The frequency of eye examinations in people with diabetes

The findings of the National Eye Health Survey are encouraging, but also identify areas for improvement

This article by Foreman and colleagues in this issue of the MJA1 explores how well people with diabetes are adhering to the 2008 recommendations by the National Health and Medical Research Council (NHMRC)2 about regular eye examinations. Their report analyses data for a recent population-based sample, the National Eye Health Survey (NEHS),3 which included sizable samples of both Indigenous and non-Indigenous Australians. The authors found a moderate level of non-compliance with the NHMRC recommendations, particularly among Indigenous people, and they argue that a carefully integrated and accessible diabetic retinopathy screening service for Indigenous Australians is needed.

A number of questions arise. Are the current NHMRC guidelines, which recommend more frequent eye examinations for Indigenous than non-Indigenous Australians, still appropriate in 2017? Are these new data sufficiently robust to reflect current practice in these two groups of Australians? And how would the proposed screening service work in practice, and what models are available for implementing it?

First, there is a concern whether the evidence available in 2017 still supports the difference in recommended eye examination frequency for Indigenous (yearly) and non-Indigenous (2-yearly) Australians. While the age-specific rates of vision impairment were substantially higher among Indigenous than non-Indigenous NEHS participants, the study also found that the proportion of vision impairment attributable to diabetic retinopathy was relatively low, but similar, in both groups. The report,2 however, provides only preliminary data on the prevalence of retinopathy, and its prevalence among Indigenous people with diabetes is currently unknown. Earlier studies had found an earlier onset and higher prevalence of diabetic retinopathy in Indigenous Australians. More detailed information needs to be collated to determine whether differing recommendations for these two groups of Australians remain appropriate.

Second, the NEHS study was a major and welcome undertaking, providing the first large scale data (30 representative sites) on the prevalence of vision impairment in middle-aged to older Australians since two major population-based studies conducted about 25 years ago, the Blue Mountains Eye Study4 and the Melbourne Vision Impairment Project,5 each of which assessed more limited samples than the NEHS.

The overall clinical eye examination rate in the NEHS was 71.5%,3 lower than the 82–83% in the earlier studies.4,5 The rate was substantially higher among Indigenous (77.6%) than non-Indigenous participants (68.5%); the non-Indigenous examination rate was 65.7% in New South Wales and 52% in Victoria. A further concern was the low proportion of dwellings where somebody was found to be home when contacted by the NEHS (overall, 51.1%; Indigenous, 77.9%; non-Indigenous, 46.2%). The combination of this low contactability — perhaps resulting from the short time spent at each of the 30 NEHS sites, all seen within less than one year — with the relatively low examination rates of those contacted may have led to selection and other biases.

The higher contact and examination rates for Indigenous participants provide greater confidence that the estimated yearly examination adherence of 52.7%3 accurately reflects that of the entire Australian Indigenous population, although documenting their rates of 2-yearly examinations would also have been valuable. The adherence rate for the non-Indigenous group of 77.5%3 (2-yearly tests) could have been affected by selection bias linked with its lower contactability and examination rates.

The prevalence of diabetic retinopathy has not yet been reported from the NEHS. However, some limitations of the study are already evident, particularly when compared with earlier studies.4,5 A non-mydriatic camera examination was used, for which only 14% of participants had pupil dilation;3 further, photographs from both fields of each eye were gradable for only one-third of participants, and no clinical fundus examination was performed.3

Third, the main recommendation by Foreman and colleagues is that a carefully integrated and accessible diabetic retinopathy screening service for Indigenous Australians be established.1 While integrating eye examinations with diabetes medical assessments is critical, screening could also be linked with telemedicine procedures for rural and remote communities, and with general practitioner annual cycle of care programs for people with diabetes.

Regular eye examinations could also be linked with the recently introduced Medicare item number for practitioners performing non-mydriatic retinal photography in people with diagnosed diabetes.6 The Australian government has expressed some commitment to funding supporting infrastructure for this program. The potential for improved diabetes screening facilitated by the new item number is a positive development, but general practitioners will need to be trained in the procedure. The problem of mandated yearly screening for Indigenous patients with diabetes also remains, as the new Medicare item can only be claimed every 2 years. A potential pitfall is the frequency of low quality or ungradable images, as was found by the NEHS.3 Reminders linked to Medical Benefits Schedule claims (for optometric or ophthalmic consultations, for instance) by people with diabetes are also now being explored.

The findings of the NEHS regarding adherence of Australians with diabetes to NHMRC recommendations for eye examinations are encouraging, but the optimal timing of screening for retinopathy in Indigenous and non-Indigenous populations remains unresolved, and further work is needed to establish whether the current guidelines are still appropriate.