Tonometry in keratoconic eyes before and after riboflavin/UVA corneal collagen crosslinking using three different tonometers

    loading  Checking for direct PDF access through Ovid

Abstract

Purpose.

To evaluate the Goldmann applanation tonometer (GAT), the Pascal dynamic contour tonometer (PDCT), and the ocular response analyzer (ORA) tonometer in measuring intraocular pressure (IOP) in keratoconic eyes before and after riboflavin/ultraviolet A corneal collagen crosslinking (CXL), to assess agreement among devices and to analyze the impact of some ocular parameters on their measurements.

Methods.

Fifty keratoconic eyes were included. Intraocular pressure was measured with GAT, PDCT, and ORA before and after CXL. Fifty nonkeratoconic eyes served as controls. Device agreements were calculated by Bland-Altman analysis. The effect of some ocular characteristics on IOP measurement differences between tonometers was determined.

Results.

Between the 2 groups, there were statistically significant differences in all examined parameters. Preoperatively, in both groups a statistically significant difference was found in IOP measurements among devices (p<0.05). Bland-Altman analysis showed a bias among devices. On average, PDCT overread GAT and ORA. The IOP measurement differences were better predicted by corneal resistance factor. Postoperatively, in keratoconus eyes, there was no statistically significant difference in IOP measurements among the 3 tonometers (p>0.05). The IOP readings with all tonometers after treatment were higher than those obtained preoperatively; however, not to a statistically significant level, with the exception of PDCT and ORA readings at the first month postoperatively. Corneal resistance factor measurements have no significant change after CXL.

Conclusions.

Pascal dynamic contour tonometer could provide more consistent and closer to the true IOP readings than GAT and ORA in healthy eyes with corneal thickness outside the 520-550 μm range, in keratoconus patients and after CXL. Corneal resistance factor was associated significantly with agreement among devices.

Related Topics

    loading  Loading Related Articles