Background: The Katherine Region Diabetic retinopathy Study (KRDRS) was carried out in the Lower Top End of the Northern Territory of Australia between 1993 and 1996 as part of the Northern Territory Eye Health Program. It investigated diabetic eye conditions and their determinants in the Australian Aboriginal population of the region.
Methods: The KRDRS was comprised of two cross-sectional surveys, in 1993 and in 1996, and involved a total of 477 subjects with diabetes. Ninety-six subjects were seen in both surveys and met the criteria for inclusion in the longitudinal study. Each subject underwent a general eye examination and retinal photography.
Results: The annual incidence rates of retinopathy were 5.6% (95% CI 3.0-10.0%) and 4.2% (95% CI 2.5-7.0%) in subjects and eyes, respectively. This compares with 8% (in subjects) of the overall Australian diabetic population found in the Newcastle Study of Diabetic Retinopathy. However, each year 1.2% of subjects (95% CI 0.2-3.8%) and 1.3% of eyes (95% CI 0.2-4.5%) with no pre-existing retinopathy developed vision-threatening retinopathy. The respective findings for maculopathy were 2.2% (95% CI 0.7-4.6%) and 1.7% (95% CI 0.7-3.2%), and for clinically significant macula oedema (CSME) 1.1% (95% CI 0.2-3.1%) and 0.9% (95% CI 0.3-2.2%). The annual incidence of CSME in those with no maculopathy when first seen was 1.1% (95% CI 0.2-3.3%) and 0.9% (95% CI 0.3-2.2%) in subjects and eyes, respectively. The annual progression to CSME in those with maculopathy was 4.8% (95% CI 0.1-26.5%) and 6.1% (0.7-21.9%) in subjects and eyes, respectively.
Conclusion: Results of the KRDRS show that despite a lower overall incidence of diabetic retinopathy among Aboriginal diabetics compared to the general Australian diabetic population, there are important reasons to consider Aboriginal diabetics at special risk. These reasons include the highest reported incidence of vision-threatening retinopathy in Australia, one of the highest ever reported incidences of CSME in the world and the likelihood that the severity of the problem may be underestimated because of the relatively short observation time and the low average time since diagnosis. It is also acknowledged that the small numbers in the study limit our ability to reliably detect progression between subsets of the population and much larger studies are required to make statistically significant comparisons.