Comparison of intraocular pressure measurements between Icare PRO Tonometer , Goldmann Applanation Tonometer and non-contact tonometer in healthy and glaucomatous eyes

Received for publication 04/06/2018 Accepted for publication 03/08/2018. The authors declare no conflicts of interests. Comparação da medição da pressão intraocular entre o tonômetro Icare PRO, o tonômetro de aplanação de Goldmann e o tonômetro de não-contacto em olhos saudáveis e olhos glaucomatosos Comparison of intraocular pressure measurements between Icare PRO Tonometer, Goldmann Applanation Tonometer and non-contact tonometer in healthy and glaucomatous eyes


G
laucoma, a chronic and neurodegenerative optic neuropathy, is a main global cause of irreversible blindness.The disease management is directed to reduce intraocular pressure (IOP) by pharmacological or surgical means.Despite the serial monitoring and the different therapeutic options, some patients maintain a progressive visual field loss. (1,2) outine evaluation of IOP is essential in the management of patients with glaucoma, since its reduction is the only proven method of slowing down disease progression. (3)To obtain accurate measurements of IOP, various diagnostic devices are available in clinical practice.9)(10)(11)(12)(13) Among other alternatives, the Icare PRO tonometer (IPT) provides a new type of portable reading using the principle of impact or rebound to measure IOP.The calculation is made by parametric variation of the motion after a magnetic probe reaches the cornea.This probe uses a motion-and-impact detection solenoid that, at the time of collision, induces voltage and tension, allowing measurement of IOP. (6,7) he Non-Contact tonometer (NCT) is another form of flattening tonometry that uses a calibrated column of compressed air to flatten the cornea. (8,14,15) T significance of central corneal thickness (CCT) in clinical diagnosis and management of glaucoma is well recognized, however its importance remains a matter of debate.18)(19)(20) Another factor that may have an impact on the value recorded by tonometry, and which is currently the focus of much debate, is the position in which the measure is performed.Recently, several studies have shown that IOP is higher in the supine position compared to measurements in the sitting position, so these changes have been associated with an increase in nocturnal IOP and could be related to the progression of glaucoma. (7,10,11,21) Th is the first study that simultaneously compares the validity of IPT with respect to NCT and GAT in the measurement of IOP in healthy eyes and in eyes with diagnosis of primary open angle glaucoma (POAG), and that also compares its measures in different groups of CCT, evaluating its applicability in sitting and dorsal decubitus positions.

Design of the study and Population Study
The study was approved by the Ethics Committee from Hospital de Braga and School of Medicine -University of Minho.Signed informed consent was obtained from all patients prior to data collection.
This study follows a prospective and comparative design.The study was performed in Hospital of Braga and data was collected from individuals that attended the Ophthalmology and Glaucoma Department, from June to November 2017, and filled the inclusion criteria after ophthalmologic observation.The study population was divided into two different groups: a healthy adult population group (Healthy Group) and an adult population group with diagnosis of POAG (POAG Group).In the Healthy Group, we included adults with 18 years or more, while the POAG Group included adults with 18 years or more with diagnosis of unilateral or bilateral POAG.The following criteria were considered for POAG diagnosis: typical optic disc changes characterized as focal or diffuse thinning of the retinal nerve fibre layer with consistent glaucomatous visual field defects, and open angle on gonioscopy.Glaucomatous visual field defects will be confirmed if 2 of the following 3 conditions are met: presence of a cluster of 3 points on a pattern deviation probability plot with P < 5%, 1 of which had P <1%; a pattern standard deviation with P<5%; or a glaucoma hemifield test result outside normal limits.For both groups, we considered the following exclusion criteria: any pathological change detected in the anterior or posterior segment; family history of glaucoma (for the Healthy Group); previous history of trauma or eye surgery; eyelid/corneal changes that made measurements impossible to perform; and, no collaboration during measurements.
Prior to IOP measurement, all elected individuals that came to Ophthalmology appointment were submitted to an ophthalmologic observation by an Ophthalmologist, with baseline assessment and routine procedures performed, including best-corrected visual acuity evaluation, slit-lamp examination, gonioscopy, and fundus biomicroscopy with a 90-diopter lens.

CCT measurements
CCT values were obtained by ultrasonic pachymetry (Alcon ® OcuScan RxP) before the IOP measurements.The pachymeter probe was placed on the center of the cornea, and the mean of 3 readings was calculated for each eye.

IOP measurements
The IOP values were obtained by the three tonometers (IPT, GAT and NCT), and the order of execution was randomly selected, with a 15-minute time interval between readings.The IOP measurements were conducted by independent observers.Each observer was responsible for using a single device.NCT (Topcon CT-80 Topcon corporation, Tokyo, Japan) automatically recorded three IOP readings, with their average per eye being recorded for the study.GAT (Haag Streit, Koeniz, Switzerland) values were recorded after 3 consecutive readings and a mean value was calculated for each eye observed.IPT (Tiolat Oy, Helsinki, Finland) was positioned near the patient's eye with the forehead being used as a base support.IOP measurements were taken with the tip of the probe maintained at a distance of approximately 3 to 7 mm from the center of the cornea, according to the manufacturer's instructions.The mean of six consecutive sets of acceptable final measurements was used for subsequent analyses.The measurements with IPT were made firstly in sitting position and after 15 minutes in dorsal decubitus.

Statistical analysis
Data analysis was performed with Statistical Practice for the Social Sciences (SPSS, version 23, IBM Corporation).Exploratory analysis for categorical variables are presented as frequencies and percentages.For quantitative variables, statistical analysis involved measures of descriptive statistics (absolute and relative frequencies, means, standard deviations and inferential statistics), after checking out for distribution's symmetry.The coefficient of flattening (kurtose) and skewness were also analyzed.The level of significance to reject the null hypothesis was set at p<0.05.Paired samples t-test was used for mean comparison between IPT and the other two instruments, including the comparison in each group Rev Bras Oftalmol.2018; 77 (5): 248-54 Comparison of intraocular pressure measurements between Icare PRO Tonometer, Goldmann Applanation Tonometer and non-contact tonometer ... had a higher mean IOP difference between IPT and the NCT than healthy eyes (0.52 and 0.17 mmHg, respectively).Statistical analysis between IPT and NCT revealed an agreement between IPT and NCT readings in healthy [t(92)=0.703,p=0.484, d=0.112] and glaucomatous eyes [t(72)=1.177,p=0.234,d=0.124].The Bland-Altman plots showed an overall agreement between IPT and NCT measurements, in both healthy and POAG eyes (Figure 3).In Figures 3 we can see the plot for healthy and POAG eyes, respectively, performed by the IPT and NCT instruments.The group mean (red line) of the differences was 0.34 mmHg in healthy eyes and 1.05 mmHg in POAG eyes.Bland-Altman analysis revealed no biases between IPT and NCT with 95% limits of agreement of -4.46-4.80 in healthy eyes (Figure 3A) and -6.80-7.85 in glaucomatous eyes (Figure 3B).In table 3 we analyze the potential bias and as B unstandardized coefficient is close to 0 and we have no statistical significant results in both healthy and POAG eyes, we can conclude an inter-device agreement between the two instruments.of CCT and for comparison of IPT measurements in siting position and dorsal decubitus position.Student's t-test for independent samples was used to compare CCT in healthy and POAG eyes.The Bland-Altman analysis was used to test the agreement between the measuring instruments.One-sample t-test was used previously to verify if we could proceed with the plot construction.In the Bland-Altman graphs, the difference between each IOP measurement (e.g.IPT-NCT) was plotted against the mean of the 2 measurements [(IPT-NCT)/2].Linear Regression analysis was used to verify bias.Pearson's correlation coefficient was used to assess correlation between the tonometers for the IOP measurement.

Descriptive Statistics
A total of 168 eyes (74 eyes with diagnosis of POAG and 94 healthy) were included in this study.Demographic characteristics of the sample population are shown in table 1.The sample consisted of 46 (54.8%) males and 38 (45.2%) females, with a mean age of 63.8 (±17.2 years).

Inter-device agreement: Agreement between IPT and GAT
Figure 1 presents the mean comparison between IPT and GAT in healthy and POAG eyes.Overall, we found a statistically significant difference in both healthy eyes [t(92)=10.942,p<0.001, d=0.320], and in eyes with POAG [t(72)=8.337,p<0.001, d=0.124].Thus, there is no inter-device agreement between the two instruments.

Sitting position vs. dorsal decubitus
Healthy eyes registered a mean IOP of 15.96±2.51mmHg and 18.93±2.83mmHg in sitting and dorsal decubitus position, respectively.In healthy patients, measurements of IOP obtained with IPT are significantly higher in dorsal decubitus [t(26)=4.155,p<0.001, cohen's d=1.110].POAG eyes had higher measurements in both postures with IOP mean of 18.12±4.43mmHg and 19.78±4.27mmHg in sitting and dorsal decubitus positions, respectively.In this group, the values are also higher in dorsal decubitus although the difference is not statistically significant [t(13)=1.045,p=0.305, cohen's d=0.382].

Correlation with CCT values
In healthy eyes, the mean of CCT was 531.88±35.45μm.In POAG eyes, the mean registered was 525.86±39.88μm.The CCT is higher in healthy eyes, although the difference is not statistically significant [t(166)=1.102,p=0.313, cohen's d=0.160].The linear regression analysis shows a positive correlation of IOP values measured with the three tonometers and CCT.The correlation coefficients between tonometers and CCT are all statistically significant, positive and moderate or low (Table 4).

Comparison of IPT in different groups of CCT
By grouping the CCT values into distinct categories (<470 μm; 470-550 μm; and ≥550 μm) and comparing IOP means by instrument in healthy and POAG eyes, we verified that for thin

Table 3 Linear regression: analysis of bias
corneas (CCT<470 μm) there are no statistically significant differences between the IPT and the GAT (p=0.518 and p=0.971 for healthy and POAG eyes, respectively).However, we were able to find statistically significant differences between IPT and GAT, in the group of 470-550 μm and ≥550 μm, either in healthy eyes or with POAG (Table 5).The same analysis was applied for the comparison with the NCT and we found statistical significant differences between IPT and NCT in a CCT ranging from 470-550 in healthy eyes (Table 6).

dIscussIon
Tonometry is a fundamental procedure in routine ophthalmologic examination.Accurately measure IOP is essential when diagnosing and managing glaucoma.The IPT has gained much interest recently, due to its practical use, comparable measurements with other forms of tonometry, good reproducibility and a relatively more comfortable experience for the patients since it is minimally invasive and a non-anaesthetic procedure. (12,14,22,23) Th clinical necessity of such tonometer is justified in people who have difficulties in positioning their head on the slit lamp (children, wheelchair, obese) and in cases where eye drops need to be avoided. (14,24,25) Hwever, doubts have arisen whether the IPT would produce comparable results with the GAT and NCT over a wide range of CCT. (18,19,26) ost of the previous comparative studies between instruments have concluded that IPT would produce slightly higher readings than the GAT. (15,16,25,27,28) On he other hand, Brusini et al. found agreement for glaucomatous eyes between IPT and GAT, using as a comparison term the GAT values corrected for the CCT.(5) Additionally, Gao et al. considered IPT a reliable alternative to GAT for patients in a low to moderate IOP range but not in patients with high IOP values.(24) Another study conducted by Tamçelik et al. showed that IPT tends to overestimate in the low GAT-measured IOPs, whereas it underestimates in high GAT-measured IOPs.(8) Our findings indicate that although IPT values cannot be considered similar and interchangeable with IOP measures determined by GAT, the behaviour of this tonometer is more similar to that of the NCT.We found an overall agreement between IPT and NCT measurements, in both healthy and POAG eyes. The interest in developing a tonometer that minimizes the limitations of GAT and can be used in an easy and interchangeable way has increased over the years.Despite the use of NCTs in clinical routine for such purpose, there is a lack of consensus with respect to the correlation observed between measurements obtained using this instrument and GAT.Indeed, while some studies demonstrated sub-optimal correlation values, (16) others demonstrated that NCT present an excellent agreement with GAT measurements, both in normal subjects and patients with glaucoma.(26) The main purpose of this study was to validate the use of IPT, not only comparing its agreement with NCT and GAT but also by assessing how body position could influence its measurements in both healthy and POAG eyes.The mechanisms underlying the changes in IOP induced by alterations of the body posture are not fully understood.Compression or altered geometry of neck vessels, changes in episcleral venous pressure, choroidal vascular congestion, gravity, and shift of body fluid have been suggested to play a role.(9-11, 29, 30) This is the first study which splits sitting and dorsal decubitus position simultaneously in healthy and glaucomatous eyes using IPT to measure IOP.We found higher IOP values in dorsal decubitus in both groups of eyes, although the differences were not statistically significant in POAG eyes.0) According to our results, CCT seems to influence IOP readings with IPT, as it does in GAT and NCT, and should be considered when interpreting measurements in clinical routine.This positive and significant correlation noted between the GAT, NCT and IPT measurements and CCT indicates that as the value of CCT increases, IOP increase with all tonometers in study, being this relation more relevant with NCT measurements.A possible explanation could be the viscoelastic property of the cornea, which is deformed over about eight milliseconds in NCT, resulting in relatively greater stiffness than the GAT, where IOP measurement is relatively static.(16,20,26) This is the first study that introduces the comparative analysis of IPT values with the other two tonometers over different groups of CCT based on its reference values.We concluded that for thin corneas (CCT<470 μm) there are no statistically significant differences between the IPT and the GAT in both groups of eyes.However, this may be due to the reduced sample size, and thus further studies may be required to clarify this issue.
The comparison of IPT with NCT, under the same analysis, showed statistical significant differences between the instruments in CCT values from 470-550 μm in healthy eyes.There was a systematic trend of IPT to give higher readings, showing higher means of IOP over a wide range of CCT and this difference was more notorious in relation with GAT.
The main study limitation is related to the fact that only healthy subjects or individuals with POAG were included.Therefore, our findings may not be applicable to eyes with other forms of glaucoma or ocular hypertension or other ocular abnormalities that may cause an elevation in IOP.Besides, each instrument used in this study was performed by a single observer, which means that three observers were responsible for the measurements and could not switch between them for using other tonometer.As so, although we decreased the risk of influencing measurements according to the previous values observed, we could not control and avoid potential bias induced by the interobserver variability.Another factor that possibly affected IOP overestimation of measures in comparison with the other two tonometers may be the no use of topical anesthesia for IPT measurements and the sequential measurements with the other instruments in a random sequence.Although IPT does not require topical anesthesia, unfamiliarity or anxiety related with the measurements in two different body positions with this tonometer may induce eyelid squeezing and increase IOP values.However, we believe the possibility of such anxiety-related bias was reduced by obtaining the other measurements in a randomized sequence.
In future studies, larger sample sizes could improve our knowledge and allow comparison with other groups (e.g.different ages, patients with previous corneal surgeries or ocular pathologies, or different subtypes of glaucoma).Additionally, more investigation is required to access whether posture-induced IOP changes are related to the onset or progression of glaucoma damage.As so, large population studies and meta-analysis are necessary to validate the IPT and to enable its standardization.
In conclusion, in our analysis we compared three instruments of tonometry, using as main goal the validation of IPT.The comparison with the gold-standard tonometer, the GAT, showed no interchangeable measures and we could not establish an agreement between the instruments.However, we found that measurements taken with IPT are quite in accordance with those obtained by NCT and we demonstrated an inter-device agreement.The measurements differ according to CCT variations, and thus pachymetry should always be taken into consideration.

Figure 2
Figure 2 presents the mean comparison between IPT and NCT in healthy and POAG eyes.In our study, eyes with POAG

Figure 3 :
Figure 3: Bland-Altman plots showing the distribuition of differences in IOP (IPT value minus NCT value, mmHg) (y-axis) and the mean IOP value of tonometers (x-axis) for each eye measured in healthy (n=94) and |POAG eyes (n=74) in (A) and (B), respectively.The red line indicates the mean of the differences (identity line) and the green lines lower and upper bounds of agreement for a 95% confidence interval.