ORIGINAL STUDY
Clinical Comparison of the Icare Tonometer and
Goldmann Applanation Tonometry
Nima Pakrou, MBBS, Tim Gray, MBBS, Richard Mills, FRANZCO,
John Landers, MBBS, MPH, and Jamie Craig, MBBS, PhD, DPhil, FRANZCO
Purpose: To compare a new method of intraocular pressure
(IOP) measurement, using the Icare tonometer, with Goldmann
applanation tonometry (GAT).
Patients and Methods: Two observers obtained IOP readings in
292 eyes (143 right and 149 left) of 153 subjects, using the Icare
without topical anesthetic. A GAT reading was subsequently
obtained by a consultant ophthalmologist, without the knowledge of the Icare readings. Central corneal thickness (CCT) was
obtained on all eyes with ultrasound pachymetry. Patient
comfort after IOP measurement was assessed in a consecutive
subset of patients.
Results: The intraclass correlation coefficient between the 2
modalities of IOP measurement was r = 0.95 for the right and
r = 0.93 for the left eye. The mean difference (Icare GAT)
between the IOP measured by the 2 methods was 0.4 mm Hg in
the right eye (SD 3.0, 95% confidence interval 5.5 to 6.3), and
0.8 mm Hg in the left eye (SD 3.0, confidence interval 4.7 to
6.2). GAT measurements did not vary with CCT [correlation
coefficient = 0.09 (P = 0.25) right and 0.14 (P = 0.09) left eyes].
However, IOP measured with Icare tonometry increased with
increasing CCT [correlation coefficient = 0.16 (P = 0.05) right
and 0.21 (P = 0.01) left eyes]. For every 100-mm increase in
CCT, the difference (Icare GAT) increased by 1 mm Hg. Of
the 38 consecutive patients surveyed, 28 (73.7%) rated the Icare
more comfortable than GAT, with only 2 (5.3%) rating it less
comfortable (P<0.001).
Conclusions: There is good correlation between the 2 methods of
IOP measurement, even at extremes of IOP. The Icare
instrument was easy to use and recorded rapid and consistent
readings with minimal training. It seems to be more comfortable
than GAT and obviates the need for topical anesthesia.
Key Words: intraocular pressure, impact tonometer, rebound
tonometer, Icare tonometer, Goldmann applanation tonometer
(J Glaucoma 2008;17:43–47)
Received for publication September 15, 2006; accepted June 2, 2007.
From the Department of Ophthalmology, Flinders Medical Centre,
Bedford Park, South Australia.
Competing interest statement: The authors would like to declare no
competing interests.
Reprints: Dr Nima Pakrou, MBBS, Department of Ophthalmology,
Flinders Medical Centre, Bedford Park, South Australia 5042
(e-mail: [email protected]).
Copyright r 2008 by Lippincott Williams & Wilkins
J Glaucoma
Volume 17, Number 1, January/February 2008
A
ccurate and consistent measurement of intraocular
pressure (IOP) remains a key factor in the diagnosis
and management of glaucoma. Despite numerous portable devices available, the majority of ophthalmologists
rely solely on the Goldmann applanation tonometry
(GAT) as the gold standard1 method of obtaining IOP,
using it to guide decisions regarding patient management.
However, accurate use of this instrument requires
significant training, and inaccuracies are common in the
hands of inexperienced examiners. Furthermore, GAT
requires a slit lamp microscope, and the use of topical
anesthetic with fluorescein dye. Topical anesthetic is
unpleasant for the majority of patients due to stinging,
and reflex blepharospasm makes accurate measurements
difficult to obtain in a subset of patients.
Given the increasing time restraints placed upon
many ophthalmologists, IOP measuring devices, which
are accurate, portable, and can be easily used by various
operators with minimal interoperator variability, are
appealing. In situations where rapid assessment or
screening is desirable, such as in remote community
settings, the use of the GAT is impractical. Furthermore,
in circumstances (such as within the emergency department) in which identification of either very low or high
pressures can have significant management implications,
accessibility to a simple reliable instrument is valuable.
The Icare tonometer (TA01; Tiolat Oy, Helsinki,
Finland) is a recently developed hand-held portable
tonometer, which relies on the induction rebound or
impact principle. It is able to rapidly obtain IOP
measurements without the need for topical anesthesia.2
Rebound or impact tonometry is based on making a
moving object collide with an eye while monitoring the
motion parameters of the colliding object.3 At low IOP,
the deceleration of the probe is less than that observed at
high IOP. Consequently, the higher the IOP, the shorter
the duration of the impact.2,4 IOP is calculated from the
measurement of impact duration and/or maximum
deceleration.2 Use of the device is simple and can be
rapidly taught. It may have a role in the setting of
community screening, or in situations where GAT is
difficult or not practical provided it can be shown to be
accurate.
The purpose of this study was to compare the
accuracy of IOP measurement by Icare impact tonometry
with that of GAT in an ophthalmology outpatient setting
of a teaching hospital. The subjective comfort of both
43
Pakrou et al
J Glaucoma
methods of applanation was also assessed in a subset of
patients to determine patient acceptability.
participants in the study. Subjects were required to
indicate whether they found:
1. The Icare much more comfortable (score 1)
2. The Icare more comfortable (score 2)
3. No difference between the 2 methods (score 3)
4. The Goldmann tonometer more comfortable (score 4)
5. The Goldmann tonometer much more comfortable
(score 5)
A mean score was calculated based on the
responses.
Statistical Analysis System 6.12 (SAS Institute Inc,
Cary, NC) was used for statistical analysis including
descriptive statistics, Student t test, paired t test,
Wilcoxon signed rank test, intraclass correlation coefficient, and simple linear regression for univariate and
multivariate analyses. In linear regressions, IOP and CCT
were used as continuous variables. Intraclass correlation
coefficient was used to assess the correlation between the
Icare and Goldmann tonometry measurements. Paired
differences with limits of agreement were calculated to
evaluate the concordance of measurements obtained with
the 2 instruments.
The difference in IOP measured with the Icare and
GAT was plotted against the IOP measured with GAT.
This is a slight modification of the Bland-Altman plot.5 A
P value of <0.05 was considered statistically significant.
MATERIALS AND METHODS
Two hundred ninety-two eyes (143 right and 149
left) from 153 patients aged 18 to 86 years (mean 59.6, SD
21.2 y) were selected from the ophthalmology clinic at the
Flinders Medical Centre, a South Australian teaching
hospital. Participants included patients with and without
glaucoma, and some patients with suspected extreme
values of IOP were selectively included to assess Icare
performance across the full range of IOPs. Informed
consent was obtained from each participant. Patients with
corneal pathology (eg, corneal epithelial defects, corneal
stromal scarring, and corneal edema) were excluded.
Icare measurements were obtained by 1 of 2
ophthalmology residents (N.P., T.G.). To obtain a
measurement, single-use probes were fitted into the
apparatus. Subjects were instructed to fixate on a distant
target while the probe of the Icare was held at a distance
of 4 to 8 mm, and perpendicular to the central cornea.
The measurement was initiated by the operator, with the
probe being propelled against the central cornea.
Individual measurements are displayed digitally in
mm Hg. Six rapidly consecutive measurements are
obtained in each eye, yielding a mean reading with a
SD on the LCD screen. As per manufacturer instructions,
a mean reading displayed with a static P and no error
bars (P) indicates a low SD of the 6 measurements. A
mean with a flashing P and error bar indicates a less than
optimal SD. An inferior error bar (P ) implies that the
different measurements have a SD slightly larger than
normal (judged acceptable according to the manufacturer); a mean with a middle error bar (P ) indicates
a SD clearly greater than normal (a new measurement
is recommended); and a mean with a superior error
bar (P ) signifies that the SD is too large and a new
measurement is required.
In our study, consecutive measurements were
obtained from each eye until either a mean reading
displaying a P with no bar (P) was acquired, or a total of
3 readings were performed, regardless of SD. The total
number of attempts needed to achieve a P reading was
recorded for each patient.
After the instillation of Fluorescein-Lignocaine
solution (Minims; Chauvin Pharmaceuticals Ltd, Kingston-Upon-Thames, England), GAT (Haag Streit AG,
Bern, Switzerland) was performed by 1 of 2 consultant
ophthalmologists (J.C., R.M.), who were blinded to the
IOP measurement obtained by the Icare. Central corneal
thickness (CCT) measurements were obtained using
a Tomey pachymeter SP-2000 (Tomey Corporation,
Nagoya, Japan). A total of 5 measurements were
obtained and the mean value used for further analysis.
Immediately after the completion of IOP evaluation, a subset of 38 consecutive patients were asked to
indicate which of the 2 methods they found more
comfortable. These subjects represented the last 38
44
Volume 17, Number 1, January/February 2008
RESULTS
There were 153 patients in the sample, 81 males and
72 females (143 right eyes and 149 left eyes). The average
age of patients was 60 years (SD 21 y, range 18 to 86 y). In
286/292 eyes (97.9%), an Icare reading with no error bar
(P) was achieved within 3 attempts at IOP measurement.
In these subjects, the mean number of measurements
required to reach a no error bar reading was 1.36 (SD
0.6). In all eyes, an acceptable mean reading showing
either P or P was obtained within 3 attempts.
Mean IOP measured using GAT was 18.2 mm Hg
(SD 8.7 mm Hg) and using Icare tonometry 17.6 mm Hg
(SD 8.7 mm Hg) (Table 1). There was a high degree of
correlation between the 2 methods with an intraclass
correlation coefficient of r = 0.95 for right eyes and
r = 0.93 for left eyes. The mean CCT was 537.5 mm (SD
43.9).
The mean of the difference (Icare GAT) in IOP
measurements was +0.4 mm Hg in right eyes [SD 3.0,
95% confidence interval (CI)
5.5 to +6.3] and
+0.8 mm Hg in left eyes (SD 3.0, CI
4.7 to +6.2),
TABLE 1. IOP Measured by GAT and Icare Tonometer, and
CCT
IOP (mm Hg)
Icare
Goldmann
CCT (mm)
Minimum
Maximum
Mean
SD
1
0
364
60
63
694
18.2
17.6
537.5
8.8
8.7
43.9
SD indicates standard deviation.
r
2008 Lippincott Williams & Wilkins
J Glaucoma
Volume 17, Number 1, January/February 2008
Icare Tonometer Compared to Goldmann Tonometry
(85.0%) of these recordings were within the same range.
The 3 exceptions recorded IOPs of 28, 28, and 30 mm Hg
on GAT, which corresponded to readings of 25, 26, and
26 mm Hg, respectively, with the Icare. Similarly to assess
Icare accuracy in the lower ranges of IOP, the 22 patients
with GAT measurement of <10 mm Hg were analyzed.
Of these patients, 18/22 (81.9%) also recorded a reading
within the same range, using the Icare. The 4 exceptions
measured 8, 8, 8, and 9 mm Hg on GAT, corresponding
to 10, 10, 13, and 11 mm Hg, respectively, using the Icare.
The mean response score for subjective patient
comfort of IOP measurement method was 2.03 (ie, in
favor of Icare). This was a significant result (P<0.001).
Only 2/38 (5.3%) patients preferred the GAT, and 8/38
(21%) indicated no preference for either method. GAT
measurements did not vary with CCT [correlation
coefficient = 0.09 (P = 0.25) for right eyes and 0.14
(P = 0.09) for left eyes]. However, IOP measured with
Icare tonometry increased with increasing CCT [correlation coefficient = 0.16 (P = 0.05) for right eyes and 0.21
(P = 0.01) for left eyes]. This resulted in an increase in the
I-Care-Goldmann difference as CCT increased (t = 2.32;
P = 0.022 for right eyes and t = 2.11; P = 0.032 for left
eyes). For every 100-mm increase in CCT, the difference
increased by 1 mm Hg.
FIGURE 1. A and B, Bland-Altman plots presenting the
differences in IOP (Icare GAT), as a function of IOP measured
with GAT for left eyes (A) and right eyes (B).
and was not significantly different between eyes (t = 1.68;
P = 0.095) (Figs. 1A, B). This difference varied as GAT
IOP increased. There was a significant difference between
mean Icare GAT for GAT IOPs <21 mm Hg compared with those when GAT IOP was Z21 mm Hg
(Table 2).
The Icare GAT difference did not change significantly throughout the study, suggesting no significant
learning effect associated with the Icare. Compared with
the first third of subjects, the last third showed a similar
mean difference (Icare GAT) [0.7 mm Hg (SD 3.0) vs.
0.2 mm Hg (SD 2.9) (t = 0.79, P = 0.34)] for right eyes
and (0.8 mm Hg (SD 3.1) vs. 0.8 mm Hg (SD 2.7)
(t = 0.01, P = 0.99)] for left eyes.
To assess the performance of the Icare in the high
ranges of IOP, we studied the 20 eyes with IOP readings
between 28 and 63 mm Hg, by GAT. Seventeen of 20
TABLE 2. Mean Difference (Icare GAT) in IOP
Measurements Stratified for IOP Measured With GAT
Eye
Right
Left
Both
r
IOP <21 mm Hg
+0.7 (SD 2.3)
+1.1 (SD 2.5)
+0.9 (SD 2.4)
IOP Z21 mm Hg
0.7 (SD 4.5)
0.4 (SD 3.6)
0.5 (SD 4.1)
2008 Lippincott Williams & Wilkins
Significance
t = 1.80, P = 0.081
t = 2.04, P = 0.048
t = 2.71, P = 0.008
DISCUSSION
The Icare tonometer has been marketed as a highly
reliable and accurate portable tonometer, which uses a
novel mechanism to obtain an IOP measurement.
However, there have not been many studies to critically
appraise its reliability in the clinical setting. Our aim was
to compare the IOP measurements obtained by the Icare
tonometer with those acquired using the gold standard,
GAT. There is widespread consensus that accurate and
reliable IOP measurement is extremely important in the
management of patients with glaucoma and has significant prognostic implications. Portable and easy to use
devices would also be of value as screening tools or in
situations where the use of GAT is impractical.
Our results demonstrate that the Icare tonometer is
in close agreement with the GAT, in the majority of
patients. Furthermore, the Icare was consistently able to
identify cases in which the IOP was well outside what is
considered normal. However, it does appear that the Icare
overestimates the GAT in the lower IOP ranges and
underestimates in the higher ranges. Despite this, these
discrepancies are small and may not be clinically
significant, especially if the instrument is used for screening purposes.
The 95% limits of agreement for the Icare versus
GAT were approximately between 5.0 and +6.0 mm Hg,
(90% CI of ± 4.8 mm Hg) which is comparable with
those reported for repeat measurements with GAT.
A repeat test with GAT was shown to have a 90% CI
of ± 4.5 mm Hg for a single reading and ± 3 mm Hg for
the average of 2 readings.6
45
Pakrou et al
J Glaucoma
Our results using the Icare are more favorable
compared to studies comparing other portable tonometers with the GAT. Studies looking at the TGDc-01
portable tonometer have found this device to have
significant fluctuations and variability compared with
the GAT.7–10 In comparing the Tonopen with GAT, some
studies reported significant differences between the 2
devices,11–13 whereas others failed to show significant
differences.14,15
There are few studies comparing the Icare impact
tonometer with the GAT. Kontiola2 compared a predecessor of the Icare with GAT in pig’s eyes that were
pressurized at different levels. They concluded that the
impact tonometer underestimated the IOP measured by
the GAT by 4.8 mm Hg. Another small study of subjects
with normal IOPs concluded that the Icare overestimates
the GAT, although with an absolute bias of <3 mm Hg
in approximately 80%.16 A study of 103 subjects
compared the Icare with the GAT and found no
significant difference between the 2 modalities.15 Recently
Martinez-de-la-Casa et al17 reported a good correlation
between IOP readings obtained by the Icare and the GAT
(r = 0.865). Unlike our findings, this study found that the
Icare readings were consistently higher than GAT
measurements, throughout the IOP ranges. The number
of eyes tested in this study, however, were approximately
half that of our study.
A more recent study looking at the Icare and GAT
readings in 42 healthy subjects concluded that IOP
measurements using the Icare were not significantly
different to the GAT. However, intersessional repeatability of IOP was poorer with the Icare compared with
GAT.18 Nakamura et al19 compared IOP measurements
using the Icare with GAT, Tonopen XL, and noncontact
tonometer, and the influence of CCT in 45 (12 control
and 33 glaucomatous or ocular hypertensive) subjects.
They reported a mean difference (95% limits of agreement) in IOP readings between Icare and GAT of
1.40 ± 4.29 mm Hg. Similar to our results as the CCT
got thicker, Icare was found to overestimate GAT.
A large study of 178 open-angle glaucoma patients
by Brusini et al,20 again demonstrated good agreement
between GAT and Icare IOP measurements. The CCT
was once more found to impact on the Icare IOP reading,
although its influence was more striking than was
observed in our study. Increasing CCT resulted in an
overestimation of the IOP measurement, with a change of
10 mm in CCT yielding a deviation of 0.7 mm Hg in the
Icare readings.
We feel that the findings with regard to the influence
of CCT on IOP measurements obtained by the Icare are
unlikely to be of significance in most clinical situations,
especially if the Icare is used as a screening instrument. In
our study, the Icare overread the Goldmann by 1 mm Hg
for every 100-mm increase in CCT measurements. Possibly,
this may be of significance in conditions in which there is
marked corneal thickening (eg, Corneal edema). Martinezde-la-Casa et al17 found that in terms of pachymetry, the 2
tonometers behaved in a similar manner.17
In our experience to obtain more accurate readings
(ie, no error bars) or readings more consistent with the
GAT, extra attention was needed to ensure the probe was
held perpendicular to the cornea and the 6 measurements
taken as quickly as possible, with minimal movement of
either the operator or the patients’ eyes. It was necessary
for the operator to stand directly in front of the subject to
ensure that the probe is not tilted either in the horizontal
or vertical plane. We felt that failure in ensuring optimal
positioning often resulted in readings with either error
bars or readings, which differed considerably from repeat
measurements, or GAT recordings. However, our results
indicated that there was no apparent learning effect when
the first third of patients were compared with the last
third.
In conclusion, the Icare seems to record accurate
readings in most situations when compared with GAT. It
is easy to use, requires little training, does not require the
use of topical anesthetic, and is very well tolerated by
patients. It may be of particular benefit in community
screening programs, inpatient settings, and emergency
departments in which the routine and accurate use of the
GAT may be unrealistic. The Icare may also be useful in
obtaining IOP measurement in children or patients in
which Goldmann applanation is technically challenging
(eg, blepharospasm). Although the Icare may be advantageous in many situations, it is unable to be used with the
patient in supine position. In addition, further studies are
required to assess its reliability over time.
46
Volume 17, Number 1, January/February 2008
REFERENCES
1. Kass MA. Standardizing the measurement of intraocular pressure
for clinical research Guidelines from the Eye Care Technology
Forum. Ophthalmology. 1996;104:184–185.
2. Kontiola AI. A new induction-based impact method for measuring
intraocular pressure. Acta Ophthalmol Scand. 2000;78:142–145.
3. Kontiola AI. Development Impact Tonometers for Clinical Use and
Glaucoma Research. Finland: Academic Dissertation, Department of Ophthalmology, University of Helsinki; 2003:39, 54, 55
(1–97).
4. Kontiola AI, Puska P. Measuring intraocular pressure with the
Pulsair 3000 and Rebound tonometers in elderly patients without
anaesthetic. Graefe’s Arch Clin Exp Ophthalmol. 2004;242:3–7.
5. Bland JM, Altman DG. Statistical methods for assessing agreement
between two methods of clinical measurement. Lancet. 1986;1:
307–310.
6. Motolko MA, Feldman F, Hyde M, et al. Sources of variability in
the results of applanation tonometry. Can J Ophthalmol. 1982;17:
93–95.
7. Amm M, Hedderich J. Transpalpebral tonometry with digital
tonometer in healthy eyes and after penetrating keratoplasty.
Ophtholmologne. 2005;102:70–76. German.
8. Losch A, Scheuerle A, Rupp V, et al. Transpalpebral measurement
of intraocular pressure using the TGDc-01 tonometer versus
standard Goldmann applanation tonometry. Graefes Arch Clin
Exp Ophthalmol. 2005;243:313–316.
9. Muller A, Godenschweer L, Lange GE, et al. Prospective
comparison of the new indentation tonometer TGDc-01, the noncontact tonometer PT100 and the conventional Goldmann applanation tonometer. Klin Monatsbl Augenheilkd. 2004;221:762–768.
German.
10. Rombld F, Thiel MJ, Neubauer AS, et al. Evaluation of TGDc-01
compared with Goldmann applanation tonometer. Ophthalmologe.
2005;102:158–162. German.
r
2008 Lippincott Williams & Wilkins
J Glaucoma
Volume 17, Number 1, January/February 2008
Icare Tonometer Compared to Goldmann Tonometry
11. Iester M, Mermoud A, Achache F, et al. New TonompenXL:
comparison with the Goldmann tonometer. Eye. 2001;15:52–58.
12. Armstrong TA. Evaluation of the Tono-Pen and the Pulsair
tonometers. Am J Ophthalmol. 1990;109:716–720.
13. Midelfart A, Wigers A. Clinical comparison of the ProTon and
Tono-Pen tonometers with the Goldmann applanation tonometer.
Br J Ophthalmol. 1994;78:895–898.
14. Bafa M, Lambrinakis I, Dayan M, et al. Clinical comparison of the
measurement of the IOP with the ocular blood flow tonometer, the
Tonopen XL and the Goldmann applanation tonometer. Acta
Ophthalmol Scand. 2001;79:15–18.
15. Van der Jagt LH, Jansonius NM. Three portable tonometers, the
TGDc-01, the ICARE and the Tonopen XL, compared with each
other and with Goldmann applanation tonometry. Ophthal Physiol
Opt. 2005;25:429–435.
16. Fernandes P, Diaz-Rey JA, Queiros A, et al. Comparison of the
ICare rebound tonometer with the Goldmann tonometer in a
normal population. Ophthalmic Physiol Opt. 2005;25:436–440.
17. Martinez-de-la-Casa JM, Garcia-Feijoo J, Castillo A, et al.
Reproducibility and clinical evaluation of rebound tonometry.
Invest Ophthalmol Vis Sci. 2005;46:4578–4580.
18. Davies LN, Bartlett H, Mallen EA, et al. Clinical evaluation of rebound tonometer. Acta Ophthalmol Scand. 2006;84:
206–209.
19. Nakamura M, Darhad U, Tatsumi Y, et al. Agreement of rebound
tonometer in measuring intraocular pressure with three types of
applanation tonometers. Am J Ophthalmol. 2006;142:332–334.
20. Brusini P, Salvetat ML, Zeppieri M, et al. Comparison of ICare
tonometer with Goldmann applanation tonometer in glaucoma
patients. J Glaucoma. 2006;15:213–217.
r
2008 Lippincott Williams & Wilkins
47