The Proview phosphene tonometer fails to measure ocular pressure accurately in clinical practice

To evaluate the Proview Eye Pressure Monitor as a medical instrument and as a technique for enabling a patient to obtain an accurate measure of his or her intraocular pressure (IOP).

An experimental laboratory evaluation and an independent prospective clinical study to test the reproducibility and accuracy of the Proview technique relative to Goldmann applanation tonometry.

For the laboratory study, we analyzed 3 tonometers, each packaged as a Proview Eye Pressure Monitor by Bausch & Lomb. In the independent prospective experimental study, 137 subjects participated, consisting of healthy volunteers and glaucoma patients.

For laboratory testing, we held each tonometer with a micrometer to assure controlled positioning and pressed its sensing tip against a force meter that produced a calibrated, digital force reading. For clinical testing, we taught subjects (n = 137) to use the Proview technique in accordance with the manufacturer's instructions. Each subject obtained 5 measurements with each of the 5 different Proview devices. A clinician measured the IOP using Goldmann applanation tonometry.

We measured the absolute value, linearity, and repeatability of the force meter readings on the tonometers during the instrument laboratory evaluation. The accuracy was evaluated by comparing the Proview measurements to the Goldmann applanation measurements. Reproducibility of clinical Proview measurements was also measured. All measurements were in mmHg during the clinical evaluation.

Laboratory: There was a linear relationship between the pressures read by the Proview tonometers and known forces. The Proview tonometers read the maximum pressure applied. Clinical: The Proview technique is simple to use because it was comfortable and reproducible, with an average variance of the measurements by the same patient of 3.4 mmHg(2). Other variables besides IOP seem to affect the Proview pressure measurements, as seen in the large scatter in our data, measured by our correlation coefficient of r = 0.41. The sensitivity of the Proview technique to detect patients with high IOP (which we defined as a Goldmann pressure of >/=22 mmHg) is low; the Proview pressure identified only 18% (4/22) of these patients.

The Proview instrument and technique were reproducible. However, the Proview tonometer seems not to be reliable as an indicator of IOP. The sensitivity for detecting high IOP was low in this cohort, and the agreement with Goldmann applanation was poor for some individuals. This brings into question the underlying assumption that a force proportional to the IOP generates phosphenes.