Identifying narrow angles

In the recent research paper published in the British Journal of Ophthalmology,¹ Dr Friedman and colleagues looked at three non-contact imaging techniques to determine whether or not sequential testing with two devices may provide sufficient sentivity with a low enough false positive rate to enable screening and follow up in a location with limited resources. "Some imaging devices," said Dr Friedman, "such as the IOLMaster, the scanning peripheral anterior chamber depth analyser (SPAC) and anterior segment optical coherence tomography (ASOCT) aid in the detection of narrow angles."

Although these newer devices alone have not been capable of achieving high sensitivity and specificity, it was hoped that a combination of two of these techniques would offer the specificity required in identification of narrow angles, which is vital in the progression of screening programmes.

"Gonioscopy is the current gold standard in detecting narrow angles," asserted Dr Friedman. However, he noted that this technique requires not only a large amount of expertise but also subjective assessment and it involves contact with the cornea. "All this, makes it a suboptimal technique," he said.

"In contrast," he added, "IOLMaster, SPAC and ASOCT assess ocular parameters in a noninvasive, fast and quantitative fashion, and can be used by a trained technician."

To compare the efficacy and sensitivity of the noncontact methods Dr Friedman and team studied 2047 phakic right eyes that underwent a full ophthalmic examination in one visit [including autorefraction, Goldmann applanation tonometry, IOLMaster (Carl Zeiss, Jena, Germany), SPAC (Takagi, Nagano, Japan) and ASOCT (Visante, Carl Zeiss, Dublin, California, USA), nonindentation and indentation gonioscopy]. A single trained ophthalmologist performed the gonioscopy, which was standardized using another glaucoma specialist.

Sequential testing

Performance of sequential testing was assessed for all pairwise combinations. If the first device gave a negative result for narrow angles, the eye was classified as having an open angle and no further testing was performed. For a positive result with the first device the eye would be further evaluated. Only an eye that had a positive results from both devices was considered to have a narrow angle.

"Theoretically, the order of sequential testing does not affect the overall diagnostic performance," said Dr Friedman. "For logistical reasons, however, it makes sense to start with SPAC because it is cheaper, easier to operate, faster and more people can be successfully tested with it."

The prevalence of narrow angles in the study was used to calculate predictive values for each single test and pairbased test. To compare the sensitivities and specificities of the methods the researchers used the McNemar test and validated the results internally with bootstrapping techniques.