Repeatability and agreement of intraocular pressure measurement among three tonometers.

Abstract

Background: Elevated intraocular pressure (IOP) is one of the causes of irreversible optic nerve head damage and visual field loss. It is often measured with applanation tonometers but the use of rebound tonometry devices has been proposed as an alternative to assess IOP. Rebound tonometers have also been proposed as a method for patients to measure their own intraocular pressure (that is, self-tonometry). The purpose of this study was to determine the intrasession repeatability and the agreement of the IOP measurement with two rebound measuring principle tonometers, ICare ic100 and ICare Home with Perkins tonometer.

Methods: This study involved 27 healthy volunteers (18 to 30 years old). We performed three consecutive IOP measurements with ICare Home, ICare ic100, and Perkins. The means of the three measurements from each device were calculated. Repeatability and agreement were defined according to the British Standards Institute and the International Organization for Standardization. The agreement was assessed using the method described by Bland and Altman, where 95 per cent of the differences or limits of agreement were between 1.96 standard deviations of the mean difference.

Results: All tonometers showed close measurements (Perkins 15.34 ± 3.45 mmHg, range 10.00-24.00; ICare ic100 15.40 ± 4.06 mmHg, range 9.67-23.33; and ICare Home 14.22 ± 4.72 mmHg, range 7.33-24.00). The co-efficient of variation (CV) and within-subject standard deviation (Sw) was low for ICare ic100 and Perkins (close to 6.30 per cent and one) with higher values for ICare Home (CV = 9.55% and Sw = 1.33). The intraclass correlation co-efficient showed values higher than 0.96 for all tonometers. The difference between both rebound tonometers was statistically significant (p = 0.03).

Conclusion: The ICare ic100 tonometer provides repeatable IOP measurements close to the measurements of the Perkins IOP (good agreement); however, ICare Home provides less repeatable values, showing worse agreement with the Perkins tonometer in healthy subjects.