient were used for statistical analysis. A P value less than 0.01 was considered significant.

(Figure 1). Fifty-four noncontact readings and 32 applanation readings were greater than 21 mm Hg. The difference between methods was statistically significant (P ⬍ .01). The mean corneal thickness in all patients was 546.14 ⫾ 44.33 ␮m. The group was first divided into 2 subgroups and then 4 subgroups according to the median corneal thickness. Table 1 shows the IOP measurements with noncontact and applanation methods and the statistical significance between measurements in each subgroup. The difference in IOP measurement between methods was statistically significant (P ⬍ .01) except in cases with a corneal thickness between 513 ␮m and 539 ␮m (P ⬎ .01). The difference in readings between methods was 3 mm Hg or more in 70 eyes (58.33%). Linear regression analysis of the relation of corneal thickness and differences in IOP measurements showed a significant correlation overall (P ⬍ .01). The correlation in the subgroup with a corneal thickness greater than 539.0 ␮m and in the subgroup with a corneal thickness greater than 574.5 ␮m was also statistically significant (P ⬍ .01). There were no statistically significant differences in the other subgroups (P ⬎ .01) (Table 1). Table 2 shows the distribution of IOP differences in the entire group and the subgroups.

Results In the entire group, the mean IOP was 20.38 ⫾ 4.97 mm Hg (range 11 to 37 mm Hg) with the noncontact tonometer and 18.84 ⫾ 4.47 mm Hg (range 11 to 43 mm Hg) with the Goldmann applanation tonometer 1788

Figure 1. (Recep) Intraocular pressure measurements by noncontact (NT) and Goldmann applanation (AT) tonometry.

J CATARACT REFRACT SURG—VOL 27, NOVEMBER 2001

RELATION OF CORNEAL THICKNESS TO IOP

Table 1. Statistical analysis of data. Mean IOP ⴞ SD Corneal Thickness (␮m)

Noncontact Tonometer (mm Hg)

Goldmann Tonometer (mm Hg)

P Value*

Correlation Coefficient

Significance of Correlation

Entire group

20.38 ⫾ 4.97

18.84 ⫾ 4.47

⬍.01

0.352

⬍.01

⬍539.0

18.72 ⫾ 5.08

17.78 ⫾ 5.20

⬍.01

–0.124

⬎.01

⬎539.0

22.05 ⫾ 4.28

19.90 ⫾ 3.33

⬍.01

0.502

⬍.01

⬍513.0

18.27 ⫾ 4.29

17.30 ⫾ 3.90

⬍.01

–0.106

⬎.01

513.0–539.0

19.14 ⫾ 5.81

18.27 ⫾ 6.27

⬎.01

–0.345

⬎.01

539.0–574.5

21.70 ⫾ 3.46

20.33 ⫾ 3.09

⬍.01

–0.046

⬎.01

⬎574.5

22.40 ⫾ 5.00

19.47 ⫾ 3.55

⬍.01

0.634

⬍.01

IOP ⫽ intraocular pressure *Difference between methods

Table 2. Distribution of IOP differences in the entire group and the subgroups. Number of Eyes (%) Corneal Thickness (␮m)

IOP in Range of 2 mm Hg

Noncontact Readings >2 mm Hg

Goldmann Readings >2 mm Hg

Entire group

70 (58.33)

41 (34.17)

9 (7.50)

⬍539.0

38 (63.33)

17 (28.33)

5 (8.33)

⬎539.0

32 (53.33)

24 (40.00)

4 (6.67)

⬍513.0

23 (76.67)

6 (20.00)

1 (3.33)

513.0–539.0

15 (50.00)

11 (36.67)

4 (13.33)

539.0–574.5

20 (66.67)

8 (26.67)

2 (6.67)

⬎574.5

12 (40.00)

16 (53.33)

2 (6.67)

IOP ⫽ intraocular pressure

Discussion The Imbert-Fick Law states that an external force against a sphere equals the pressure in the sphere times the area flattened by the external force. The validity of the law requires that the sphere is perfectly spherical and flexible, dry, and infinitely thin. The cornea fails to satisfy any of those requirements as it is aspherical and wet and neither perfectly flexible nor infinitely thin. The moisture creates surface tension, while the lack of flexibility requires force that is independent of the internal pressure to bend the cornea. In addition, because the cornea has a central thickness of approximately 550 ␮m,

the outer area of flattening is not the same as the inner area. Goldmann based his concept of tonometry on a modification of this law and found that ocular rigidity does not significantly influence the measurement of IOP. However, the semicircles of Goldmann tonometry, corneal variables, contact duration, and calibration may affect the results. Corneal thickness influences IOP measurement, with thin corneas producing falsely low readings. A thick cornea causes a falsely high measurement if the thickness is the result of increased collagen fibrils; low readings occur if the thickness is a result of corneal edema.1

J CATARACT REFRACT SURG—VOL 27, NOVEMBER 2001

1789

RELATION OF CORNEAL THICKNESS TO IOP

In our routine clinical practice, we observed a difference between IOP measurements with noncontact and applanation methods in almost every case. First, we wanted to know whether this difference was statistically significant and then whether there was a correlation with corneal thickness. We included both eyes of each patient because consensual IOP-lowering by the tonometer is not an immediate reaction.2 Previous studies have found IOP measurements with noncontact tonometers to be reliable.3–7 Comparisons with Goldmann applanation tonometers indicate that the noncontact tonometer is reliable within the normal range, although the reliability is reduced in higher pressure ranges and is limited by an abnormal cornea or poor fixation.1 However, this is not the case every time. Some studies found applanation tonometer measurements to be higher5,8 and others,9 the results of the noncontact tonometer. In our study, the measurements with the noncontact tonometer were higher. The difference between the IOP measurements with the noncontact tonometer and those with the applanation tonometer was statistically significant. When patients were divided into 2 and then 4 subgroups, the difference was statistically significant in every subgroup except in that with pachymetry ranging from 513 ␮m to 539 ␮m. Among all cases, 41.67% of measurements were out of the ⫾2 mm Hg range. Differences of more than 2 mm Hg between 2 measurements decrease the reliability of the noncontact tonometer if we accept the Goldmann applanation method as reliable. Several factors may influence the measurement of IOP with different instruments, with corneal parameters probably the most important. When we performed linear regression analysis, we found a statistically significant positive correlation between corneal thickness and the differences in IOP measurements with noncontact and applanation methods in the entire group, in the second half of the cases, and in cases with a corneal thickness greater than 574.5 ␮m. The correlation coefficient was small in the entire group (r ⫽ 0.352) and slightly larger in the second half of the group (r ⫽ 0.502). The highest value was detected in eyes with a corneal thickness greater than 574.5 ␮m (r ⫽ 0.634), showing a significant correlation between corneal thickness and differences in IOP measurements, especially with regard to higher corneal thickness values (⬎574.5 ␮m). 1790

To the best of our knowledge, this is the first study that presents a correlation between differences in IOP measurements with contact and applanation tonometry and corneal thickness. A study of the significance of corneal thickness in noncontact tonometry found it to be less influenced by corneal thickness than applanation tonometry; that is, thicker corneas tend to give higher IOP readings by applanation tonometry than by noncontact tonometry.10 This is in contrast to our findings. Several studies suggest a correlation between corneal thickness and false IOP readings. Johnson and coauthors11 report a 17-year-old girl who had an IOP of 30 to 40 mm Hg with normal visual fields and optic nerve heads in both eyes. Medical treatment was unsuccessful in substantially lowering IOP. The central corneal thickness was 900 ␮m in both eyes in the absence of corneal edema. Cannulation of the left anterior chamber showed an IOP of 11 mm Hg; the IOP with the Perkins and Schiotz tonometers was 35 and 34 mm Hg, respectively. Whitacre and coauthors12 correlated IOP with Perkins tonometry in vivo in 15 eyes. The eyes were cannulated before intraocular surgery, and IOP was controlled manometrically to 10, 20, and 30 mm Hg. There was a statistically significant relation between corneal thickness and errors in Perkins tonometry. Thin corneas produced IOP underestimations by as much as 4.9 mm Hg, whereas thick corneas produced overestimations by as much as 6.8 mm Hg. Argus13 reports that increased corneal thickness leads to artificially high estimations of IOP in ocular hypertension patients, and studies of eyes having excimer laser procedures show IOP changes postoperatively.14 –19 The results show that there is a single corneal thickness for accurate measurement of IOP for each tonometer. Ehlers and coauthors20 performed simultaneous manometry and Perkins or Draeger handheld tonometry in 29 eyes scheduled for cataract or glaucoma surgery. All eyes had normal corneas, and none had astigmatism greater than 1.5 diopters. Three tonometer readings were taken at manometrically determined IOPs of 10 mm Hg and 30 mm Hg. A systematic error in the accuracy of applanation tonometry proportional to the true IOP and the central corneal thickness was present, with a highly significant correlation coefficient (P ⬍ .001) of 0.707 at 10 mm Hg

J CATARACT REFRACT SURG—VOL 27, NOVEMBER 2001

RELATION OF CORNEAL THICKNESS TO IOP

and 0.737 at 30 mm Hg. The authors extrapolated that at a true IOP of 20 mm Hg, a corneal thickness of 450 ␮m would produce a mean underestimation of 5.2 mm Hg and a corneal thickness of 590 ␮m would produce a mean overestimation of 4.7 mm Hg. The Goldmann tonometer was believed to give accurate readings when the central corneal thickness was 520 ␮m. We suggest that if the IOP measured with a noncontact tonometer is significantly different from that taken with an applanation tonometer, the patient’s corneal thickness be measured. Corneal thickness is indirectly important in the diagnosis and follow-up of glaucoma as it may lower the incidence of misdiagnosis. We also suggest that new tonometers not affected by corneal thickness be manufactured. The main conclusion from our study is that when corneal thickness is greater than 539 ␮m, noncontact tonometry is significantly less accurate than applanation tonometry.

10.

11.

12.

13. 14.

15.

16.

References 1. Shields MB. Textbook of Glaucoma, 3rd ed. Baltimore, MD, Williams & Wilkins, 1992; 53– 83 ¨ F, Hasıripi H, Vayısog˘ lu E, et al. Accurate time 2. Recep O interval in repeated tonometry. Acta Ophthalmol Scand 1998; 76:603– 605 3. Vernon SA. Reproducibility with the Keeler Pulsair 2000 non-contact tonometer. Br J Ophthalmol 1995; 79:554 –557 4. Lagerlo¨ f O. Airpuff tonometry versus applanation tonometry. Acta Ophthalmol 1990; 68:221–224 5. Moseley MJ, Evans NM, Fielder AR. Comparison of a new non-contact tonometer with Goldmann applanation. Eye 1989; 3:332–337 6. Fisher JH, Watson PG, Spaeth G. A new handheld air impulse tonometer. Eye 1988; 2:238 –242 7. Sponsel WE, Kaufman PL, Strinden TI, et al. Evaluation of the Keeler Pulsair non-contact tonometer. Acta Ophthalmol 1989; 67:567–572 8. Moseley MJ, Thompson JR, Deutsch J, et al. Comparison of the Keeler Pulsair 2000 non-contact tonometer with Goldmann applanation. Eye 1993; 7:127–130 9. Eisenberg DL, Sherman BG, McKeown CA, Schuman JS. Tonometry in adults and children; a manometric evaluation of pneumotonometry, applanation, and TonoPen

17.

18.

19.

20.

in vitro and in vivo. Ophthalmology 1998; 105:1173– 1181 Doughty MJ, Zaman ML. Human corneal thickness and its impact on intraocular pressure measures: a review and meta-analysis approach. Surv Ophthalmol 2000; 44: 367– 408 Johnson M, Kass MA, Moses RA, Grodzki WJ. Increased corneal thickness simulating elevated intraocular pressure. Arch Ophthalmol 1978; 96:664 – 665 Whitacre MM, Stein RA, Hassanein K. The effect of corneal thickness on applanation tonometry. Am J Ophthalmol 1993; 115:592–596 Argus WA. Ocular hypertension and central corneal thickness. Ophthalmology 1995; 102:1810 –1812 ¨ E, Bowman RW, Cavanagh HD, McCulley Abbasog˘ lu O JP. Reliability of intraocular pressure measurements after myopic excimer photorefractive keratectomy. Ophthalmology 1998; 105:2193–2196 Schipper I, Senn P, Thomann U, Suppiger M. Intraocular pressure after excimer laser photorefractive keratectomy for myopia. J Refract Surg 1995; 11:366 – 370 Chatterjee A, Shah S, Bessant DA, et al. Reduction in intraocular pressure after excimer laser photorefractive keratectomy; correlation with pretreatment myopia. Ophthalmology 1997; 104:355–359 Mardelli PG, Piebenga LW, Whitacre MM, Siegmund KD. The effect of excimer laser photorefractive keratectomy on intraocular pressure measurements using the Goldmann applanation tonometer. Ophthalmology 1997; 104:945–948; discussion by WA Argus, 949 Faucher A, Gre´goire J, Blondeau P. Accuracy of Goldmann tonometry after refractive surgery. J Cataract Refract Surg 1997; 23:832– 838 Fournier AV, Podtetenev M, Lemire J, et al. Intraocular pressure change measured by Goldmann tonometry after laser in situ keratomileusis. J Cataract Refract Surg 1998; 24:905–910 Ehlers N, Bramsen T, Sperling S. Applanation tonometry and central corneal thickness. Acta Ophthalmol 1975; 53:34 – 43

From Is¸ık Eye Clinic, Ankara, Turkey Presented at the Symposium on Cataract, IOL and Refractive Surgery, Boston, Massachusetts, USA, May 2000. None of the authors has a proprietary interest in the development or marketing of any product mentioned.

J CATARACT REFRACT SURG—VOL 27, NOVEMBER 2001

1791