Sir,

We are flattered to have attracted the kind of high-quality, considered response above. Ghazawy, Saldana and McKibbin have provided original data and in many ways their article surpasses our own in its contribution to the debate for this reason. Their study examined an innovative and potentially sustainable model for fast tracking suspected choroidal neovascular membrane (CNV) referrals and found that 42% of cases with distortion on Amsler grid testing had neovascular macular degeneration. Faced with a confirmed pathology in less than half of those referred, some disappointment is implicit in their use of the word ‘very’ in their subsequent statement; ‘(there were) a very high number of false positives’. It is possible to draw precisely the opposite conclusion; namely that for so simple and inexpensive a test, the proportion with genuine pathology in this group is remarkably high.

This proportion represents the positive predictive value (PPV) of the Amsler test. Unlike the sensitivity and specificity of the test, which are entirely independent of the amount of pathology in the community, the PPV is profoundly affected by the prevalence of the pathology being sought. A PPV of 42% (38% for CNV) compares favourably with the PPV of screening programmes already widely accepted, for example, 9% in breast screening for women aged between 50 and 59 years,1 1% in cervical screening of postmenopausal women on hormone replacement,2 and, closer to home, 0% for the finding of isolated field defect and subsequent confirmation of glaucoma.3

The authors comment that when the optometrist examined the fundus the sensitivity fell to 71% (it would have been interesting to know by how much it fell, but they do not give the figure derived without examination). They were able to achieve a sensitivity and specificity 90% or more with their ‘fast track and refinement’ clinic. This would undoubtedly greatly elevate the PPV of those being sent on to the medical retina specialist, as the prevalence of pathology in this population (those referred to secondary care with abnormal Amsler test results) is so much higher than in the community. We would love to hear a full report of this patient pathway or of its wider adoption and use in larger numbers.

As the gold standard remains fluorescein angiography, the need for this to take place in the hospital ophthalmic care setting is self-evident, and it demonstrates what may be achieved within the constraints of current resources. But most important of all, it achieved its primary goal: it was fast.