Current issues of ACP Journal Club are published in Annals of Internal Medicine


Review: Annual screening for retinopathy for those with high-risk macular edema or proliferative retinopathy is beneficial

ACP J Club. 1992 July-Aug;117:20. doi:10.7326/ACPJC-1992-117-1-020

Source Citation

Singer DE, Nathan DM, Fogel HA, Schachat AP. Screening for diabetic retinopathy. Ann Intern Med. 1992 Apr 15;116:660-71.



To review the literature on diabetic retinopathy pertaining to natural history, therapy, diagnosis, and the cost effectiveness and timing of screening and to assess current clinical guidelines for diabetic retinopathy screening.

Data sources

A MEDLINE search from 1986 to present, recommendations from the American Academy of Ophthalmology, review of authors' files, bibliographies of relevant articles, and review articles yielded articles for possible inclusion.

Study selection

Articles presenting results from large clinical series, formal epidemiologic studies, randomized controlled trials, and studies of diagnosis and screening were selected.

Data extraction

Clinical data from pre-eminent studies and ranges of values from multiple trials were presented; meta-analytic summary estimates were not calculated.

Main results

The 20-year risk among patients with diabetes for clinically significant macular edema and proliferative retinopathy ranged from 10% to 15% and from 10% to 50%, respectively. Risk factors for severe retinopathy included insulin-dependent diabetes mellitus (IDDM), duration of diabetes, severity of retinopathy on initial examination, poor glycemic control, proteinuria, and pregnancy (in patients with IDDM). There was a 16% risk for severe visual loss if proliferative retinopathy was left untreated for 2 years. Improving glycemic control has not yet been clearly shown to reduce this risk; only laser photocoagulation has proved beneficial (> 50% reduction in rate of developing vision loss). About 20% of treatable cases of proliferative retinopathy or macular edema were missed by ophthalmologists using dilated ophthalmoscopic examination. For 5 studies among ophthalmologists, sensitivity ranged from 28% to 96% and specificity ranged from 79% to 100%. Examinations of nondilated eyes or examinations done by medical personnel not expert in eye disease were worse. Nonmydriatic fundus photography was as accurate as dilated ophthalmoscopy. Stereoscopic fundus photography is the diagnostic standard but may not be a practical screening test. 4 cost-effectiveness studies (2 from the United Kingdom and 2 from the United States) suggested that annual screening for diabetic retinopathy prevents blindness at a relatively low cost or at a cost saving.


Annual screening for retinopathy, beginning 5 years after onset of IDDM and beginning immediately for noninsulin-dependent diabetes mellitus, and subsequent therapy for those with high-risk macular edema or proliferative retinopathy are beneficial.

Sources of funding: American College of Physicians; Agency for Health Care Policy and Research; Henry J. Kaiser Family Foundation.

Address for article reprint: Dr. D.E. Singer, General Internal Medicine Unit, Bulfinch 1, Massachusetts General Hospital, Boston, MA 02114, USA.


This article carefully reviews the relevant literature on screening for diabetic retinopathy. The authors concisely summarize the natural history, epidemiology, and treatment of this condition. They conclude that annual screening of adult patients with diabetes by eye specialists such as ophthalmologists is beneficial and is likely a cost-saving strategy for society.

Screening patients with diabetes for retinopathy is useful for 4 reasons: Visual loss from retinopathy is common; treatment is very effective (laser photocoagulation reduces the risk for vision loss by 50%); many patients who need laser treatment are asymptomatic; and visual loss from diabetic retinopathy is largely irreversible.

Stereoscopic fundus photography is the diagnostic standard and is clearly superior to the other screening test options, but this method is not readily available. Stereoscopic photography may also suffer in comparison with other options as it becomes more available outside of the research setting. Based on published studies, dilated ophthalmoscopy by eye specialists is the best practical option and is superior to dilated examinations by diabetologists, primary care physicians, and other health care providers. This option may also be impractical, however, in some practice settings (e.g., rural areas).

The best population-based study on screening rates was done about 10 years ago (1) and found that only two thirds of patients with diabetes have ever had an ophthalmologic examination. More recently, in our general medicine clinic, only two thirds of patients with diabetes have been appropriately screened within the past year. More attention needs to be focused on implementing screening guidelines in various practice settings, especially because this strategy will not only save years of vision but will also save our society money.

Arthur T. Evans
Jan A. KylstraUniversity of North CarolinaChapel Hill, North Carolina, USA


1. Witkin SR, Klein R. Ophthalmologic care for persons with diabetes. JAMA. 1984;251:2534-7.