Revised Grading

Definitions for the NHS

Diabetic Eye Screening
Programme
Version 1.0, 06 March 2012
To more closely define grading levels in the NHS Diabetic Eye Screening
Programme in England and to provide guidance notes and example images
to graders. To be introduced across England during 2012.

Project/Category

Grading

Document title

Revised Grading Definitions for the NHS Diabetic Eye Screening

Programme

Version and date

v1.0, 6 March 2012,

Release status

Final

Author

Peter Scanlon

Owner

Peter Scanlon

Type

Policy

Authorised By

Grading and Assessment Sub-Committee endorsed by the

Programme Advisory Committee and the UK National Screening
Committee

Valid from

6 March 2012

Review Date

6 March 2013

Audience

NHS DESP , Clinical Leads, Programme Managers, Graders,

Ophthalmology providers,

Distribution
Name / group

Responsibility

NHS DESP national, Clinical Leads,

Programme Managers, Graders,
Ophthalmology providers
Amendment history
Version

Date

Author

Description

Review / approval
Version

Date

v1.0 6 March 2012

Requirement

Signed

METHODOLOGY
1

Background
The NSC review of Quality Assurance in the previously named ENSPDR recommended that there is a
review of the classification of grading system of retinopathy by an expert group. The grading system
should be clear, unambiguous and simple for graders to follow. It should be focused on what is a referable
retinopathy and what is an observable retinopathy and should rigorously exclude subclassifications that
do not affect the referral decision that is the only outcome of screening. The group should include
representative input from ophthalmologists and their professional body, and should address the apparent
systematic differences of interpretation between graders and ophthalmologists.

A Grading and Assessment Committee was set up under the chairmanship of Declan Flanagan to
undertake this work. The Group consisted of:
Declan Flanagan, Chair
Peter Scanlon, National Programme Officer
Gilli Vafidis, RCOphth representative
Cathy Egan, RCOphth representative
Rob Johnston, Consultant Ophthalmologist
Lyndon Taylor, Association of Optometrists
Shelley Widdowson, Senior Grader York
Susan Carter, Optometrist
Simon Harding , Consultant Ophthalmologist
Irene Stratton, Statistician
Victoria Humble, Senior Grader, Wakefield
Fu-Meng Khaw, Director of Public Health, Newcastle PCT and Newcastle City Council
Lynne Lacey, Project Manager, ENSPDR
David Taylor, Regional QA Manager, ENSPDR
Jenny Sykes, Minute taker

Work on the multiple blot haemorrhages criteria of R2

This work had been led by Simon Harding in 2010 and the evidence from this work was reviewed.
In the 2010 exercise 20 image sets were circulated on CD and 36-40 experienced graders and clinicians
performed masked grading of images. The results from this exercise informed the recommendations in this
paper.

The website images and Expert Ophthalmology Grading Group

A website was developed by Sue Carter with Netsima Ltd and the support of Gilli Vafidis for grading of
images to help clarify and tighten the definitions when allocating R2 and M1 grades respectively, including
some challenging cases where the grade might be considered borderline.
This exercise was done remotely via a specially designed website with standard image manipulation tools
to grade around 20 selected IRMA images and 20 Groups of Exudates, and the grades compared centrally.
Approximately 20 Ophthalmology Clinical Leads of Screening programmes and 5 senior graders
participated. Steve Aldington and Irene Stratton contributed to the design of the website and Irene
Stratton undertook the analysis. The results have informed this paper

v1.0 6 March 2012

THE IMPACT OF THESE CHANGES ON REFERRAL LEVELS

Defining the R2 level:
When the grading form was originally introduced in the English Screening programme in 2003, it was
envisaged that referable retinopathy at the pre-proliferative R2 level would be defined by an approximate risk
of progression to proliferative retinopathy R3 of 11.3% over the following 12 months, as defined by ETDRS level
40 (Fundus photographic risk factors for progression of diabetic retinopathy. ETDRS report number 12. Early
Treatment Diabetic Retinopathy Study Research Group. Ophthalmology 1991; 98:823-33). However, what has
happened is that patients have been referred in some programmes at the ETDRS level 30 which only carries a
6.2% risk of progression in 12 months to proliferative retinopathy. These guidance notes will, therefore, clarify
the R2 level of referral and reduce the number of over referrals.
It is recognised that risk levels of progression in the general population of people with diabetes are less than in
the ETDRS paper quoted because of the general improvements in control of diabetes and blood pressure
(Hovind P, et al. Diabetes Care 2003; 26: 12581264. Klein, R et al. Ophthalmol. 2008; 115:18591868).
However, patients seen with pre-proliferative and proliferative DR in the Eye clinic are often in a less well
controlled group for glucose and blood pressure control and hence, the risk may be only slightly lower than the
risk from the ETDRS paper.
Defining Groups of Exudates:
When the grading form was originally introduced in the English Screening programme in 2003, those
writing the form were clinicians who had in their own head a definition of group of exudates and it has become
clear that this needed to be more clearly defined because referrals for very small groups of exudates were
being made in some programmes when these small groups were only just within the edge of the macular circle
and there was not a threat to vision.
Some recent work by Mr N Dhingra, Consultant Ophthalmologist, Pinderfields Hospital, Wakefield that
was presented at the Manchester Royal College of Ophthalmology seminar confirmed the view of the
committee that the group of exudates needed to be a certain size when it was more than 1 disc diameter away
from the centre of the fovea in order to pose a threat to vision.
The definition of group of exudates based on size as defined in this document will reduce the number of over
referrals.
Defining a stable treated R3s grade
Currently there is a wide difference between programmes about how patients should be graded who have
received peripheral scatter laser treatment in the past and are considered to be stable.
Introducing a stable treated grade will reduce the inconsistency between programmes and enable an easier
comparison of outcome data between programmes.
The numbers in this category are small and this change will not have any effect on the number of referrals.

v1.0 6 March 2012

Index

1.

2.
3.
4.

Title
Revised English Grading Classification
General guidance notes and specific guidance for cotton wool spots,
venous loops and photocoagulation scars
Revised Grading Classification for Pre-proliferative DR (R2)
A. Multiple blot haemorrhages
B. IRMA
C. Guidance notes on IRMA, Venous Beading, Reduplication and
Loops
D. Relationship of English Diabetic Retinopathy Classification of
Progression to Proliferative DR with ETDRS & Scottish
Classifications.
Revised Grading Classification for R3
Revised Grading Classification for Groups of Exudates
Revised Revision to Classification of Image Quality

v1.0 6 March 2012

Page
5
6

8-19
20-24
25
26-27

28
29 -33
34

Revised English Diabetic Eye Screening Programme Grading Classification

R

Retinopathy

R0
R1

None
Background

No DR
*
microaneurysm(s) or HMa *
retinal haemorrhage(s)
venous loop
any exudate in the presence of other nonreferable features of DR
any number of cotton wool spots (CWS) in
the presence of other non-referable features
of DR

R2

Preproliferative

venous beading
venous reduplication
multiple blot haemorrhages
intraretinal microvascular abnormality (IRMA)

R3

Proliferative

R3a (Active Proliferative Retinopathy)

R3s (Stable post treatment)

Maculopathy

M0
M1

No maculopathy
exudate within 1 disc diameter (DD) of the
centre of the fovea

All newly occurring R3 patients with:

new vessels on disc (NVD)
new vessels elsewhere (NVE)
pre-retinal or vitreous haemorrhage
pre-retinal fibrosis tractional retinal
detachment
Evidence of Peripheral Retinal Laser
Treatment
AND
Stable retina from photograph taken
at or shortly after discharge from the
Hospital Eye service (HES)
absence of any M1 features
M1 - list of features as now - but
individual features are not mutually
exclusive

group of exudates within the macula

retinal thickening within 1DD of the centre of
the fovea (if stereo available)
any microaneurysm or haemorrhage within
1DD of the centre of the fovea only if
associated with a best VA of 6/12 (if no
stereo)

Photocoagulation

No evidence of previous photocoagulation

focal/grid to macula or peripheral scatter

Unclassifiable

No grade is assigned
Only assigned if laser scars are
identified

An image set that is inadequate for grading

*

HMa is a term used when it is difficult to tell the difference between a microaneurysm and a dot
haemorrhage

v1.0 6 March 2012

General Guidance
This document relies on a decision being made that a lesion is definitely present.
If a grader feels that the presence of an individual lesion is questionable, it should be regarded as
absent for the purposes of grading in the English National Screening Programme.
Guidance Notes for Cotton Wool Spots
The proposal is that graders should be given the following instructions:
1.

Isolated cotton wool spots (one or more) in the absence of any microaneurysm or
haemorrhage should be counted as no DR (R0).
2. Any number of cotton wool spots (CWS) in the presence of other non-referable features of
DR should be graded as background DR (R1).
3. Once any CWS are detected, graders should exercise particular vigilance in searching for
features of referable DR in particular IRMA and early venous beading.
Guidance Notes for Venous Loops
The proposal is that graders should be given the following instructions:
1. An isolated venous loop should no longer be referred and should be regarded as a feature of
R1.

Photocoagulation scars
This remains as is currently recommended:
If there is no evidence of previous photocoagulation, no grade is assigned
If there is evidence of previous photocoagulation (focal/grid to macula or peripheral scatter) a P
grade is assigned.
A P grade is, therefore, only assigned if laser scars are identified.

Definition of the macula

The macula is defined as that part of the retina which lies within a circle centred on the centre of the
fovea whose radius is the distance between the centre of the fovea and the temporal margin of the
disc.

v1.0 6 March 2012

1. Revised Grading Classification for Pre-proliferative DR (R2)

R2
Pre-proliferative
Venous beading
(N.B. Venous beading from ischaemia in diabetic retinopathy does not occur in isolation)
Venous reduplication
Multiple blot haemorrhages
(N.B. If uncertain, refer only in the presence of IRMA that are definitely seen)
Intraretinal microvascular abnormality (IRMA)
(N.B. Check that they can still be seen on the colour image)

v1.0 6 March 2012

A.

Multiple blot haemorrhages

Guidance notes Three image sets (macula and disc centred images) are shown in photos
MBH 1, MBH 2, and MBH 3 where referral for multiple haemorrhages should be made.

Photo MBH 1 macula centred

v1.0 6 March 2012

Photo MBH 1 disc centred

v1.0 6 March 2012

10

Photo MBH 2 macula centred

v1.0 6 March 2012

11

Photo MBH 2 disc centred

v1.0 6 March 2012

12

Photo MBH 3 macula centred

v1.0 6 March 2012

13

Photo MBH 3 disc centred

v1.0 6 March 2012

14

Guidance notes Three image sets (macula and disc centred images) are shown (photos MBH 4,
5, 6). The amount of haemorrhage present in the 3 image sets does not warrant a referral.
However, a careful search for IRMA should be made when the amount of haemorrhages is equal
to that shown in the images. If IRMA is definitely seen, then a referral should be made. If IRMA is
not seen, the patient should be screened again in 12 months. It should be noted that IRMA is
not present in any of these example images.
Photo MBH 4 macula centred

v1.0 6 March 2012

15

Photo MBH 4 disc centred

v1.0 6 March 2012

16

Photo MBH 5 macula centred

v1.0 6 March 2012

17

Photo MBH 5 disc centred

v1.0 6 March 2012

18

Photo MBH 6 macula centred

v1.0 6 March 2012

19

Photo MBH 6 disc centred

v1.0 6 March 2012

20

B. Intraretinal microvascular abnormality (IRMA)

Patients with IRMA that are definitely seen should be referred into the Hospital Eye Service.
Examples of one IRMA are shown in example IRMA 1 and two IRMA in IRMA 2, 3, 4, 5

Photo IRMA 1

v1.0 6 March 2012

21

Two IRMA in examples IRMA 2, 3, 4, 5

Photo IRMA 2

v1.0 6 March 2012

22

Photo IRMA 3

v1.0 6 March 2012

23

Photo IRMA 4

v1.0 6 March 2012

24

Photo IRMA 5

v1.0 6 March 2012

25

Guidance notes.
Only IRMA that are definitely seen should be referred.
a) Once an IRMA is found, one should always return to the colour image. IRMA is considered
present if the IRMA can still be seen on the colour image as well as on the red free.
b) If an IRMA can only be seen on a red free image and not on the colour image a referral should
not be made (returned to annual screening).
a) and b) assumes screen settings, colour balance, monitor, software and camera settings are
optimal according to the recommendations of the NHS Diabetic Eye Screening Programme.
c) if there is a localised patch of possible IRMA in one area of the retina with very little other
signs of diabetic retinopathy, one needs to consider whether a small branch vein occlusion may
have occurred in this area in the past and that these might be small collaterals. If it is judged that
small collaterals are present from an old small vein occlusion instead of IRMA, this would not
warrant a referral.

C. Venous beading patients with venous beading should be referred. Venous beading does
not occur in isolation from multiple blot haemorrhages or IRMA.
D. Venous reduplication patients with venous reduplication should be referred.

v1.0 6 March 2012

26

Relationship of English Diabetic Retinopathy Classification of Progression to Proliferative DR with ETDRS & Scottish
ETDRS final
Retinopathy
Severity Scale1

ETDRS
(Final)
Grade

Lesions

No apparent
retinopathy

10
14, 15

DR absent
DR questionable

Mild NPDR

20

Micro aneurysms only

35
a
b
c
d
e

One or more of the following:

Venous loops > definite in 1 field
SE, IRMA, or VB questionable
Retinal haemorrhages present
HE > definite in 1 field
SE > definite in 1 field

Level 30 =
6.2%

43a

H/Ma moderate in 4-5 fields or

severe in 1 field or
IRMA definite in 1-3 fields

Level 41 =
11.3%

Moderate
NPDR

Risk of
progression to
PDR in 1 year
(ETDRS
Interim)

(ETDRS:
Grade 0 = no evidence of IRMA
Grade 1 = questionable IRMA
Grade 2 = IRMA present < standard photo
8A
Grade 3 = IRMA present > standard photo
8A but < standard photo 8B
Grade 4 = IRMA > standard photo 8B)

v1.0 6 March 2012

ETDRS Screening
/ Clinic follow up
intervals

English Screening
Programme levels

Scottish Grading Classification

R0
Currently screen Annually

R0
Currently screen
Annually

1 year

R1

R1

4-6 months

Screen annually

Screen annually

Background
microaneurysm(s)
Retinal haemorrhage(s) any
exudate

Background
dot haemorrhages
microaneurysms, hard exudates
cotton wool spots, blot haemorrhages superficial/
flame shaped haemorrhages

R2
Refer to ophthalmologist
Pre-proliferative
multiple blot haemorrhages
intraretinal microvascular
abnormality (IRMA)
venous beading
venous reduplication
It is recommended that
venous loop is removed from

R2
Background diabetic retinopathy BDR
observable

3-6 months

Rescreen 6 months
Four or more blot haemorrhages (i.e. _AH
standard photograph 2a in one hemi-field
only

27

Moderately
severe NPDR

47
a

b
c
d

Both level 43 characteristics

H/Ma moderate in 4-5 fields or severe
in 1 field and IRMA definite in 1-3
fields
or any one of the following:
IRMA in 4-5 fields
HMA severe in 2-3 fields
VB definite in 1 field

Level 45 =
20.7%
4 months

the English Diabetic Eye

Screening Programme
referral criteria.

R3 Background diabetic retinopathy BDR

referable
Any of the following features:
Four or more blot haemorrhages (i.e. _AH
standard photograph 2a in both inferior and
superior hemi-fields
Venous beading standard photograph 6a
IRMA standard photograph 8a

1. Fundus photographic risk factors for progression of diabetic retinopathy. ETDRS report number 12. Early Treatment Diabetic Retinopathy Study Research Group.
Ophthalmology 1991; 98:823-33.

v1.0 6 March 2012

28

2. Revised Grading Classification for Proliferative DR (R3)

R3 will be split into
R3a (Active Proliferative Retinopathy)
And
R3s (Stable Treated Proliferative Retinopathy)
This allows a logical progression from R3 for patients and a robust monitoring pathway for discharged patients.
All newly occurring R3 patients will remain R3a (Active)
The Definition of R3s (Stable) will be
Evidence of Peripheral Retinal Laser Treatment
AND
Stable retina from photograph taken at or shortly after discharge from the HES
Guidance to the discharging ophthalmologist must make it clear that the ENSPDR only operates an annual
screening programme and that they should only discharge patients who they assess are at sufficiently low risk to
receive 12 monthly photographic screening. Assuming that the screening programme has an OPDR pathway it is
recommended that stable treated retinopathy is kept in the OPDR pathway.
Pathway
On discharge, the hospital must either place a discharge set of images on the Screening Service software, supply a
set of images electronically for the service to import or arrange for a set of discharge images to be taken by the
Screening Service within 3 months.
When such patients are screened subsequently, their images must be compared with the baseline discharge
image before deciding the grade. Assuming that the screening programme has an OPDR pathway it is
recommended that stable treated retinopathy is kept in the OPDR pathway.
The only R grades that will be allowed for such patients are R3a and R3s.
The Grading would be R3s if there are no significant changes from the baseline discharge images.
If there are significant changes then the patient would revert to R3a (Active) and be urgently referred back to the
HES. It is accepted that not all changes will in fact be clinically urgent but the committee feel that it is better to
keep things simple and not introduce the concept of routine referral of R3.
Guidance on Significant changes would include:Urgent re-referral:
Signs of active neovascularisation including active new vessels, pre-retinal or vitreous haemorrhage.

v1.0 6 March 2012

29

3. Revised Grading Classification for Groups of Exudates

Groups of Exudates should be defined as:

A group of exudates is an area of exudates that is greater than or equal to half the disc area
and
this area (of greater than or equal half the disc area) is all within the macular area

Guidance:
In order to ascertain the area of exudates the outer points of the exudates are joined and compared to half the
area of the optic disc.

Examples of referable groups of exudates are given below are given below as well as photographic images that
are not referable.

v1.0 6 March 2012

30

Example of an area of exudates that is less than half a disc area is given in photo GE 1 and would not be referred.
Photo GE 1

v1.0 6 March 2012

31

Example of an area of exudates that is less than half a disc area which is borderline in size but there is less than
half a disc area within the macular area is given in photo GE 2 and would not be referred.
Photo GE 2

v1.0 6 March 2012

32

Example of an area of exudates that is greater than half a disc area is given in photo GE 3 and this would be
referred.
Photo GE 3

v1.0 6 March 2012

33

Example of an area of exudates that is greater than half a disc area is given in photo GE 4 and this would be
referred.
Photo GE 4

v1.0 6 March 2012

34

4. Revision to Classification of Image Quality

Changes:
a) the main change will be that there will no longer be a Good category, just Adequate and Inadequate.
b) The concept of jig sawing is introduced, which is where a series of images can be combined to provide retinal
views of the same areas as adequate macular and nasal images.

Hence the Definition of Adequate Image Quality will be:

Photographers should capture 2 x nominal 45 fields per eye (1 x fovea centred, 1 x disc centred).
A combined assessment of field position and image quality should be made for each eye.
Images must be graded for diabetic eye disease only if the grader is confident the quality is sufficient.
All grading is to be performed by trained and accredited staff.

A combined assessment of field position and image quality is made in the software as follows:
ADEQUATE
Macular image
centre of fovea >2DD from edge of image
& vessels visible within 1DD of centre of fovea
and
Disc image
complete optic disc >2DD from edge of image
& fine vessels visible on surface of disc

INADEQUATE
Decision process for allocating images of an eye inadequate image quality is the failure to meet definition of
adequate above with the provisos given below:
1. If sight threatening retinopathy (STDR) is present on any image, the eye should be graded as adequate
and patient referred to HES.
2. Make sure there is a macula and disc / nasal image for the eye; absence of either view means the fields
are inadequate for grading except if criteria in point 1 is fulfilled.
3. Look at all the images available for the eye; if fine vessels are visible within 1DD of centre of fovea on any
image available and fine vessels are visible on the surface of disc on any image available, the eye can be
graded for R and M level. This could be by jig sawing a series of images so that an adequate view is
obtained of critical areas.

v1.0 6 March 2012

35