Design: Population-based cross-sectional survey, RAAB with diabetic retinopathy module
Population: 7,835 adults aged ≥50 in rural-tribal Tapi district, Gujarat
Primary outcome: Prevalence of diabetes and diabetic retinopathy
Effect: Diabetes 4.9% (4.2–5.6); any DR 16.3%, maculopathy 9.2%, sight-threatening DR 2.1% among diabetics

A Rapid Assessment of Avoidable Blindness survey with a diabetic retinopathy module sampled 139 clusters of people aged 50 and over in the rural-tribal Tapi district of Gujarat. Of 8,344 eligible, 7,835 (94%) were examined; diabetes was defined as known diagnosis or random glucose of at least 200 mg/dL at the doorstep.

Diabetes prevalence was 4.9% (95% CI 4.2–5.6), lower than national estimates. Almost a third of cases were new diagnoses from the spot glucose test, and 55% of those already diagnosed had random glucose of 200 mg/dL or more. Among people with diabetes, 16.3% had any retinopathy, 9.2% maculopathy and 2.1% sight-threatening retinopathy.

For eye services working in tribal and rural India, the message is that diabetes and retinopathy are present and rising, and that poor glycaemic control is common even after diagnosis. Combined glucose and retinal screening camps reach people who are not yet diagnosed.

Offer a random glucose check alongside eye screening in rural and tribal outreach camps.
Examine the fundus of every person found to have diabetes, including new diagnoses.
Link patients with uncontrolled diabetes to primary care for glycaemic management.
Look for maculopathy specifically; it was more than four times as common as sight-threatening retinopathy.

Why it matters

It gives rare population data on diabetic eye disease in an underserved Indian tribal community.

The statistics, in plain English

The 16.3% retinopathy figure has an interval of about 12.5–20%, because it is based on only 381 people with diabetes. Random glucose is a crude diagnostic test, so some diabetes will have been missed and some over-called.