A number of comparative analyses of tonometers have been conducted on adult eyes, both normal and glaucomatous or hypertensive [9–12]. Fewer studies have conducted similar comparative analyses with the eyes of children. The ICare tonometer was developed for use on patients who present with the sort of measurement problems that are often associated with children. The instrument is easy to handle, portable, light weight, and does not require anaesthesia or fluorescein. Some of these advantages are also pertinent to the NCT. Comparison of the ICare with an NCT on children from 6 months to 15 years of age showed comparability between instruments but also demonstrated the greater tolerance for measurements taken with the ICare tonometer compared with the NCT, on younger children . High levels of tolerance for, as well as reproducibility of, IOP measurements were also found with the ICare tonometer, on a group of healthy infants  and schoolchildren . Ease of use and tolerance, however, need to be balanced against effectiveness and accuracy. For many years, applanation tonometry was considered to set a standard against which other instruments were measured. However, recent reviews that have compared a number of instruments suggest that depending on the population and the application, different instruments may be preferred [15, 16].
A previous study of 460 children aged from 0 to 16 years, found that IOP increases from birth to around 7–8 years (slightly earlier for males than for females) and then stabilises . The results of this study do not show any correlation or trends with age but the population used in this study was largely beyond the age range over which the rapid increase in IOP was observed .
The findings indicate that radius of curvature is correlated with IOP only for the ICare tonometer; this applies for the entire cohort and for the subgroup of subjects with thicker corneae. The only other comparable study, on Turkish schoolchildren, aged between 7 and 12 years, found no correlation between IOP and radius of curvature . Significant differences between measurements obtained with the ICare tonometer and GAT that exist for the entire data set are evident for subsets of both lower and higher radii of curvature ranges. Given that only results obtained using the ICare tonometer showed a statistically significant correlation with radius of curvature and that the most significant differences between tonometers for the whole data set are evident between the ICare tonometer and GAT, it is not surprising that measurements vary significantly between these two instruments in cohorts with lower and higher radii of curvature ranges. Previous findings on adult eyes have found radius of curvature to influence readings with GAT  and with ICare (for CCT > 556 μm) .
Correlation with CCT and IOP was found for all tonometers used and CCT was found to affect correlations between the IOP values obtained with the different instruments. These correlations applied whether or not a correction factor for the GAT was used. This correction factor, derived for an adult eye , may not be appropriate for the eyes of children.
It should be noted that the extent of the effect of CCT on IOP measurements is not clear even for the adult eye. Originally, Ehlers and colleagues proposed a correction factor for CCT variations based on finding a significant correlation between CCT and IOP . However, subsequent studies have not been consistent in their findings about how, and indeed whether at all, CCT should be corrected. Whilst a number of studies have found variations in IOP with CCT [19–23], correction factors have varied [14, 23] and the use of any correction factor for the healthy adult cornea has been questioned [24, 25]. Paucity in the understanding of the rheological properties of the cornea prevents definitive conclusions from being made about the effect of CCT on IOP measurements in the adult eye.
This notwithstanding, the findings in this study indicate that the thickness of the cornea in children can have an effect on correlation of IOP measurements between instruments. The ICare tonometer and the GAT gave readings that were significantly different statistically regardless of CCT; the NCT gave comparable readings with the ICare tonometer for the subgroup with thicker cornea and for the group as a whole and was comparable with the GAT for the subgroup of CCT < 563 μm.
This is the first study of IOP measurements on healthy Polish schoolchildren. It contributes to the literature on IOP in the young eye and provides a comparative study to those conducted on other ethnic groups. Findings from this study show that average CCT values were higher than those found in controls aged between 5-17 years from the USA (mean CCT = 555 ± 37 μm) . This included Caucasian and African ethnicities. For the Caucasian (‘white’) group alone, the average CCT was 564 ± 28 μm, which is very close to the mean value found in this study (563 ± 30 μm). The IOP for controls (Caucasian and African) was found to be 14.9 ± 2.7 mmHg for the Goldmann and 15.1 ± 2.4 mmHg for the Tono-Pen . This compares to 14.7 ± 2.9 mmHg (GAT) in this paper. A study of 460 Italian subjects, aged 0 to 16 years of age, (with 282 subjects aged 5–16 years), showed a mean IOP of 13.9 ± 2.3 mmHg and 14.88 ± 2.39 mmHg for 10–16 year old males and females respectively using an NCT. These values are slightly lower than the IOP values found with the NCT in this study (15.9 ± 3.5 mm Hg, males and females combined) but are within the same range. No statistically significant differences between males and females were found in this study.
Previous investigations on 165 Turkish schoolchildren, aged 7–12 years, showed mean IOP values of 16.81 ± 3.1 mmHg using rebound tonometry which is very close to the 16.9 ± 3.4 mmHg found for the ICare in this study. IOP measurements on Japanese subjects aged from 6 months to 15 years of age reported mean IOP values of 15.9 ± 2.3 mmHg and 15.1 ± 2.6 mmHg in right and left eyes of 130 subjects respectively for the NCT with 15.1 ± 2.6 mm Hg and 13.9 ± 2.9 mmHg in right and left eyes respectively of 160 subjects for the ICare tonometer . No explanation was given for the greater difference between right and left eyes using the latter instrument but the authors indicate that ICare is the more suitable instrument for children under 6 years of age .
A study that uses young subjects, notwithstanding the steps taken to explain the procedure and to take carefully controlled measurements, can be limited by the extent of subject participation and parental involvement. Collaboration of subjects, such as children, is greatly improved if the subjects feel comfortable in the environment where measurements are taken and if there is no or minimal parental pressure and/or anxiety. The University clinic where the study was conducted is one which routinely deals with large numbers of young subjects. It is purposely equipped to create an appropriate ambience and a relaxed atmosphere for these age groups. In addition, the subjects who participated in this study were all familiar with the clinic, parents/guardians had confidence in the investigators rendering high both subject and parental levels of cooperation.
The differences in IOP measurements from the three tonometers, presented in pairs (Figure 4), varied, for most measurements taken, by up to 2 mmHg. However, from a fifth to a quarter of the measurements varied by ≥ 4 mmHg. As there is no gold standard for tonometry measurement in children it is not possible to say which measurement can be taken as the most reliable. Bland-Altman plots, which are commonly used to check agreement between methods, often depend on one technique being the standard against which the other is measured . Given the effect of CCT on IOP measurements and the paucity of knowledge about the growing eye or how to correct or adjust for CCT in such cases, it is prudent to avoid reliance on any one method. Many more studies on children from different cohorts are required before consistent trends can be established and expected age-related norms derived.