Article Text

Download PDFPDF

Should diabetic patients be screened for glaucoma?
  1. S A VERNON, Consultant ophthalmologist and chairman, Royal College Working Party responsible for the 1997 Guidelines on Ocular Hypertension and Glaucoma
    1. J ELLIS,
    2. A D MORRIS,
    3. C J MAcEWAN
    1. University Department of Ophthalmology, Ninewells Hospital and Medical School, Dundee DD1 9SY

      Statistics from

      Editor,—In their article, Elliset al claimed to have “reviewed all potentially relevant English language articles from 1966 to 1997.”1 It is unfortunate that they did not define the method of detecting such articles as it may have cast light upon the omission of perhaps the most complete study yet performed on the subject.2 This was published in April 1997, nearly 2 years before the publication of the perspective and approximately 6 months before the Royal College Guidelines on Ocular Hypertension and Glaucoma (although after the “last revision” date). The inclusion of other articles from 1997 in the reference list would appear to indicate that their quoted dates of 1966 to 1997 were inclusive rather than exclusive. The findings of a positive association of open angle glaucoma with diabetes was a result of an analysis of data from the Blue Mountains eye study, a carefully controlled study of a large population of people aged 49 and over. This study would appear to come closest to Ellis et al’s ideal in view of the non-IOP based nature of the glaucoma diagnosis and the use of fasting glucose levels in those not already known to be diabetic.

      I would encourage all those who read the perspective1 to study Mitchell et al’s paper and thereby come to their own conclusions concerning the question “Should diabetic patients be screened for glaucoma?”



      Editor,—We are grateful to the chairman of the Royal College Working Party responsible for the 1997 guidelines on ocular hypertension and glaucoma, Mr S Vernon, for highlighting our inadvertent omission of the Blue Mountains eye study paper which was available to the working party before the publication of the guidelines. The omission of this paper from our review was unintentional and unfortunate.

      The strengths of this rigorous study are its population based sampling frame, its careful definition of diabetes, and its definition of glaucoma to include field loss matching disc changes regardless of intraocular pressure. The overall prevalence of diabetes was 7.0% of whom 98% had type 2 diabetes and 15% were newly diagnosed during the course of the study (by fasting glucose test). The prevalence of POAG was 3.0% (108 patients).

      Exactly half the glaucoma cases were newly diagnosed by the researchers. The prevalence of diabetes in the 54 pre-study diagnosed POAG patients was almost twice that in the 54 “newly diagnosed” cases (nine cases in the former compared with five in the latter; age/sex adjusted odds ratio (OR) 2.12). The prevalence of diabetes in those with POAG differed significantly from that of the general population without glaucoma (6.9%) only in the latter group (age/sex adjusted OR 2.82, 95% CI 1.35–5.87). However tempting to suggest that unmasking bias may be operative in this group, it is noteworthy that six of the nine patients did not know they had diabetes at the time of diagnosis of POAG. Unmasking bias will only have been operative if the eye clinic is as likely to diagnose diabetes in referrals with possible POAG as it is to diagnose unsuspected POAG in referrals for possible retinopathy or diabetes related cataract. This is unlikely since random blood glucose estimation is probably not a routine test in a busy eye clinic.

      Our review asked whether the published information available to clinicians and health planners could support targeted screening of diabetic patients. The suggested method was measurement of intraocular pressure. It is informative that in the Blue Mountains eye study only 25.5% of POAG patients had raised pressure. This then would seem an inappropriately insensitive means for screening.

      The paper by Mitchell et al adds convincing support to the existence of an association between diabetes and POAG. The problem of how best to cater for this group remains. In light of the available evidence the advice of the Royal College of Ophthalmologists seems apposite; to be aware of the increased risk of POAG in this group of patients who attend for retinopathy screening.


      Request Permissions

      If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.