Update on Tonometry Falls Short

Glaucoma Today published an article titled "An Update on Tonometry". It's interesting reading, and I highly recommend it. (In fact, I'm mirroring the pdf version of the article here.) However, the article leaves out some very important tonometers as well as critical information about maintaining the accuracy of Goldman tonometers. I address some of these points below.

First, the newest non-contact tonometers have very good clinical accuracy. An instrument like the Reichert AT555 should have been reviewed for the main article rather than relegated to the table. These modern NCTs need not be relegated solely to screening applications as the article states. The latest study (1) that I am aware of on this topic concludes that, "IOP measurements using the NCT and OBF-pneumotonometer closely agree with those of GAT even in high IOP ranges." I have seen several other studies that reached the same conclusion.

In fact, Bernard Grolman, the inventor of the NCT, notes that comparing one Goldman Applanation Tonometer (GAT) to another GAT has been shown to produce a larger discrepancy than obtained by comparing an NCT and a GAT. Specifically, In studies where the measurement protocols were GAT1 > NCT > GAT2, it can be seen that NCT vs. GAT1 and NCT vs. GAT2 are each in closer agreement than GAT1 vs. GAT2, according to Grolman.

Furthermore, in practice the majority of Goldman tonometers are out of calibration (2). GATs need to be calibrated at least every month -- and ideally calibration should be checked before each session. (The abstract of the study from which I draw this information is reproduced below.) The newest (Reichert) non-contact tonometers typically need to be calibrated much less frequently (possibly only once every 3 years). In practice, this could be an important accuracy advantage in favor of the NCT.

Another new instrument the Glaucoma Today article neglected (in the article body) is the Reichert Ocular Response Analyzer. The author, Dr. Sarkisian, told me in a private email conversation that he agrees that the Ocular Response Analyzer (ORA) is an interesting new development in tonometry. The ORA utilizes a "dynamic bi-directional applanation process to measure the biomechanical properties of the cornea and the intraocular pressure of the eye." The morphological signal produced by the ORA certainly contains at least as much clinically significant information (although of a different type) as does the ocular pulse amplitude available from the Pascal Dynamic Contour Tonometer mentioned in the article. They are different instruments and they will ultimately be used to reveal different information, but I believe there is no question that both produce more clinical information than the GAT. Both of these new instruments should be of interest to clinicians, and in my opinion the ORA is the more promising of the two.

In terms of research, the NCT has several advantages. The potential IOP altering effects of corneal anesthesia are avoided. Operator error is much less of a problem with modern NCTs. NCTs eliminate the potential risk of slow virus disease transmission. And finally, a researcher can collect a much larger volume of data with an NCT. In my experience it is possible to collect 100 or more IOP measurements per day per subject without any discomfort. Doing this with a GAT would just not be possible or safe. The larger research data sets enabled by an NCT can lead to new discoveries and insights. This makes NCTs a valuable research tool that the GAT cannot equal.

References:

1. Gupta V, Sony P, Agarwal HC, Sihota R, Sharma A. Inter-instrument agreement and influence of central corneal thickness on measurements with Goldmann, pneumotonometer and noncontact tonometer in glaucomatous eyes. Indian J Ophthalmol [serial online] 2006 [cited 2006 Dec 13];54:261-265. Available from: http://www.ijo.in/article.asp?issn=0301-4738;year=2006;volume=54;issue=4;spage=261;epage=265;aulast=Gupta

2. Sandhu, Sukhpal Singh MRCOphth; Chattopadhyay, Sandip FRCS Edinburgh; Birch, Michael Kevin FRCOphth; Ray-Chaudhuri, Neeta FRCOphth. Frequency of Goldmann Applanation Tonometer Calibration Error Checks. Journal of Glaucoma. 14(3):215-218, June 2005. http://www.evidence-based-ophthalmology.com/pt/re/ebeye/abstract.00061198-200506000-00009.htm;jsessionid=F1hQBG1wzQPjMNml58VmYtxQQQWspf4dn2Q2YPy2QjzxwGvZkppr!-2066534216!-949856144!8091!-1?nav=forward&basedoc=00132578-200301000-00007&article=9

Frequency of Goldmann Applanation Tonometer Calibration Error Checks.

• Month zero: only 5.9% (2 of 34) tonometers fell within the manufacturer's recommended calibration range of +/-0.5 mm Hg.
• Month one: only 10.3% (3 of 29) tonometers fell within the manufacturer's recommended calibration range of +/-0.5 mm Hg.
• Month four: zero percent (0 of 33) tonometers fell within the manufacturer's recommended calibration range of +/-0.5 mm Hg.

• Month zero: 14 of 34 (41.2%) tonometers had calibration errors greater than +/-2.5 mm Hg.
• Month one: 10 of 29 (34.5%) tonometers had calibration errors greater than +/-2.5 mm Hg.
• Month four: 17 of 33 (51.5%)  tonometers had calibration errors greater than +/-2.5 mm Hg.