LAUSR

To the Editor: I am writing to express my concerns about the conclusions reached by the authors in a recent paper that appeared online ahead of print in Journal of Glaucoma: Mathematical discrepancies of the Tono-Pen applanation Tonometer.1 The stated purpose of the study was to determine if Tono-Pen tonometers use simple average and coefficient of variation (CV) algorithms to calculate intraocular pressure (IOP). The answer is: no, Tono-Pens do not use simple averaging to compute the displayed IOP and CV values. The device uses “smart” processing of the data to ensure that displayed results are highly correlated with Goldmann tonometry values. Therefore one should not expect that the individual values, which appear on the screen with each “tap” of the cornea, are all used in the final displayed IOP. A simple average, calculated from each individual measurement, will not agree with the final displayed IOP in many cases because certain measurements (outliers, etc.) are disregarded as part of the Tono-Pen averaging process. The authors calculated their own IOP and CV values using simple averaging of the individual measurement results and then determined that “Calculated average IOP and CV differ significantly from values displayed by the instrument, especially at higher IOPs. A difference of >5mm Hg between calculated and displayed average IOP seen in 6% of cases may impact clinical judgment. Displayed CV<5% does not correlate with accurate IOP measurement based on individual results.” The assertion that Tono-Pen “does not correlate with accurate IOP” is unfounded. The authors did not compare the Tono-Pen results to another method of tonometry. Nor did they perform penetrating manometry in this study. So there is no means by which to define what “accurate IOP” is. The fact that the Tono-Pen XL correlates well with Goldmann applanation tonometry over a wide range of IOP values has been established through dozens of publications over the past 30+ years and should serve as evidence of the reliability of TonoPens averaging technique.2–5 It is also concerning to me that the authors mention various models of TonoPen (XL, Vet, and AVIA) but do not differentiate how these operate. The AVIA is a newer model that uses a significantly different methodology in its calculation of IOP values. Lumping the XL and AVIA together, and then arriving at a negative conclusion regarding the accuracy of the IOP values produced by the Tono-Pen XL, indirectly condemns the AVIA model as well. The authors state that they attempted to contact Reichert via mail and phone regarding Tono-Pen’s averaging technique. It is probable that customer service or technical support personnel at Reichert responded, as they are trained to, that this is confidential information. It is unfortunate that the attempted communication never reached my desk that I am aware of. As the Product Manager for our tonometry devices, I would have been willing to share certain information to assist with the authors’ investigation. In summary, the authors assert that since the displayed final values on the Tono-Pen do not agree with their calculated “simple averages” that the displayed results are “inaccurate” in some or many cases. The authors should have simply concluded that Tono-Pen’s displayed IOP value does not always agree with a simple mathematical average of the individually obtained values. It is not possible or reasonable to draw conclusions about accuracy from this observation.