Skip to Content

Retinopathy of Prematurity (ROP): Screening Windows, Zones, Stages & Treatment (2026 Guide)

A premature baby's retina is still being built at birth. Why ROP happens in two opposite phases, how examiners read zone, stage and plus disease, why India screens a wider net of babies than the West, and what the ETROP, BEAT-ROP and RAINBOW trials settled about when and how to treat.
Retinopathy of Prematurity (ROP) 2026: Screening, Zones, Stages & Treatment | Agaaz Ophthalmics

Paediatric Retina · Surgeon Series

The retina of a
premature baby is
still being built.

Retinal blood vessels finish growing only around full term. A baby born early has an unfinished retina, and what happens next — quiet completion or a wave of abnormal vessels that peels the retina off — is decided over a handful of weeks, visible only on a dilated examination nobody outside the nursery ever sees.

~32,000infants worldwide blind or
severely impaired by ROP (2010)
72 hrsconventional window to treat
Type 1 disease
Zone Ithe most posterior and
most dangerous territory
14 minreading time

Section 01 — The Mechanism

Two phases, and only
the second one
does the damage.

Retinopathy of prematurity is often described as "eye damage from oxygen in incubators." That description is old, partial and misleading. ROP is a developmental disorder of retinal blood vessels, and understanding it requires holding two opposite events in mind, separated by weeks.

In normal gestation, retinal vessels sprout from the optic disc at around the fourth month and creep outward across the retina, reaching the nasal edge earlier and the temporal edge last, at roughly term. A baby delivered at 27 weeks therefore arrives with a large rim of peripheral retina that has no blood supply yet and was never supposed to need one at that point.

Phase 1 — vaso-obliteration. After birth the baby moves from the low-oxygen environment of the womb into room air or supplemental oxygen, which is comparatively hyperoxic for that retina. Relative hyperoxia suppresses the growth factors driving vessel extension. The advancing vascular front stalls, and some already-formed vessels regress. The retina is now not just unfinished but arrested.

Phase 2 — vaso-proliferation. As the infant grows, the metabolic demand of that avascular peripheral retina rises while its blood supply does not. It becomes genuinely hypoxic and responds the way ischaemic retina always responds: by releasing vascular endothelial growth factor (VEGF) and related mediators. Vessels now regrow, but disorganised — piling into a ridge at the junction with vascularised retina and, in severe disease, breaking out of the retinal plane into the vitreous as fragile fronds that bleed, contract and drag the retina off its wall.

The core answer, in 120 words

ROP happens because preterm birth interrupts retinal vessel growth mid-way. First the relatively oxygen-rich postnatal environment halts that growth and leaves peripheral retina without blood supply; then, as the baby grows, the starved retina floods the eye with VEGF and vessels return in a destructive, disorganised form. Most ROP regresses on its own. A minority progresses to a stage where abnormal tissue contracts and detaches the retina, and that outcome is irreversible. Because the dangerous phase is silent, painless and invisible to parents and to an undilated look at the eye, the entire clinical strategy rests on one thing: examining the right babies, on schedule, until the retina is finished.

Why prematurity matters
Retinal vessels normally reach the peripheral retina only near term. The earlier the delivery, the larger the avascular rim and the greater the later hypoxic drive.
Why oxygen matters
Supplemental oxygen contributes to phase 1 vessel arrest. It is one risk factor among several, not the whole disease — and reducing it carries its own neonatal cost.
Why VEGF matters
The same growth factor that normally guides orderly vessel extension drives disorganised proliferation when released by ischaemic retina. That is why anti-VEGF drugs work.
Why timing matters
Phase 2 follows the retina's developmental clock, so screening is scheduled against postmenstrual age rather than days since birth.

Section 02 — Interactive

How an examiner reads
a retina: zone, stage, plus.

Every ROP examination produces three pieces of information, and the treatment decision falls out of their combination. Set the zone, stage and plus status below and watch the classification — and the treatment threshold — resolve.

Interactive: ICROP zone & stage classifier
A schematic right fundus seen from in front, centred on the optic disc. Zone boundaries, ridge morphology and vessel tortuosity respond to the controls. Illustrative teaching model — not a diagnostic tool and not to anatomical scale.
Zone Plus disease
Stage 3
drag to tilt
Vascularised retina & vessels Ridge / proliferation Active zone boundary

Zone answers how far did normal vessels get? — measured as a circle centred on the optic disc. Zone I is a small posterior circle with a radius of twice the disc-to-fovea distance; Zone II is the annulus extending out to the nasal ora serrata; Zone III is the residual temporal crescent beyond it. Posterior disease is worse disease, because a Zone I eye has almost no vascularised retina and an enormous ischaemic drive behind whatever is seen at the ridge.

Stage answers what does the junction look like? — a flat white demarcation line is stage 1, a ridge with volume is stage 2, a ridge with fibrovascular tissue growing off it into the vitreous is stage 3, partial retinal detachment is stage 4 (4A sparing the fovea, 4B involving it) and total detachment is stage 5, which the third edition of the international classification subdivides further by the configuration of the funnel.

Plus disease answers how active is it right now? — and it is recorded at the posterior pole, not at the ridge. Dilated, tortuous arterioles and venules near the disc indicate high blood flow driven by high VEGF. Pre-plus describes abnormality insufficient for plus. The third edition explicitly reframes this as a continuous spectrum rather than a yes/no photograph comparison, which matters because plus disease is the variable that most often converts an observation decision into a treatment decision.

The one term that has changed recently. What used to be called aggressive posterior ROP (AP-ROP) is, in the 2021 third edition of the international classification, simply aggressive ROP (A-ROP). The word "posterior" was dropped because the same rapid, flat, ill-defined disease with severe plus — progressing to detachment without passing tidily through stages 1, 2 and 3 — is well documented in more peripheral zones and in larger babies, particularly in low- and middle-income settings. It does not wait, and it does not look like textbook staged disease.

Section 03 — Who Gets Screened

Different countries
draw the line
in different places.

Screening criteria are not a universal biological constant. They are a bet about where the disease will show up in a given population's babies, and the bet differs between health systems with different neonatal survival profiles and different oxygen-monitoring capability.

The 2018 joint policy statement from the American Academy of Pediatrics, the American Academy of Ophthalmology and AAPOS recommends screening infants with a birth weight of 1500 g or less, or a gestational age of 30 weeks or less, together with selected larger and more mature infants whose neonatal course was unstable in the judgement of the treating team. That threshold reflects a setting where very preterm infants receive tightly titrated oxygen and continuous saturation monitoring.

India's national newborn eye-screening guidance deliberately casts a wider net — commonly cited as screening babies born at about 34 weeks or below, or weighing about 2000 g or below, with further widening for risk factors. The reason is empirical rather than precautionary: Indian series have repeatedly described severe, treatment-requiring ROP in bigger and more mature babies than Western criteria would ever have captured, a pattern often shorthanded as "bigger baby ROP" and generally attributed to variability in oxygen delivery and monitoring across a very large and heterogeneous neonatal care landscape.

ElementAAP / AAO / AAPOS 2018 policyIndia — national newborn eye-screening guidance
Birth weight trigger≤1500 gApproximately ≤2000 g
Gestational age trigger≤30 weeksApproximately ≤34 weeks
Larger/older babiesScreened selectively, on the neonatal team's judgement of unstable courseScreened where risk factors are present; wider net by design
First examinationConventionally around 4 weeks chronological age or 31 weeks postmenstrual age, whichever is later, with the exact week set by gestational age at birth
Endpoint of screeningFull retinal vascularisation, or unambiguous regression, or completion of treatment follow-up — not a fixed number of visits
Why the wider net is the safer error. Missing a baby who needed screening produces an irreversible outcome. Screening a baby who did not need it produces one uncomfortable but safe examination. The asymmetry is total, and it is the single strongest argument for erring wide wherever neonatal oxygen practice varies between units — which is most of the world.

The logistics problem, and telemedicine

Screening is not limited by knowledge; it is limited by there being an ophthalmologist trained in indirect ophthalmoscopy physically present in a neonatal unit on the right week. That is a hard constraint in rural and district hospitals everywhere. Wide-field retinal imaging by trained non-physician technicians, with images read remotely, has been deployed at scale to close exactly that gap — the KIDROP programme in Karnataka being one of the better-documented models, reporting screening of infants in outreach centres far from any retinal specialist. The model does not replace an examiner; it relocates the examiner's eye.

Section 04 — The Treatment Threshold

Type 1 disease, and
the trial that moved
the line earlier.

For most of the modern history of ROP, treatment waited for "threshold" disease as defined by the CRYO-ROP trial of the late 1980s — the first study to show that ablating the avascular peripheral retina reduced unfavourable outcomes at all. The question that took another fifteen years to answer was whether waiting that long was itself costing vision.

The Early Treatment for Retinopathy of Prematurity (ETROP) randomised trial answered it. Published in 2003, it compared treatment at the older threshold against earlier treatment of high-risk pre-threshold eyes, and found better structural and visual acuity outcomes with the earlier intervention. Its practical legacy is the Type 1 / Type 2 vocabulary still used at every cotside today.

CategoryFindingsConventional action
Type 1 ROPZone I, any stage, with plus disease — or Zone I stage 3 without plus — or Zone II stage 2 or 3 with plus diseaseTreat, conventionally within about 72 hours
Type 2 ROPZone I stage 1 or 2 without plus — or Zone II stage 3 without plusDo not treat yet; re-examine at a short interval and treat on progression to Type 1
Aggressive ROP (A-ROP)Rapid, flat, ill-defined disease with prominent plus, not progressing through orderly stagesTreat urgently — conventional stage-based waiting is unsafe here

The 72-hour figure is a clinical convention drawn from the pace at which Type 1 disease actually moves, not a regulatory rule or a number the trial randomised against. The underlying point stands regardless of the exact figure: this is one of the few situations in ophthalmology where an appointment slipping by a week can change a child's lifelong visual outcome.

0conventional window
to treat Type 1
0postmenstrual age anchor
for the first exam
0in the international
classification

Section 05 — Laser or Injection

Burn the periphery,
or block the signal.

Once treatment is indicated there are two established options, and they attack the same disease from opposite ends. Laser removes the tissue producing the growth-factor signal. Anti-VEGF neutralises the signal and leaves the tissue in place. Neither is a universally correct answer.

Laser photocoagulation ablates the avascular peripheral retina, typically under near-confluent burns delivered by indirect laser in a single sitting. It ends the ischaemic drive permanently, which is its great strength: once the peripheral retina is ablated it cannot restart the process. The costs are equally permanent — peripheral visual field is sacrificed by design, and treated eyes tend toward myopia, often high myopia, in later childhood.

Intravitreal anti-VEGF injects a very small dose of an anti-VEGF agent into the vitreous, suppressing the proliferative drive and allowing peripheral vessels to resume growing in a more normal pattern. The BEAT-ROP trial, published in the New England Journal of Medicine in 2011, randomised infants with stage 3+ ROP to intravitreal bevacizumab or conventional laser and found a significant advantage for bevacizumab specifically in Zone I disease — the very population where laser performs worst because so much retina must be destroyed. The RAINBOW trial, published in The Lancet in 2019, compared ranibizumab against laser in very low birthweight infants and supported ranibizumab as a treatment option, with ranibizumab subsequently gaining regulatory approval for ROP in several jurisdictions.

DimensionLaser photocoagulationIntravitreal anti-VEGF
MechanismDestroys the ischaemic retina producing VEGFNeutralises circulating intraocular VEGF
Peripheral retinaPermanently ablated — field loss by designPreserved; vascularisation can continue outward
Zone I diseaseDifficult — a very large area must be treatedTrial evidence favours anti-VEGF in this specific subgroup
ReactivationUncommon once ablation is adequateDocumented, sometimes weeks to months later — demands longer, disciplined follow-up
Refractive outcomeTendency toward myopia, frequently highGenerally less myopigenic in comparative reports
Open questionsField loss quantification in later lifeSystemic VEGF suppression in a developing infant; optimal dose
Practical demandsLaser equipment, anaesthetic support, one definitive sittingSterile injection setting; mandatory extended surveillance
Definitiveness of a single adequate treatmentLaser favoured
Outcome in Zone I stage 3+ diseaseAnti-VEGF favoured
Peripheral retina preservedAnti-VEGF favoured
Length and intensity of follow-up requiredAnti-VEGF higher burden

Illustrative visualisation of the direction of difference described across the trial literature and subsequent reviews (BEAT-ROP 2011, RAINBOW 2019 and related work) — see References. Not pooled meta-analytic effect sizes, and not a substitute for a clinician's judgement in an individual eye.

The follow-up caveat is the whole caveat. Anti-VEGF's advantage assumes the child comes back. Where families travel long distances, where a discharge from the neonatal unit is treated as the end of the episode, or where no system exists to recall an infant weeks later, the reactivation risk is not a footnote — it is the deciding variable, and it can make laser the safer choice in that setting even where a trial would favour the injection.

Section 06 — The Oxygen Trade-off

Less oxygen means
less ROP. It also
meant more deaths.

Because phase 1 of ROP is driven by relative hyperoxia, the obvious prevention strategy is to target lower oxygen saturations in preterm infants. This was tested properly, and the result is one of the most uncomfortable findings in neonatal medicine — and an important corrective to any account of ROP that treats oxygen as a simple dial to turn down.

The SUPPORT trial, published in the New England Journal of Medicine in 2010, randomised extremely preterm infants to a lower (85–89%) or higher (91–95%) oxygen saturation target. The lower-target group had substantially less severe retinopathy. The same group also had higher mortality. The NeOProM collaboration's prospective meta-analysis of the individual trials, published in JAMA in 2018, pooled this evidence and reported the same direction: lower saturation targeting was associated with a higher risk of death, alongside a lower risk of treated retinopathy.

What this means in practice — and what it does not. It does not mean oxygen is harmless or that titration does not matter; careful, monitored oxygen delivery remains central to reducing severe ROP. It means the target range is a neonatal decision balancing survival against retinopathy, made by the neonatal team, not an ophthalmic preference. Any guide that tells parents ROP is simply "caused by too much oxygen in the incubator" is describing a trade-off as if it were a mistake.

The practical consequence for ophthalmology is unglamorous and unchanged: because oxygen exposure cannot be reduced to zero risk without cost elsewhere, ROP will keep occurring in surviving preterm infants, and screening remains the intervention that actually prevents blindness.

Section 07 — Afterwards

Regression is not
the same as discharge.

Most ROP regresses. That is the single most reassuring fact in this subject and the one most often mistaken for the end of the story. An eye that had ROP — treated or spontaneously regressed — remains a different eye, and the follow-up schedule should reflect that.

Myopia, often high
Strongly associated with a history of ROP, and more marked after laser. Uncorrected high myopia in a young child is itself an amblyopia risk.
Strabismus & amblyopia
Both occur at higher rates after ROP, and both are treatable — but only within the visual-development window of early childhood.
Reactivation after anti-VEGF
Documented weeks to months after injection. This is the specific reason anti-VEGF treatment obliges a longer surveillance schedule than laser.
Late structural risk
Dragged discs, macular ectopia, angle-closure glaucoma and late retinal detachment are described years later in eyes with a significant ROP history.
Illustrative Scenario — Composite, Not an Individual Patient
The baby who was never on the list

A baby born at 33 weeks weighing 1750 g, ventilated briefly for respiratory distress and discharged home in good condition, would fall outside a screening criterion set at 30 weeks or 1500 g. Under India's wider criteria the same baby is screened, and it is precisely in this band that Indian series have described treatment-requiring disease — including aggressive forms that do not wait for an orderly progression through stages.

The failure mode here is not misdiagnosis. It is that the examination never happens, because nobody was told the baby needed one. By the time a parent notices something wrong with a two-year-old's vision, phase 2 finished eighteen months earlier.

Illustrative composite constructed from published screening criteria and the "bigger baby ROP" pattern reported in Indian literature — not a specific patient record.

For parents, the actionable version is short: if a baby was born preterm, ask explicitly whether a retinal examination has been scheduled, get the date in writing before discharge from the neonatal unit, and keep it even if the baby looks perfectly well — because at the stage where it is still fixable, the baby always does.

Section 08 — FAQ

Frequently asked questions
about retinopathy of prematurity.

Retinopathy of prematurity (ROP) is a disorder of the developing retinal blood vessels in babies born preterm. In the womb, retinal vessels grow outward from the optic nerve and normally reach the edge of the retina only near full term. A baby born early has an incompletely vascularised retina. After birth the relatively oxygen-rich environment first slows or halts that vessel growth, leaving peripheral retina without a blood supply. As the baby grows, that starved retina releases growth factors including VEGF, and vessels then regrow abnormally — as a ridge of tissue and, in severe disease, as fragile vessels that grow into the vitreous, scar and pull the retina off. It is a two-phase disease, and the damaging phase is the second one.

Thresholds differ by country. The American Academy of Pediatrics joint 2018 policy recommends screening infants with a birth weight of 1500 g or less or a gestational age of 30 weeks or less, plus selected larger or more mature infants whose clinical course has been unstable. India's national newborn eye-screening guidance uses a deliberately wider net — commonly cited as all babies born at about 34 weeks or less, or weighing about 2000 g or less — because Indian series have repeatedly reported severe ROP in bigger, more mature babies than Western criteria would capture. Any premature baby who received prolonged supplemental oxygen or had a stormy neonatal course should be examined regardless of where the numbers fall.

The first dilated examination is conventionally timed at about four weeks of chronological age or about 31 weeks postmenstrual age, whichever comes later, with the exact first-visit week set by gestational age at birth. The reason for anchoring to postmenstrual age is that ROP follows the retina's developmental clock more closely than the calendar since delivery. Follow-up intervals — typically one, two or three weeks — are then set by what the examiner finds, and screening continues until the retina is fully vascularised or the disease has clearly regressed.

Zone describes how far normal vessels have grown, measured from the optic disc. Zone I is the small posterior circle centred on the disc, Zone II the annulus beyond it, and Zone III the remaining temporal crescent. The more posterior the disease, the more dangerous it is, because less retina has been vascularised. Stage describes what the junction between vascularised and non-vascularised retina looks like: stage 1 a flat demarcation line, stage 2 a raised ridge, stage 3 a ridge with extraretinal fibrovascular proliferation, stage 4 partial retinal detachment and stage 5 total detachment. Plus disease — abnormal dilation and tortuosity of the posterior pole vessels — is recorded separately and is the single strongest indicator of activity.

Type 1 ROP is the treatment threshold defined by the Early Treatment for Retinopathy of Prematurity (ETROP) randomised trial: Zone I with any stage plus plus disease, Zone I stage 3 without plus disease, or Zone II stage 2 or 3 with plus disease. ETROP showed that treating at this earlier point produced better visual and structural outcomes than waiting for the older, more advanced threshold criteria. Treatment is conventionally advised within roughly 72 hours of the diagnosis being made because Type 1 disease can progress to retinal detachment over days, not months.

Both are in routine use and the choice is case-specific. Laser photocoagulation ablates the avascular peripheral retina, removing the source of the growth-factor drive; it is definitive, it does not depend on a drug reaching the eye, and its main cost is a permanently ablated peripheral retina and a tendency toward myopia. Intravitreal anti-VEGF blocks the drive chemically and allows peripheral vessels to continue growing; the BEAT-ROP trial found bevacizumab advantageous over laser specifically for Zone I stage 3+ disease, and the RAINBOW trial supported ranibizumab as an alternative to laser. The trade-off with anti-VEGF is that disease can reactivate weeks later, so follow-up must be longer and more disciplined, and questions about systemic VEGF suppression in a developing infant have not been fully closed.

Many do. Mild ROP frequently regresses spontaneously with no lasting visual consequence. But a history of ROP — treated or regressed — carries a higher lifetime rate of myopia, often high myopia, and also of strabismus, amblyopia, glaucoma and late retinal detachment, so these children need ophthalmic follow-up through childhood rather than a single discharge visit. Advanced disease that reaches stage 4 or 5 has a far worse prognosis; surgery for a detached premature retina can restore anatomy in some cases but rarely restores good functional vision, which is exactly why the screening window matters so much more than the surgical one.

References & Evidence Base

Peer-reviewed
citations.

Chiang MF, Quinn GE, Fielder AR, et al. "International Classification of Retinopathy of Prematurity, Third Edition." Ophthalmology. 2021;128(10):e51–e68. PubMed 34247850
Early Treatment for Retinopathy of Prematurity Cooperative Group. "Revised indications for the treatment of retinopathy of prematurity: results of the Early Treatment for Retinopathy of Prematurity randomized trial." Arch Ophthalmol. 2003;121(12):1684–1694. PubMed 14662586
Cryotherapy for Retinopathy of Prematurity Cooperative Group. "Multicenter trial of cryotherapy for retinopathy of prematurity. Preliminary results." Arch Ophthalmol. 1988;106(4):471–479. PubMed 2895630
Mintz-Hittner HA, Kennedy KA, Chuang AZ; BEAT-ROP Cooperative Group. "Efficacy of intravitreal bevacizumab for stage 3+ retinopathy of prematurity." N Engl J Med. 2011;364(7):603–615. PubMed 21323540
Stahl A, Lepore D, Fielder A, et al. "Ranibizumab versus laser therapy for the treatment of very low birthweight infants with retinopathy of prematurity (RAINBOW): an open-label randomised controlled trial." Lancet. 2019;394(10208):1551–1559. PubMed 31522845
Fierson WM; AAP Section on Ophthalmology, AAO, AAPOS, AACO. "Screening Examination of Premature Infants for Retinopathy of Prematurity." Pediatrics. 2018;142(6):e20183061. PubMed 30478242
SUPPORT Study Group of the NICHD Neonatal Research Network. "Target ranges of oxygen saturation in extremely preterm infants." N Engl J Med. 2010;362(21):1959–1969. PubMed 20472937
Askie LM, Darlow BA, Finer N, et al. (NeOProM Collaboration). "Association Between Oxygen Saturation Targeting and Death or Disability in Extremely Preterm Infants in the Neonatal Oxygenation Prospective Meta-analysis Collaboration." JAMA. 2018;319(21):2190–2201. PubMed 29872859
Blencowe H, Lawn JE, Vazquez T, Fielder A, Gilbert C. "Preterm-associated visual impairment and estimates of retinopathy of prematurity at regional and global levels for 2010." Pediatr Res. 2013;74(Suppl 1):35–49. PubMed 24366462
Vinekar A, Jayadev C, Mangalesh S, Shetty B, Vidyasagar D. "Role of tele-medicine in retinopathy of prematurity screening in rural outreach centers in India — a report of 20,214 imaging sessions in the KIDROP program." Semin Fetal Neonatal Med. 2015;20(5):335–345. PubMed 26092301 — see also the KIDROP model paper, Indian J Ophthalmol. 2014;62(1):41–49, PubMed 24492500
Ministry of Health & Family Welfare, Government of India. Guidelines for Universal Eye Screening in Newborns including Retinopathy of Prematurity — the national guidance underlying the wider Indian screening criteria described above. Consult the current published edition for exact thresholds, which are periodically revised.
American Academy of Ophthalmology. "Retinopathy of Prematurity" — EyeWiki clinical reference on classification, screening and management. eyewiki.aao.org

Continue Reading

Related guides
from Beyond Vision.

Start writing here...