SEARCH

SEARCH BY CITATION

Keywords:

  • central IOP;
  • Goldmann tonometry;
  • ICare;
  • intraocular pressure;
  • peripheral IOP;
  • precision and accuracy;
  • glaucoma;
  • rebound tonometry

Abstract.

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. References

Purpose:  The purpose of this study was to evaluate the ICare tonometers precision and accuracy and the extent to which intraocular pressure (IOP) measurements are influenced by measuring position.

Methods:  This was carried out by comparing the central and peripheral ICare-IOP readings and comparing ICare- with the Goldmann applanation tonometer (GAT)-IOP readings. IOP was measured using the ICare rebound tonometer on the right eye of 40 subjects, straight at the centre of the cornea (CS), straight 2 mm from the nasal and temporal limbus (NS and TS), and in 10 degrees nasally and temporally angled positions measured from the same location as CS (NA and TA). The IOP was also assessed with the GAT.

Results:  Central IOP (CS) was significantly (p < 0.001) greater than peripheral measurements (NS, TS, NA and TA) by approximately 3–4 mmHg. Centre IOP (CS) significantly overestimated by mean 2 mmHg and the peripheral measurements significantly underestimates approximately 1.4–2 mmHg compared with GAT readings.

Conclusion:  The ICare tonometer may be useful in a routine clinical setting for IOP screening, but the ICare measurement is not a substitute for the GAT measurement, when a precise and accurate IOP is desired.


Introduction

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. References

Statistics gathered by WHO in 2002 (Kingman 2004) show that globally glaucoma is now the second leading cause of blindness after cataract. As intraocular pressure (IOP) is a risk factor for glaucoma, the assessment of a correct IOP is of major importance in glaucoma diagnosis and follow-up. The measurement of the IOP is also a fundamental part of the basic routine ophthalmologic examination.

IOP measurements obtained by Goldmann applanation tonometry (GAT) is worldwide considered as the “gold” standard. However, it is only accurate when measurements are taken by experienced individuals (Abraham et al. 2008). Also, the time needed to apply anaesthetics and fluorescein, to make regular calibration checks and to set up the tonometer can add up. There are many alternatives to the GAT (Knecht et al. 2009). A new mobile easy-to-use alternative to the GAT is the ICare rebound tonometer, developed and patented by Kontiola in the 90s and based on the basic principles by Dekking & Coster (1967). This handheld device requires no instillation of anaesthetics and is still painless and comfortable for the patient. Readings may be obtained by inexperienced individuals (Abraham et al. 2008).

The ICare tonometer projects a small, light in weight and permanent magnetized probe onto the corneal surface. The probe is launched by a voltage-pulse induced by a coil inside which the probe moves. The tip of the probe-ball hits the cornea and rebounds. The probe movement is monitored by a sensing coil. Several motion parameters of the probe are determined from the voltage that the moving probe induces in the sensing coil. The motion parameters (especially the deceleration time), which vary according to the IOP, are used to estimate the IOP. The exact conversion from voltage to IOP remains a patented secret (Kontiola 1996–1997, 2000, 2001).

Recently, ICare has been compared with other tonometers, but only few studies (Roukonen et al. (2007), Munkwitz et al. (2008) and Queiros et al. (2007)) evaluate the ICare tonometers precision and accuracy and the extent to which ICare-IOP readings are influenced by measuring position. This was therefore the purpose of this study and was carried out by determining differences in ICare-IOP measurements at different positions (between straight central IOP and straight peripheral IOP and between straight central IOP and angled peripheral IOPmeasurements) in a normotensive sample. These measurements were compared to the GAT readings taken in the same session.

Materials and Methods

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. References

Forty-six volunteers from the hospital-staff, ages ranging from 21 to 59 years (mean 36.5), nine men and 37 women, participated in this study. The nonequal proportion between women and men is not considered to affect the study results. For the following reasons, six of the 46 volunteers (five women and one man) were excluded: wearing hard contact lenses, present eye infection or eye infection within the last 3 months, subject to ocular injury or trauma, intraocular surgery, corneal surgery, astigmatism and refraction with more than ± 3 D. The study followed the instructions by the ethical committee.

IOP values were only obtained for the right eye. The ICare tonometer was fixed to a slitlamp as showed in Fig. 1. A pilot study, which showed fixed and handheld settings were comparable, was already performed. All the measurements were taken, when the participant was sitting relaxed and looking straight forward, and so did not activate the extraocular eye muscles, which might elevate the IOP (Hofer et al. 1995). All the measurements were taken from a distance of 4–8 mm from cornea. ICare measurements were taken at five different positions (Fig 2).

image

Figure 1.  ICare tonometer fixed to the slit-lamp.

Download figure to PowerPoint

image

Figure 2.  The five measuring positions with the ICare tonometer.

Download figure to PowerPoint

Central straight measurements (CS) were measured when the probe was held perpendicular (90°) to the central cornea.

Peripheral ICare measurements: Nasal/temporal straight measurements (NS/TS) were taken at a constant straight distance about 2 mm from the limbus in the nasal and temporal regions of the horizontal meridian of the cornea. The nasal/temporal-angled measurements (NA/TA) were taken by turning the tonometer 10° temporally and nasally, when the tonometer was held at the same place as for CS measurements.

At each position, six readings were taken and the mean value calculated. The order of the different positions was randomised to minimise the potential effect of the first measurement on subsequent measurements. All the ICare measurements were taken first, and afterwards the GAT measurements were taken, as it is known that repeated applanation measurements with GAT decrease IOP (Stocker 1958). Similar effects have not been found for the ICare in the literature. As ICare is not an applantion tonometer and the area of the probe-ball-tip that touches cornea is several times smaller than the area of the GAT (7.44 mm2), it is most likely that GAT’s influence on IOP is greater, than ICare’s influence. GAT readings were taken by using 0.4% oxybuprocain and fluorescein. Three GAT measurements were assessed, and the median value was used (Dielemans et al. 1994). The study time per volunteer did never exceed 10 min.

Statistics

The mean differences, between central and peripheral measurements and between ICare and GAT measurements, were compared to zero in a paired two-tailed sample t-test. The 95% confidence interval (CI) was also calculated for these differences. To assess agreement between the methods, Bland–Altman plots (Altman & Bland 1983) were constructed (Fig. 3a,b). For these plots, simple linear regression was performed. The regressions coefficients were tested for being significant. The r2 coefficient was also calculated. All data are shown in Table 1.

image

Figure 3.  The Bland–Altman plots describe the relationship between GAT and CS measurements (A) and between CS and NS measurements (B). The 95% CI is shown for each point at the plot. Each point shows the mean of the six obtained values. The regression line, the p-value for the regression coefficient and r2 are also displayed.

Download figure to PowerPoint

Table 1.   Statistical data for the difference between two measuring methods as described under statistics.
ComparisonsMean difference ± SD mmHgCI95% of difference mmHgp-valueRegression coefficientp-value regressions coefficientR2
CS-GAT2.10 ± 2.56[1.31; 2.90]p < 0.0010.592p < 0.0010.271
NS-GAT−1.40 ± 3.18[−2.39; −0.42]p = 0.0040.166p = 0.0090.137
TS-GAT−1.53 ± 3.48[−2.61; −0.45]p = 0.0040.059p = 0.0010.001
NA-GAT−1.62 ± 3.42[−2.68; −0.56]p = 0.0050.133p = 0.0140.008
TA-GAT−2.00 ± 3.58[−3.11; −0.89]p = 0.0010.291p = 0.0820.032
CS-NS3.34 ± 1.42[3.02; 3.97]p < 0.0010.264p < 0.0010.265
CS-TS3.49 ± 1.54[2.90; 3.78]p < 0.0010.216p = 0.0150.147
CS-TA4.09 ± 3.35[3.05; 5.13]p < 0.0010.105p = 0.5480.011
CS-NA3.56 ± 3.13[2.59; 4.52]p < 0.0010.024p = 0.6110.008
TS-NS0.15 ± 2.07[−0.49; 0.79]p = 0.347   

Results

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. References

Table 1 presents the mean difference of IOP obtained by ICare (CS, NS, TS, NA and TA) compared to the GAT. Compared to GAT, the CS significantly overestimated the IOP by approximately 2 mmHg, while the peripheral measurements significantly underestimated IOP with 1.4–2.0 mmHg. Also, a greater overestimation (CS-GAT) with higher average IOP (CS+GAT)/2 was found, and an underestimation was found, when the average IOP (CS+GAT/2) was less than ca. 11.6 mmHg (Fig. 3a). There was also a significant positive regression coefficient for NS-GAT, NT-GAT and NA-GAT, but no significant regression coefficient was found for the angled measurement TA-GAT (Table 1).

There was a significant difference between central and peripheral corneal measurements (Table 1).The straight peripheral measurements (TS/NS) were about 3.3–3.5 mmHg less than the CS measurements. As it is shown in the Bland–Altman plot (Fig. 3b), the difference tends to be greater with a higher IOP (the regression coefficient is significant).

The angled peripheral measurements at cornea (TA/NA) were about 3.5–4 mmHg less than CS measurements and there were not found any linear relationship for these.

It is also useful to note the precision (repeatability) of the IOP readings. This was performed by looking at the 95% CI and the standard deviation (SD) shown in Table 1. The SD was much higher for peripheral ICare IOP-GAT than for CS-GAT. The SD was highest for the angled IOPs, when these were compared with the CS. The difference between the maximum and minimum IOP value among the six obtained measurements were also calculated. The greatest differences were calculated for the angled measurements (TA = 2.29 ± (SD) 1.36 mmHg and NA = 2.23 ± 1.19 mmHg). CS was 2.01 ± 1.06 mmHg and the lowest was GAT 1.06 ± 0.354 mmHg. The most precise ICare measurement is the centrally straight (CS) taken measurement.

Discussion

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. References

In this study, we compare the new (ICare) with the old classical instrument (GAT). The principle of the GAT is well known (Goldmann & Schmidt 1957). GAT provides the standardized and accepted IOP, even though it is well debated that GAT-IOP is influenced by e.g. the central corneal thickness (Ehlers et al. 1975a,b), refractive surgery (Kohlhaas et al. 1996), tear film (Zeng et al. 2008) and irregular cornea-like keratoconus (Brooks et al. 1984). The present study shows significant overestimation by CS ICare measurements and significant underestimation by peripheral ICare measurements compared to GAT measurements. Hence, ICare’s accuracy is doubtful.

This is the only study known so far that fixes the ICare tonometer to obtain the IOP, but there are several studies (Martinez-de-la-Casa et al. (2009), Jóhannesson et al. (2008), Munkwitz et al. (2008), Pakrou et al. (2008) and Roukonen et al. (2007)) that do show an overestimation by central ICare tonometry compared to GAT measurements. This is shown in Table 2. Munkwitz et al. (2008) found moderate agreement between central ICare and GAT in normal to moderate elevated IOP and a poor agreement in the higher IOPs. This is also in agreement with Roukonen et al. (2007) and our data, which showed greater difference GAT-CS at greater IOP (Fig. 3a). Roukonen et al. (2007) and Martinez-de-la-Casa et al. (2009) concluded that ICare did not fulfil the international standards for tonometers ISO 8612.

Table 2.   Presents the mean difference and the 95%CI between ICare and GAT measurements (ICare-GAT) in different cited articles.
 Our study CS-GATMartinez-de-la-Casa et al. (2009)Jóhannesson et al. (2008)Munkwitz et al. (2008)Pakrou et al. (2008) left/right eyeRoukonen et al. (2007)
Mean difference, mmHg ICare-GAT2.100.972.020.790.8/0.42.5
95%CI of difference mmHg[1.31; 2.90][0.2; 1.7][−4.06; 8.10][−8.67; 10.25][−4.7; 6.2]/[−5.5; 6.3][0.3; 4.7]

Our study found a greater overestimation (CS-GAT) than most of the previous studies. One explanation is that the ICare measurements in those studies were performed handheld, and therefore, they may have angled or measured a little peripheral. This tends to reduce the overestimation. Thereby, we suggest, as the instrument is handheld, it is likely to measure off-centre and angled.

There is only one study that describes the relationship between central and peripheral ICare measurements. Queiros et al. (2007) only found significant differences between central and nasal readings (the central – nasal was 0.75 mmHg, CI95% [−1.6; 3.07], p = 0.045). The problem in that study is that all the peripheral measurements were taken by asking the participant to look to the side. So, when they obtained the peripheral measurement, they probably increased the IOP by activating the extra-ocular muscles (Hofer et al. 1995). In the present study, there was a statistical significant difference between central and peripheral measurements. There are many biomechanical factors in the eye, which could explain this difference for instance the corneal thickness. As the cornea is about 25% thicker at the peripheral compared to the centre (Martola & Baum 1968), this may cause less rebound of the probe, which explains the smaller peripheral IOP readings.

The IOP differences in the present study are greater than in other studies probably because we have been able to control the angle and the location at cornea where the measurements were taken. In this study, we found, that the peripheral readings with ICare, especially the angled readings (TA/NA) were less precise (repeatable), while the most precise was the CS ICare reading. One explanation is that ICare as a rebound tonometer registers motion parameters. In the CS measuring, the force induced back from cornea is transmitted through central part of the probe-head, but in peripheral measurements, the force is transmitted through the peripheral parts of the probe-head. The force is spread perpendicular to the contact place at cornea, which in CS measurements are along the axis of the probe. In peripheral measurements, the axis of the probe is no longer at the perpendicular line to the contact location at cornea. Therefore, the force F (F = m*acceleration) transmitted may easily be less and even varying for each measurement. As force F is less, and the mass is the same, the deceleration (rebound) is less. A second explanation could be that the corneal area hit by the probe is more varying and greater peripherally than straight centrally (CS). As Pressure = Force/Area, and the force is same, the measured IOP would be lowered. A third important explanation is that the distance from the tip of the probe-ball to cornea is less controlled (4–8 mm). This could influence how off-centre the angled measurements will be.

Easy-to-use, no need for anaesthesia, portability, and a small contact area at cornea are some of the advantages of the rebound tonometer, making the ICare measurement suitable for patients with disability, children, patients with corneal scarring and the routine screenings. However, CS ICare significantly overestimates, while ICare measured peripherally significantly underestimates compared to the GAT. Also, its repeatability is doubtful, especially for the peripheral measurements.

Hence, it is not advisable to use it, when a precise and accurate IOP is necessary, as in the glaucoma treatment. Here, the GAT is far more appreciable.

References

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. References