Guidelines for screening intervals
Yasmin was diagnosed with type 2 diabetes 9 years ago when she was 42 years old. She had an eye test done about 5 years ago at a local hospital while she was getting her diabetes checked. At that time, she was told there were some small changes at the back of her eye “but she should not worry as long as she looked after her diabetes”. Yasmin explained that it has always been difficult to manage her diabetes with regular oral medication and that sometimes she had missed taking her drugs due to lack of money to buy them.
3 months ago, she woke up with very poor vision in her right eye and when she was seen by the eye doctor, they informed her that she would urgently require treatment with a laser or she will go blind. “I did an eye test as I was told, so why am I going blind now?” she thought to herself.
The aim of screening is early detection of sight threatening diabetic retinopathy. The key question about screening for a person with diabetes is “How often do my eyes need to be checked?” For a health care provider the questions are different, ”What is the cost-effective screening programme?” and “Does everyone with diabetes need an eye examination at the same frequency?”
How often should a person with diabetes be screened?
The Liverpool Diabetic Eye Study was established in 1991 to study various aspects of systematic screening for diabetic retinopathy (DR). It also investigated the incidence (rates of new cases) and progression rates of sight threatening diabetic retinopathy amongst the people with type 2 diabetes enrolled in the study.
The data collected was also used to calculate optimum screening intervals by DR grade amongst people with type 1 and type 2 diabetes (Younis et al. 2003).
The study assessed 501 people with type 1 diabetes and 2,742 screening events and concluded that screening at 2 to 3 year intervals, rather than every year, for people with type 1 diabetes but with no retinopathy was feasible because of the low risk of progression to sight-threatening DR.
4,770 people with type 2 diabetes and 20,570 screening events were also assessed. The yearly incidence of sight-threatening DR in the 3,743 patients without retinopathy at the start of the study (baseline) was 0.3%, rising to 1.8% in the fifth year. The total (cumulative) incidence after 5 years was 3.9%. The study concluded that a 3-year screening interval could be safely adopted for people with type 2 diabetes with no retinopathy.
The Liverpool study also highlighted several risk factors (duration, Hb1AC levels, blood pressure, level of retinopathy at first screening visit) which also affect the incidence of diabetic retinopathy. Since then, several other studies have used modelling and algorithms techniques to further establish appropriate guidance on the ideal screening interval.
What are the most cost effective screening intervals?
Cost is a major concern when trying to set up screening programmes. One study (Vijan 2000) modelled the cost effectiveness of screening intervals based on the level of diabetes control (glycaemic levels) and the risk of developing sight threatening retinopathy. It concluded that annual screening for all patients without previously detected retinopathy is not cost effective and that frequency screening intervals should be tailored to medical factors - such as poor diabetes control - which substantially increase the risk of developing sight threatening retinopathy. The practicalities of implementing a risk based screening programme are context specific and need to be understood based on local practice within the health system.
Do longer screening intervals affect coverage?
Questions remain about the potential consequences of long intervals between screening events (Klein 2003). Does a long interval make it difficult for a screening programme to maintain contact with a person with diabetes? Does it also send the wrong message to people with diabetes that vision testing is not important? There is a need to undertake more studies to understand the factors that influence attendance at screening services.
International Council of Ophthalmologist (ICO) guideline on DR screening intervals
On this course, we use the ICO’s current guideline for screening intervals (ICO 2017). The guidelines are for both high and low/intermediate resource settings and take a balanced approach to the key areas of debate about screening intervals as summarised in the table below.
Table. Follow-up schedule and management based on diabetic retinopathy severity
|Diabetic retinopathy (DR) severity||Follow-up schedule for management by ophthalmologist|
|No apparent DR||Re-examine* in 1-2 years|
|Mild non-proliferative DR||Re-examination* in 6-12 months is ideal but can be extended to 1-2 years in low resource settings|
|Moderate non-proliferative DR||Re-examine in 3-6 months|
|Severe non-proliferative DR||Re-examine in <3 months. Consider early pan-retinal photocoagulation|
|Proliferative DR||Re-examine in <1 month. Consider pan-retinal photocoagulation|
|Stable (Treated) PDR||Re-examine in 6-12 months|
|* Re-examination may not require an ophthalmologist in these cases|
|Diabetic macular oedema (DME) severity||Follow-up schedule for management by ophthalmologist|
|Non central-involved DME||Re-examine in 3-6 months. Consider focal laser photocoagulation|
|Central-involved DME||Re-examine in 1-3 months. Consider focal laser photocoagulation or antiVEGF therapy|
|Stable DME||Re-examine in 3-6 months|
On reflection, is this a practical guideline to adopt locally in your setting?
© London School of Hygiene & Tropical Medicine CC BY-NC-SA 4.0