National screening programme for diabetic retinopathyBMJ 2002; 324 doi: https://doi.org/10.1136/bmj.324.7341.849 (Published 06 April 2002) Cite this as: BMJ 2002;324:849
Digital image may be better for screening
- Colin Clements, ophthalmic photographer.
- Department of Ophthalmology, King's College Hospital, London SE5 9RS
EDITOR—I work as an ophthalmic photographer, but I have no involvement in screening. Prasad et al assert that screening for diabetic retinopathy by optometrists is superior to photographic screening.1 This is questionable for several reasons.
Firstly, biomicroscopy detects macular oedema, and more of the retina can be examined. Both are true, but macular oedema will be accompanied by reduced visual acuity, which is easily measured. Treatable retinopathy is most likely to be found around the posterior pole. Is there an obvious advantage in examining the periphery?
Secondly, photography has a high technical failure rate compared with ophthalmology. Careful reading of the paper cited as evidence for this does not give a failure rate for photography of 14%. Of 326 patients, six were ungradable by any means. Twelve of the remaining 320 patients could not be photographed because of posture or tremor. A further 34 patients were ungradeable; 18 of them had opacities preventing photography. True failure rate is not 46 out of 320 (14%) but 16 out of 320 (5%). Prasad et al have added the 12 patients who had posture and tremor difficulties to the 34 who had ungradeable pictures, giving 46 out of 320. Failures were 46 minus 30, equivalent to 16 out of 320, or 5%. A single technique of examination is not appropriate for all patients, and recording a technical failure when it is impossible for the technique to succeed results in a distortion of the facts.
Thirdly, the choice of digital photography for …