Predictors of attendance at diabetic retinopathy screening among people with type 2 diabetes: Secondary analysis of data from primary care

Highlights

• First study of uptake to examine a mix of private and public screening services.

• Most people managed in a diabetes care programme attended retinopathy screening.

• One third of those who attended screening attended other screening services only.

• Attendees more likely to use oral hypoglycaemic agents or injectables.

• Attendees more likely to have on target blood glucose and attend DNS.

Abstract

Aims

To estimate the uptake of diabetic retinopathy screening among adults with type 2 diabetes and to identify and compare factors associated with attendance at the national diabetic retinal screening programme, and other screening services provided by ophthalmic surgeons, community ophthalmologists and optometrists.

Methods

An observational retrospective cohort study was carried out using data from the 2016 audit of care delivered by general practices (n = 30) enrolled in a structured diabetes care initiative.

Attendance at any screening in the previous 12 months, and attendance across different types of service (national programme and other screening services) was calculated. Logistic regression was performed to examine predictors associated with (1) attendance at any screening and (2) attendance at the national programme (RetinaScreen). Sociodemographic, clinical, and lifestyle factors were examined as predictors.

Results

Data were available for 1106 people with type 2 diabetes aged ≥18 years.

Overall, 863 (78%) of patients had a record of screening attendance in the previous 12 months. Of those screened, 494 (57.2%) attended RetinaScreen only, 258 (28.7%) attended other screening services only, and 111 (12.9%) attended both services.

Statistically significant predictors of attendance at any screening were tablet/injectable controlled diabetes, attendance at a diabetes nurse specialist (DNS) in the past 12 months and a blood glucose level which was not on target (HbA1c >7.0% or >53 mmol/mol). In addition to these factors, when examining predictors of attendance at the national screening programme specifically, females were less likely to attend.

Conclusions

Most patients managed in a structured diabetes care programme in primary care attended screening. Those with on target blood glucose control, those who were on oral or injectable medication or had been seen by a DNS were more likely to attend for annual screening. Of those who attended screening, almost one-third attended other screening services and so were not availing of the national programme, which is free, quality assured and has an integrated treatment arm.

Introduction

Worldwide, the prevalence of type 2 diabetes is increasing due to an aging population and increasing obesity levels [1]. As a consequence, it is estimated that the prevalence of diabetic retinopathy, a common microvascular complication of diabetes [2], will grow from 126.6 million in 2011 to 191.0 million by 2030 [3,4]. Diabetic retinopathy can cause vision problems including vitreous haemorrhage, retinal detachment, glaucoma and blindness [2]. Regular diabetic retinopathy screening leads to the earlier detection of retinopathy and treatment that can prevent or delay the development of diabetes-related blindness [5]. National and international guidelines recommend retinopathy screening every 12 months if no or minimal unchanged retinopathy is present at the previous screening [6,7].

In Europe, few countries have established systematic (organised) screening programmes designed to reach all people in the population with diabetes. The countries that do include Denmark, England, Finland, Iceland, Scotland, Wales, Northern Ireland, and the Republic of Ireland. Other European countries have taken a stepwise approach to the implementation of screening, initially establishing city-wide or smaller regional screening programmes, with the aim of eventually developing national programmes. In part due to this incremental approach and perhaps also reflecting the wider health system, some countries have a mix of public and private screening services [8]. In Ireland, prior to the introduction of the national screening programme in 2013, the availability and type of screening services varied across regions. Screening was provided publicly and privately in community and secondary care settings by a variety of healthcare professionals including through community ophthalmologists, community ophthalmic physicians and private ophthalmic surgeons and optometrists [9]. These other screening services are still available. This reflects the complex mix of public and private care in the wider healthcare system. Ireland’s healthcare system is predominantly tax financed, but it does not offer universal coverage for primary care and there are multiple tiers of eligibility depending on the service [10]. Almost half of the population have private health insurance (PHI) (46% in 2019) [11]. PHI is voluntary and duplicative, that is, it covers services already provided by the public health system but also provides access to other providers (e.g., private consultants) or to different levels of service (e.g., faster access).

Despite routine retinopathy screening being an internationally accepted standard of care for people with diabetes [12], studies suggest variable and sub-optimal uptake ranging from 50 to 81% [[13], [14], [15], [16], [17]]. A recent systematic review (2003–2017) of predictors of attendance at retinopathy screening found that socio-economic deprivation and younger age were associated with non-attendance [18]. Other factors associated with non-attendance included sex, ethnicity, birth country, type of diabetes, use of insulin, education levels, smoking, and other poorly controlled systemic conditions [18]. Hyperlipidaemia and hypertension may be a marker for other poorly controlled systemic conditions; however, they have previously not been examined as predictors of attendance at screening. Previous studies of attendance did not include health service use indicators such as attendance at a Diabetes Nurse Specialist (DNS) however studies suggest attendance at other diabetes services may play a part in whether patients attend a new service [19,20]. Many of these studies also relied on self-report of attendance at screening which is more prone to bias [21]. This study will use general practice records to provide a more accurate picture of attendance at screening and potential predictors of attendance.

The aim of this study is (1) to estimate the rate of attendance at different diabetic retinopathy screening services among adults with type 2 diabetes managed in primary care, and (2) to identify factors associated with attendance at screening in general, and at the national screening service RetinaScreen. To date few studies have examined predictors of attendance at diabetic retinopathy screening in an Irish population [19,20] and only one has looked at predictors of attendance at the national population-based screening programme [20].