The DR evaluation procedures for general diabetes care practice have been developed for clinical physicians’ daily practice, adapted from the “Early Treatment of Diabetic Retinopathy Study (ETDRS)” classification system. Quality control process plays a key role in running these programs.

There are three types of DR evaluation procedures for general health care practice, the screening reading, the diagnostic reading and the quantitative diagnostic reading.

DR Screening Reading

The screening reading procedure (Level I DR image reading) is designed for the ease of practice for some of the primary care providers to start their work immediately without difficulty. Screening readers will do the same tasks for the photograph reading as the general diagnostic readers, except that the screening readers will not be asked to differentiate the morphology of some subtle retinal changes, e.g., microaneurysms vs. spot hemorrhages, hard exudates vs. drusen, for the ease of the readers. 

Screening reading should be carried out by the clinical physician him/her self for their own patients. Photographs with any positive finding from the screening reading will be automatically transferred to an appointed DR reading center for general diagnostic reading.

The main purpose of screening reading is to detect the existence of diabetic retinal abnormality. It is also expected that the screening readers will be able to perform (and transformed to) the general reading after a while with accumulated knowledge and skill for DR reading.

Screening reading is applied only under the circumstance that general diagnostic reading is not possible for the clinical physician.  

DR General Diagnostic Reading

General diagnostic reading (Level II DR image reading) is developed for general clinical practice. The main task of general diagnostic reading procedure is to detect the existence of the pathological elements of DR and observe their changes for diabetes care follow-up.

General diagnostic reading may be performed either by clinical physicians him/her self (named physician’s reading, i.e., the physicians perform the reading for his/her own patients), or by a third-party DR reading center service (named third-party reading). 

In cases that a sight-threatening DR pathology is noticed by the general diagnostic reading, the clinical physicians may either refer the patient to an eye specialist for further mydriatic fundus photography (if the current photographs covers a fundus field less than 90 degree), or send the photographs to a third-party DR reading center for quantitative reading (if the photographs covers a fundus field above 90 degree). Further ophthalmic examination and treatment may then be recommended. Patients with non-diabetic retinopathy, if detected, and those with unreadable fundus images, are also indications for ophthalmic referral.

DR Quantitative Diagnostic Reading

The procedure for grading the severity of diabetic retinopathy developed by ETDRS, which is recognized as the “gold standard” for grading the severity of diabetic retinopathy in clinical trials, is adapted as the quantitative diagnostic DR reading in the present project for serving general diabetes care practice. The quantitative diagnostic DR reading could be used for observation and clinical follow-up of the subtle changes of the DR pathology in diabetic patients. For serving the eye care practice at the same time, result of DR severity score by ETDRS standard will also be converted to International DR scale.

Theoretically, quantitative diagnostic DR reading might be performed for any patient with diabetes. It is the best of choice in terms of DR image diagnosis itself, which may provide a complete picture of the patient’s DR conditions. 

However, the present program focuses mainly on practical issues of the diabetes care in clinical practice. Therefore, from the practical point of view, quantitative diagnostic DR reading is recommended only for those cases that a sight-threatening DR pathology is detected/suspected with the general diagnostic procedure.

It is considered impractical and unnecessary to perform this procedure as a clinical routine for all diabetes patients, as the following:

  1. Quantitative diagnostic DR reading requires fundus photographs covering 90 degree fundus field. When conventional optical fundus camera is used (single field covers less than 50 degree), a mydriatic ETDRS 7-field fundus photography is required. Contraindications of mydriasis should be evaluated and eliminated before mydriatic eye drops applied to the patient’s eyes, which should be done in an eye care facility, which will create inconvenience for the patients,  increased medical risks for the patients, and increased social cost of manpower (eye doctors), and the added financial cost to the patients.
  2. Non-mydriatic photography have high enough sensitivity and specificity for detection sight-threatening DR pathology.  In the cases that a sight-threatening DR pathology is detected/suspected in the general diagnostic procedure, a mydriatic/wide-field photography and quantitative diagnostic DR reading will then be indicated. [For those with unreadable fundus images, referral to an eye specialist for mydriatic photography is also indicated.]
  3. A wide-field laser scanning fundus photograph may be used for achieving wide-enough fundus photograph without mydriasis, but high cost of this instrument is the barrier for the application of this technology in most of the countries. [In cases that non-mydriatic photographs were taken with wide-field  (>90 degree) laser scanning funds camera, the photographs may directly be sent to a third-party DR reading center for quantitative diagnostic DR reading. ]
  4. The ETDRS DR grading system involves very complicated procedures, which is impractical for clinical physicians to have it done by themselves. Image diagnosis service from a specialized third-party independent DR reading center, therefore, is mandatory.


* Quantitative diagnostic DR reading is the choice of DR image diagnosis for clinical trials.