High intraocular pressure (IOP) values are the most important risk factors for developing glaucomatous optic neuropathy (GON). Furthermore, IOP lowering is essential to prevent optic nerve damage or visual field deterioration     .
Therefore, accurate assessment of IOP is essential in the routinely clinical practice.
The Goldmann applanation tonometer (GAT) is still considered the gold standard instrument for IOP measurement  . This device infers IOP from the force required to flatten a predetermined area of the central cornea. It is already known GAT reading is influenced by modifications of the structure, the curvature and the central thickness of the cornea  .
An ideal tonometer should have an excellent accuracy in intraocular pressure measurement in both normal and pathological conditions (e.g. corneal irregularity and post-keratoplasty). In addition, it should be minimally invasive, easily usable for all patients, including children and uncooperative patients.
Several new instruments have been introduced in order to overcome some of the GAT limitations such as the ICare® rebound tonometer (RT). The latter is portable, ease of use and even an operator with little experience can use it with confidence, since there is no need for any anaesthetic or fluorescein   .
RT device works on the principle of a rebound tonometry using an induction-based coil system to measure the motion parameters of a bouncing probe after colliding with the cornea   . An advanced algorithm combined with a special software analyzes deceleration and the contact time of the probe while it touches the cornea.
The primary aim of this study was to evaluate the effect of cataract surgery on the IOP measurements performed with the rebound tonometer, as well as to assess the agreement between GAT and RT. Moreover, the effect of central corneal thickness (CCT) on IOP values was evaluated.
Finally, the influence of GAT in subsequent RT measurements was studied.
2. Materials and Methods
This was a retrospective and comparative study that adhered to the tenets of the Declaration of Helsinki. The study was approved by our Institutional Review Board of the Department of Medicine, Surgery and Health Sciences of Trieste and Trieste University Hospital, Italy.
We conducted a retrospective analysis of 472 medical records of 472 patients (60% female, 40% male) who were recruited at the Eye Clinic of Trieste, AOUTS Ospedali Riuniti of Trieste between December 2015 and December 2016. Their IOP was checked with the two commercially available devices: Goldmann applanation tonometer (AT900, Haag-Streit, Koeniz, Switzerland) and ICare® rebound tonometer (TAO1i-Tiolat Oy, Helsinki, Finland), as our routinely clinical practice. Data of one eye per subject was enrolled in the study, choosing randomly between the right and the left eye.
Mean age of all subjects was 75 years (range: 53 - 92) (Table 1).
Inclusion criteria were:
- age < 95 years
- absence of anterior segment surgery or refractive surgery
- absence of corneal pathology
- absence of high refractive errors (myopia > 5D, astigmatism > 2D)
Exclusion criteria were:
- corneal opacity
- glaucomatous and ocular hypertensive eyes
- history of previous retinal surgery
- history of previous ocular trauma
Intraocular pressure was measured with the two different tonometers: RT and GAT.
The tonometers used were regularly calibrated according to manufacturer’s instructions. All measurements were performed by the same experienced examiner.
IOP readings were obtained as follows:
1) IOP was first measured by ICare rebound tonometry (RTa). This device provides an automatically average (d) IOP value, visible on the display after six consecutive measurements;
2) GAT was performed five minutes later, preceded by topical anaesthesia with 0.4% oxybuprocaine hydrochloride and the application of fluorescein paper - Haag-Streit, Koeniz, Switzerland. Two consecutive readings were obtained and a mean IOP value was calculated.
3) After 5 minutes GAT performing, a further measurement with the ICare (RTb = Rta + 10 min) was taken.
For each patient CCT was measured by the same examiner. The CCT value was obtained for all subjects as an average of 5 consecutive readings taken with the Corneo-Gage Plus pachymeter (Sonogage, Cleveland, OH).
Table 1. Patient demographics.
In 106 eyes of 106 enrolled patients, who were to undergo cataract surgery the following day, IOP was measured by means of RT and GAT before (RT1 and GAT1) and after (RT2 and GAT2) the surgical procedure using the same scheme of tonometry as above described.
Patients with cataract surgery-related complications were excluded from the study.
Bland-Altman analysis was used to assess the clinical agreement of the IOP measurements obtained with the two different tonometers. Pearson correlation coefficient was determined to evaluate the relationship between CCT and IOP. The differences between the two tonometers in terms of IOP readings were compared with a paired t-test. The level of significance for each test was set at p < 0.05.
Shapiro-Wilk’s test was performed to evaluate the departures from normality distribution for each variable.
Statistical analysis was performed using the statistical package SPSS Statistics v 17.0 software (SPSS Inc., Chicago, Illinois, USA).
Bland-Altman plot analysis (Figure 2) revealed that the mean difference (±1 Standard Deviation, SD) between RT and GAT measurements was −0.35 ± 2.82 mmHg.
The 95% limits of agreement between RT and GAT were −5.88 and 5.18 mmHg.
Although a good correlation between the values measured using the two different devices was found (Pearson correlation r = 0.72, r2 = 0.51, p < 0.001), the difference between RT and GAT was statistically significant (p < 0.05, t-test). RT IOP values > 15 mmHg were overestimated, while RT IOP values < 15 mmHg were underestimated (Figure 3).
The difference between RT1 and RT2, GAT1 and GAT2, RT1 and GAT1 was not statistically significant, while the difference between RT2 and GAT2 was statistically significant (p < 0.04) (Figure 4).
The results of the IOP mean values of RT1, RT2, GAT1 and GAT2 (±SD) are described in Table 3.
Table 2. Mean and Standard deviation of central corneal thickness and intraocular pressure using rebound and Goldmann applanation tonometers.
a. Abbreviations: CCT = central corneal thickness; RT: rebound tonometer; GAT: Goldmann applanation tonometer.
Figure 1. Box plot. Distribution of central corneal thickness and intraocular pressure values. a. Abbreviations: CCT = central corneal thickness; RT: rebound tonometer; GAT: Goldmann applanation tonometer.
Figure 2. Bland-Altman plot: mean IOP[(RT + GAT)/2] versus their difference (RT- GAT). a. The mean difference ± SD was −0.35 ± 2.82 mmHg. b. The 95% limits of agreement between RT and GAT were −5.88 ~ 5.18 mmHg.
Figure 3. The graph shows the linear correlation of values obtained with RT and GAT. a. A good correlation (r = 0.72) between RT and GAT was found. b. The linear function is GAT = 0.60 + 6.10 × RT (red line), RT = 0.85 × GAT +1.80 (black line).
Figure 4. Comparison between IOP readings using RT and GAT before and after cataract surgery. a. t-test. b. Top left: RT1 vs RT2 (p = 0.16; RT2 = RT1 × 0.26 + 10.19; RT1 = RT2 × 0.18 + 12.17; correlation coefficient = 0.22); top right: GAT1 vs GAT2 (p = 0.4; GAT2 = GAT1 × 0.27 +10.5; GAT1 = GAT2 × 0.21 + 11.87; correlation coefficient = 0.24); bottom left: RT1 vs GAT1 (p = 0.5; GAT1 = 0.48 × RT1 + 7.81; correlation coefficient = 0.62); bottom right: RT2 vs GAT2 (p = 0.034, GAT2 =0.59 × RT2 + 6.38; correlation coefficient = 0.78).
Table 3. Mean and Standard deviation of intraocular pressure using rebound and Goldmann applanation tonometers before and after cataract surgery.
a. Intraocular pressure mean values of rebound tonometer before cataract surgery (RT1), rebound tonometer after cataract surgery (RT2), Goldmann applanation tonometer before cataract surgery (GAT1) and Goldmann applanation tonometer after cataract surgery (GAT2).
Moreover, a low correlation between RT reading and CCT was found (Pearson correlation r = 0.11), while the relationship between GAT reading and CCT had a very weak correlation coefficient (Pearson correlation r = 0.045). Every 100 µm of CCT an increase of 0.97 mmHg and 0.33 mmHg was found for the RT and GAT respectively (Figure 5).
The difference between RTa and RTb was not statistically significant (p > 0.13, t-test), whereas the difference between RTa and GAT and RTb and GAT was statistically significant (p < 0.001) and their correlation coefficients were 0.72 and 0.78 respectively (Figure 6).
Figure 5. Correlation of rebound tonometer (RT) and Goldmann applanation tonometer (GAT) with central corneal thickness (CCT). a. Pearson correlation.
Figure 6. Comparison between series of IOP measurements using GAT and RT. a. t-test b. In the first graph, top left: RTa and RTb measurements are compared (RTb = 0.81 × RTa + 2.52; r2 = 0.67; correlation coefficient = 0.82; p > 0.13). Top right: RTa vs GAT (GAT = 0.60 × RTa + 6.10; r2 = 0.51; correlation coefficient = 0.72; p < 0.001). Bottom left: RTb vs GAT (GAT = 0.66 × RTb + 5.34; r2 = 0.61; correlation coefficient = 0.78; p < 0.001).
Nowadays, RT has become an excellent alternative of GAT in clinical practice for IOP reading due to its accuracy and ease of use. Kontiola et al.  were the first who studied the Icare tonometer performance in animals suggesting that RT is a reliable and non-invasive instrument for IOP reading.
Nevertheless, the comparison between RT and GAT performed in the Fernandes’s study  showed a significant difference between the two devices because of the IOP overestimation with the ICare. Therefore, IOP values obtained with the rebound tonometer should be re-assessed or be referred to GAT evaluation due to the high number of false positives found with the RT.
The aim of the present study was to compare rebound with applanation tonometers and to identify possible elements influencing the measurements of IOP, such as cataract surgery, central corneal thickness and corneal applanation after previous applanation tonometry.
Literature provides conflicting results concerning the comparison of the two instruments, related to the different characteristics of the subjects enrolled, the time elapsing between the several tonometries and the number of measurements for the final average value considered  -  .
In the present study a good agreement of IOP readings between the two different tonometers was detected, as it has widely reported in literature    .
These findings suggested RT may be a valid alternative to GAT, especially in uncooperative patients, children or wheelchair subjects who may have difficulties in positioning correctly on the slit lamp. Nevertheless, our results showed a statistically significant difference between RT and GAT IOP measurements, with a RT slight overestimation of IOP values over 15 mmHg and an underestimation of those below 15 mmHg.
The overestimation of the rebound tonometer may be related to a systematic proportional error in the use of RT, since the difference between the two instruments increases as IOP rises.
Previous studies   have already suggested a similar trend of RT.
Concerning the comparison between RT before and after GAT measurement, the results demonstrated that the applanation tonometry does not affect subsequent measurements with rebound tonometry.
Although a weak correlation between CCT and the two different tonometers was found, CCT is one of the parameters to be considered in tonometry, especially when it is performed using RT. The latter is more affected by CCT than GAT, as already reported by other authors     .
Our study mainly focused on the effects of cataract surgery on intraocular pressure measurement and on the comparison between rebound and applanation tonometry pre and postoperatively. In literature  changes in intraocular pressure after cataract surgery have already been described, in particular the IOP-lowering effect.
In addition, cataract surgery procedure may change the biomechanical properties     of the cornea, thus influencing the postoperative IOP reading as well as it may increase the difference between IOP values obtained with applanation and rebound tonometers.
In our work the rebound tonometer appears to be not so reliable since it differed significantly from GAT IOP values. Caution should be taken on the interpretation of the values obtained using the two different instruments, especially in cataract surgery.
Our study underlines that the two tools are quite concordant but they cannot be considered interchangeable in the clinical practice, in particular in the postsurgical follow up of cataract patients.
Nevertheless, further works will be necessary to better understand and to confirm the same conclusions related to rebound tonometry accuracy in cataract surgery.