Introduction Alterations in intraocular oxygen levels are important contributors to, or indications of, ocular disease. Polarographic electrodes and fibre-optic sensors (optodes) have been used to measure oxygen and to map the distribution of oxygen in animal models and in human eyes. A recent study reported the use of a commercial electrode to compare oxygen distribution in the vitreous of patients undergoing vitrectomy related to central retinal vein occlusion, macular hole or preretinal membrane. The results of this study were at variance with previous measures of oxygen distribution in the human vitreous using polarographic or optical sensors. To resolve this discrepancy, the present study compared measurements made in vitro or in animal eyes, using the electrode employed in the previous study or a fibre-optic sensor of a different design.

Study design Comparative in vitro and in vivo measurements.

Results In vitro, the two devices reported similar levels of oxygen, although the electrode consistently detected levels above the calculated values. In rabbit eyes, the electrode had a slow response time and was unable to detect oxygen gradients that were readily measured by the smaller optode. When the electrode was inserted into an eye of similar size to the human eye, the reference thermistor measured the temperature outside the eye, not in the vitreous.

Conclusions The design of the electrode used in the previous study makes it unsuitable for measurements of oxygen distribution in the eye.