Approaching 90% sensitivity with new techniques

EDITOR,--Bob Ryder advocates screening for retinopathy by using two methods simultaneously.1 This is because none of the techniques available are sufficiently sensitive to use alone. Use of two techniques (for example, ophthalmoscopy and fundus photography) has yet to be shown to improve the overall sensitivity of the screening service, though it seems to be reasonably specific. Use of only one technique would be easier to administer and more acceptable to patients and is thus likely to have a better uptake.

Ryder points out some of the difficulties associated with non-mydriatic single field fundus photography and highlights the low sensitivity (67%) found by Buxton et al when they used this technique.2 In their study, however, ophthalmoscopy was used as the reference standard; this technique has considerable limitations when compared with slit lamp biomicroscopy or fluorescein angiography. The technique used by Buxton et al has other limitations that can be improved on. Use of mydriasiseliminates the problem of poor quality photographs in the second eye as a result of persistent pupillary constriction. Use of more than one photographic field increases the area of fundus visualised, especially at the peripheries, and the detail of film clarity can be improved by using 35 mm film. These measures are all thought to improve diagnostic accuracy.3

We used mydriatic three field fundus photography with 35 mm film in a screening programme and found that it had a sensitivity of 89% when compared with a reference standard of slit lamp biomicroscopy by a senior ophthalmologist; its sensitivity was significantly better than that of 65% achieved by a trained ophthalmologist using an ophthalmoscope.4 This figure approaches the target of 90% sensitivity for a useful screening service.

Many other screening mechanisms are under examination, notably those based on community optometrists. It is important that the effectiveness of each of these methods is assessed by comparison of the sensitivity and specificity with those obtained with a quality reference standard. In addition, each method must be shown to reach the target population. Mobile fundus photography can successfully reach the target population.5 Although mydriatic multiple field fundus photography has not yet been shown to be consistently sensitive, we believe that, with the improvements in technique being introduced in many centres, it should not yet be dismissed as an independent form of screening for diabetic retinopathy.