Changes in serum lipid concentrations during first 24 hours after myocardial infarction.

Results ofstudy ConsultantsJunior doctors (n=20) (n= 16) Glaucomatous discs (n=16) Doctors' mean scores (%)Correct (ie, sensitivity) 7-0 (44) 8-5 (53) Unsure 2-4 (15) 1-5 (9) incorrect (false negative) 6-6 (41) 6-0 (38) Scores on individual discs (no of discs) 75-l100%/o answers correct 2 4 50-74% answers correct 4 5 25-49% answers correct 5 3 0-24% answers correct 5 4 Non-glaucomatous discs (n=27) Doctors' mean scores (%) Correct (ie, specificity) 20-7 (77) 18-6 (69) Unsure 2-5(9) 3-4(13) Incorrect (ie, false positive) 3-8 (14) 5-0 (18) Scores on individual discs (no of discs) 75-1000/o answers correct 16 13 50-74% answers correct 8 7 2549%/o answers correct 3 3 0-24% answers correct 0 4used in isolation, direct ophthalmoscopy may be quite inaccurate.If ophthalmologists can miss 40% of cases ofearly to moderate glaucoma when examining patients under ideal circumstances it is likely that general practitioners miss many more cases, especially ifthey examine fundi without dilating the pupils.Not surprisingly, therefore, recent surveys ofthe referral source of patients with glaucoma have shown that nearly all early and moderate cases are referred by opticians, while general practitioners refer mainly advanced, symptomatic cases.34General practitioners do need further training in the use of the ophthalmoscope, but these findings suggest that when they identify patients at risk of developing glaucoma, as well as performing funduscopy themselves, they should also encourage these patients to be seen regularly by their optician.5Finally, ophthalmologists themselves should be aware of the potential inadequacy of direct ophthalmoscopy when examining patients with suspected or definite glaucoma.We are grateful to Dr David Appleton for his help with statistical analysis.1 Hitchings RA.Screening for glaucoma.