I now have comprehensive material. Let me compile a complete answer on the ETDRS study.
ETDRS — Early Treatment Diabetic Retinopathy Study
Overview
The Early Treatment Diabetic Retinopathy Study (ETDRS) was a landmark multicenter, randomized controlled trial sponsored by the National Eye Institute (NEI), USA. It enrolled over 3,700 patients with mild-to-severe NPDR or early PDR in both eyes, randomizing them to answer two primary clinical questions:
- When should photocoagulation be initiated in diabetic retinopathy — early (before high-risk PDR develops) or deferred?
- Is aspirin effective in altering the course of diabetic retinopathy?
It also established the ETDRS classification of diabetic retinopathy and defined clinically significant macular oedema (CSMO), both of which remain in universal clinical use today.
Study Design
| Feature | Detail |
|---|
| Study type | Randomized controlled trial |
| Enrollment | ~3,711 patients (mild-to-severe NPDR or early PDR) |
| Eye allocation | Each eye randomized independently — one eye to early photocoagulation, fellow eye deferred |
| Aspirin arm | Patients randomized to aspirin 650 mg/day vs. placebo |
| Primary outcome | Severe visual loss (≥doubling of visual angle, e.g., 6/12 → 6/24) |
| Follow-up | Up to 9 years |
Key Findings
1. Photocoagulation Timing
| Finding | Detail |
|---|
| Macular laser for CSMO | Focal/grid photocoagulation reduced the risk of moderate visual loss (loss of ≥3 lines) by ~50% compared with observation |
| When to treat CSMO | Treat immediately when CSMO is present — do not defer |
| Early scatter (PRP) in severe NPDR/early PDR | Reduced risk of severe visual loss but was not uniformly beneficial at early stages — recommended to defer PRP in eyes with less than high-risk PDR, unless close follow-up is not possible |
| High-risk PDR | Prompt PRP is indicated; reduces risk of severe visual loss by >50% (building on the DRS findings) |
| Early PRP in mild-moderate NPDR | Not recommended — risks of early treatment (mild visual field loss, slight reduction in VA) outweigh benefits at this stage |
Key conclusion: "In eyes with CSMO, initiate focal photocoagulation. In eyes with high-risk PDR, initiate PRP promptly. In eyes with less severe retinopathy, careful observation is acceptable if follow-up is reliable."
2. Aspirin
- Aspirin 650 mg/day did NOT affect the progression of diabetic retinopathy — neither beneficial nor harmful on retinopathy outcomes
- Aspirin did not increase risk of vitreous hemorrhage
- Therefore: aspirin is NOT contraindicated in diabetic patients who require it for cardiovascular indications — do not withhold it to protect the eyes
3. ETDRS Classification of Diabetic Retinopathy
The ETDRS adopted the modified Airlie House classification, graded from 7-field stereoscopic fundus photographs. This is the internationally accepted grading system:
| ETDRS Category | Key Features | Risk of Progression |
|---|
| No DR | Normal | — |
| Very mild NPDR | Microaneurysms only | Low |
| Mild NPDR | MA + haemorrhages/exudates/CWS, below moderate NPDR threshold | Low–moderate |
| Moderate NPDR | Severe haemorrhages in 1–3 quadrants, OR mild IRMA, OR venous beading in ≤1 quadrant | PDR in 26%, high-risk PDR in 8% within 1 year |
| Severe NPDR | 4-2-1 rule: ≥1 of — haemorrhages in all 4Q / VB in ≥2Q / IRMA in ≥1Q | PDR in 50%, high-risk PDR in 15% within 1 year |
| Very severe NPDR | Two or more criteria of severe NPDR | PDR in 75% within 1 year |
| Early PDR | New vessels, mild | — |
| High-risk PDR | Any NV + VH/PRH; or NVD > ⅓ disc area; or NVE > ½ disc area | Immediate risk of severe visual loss |
| Advanced diabetic eye disease | TRD, persistent VH, NVG | — |
— Kanski's Clinical Ophthalmology, p. 522
4. Clinically Significant Macular Oedema (CSMO)
The ETDRS defined CSMO — still the standard clinical definition — as any one of the following:
- Retinal thickening within 500 μm (≈ ⅓ disc diameter) of the centre of the macula
- Hard exudates within 500 μm of the centre with adjacent retinal thickening (the thickening itself may extend beyond 500 μm)
- Retinal thickening ≥1 disc area, any part of which is within 1 disc diameter of the centre of the macula
CSMO is an indication for focal/grid laser photocoagulation, regardless of the level of visual acuity. (In contemporary practice, centre-involving CSMO is preferentially treated with anti-VEGF agents — DRCR.net Protocol I and T.)
— Kanski's Clinical Ophthalmology, p. 525
The 4-2-1 Rule (Severe NPDR) — ETDRS Derived
An easy mnemonic for severe NPDR (any one criterion present):
- 4 quadrants with severe intraretinal haemorrhages
- 2 quadrants with venous beading
- 1 quadrant with IRMA
Two or more = very severe NPDR.
ETDRS Visual Acuity Chart
A secondary legacy of the ETDRS is the standardized logMAR visual acuity chart used in clinical trials:
- Uses Sloan optotypes on a 5×5 grid
- Balanced rows: each row has equivalent legibility across letters
- Calibrated for 4 metres
- 5 letters per row; each letter = 1 logMAR unit ÷ 5 = 0.02 logMAR per letter
- Allows more precise measurement than Snellen charts — preferred for clinical trials
— Kanski's Clinical Ophthalmology, p. 18
Clinical Impact & Legacy
| Contribution | Impact |
|---|
| CSMO definition | Universal — still the benchmark for treatment threshold |
| DR classification | Modified Airlie House (ETDRS) grading adopted worldwide |
| Macular laser technique | Defined focal/grid laser parameters for 30+ years |
| Aspirin safety | Confirmed aspirin can be used safely in DR patients needing it |
| Timing of PRP | Established the principle: treat high-risk PDR promptly, observe early disease with reliable follow-up |
| logMAR chart | ETDRS chart became the gold standard for VA measurement in research |
The ETDRS, together with the Diabetic Retinopathy Study (DRS) for PRP and the subsequent DRCR.net protocols, forms the evidence foundation of diabetic eye disease management. Contemporary practice has evolved beyond the ETDRS macular laser findings with anti-VEGF therapy (Protocol I, T), but the ETDRS classification, CSMO definition, and aspirin conclusions remain foundational.