This entry is short on purpose. It is one of the few medical situations where a normal person, with no training, can make the call that saves the eye — by recognising three specific symptoms and responding the same day. The action is one visit. The cost of being wrong is a few hours of your afternoon. The cost of being right and not going is permanent blindness in one eye.
The inside of your eye is filled with a clear gel called the vitreous. For most of your life it sits quietly against the retina — the thin layer of light-sensing tissue lining the back of the eyeball. With age, the gel liquefies and shrinks, and at some point in middle age or later it peels away from the retina. The medical name is a posterior vitreous detachment, and most people go through it eventually — roughly a quarter of people in their fifties, most people past eighty Bond-Taylor 2017.
Usually that peel happens cleanly and you notice a few new floaters for a couple of weeks and that's that. Sometimes — about one time in seven, when symptoms come on suddenly — the gel is stuck to the retina at some point and yanks hard enough to tear a piece of it on the way off Hollands 2009. A tear in the retina is a hole. Liquid from inside the eye can seep through that hole, get behind the retina, and lift it off the wall like wallpaper pulling away from a damp wall. That is a retinal detachment. The detached tissue has lost its blood supply; the photoreceptors in that patch start dying within hours.
The three warning symptoms are what each step of that sequence feels like from the inside.
- Flashes are the gel tugging on the retina. The pull mechanically stimulates the photoreceptors; your brain has no other language for it and renders it as light — usually a brief lightning streak off to the side, more noticeable in a dim room, in one eye only.
- New floaters — a sudden shower of specks, a big new cobweb, or what looks like a small cloud of pepper drifting across your vision — are debris released by the detaching gel: pigment cells scraped off the back of the eye, sometimes red blood cells if a small vessel tore.
- A curtain or shadow creeping in from one edge of your vision is the retina already coming away. Because the eye's optics flip the image, a curtain coming down from the top of your vision is actually a detachment of the lower retina, and vice versa. When the curtain reaches the centre of your sight, the detachment has reached the macula, and the visual outcome gets much worse, fast Greven 2019.
How much should you trust these three symptoms?
A lot. The clinical-examination evidence on flashes, floaters, and acute vision change in adults is one of the more carefully measured rules in medicine. The benchmark is a synthesis in JAMA pooling seventeen studies of adults who turned up at clinic with sudden flashes or floaters: about one in seven had a retinal tear waiting to be found Hollands 2009. That is not a rare-event probability. That is the kind of number that justifies clearing your afternoon.
Two specific symptom features push the probability much higher. If your vision is also noticeably worse — blurrier, dimmer, a missing patch — the odds of a tear jump roughly five-fold. If the eye doctor finds blood in the gel of the eye (a vitreous hemorrhage) the odds of a tear jump roughly ten-fold, and about two-thirds of patients with that finding have at least one tear Hollands 2009.
The flip side is also worth knowing: about six in seven of those urgent presentations turn out to be an uncomplicated posterior vitreous detachment with no tear. That is not a wasted visit. You have ruled out the dangerous thing — and you now have a known-clean baseline against which any new change in symptoms over the following weeks gets re-examined. Because even after a clean first look, roughly 1 in 50 people develop a tear in the following weeks, and the chance is much higher if the doctor saw any blood or pigment debris on the first visit Coffee 2007, AAO PPP 2025. That is why the standard plan is a same-day exam, then a return visit four to six weeks later, with explicit instructions to come back sooner if anything changes.
What waiting costs you
The clock starts when the macula — the small central patch of retina that handles reading, faces, screens, fine detail — comes off the wall. Before that, repair restores most of your central vision. After that, even a textbook-perfect surgery often leaves the affected eye permanently softer, dimmer, with straight lines that look subtly bent Williamson 2014.
Day one, untreated, what you notice is the curtain creeping a little further in from the side overnight. Reading is fine. Driving is fine. You think you'll see how it is in the morning.
Day two or three, untreated, the curtain reaches the centre. The phone you were holding goes from sharp to a smear. Faces stop being faces. You close the bad eye and the world is normal again, which is the part people remember — the moment you understood that one eye does almost all the work and you had been outsourcing without realising. Studies that tracked this exact window — repair done on day 1 versus day 3 of macula-off detachment — found measurable losses of final vision for every extra day before the operating room Greven 2019, van Bussel 2014.
A week later, the eye that had the detachment has been put back together by a retinal surgeon. The retina is reattached in over 85% of cases on the first surgery Williamson 2014. But "reattached" is not "restored." Many people who waited come out the other side reading the eye chart at the level of someone who needs glasses they cannot get prescribed for — the optics are fine, the wiring is the problem. Contrast is dimmer. Straight lines (door frames, the edge of a page) bow inward. Reading speed in that eye drops. Driving is still possible because the other eye carries it, but you have moved from binocular to one-eyed-with-a-backup for the rest of your life. People you know stop noticing after a while; you don't.
Months later, the second eye becomes the new clock you live by. Lifetime risk of detachment in the fellow eye, after a first one, is roughly one in ten Mitry 2011. You know the symptoms now. You know what waiting cost. The question is whether you knew them the first time.
What to do, today
If you have any new flashes, any sudden change in floaters, or any curtain or shadow in one eye, the next move is a dilated eye exam by an ophthalmologist, the same day. That is the only examination that can actually see the periphery of the retina where most tears live; it requires drops to widen the pupil and a special handheld lens. One presentation is even more urgent than this: a sudden, total, painless loss of vision in one eye — no flashes, no floaters, just the lights going out — is a possible eye stroke, and that one is emergency services now, not a same-day clinic slot. Your regular optometrist may be able to do it; if not, they will refer you on, and the time spent finding out is time you do not have.
Do not try to wait it out. Do not call your primary care doctor first — they cannot do the exam and the appointment lag costs you vision. Do not assume floaters that come on suddenly are the same as floaters you have always had; the question is change, not absolute count.
Things that get this wrong
"Floaters are just floaters." Long-standing background floaters — the ones you have noticed against blank ceilings for years — are usually nothing. The question is sudden change: a shower of new specks, a single big new cobweb, or floaters with flashes. The one-in-seven tear rate from Hollands 2009 is specifically for acute new symptoms, not chronic ones.
"That's a migraine." Maybe, but check before you decide. Migraine visual aura is almost always in both eyes (close one eye — you still see it), spreads over five to thirty minutes as a shimmering zig-zag, and is usually followed by a headache. Retinal flashes are in one eye (close the affected eye — they disappear), are brief lightning streaks lasting fractions of a second, and recur over hours or days without a headache. Quick bedside check: cover each eye in turn. If the flash is gone with one eye covered, it is in the eye, not the brain — and that means an eye exam today.
"I'll wait and see how it is in the morning." This is the cheapest mistake in the entry. The cost of being wrong about waiting — a macula that detaches overnight — is permanent central vision loss in that eye, measurable in lost lines on the eye chart for every day before surgery Greven 2019, van Bussel 2014. The cost of being wrong about going in — a normal-looking posterior vitreous detachment with no tear — is one afternoon and an Uber home.
"It will probably stop on its own." The flashes from a posterior vitreous detachment do tend to fade over a few weeks as the gel finishes peeling away. The problem is that you cannot tell from your symptoms whether anything tore on the way. Only the dilated exam can.
Who should be especially fast about this
The rule applies to everyone. But several groups walk in with a much higher baseline risk that any new symptom reflects a tear, and the threshold for "same day or sooner" should be reflexive:
- Anyone seriously nearsighted — glasses prescriptions stronger than about
-6diopters, or what your eye doctor has called "high myopia." Detachment risk in this group runs roughly 39 times the risk of someone without myopia, and risk climbs sharply with each additional diopter. The eyes are physically elongated, which puts the retina under more tension. - Anyone who has had cataract surgery, even years ago. Lifetime detachment risk is several times higher after lens replacement and stays elevated for decades.
- Anyone who has had a detachment in the other eye. Lifetime risk in the fellow eye is around 10% over ten years Mitry 2011. You already know the drill; the second eye is when it matters most.
- Anyone with a family history of retinal detachment, or with Stickler, Marfan, or Wagner syndrome, or with lattice degeneration noted on a previous eye exam.
- Anyone who has had a serious eye injury in the last weeks to months — direct blow to the eye, severe whiplash, sports impact. Tears from trauma can present immediately or delay by weeks.
- Anyone in their fifties through seventies — the peak window when the vitreous gel is finishing its lifetime separation from the retina.
If you are in one of those groups and you are still on the fence about whether your floaters are "really" new, the answer is to get the exam.
Where this goes wrong in practice
The standard failure mode is not in the medicine — it is in the gap between the symptom and the appointment. A few patterns the chart-review studies keep finding:
- The symptoms started Friday evening. People wait the weekend, the curtain advances, the Monday visit finds a macula already off. If it is the weekend, it is the emergency department, not a Monday slot.
- Primary care got called first. A general practitioner cannot do the exam that matters and the triage often understates the urgency. Skip the layer; call ophthalmology or go to the ED directly.
- The first exam was incomplete. If there is enough blood in the vitreous to obscure the view, the doctor needs an ultrasound of the eye that visit and a re-examination within days as the blood clears; the underlying tear rate in that situation is roughly 60% Hollands 2009. Make sure a re-examination is on the books before you leave.
- The first exam was clean and the patient never returned. Roughly 1 in 50 acute presentations develops a tear in the following weeks even after a clean first look Coffee 2007, AAO PPP 2025. The four-to-six-week follow-up is not optional, and any new symptom resets the clock.
- The symptoms were dismissed as a migraine by the patient or the clinician without checking which eye they were in. Cover-each-eye test, every time.
Related topics worth knowing about: routine dilated eye exams (how often to get one, and what the eye doctor looks for during them); high myopia and what it does to the back of the eye over time; the difference between the three kinds of retinal detachment and which one this entry is about (rhegmatogenous, the one started by a tear); and what an actual detachment repair looks like — scleral buckle, pneumatic retinopexy, vitrectomy — for the reader who wants to know what they would be walking into if it came to that.
- — If the loss is total and painless rather than flashes and floaters, it may be an eye stroke — that one is 911, not a same-week clinic visit.
- — The Amsler grid catches central, macular changes; flashes, floaters, and a side curtain are the peripheral retinal-tear warning — different signals, both urgent.
- — Knowing your normal from a routine dilated exam makes it easier to recognise when new flashes or floaters are the real thing.
- — These warning signs are exactly what the yearly diabetic eye exam is trying to catch before they appear.
- — Glaucoma is the silent counterpoint to these red flags: no symptom to notice, which is why a scheduled exam, not a warning sign, catches it.
Substance and claimed effects
The entry covers the three classical warning symptoms of rhegmatogenous retinal pathology — sudden flashes of light (photopsia), a sudden change in floaters (a new shower, a single large floater, or a sudden increase from baseline), and a curtain-, shadow-, or veil-like loss of peripheral vision, sometimes preceded or accompanied by sudden subjective visual blurring. The substance is symptom-recognition (action type respond): the reader who recognises these symptoms presents same-day to an ophthalmologist or emergency department, where dilated indirect ophthalmoscopy with scleral depression determines whether the underlying event is an uncomplicated posterior vitreous detachment (PVD), a retinal tear, or a retinal detachment in progress. Claimed effects scored under meta are entirely downstream of that response decision: longevity in the sight-preservation sense (saving the eye's central vision), health_short_term (preserved function in the days after presentation), small contributions to focus and mood (vision-loss anxiety and disability avoided), and a non-trivial effort_burden only at the point of recognition and travel — the protocol is one urgent visit, not an ongoing regimen.
Evidence by addressing question
Mechanism
The vitreous gel that fills the eye is anchored to the retina at the vitreous base, at the optic nerve, along major retinal vessels, and at any areas of pre-existing vitreoretinal adhesion (lattice degeneration, prior scars). With age the vitreous liquefies (synchysis) and contracts; eventually the posterior cortex separates from the retinal surface, an event called posterior vitreous detachment (PVD) Bond-Taylor 2017. PVD is near-universal with age — prevalence rises from roughly a quarter of adults in their fifties to most adults past eighty Bond-Taylor 2017. The detaching vitreous tugs on its anchor points; this traction is what the brain perceives as photopsia — brief lightning-streak flashes typically in the temporal peripheral field, often more visible in dim conditions when the contrast is high. The new floaters are the now-mobile posterior hyaloid membrane (commonly seen as a ring, the Weiss ring, when it pulls free of the optic disc), pigment cells released from the retinal pigment epithelium ("tobacco dust"), and red cells if a vessel was torn — vitreous hemorrhage. At a focal point of strong adhesion, the traction can tear the retina. Once a retinal break exists, liquefied vitreous can pass through the break into the subretinal space and detach the retina from the underlying retinal pigment epithelium — a rhegmatogenous retinal detachment (RRD), distinct from tractional and exudative subtypes Flaxel 2020, AAO PPP 2025. The detached neurosensory retina loses contact with its choroidal blood supply, photoreceptors begin apoptosis within hours, and the corresponding field of vision goes dark — perceived by the patient as a stationary curtain or shadow advancing from the periphery, because the retinal image is inverted (a detachment of the superior retina produces an inferior visual-field defect, and vice versa). When the detachment reaches the fovea ("macula-off"), central acuity collapses.
Evidence
The anchor evidence is the JAMA Rational Clinical Examination synthesis by Hollands 2009: across 17 studies of patients presenting with acute-onset floaters and/or flashes, the pooled prevalence of retinal tear was 14% (95% CI 12–16%). Among history features, subjective visual reduction carried the strongest positive likelihood ratio (LR 5.0, 95% CI 3.1–8.1); on examination, vitreous hemorrhage on slit-lamp biomicroscopy carried LR 10 (95% CI 5.1–20). Roughly two-thirds of patients found to have vitreous hemorrhage have at least one retinal break Hollands 2009. The natural history of an untreated symptomatic tear is well documented: 33–46% of retinal tears progress to retinal detachment if not treated Hollands 2009, and treatment of a discovered tear with focal laser retinopexy or cryotherapy is highly effective at preventing that progression Flaxel 2020, AAO PPP 2025.
Delayed breaks are also documented: a patient with a symptomatic PVD and no break at the first examination has a roughly 2% chance of developing one in the following weeks, but the risk rises markedly in the presence of vitreous hemorrhage or pigment cells Coffee 2007, AAO PPP 2025. Coffee 2007 reported that approximately 80% of patients who developed delayed breaks had either pigment, hemorrhage, or new symptoms at re-presentation — which is what motivates the standard 4–6 week follow-up after an initially clean acute PVD.
Population incidence of RRD itself is in the range of 10–18 per 100,000 per year in European cohorts, rising to ~26 per 100,000 in the Netherlands and increasing globally with the myopia epidemic and population aging Mitry 2010, van de Put 2013. The fellow eye in unilateral RRD carries an elevated lifetime risk of detachment — quoted around 10% over a decade in the Scottish cohort Mitry 2011.
Stakes — timing and visual outcomes
The clinically actionable variable is whether the macula is still attached at the moment of surgery. Once the macula detaches, the visual prognosis is set largely by the duration the macula has been off and by preoperative acuity. van Bussel 2014's systematic review and meta-analysis of macula-off RRD found that postoperative best-corrected visual acuity (BCVA) deteriorated with longer macula-off duration, with a discernible step in outcomes between repair within ~3 days and repair after; the typical-patient summary is that macula-off detachment repaired within roughly 72 hours yields meaningfully better central vision than repair delayed by a week or more. Greven 2019 sharpened the curve at the front end: among 324 eyes with macula-involving RRD, even one to three days of macula-off duration was associated with measurable reductions in final acuity, with each additional day producing a small but real decrement before plateauing. Williamson 2014's 1,189-eye series confirmed that preoperative macula status and duration of symptoms were dominant predictors of final acuity.
The most recent synthesis by Sothivannan et al. 2022 pooled 20 observational studies covering 1,929 patients and found that macula-on RRD repaired within 24 hours of presentation produced superior final BCVA versus repair after 24 hours (MD −0.02 logMAR, 95% CI −0.03 to −0.01) — a small absolute difference, supported by moderate-quality evidence, but consistent with the broader clinical bias toward urgent repair of macula-on detachment to prevent intra-operative conversion. The same meta-analysis confirmed the time-dependent deterioration for macula-off cases.
The competing literature is honest about magnitude: Mahmoudi 2016 argues that selectively urgent rather than emergent timing for macula-on detachments (factoring in break location and patient factors) is defensible because pre-operative conversion to macula-off within 24 hours is uncommon. The clinical reading, summarised in current practice patterns: time-to-presentation is what the patient controls and what determines macula status at presentation; time-to-surgery is what the surgeon controls and is generally optimised the same day for macula-on cases and within 3–7 days for macula-off cases, depending on system capacity and break configuration Flaxel 2020, AAO PPP 2025.
Protocol — what the symptomatic patient does
The recommendation that falls out of Flaxel 2020 and AAO PPP 2025: anyone with a sudden onset of flashes, a sudden change in floaters, or any peripheral curtain/shadow should obtain a same-day dilated fundus examination by indirect ophthalmoscopy with scleral depression. If the local ophthalmology service cannot accommodate same-day, the emergency department is the correct alternative because the examination is time-sensitive and requires equipment most primary care offices do not have. The visit's purpose is to look for retinal breaks across the full periphery; B-scan ultrasonography is used if vitreous hemorrhage prevents a clear view. A confirmed retinal tear is treated immediately with focal laser retinopexy or cryotherapy at the slit lamp (success rates well above 90% for prophylaxis against detachment). A confirmed retinal detachment is referred for surgical repair — pneumatic retinopexy, scleral buckle, or pars plana vitrectomy depending on configuration — generally within hours for macula-on and within days for macula-off cases. An acute PVD with no break at the first examination still warrants a return visit at 4–6 weeks or sooner if new symptoms occur, because of the ~2% delayed-break rate, and proportionally higher with vitreous hemorrhage Coffee 2007, AAO PPP 2025.
Misconceptions
Three common errors. First, "floaters are just floaters." Long-standing baseline floaters are usually harmless; the question is change — a new shower, a single large new floater, or any new flashes alongside them. The Hollands 2009 14% prevalence figure is for acute-onset symptoms, not chronic ones. Second, "flashes mean migraine." Migraine aura is typically binocular (visible with either eye closed), takes 5–30 minutes to spread across the visual field as a scintillating zig-zag, and is followed by headache. Retinal photopsia is monocular (visible only when the affected eye is open), lasts seconds at a time, and tends to recur over hours to days. The bedside test: if the flash moves when the eye moves, it is in the eye; if it sits still in the visual field, it is cortical. Third, "I'll wait until tomorrow." The cost of an unnecessary same-day visit is one office visit; the cost of a 48-hour delay that lets a macula-on detachment progress to macula-off is permanent central-vision loss Greven 2019, van Bussel 2014.
Audience — who is at elevated baseline risk
The standard recommendation applies to everyone, but several groups carry a much higher prior probability that new symptoms reflect a tear or detachment and should respond with even lower friction:
- High myopes (≥6 diopters of myopia, or axial length >26 mm). RRD risk rises with myopia in a steep dose-response. In a large US insurance cohort, high myopes had RRD incidence of 868.83 per 100,000 person-years versus 22.44 in non-myopes — a 39-fold elevation; each additional 6 diopters of myopia in UK Biobank multiplied RRD risk by ~7.2.
- Post-cataract-surgery patients. Pseudophakia raises lifetime RRD risk several-fold; the risk is concentrated in the first few years post-op but persists for decades.
- Prior RRD in the fellow eye. Approximately 10% over 10 years in Mitry 2011; bilateral risk is substantially higher than population baseline.
- Family history of RRD, lattice degeneration, or known Stickler syndrome / Marfan / Wagner syndrome.
- Recent significant ocular trauma. Mechanical disruption can produce immediate or delayed breaks.
- Age 50–80, peak incidence band, because PVD prevalence rises sharply through this window.
Failure modes
The dominant failure mode is delay between symptom onset and presentation. Reasons reported in clinical series: symptoms attributed to migraine or "getting older"; weekend or evening onset and reluctance to visit the emergency department; primary-care triage that downgrades the urgency. The Hollands 2009 data implies that any clinician seeing a new-onset flashes/floaters complaint must refer for same-day dilated examination — examination by a non-dilated, non-indirect technique systematically misses peripheral breaks. A second failure mode is incomplete first examination: a patient with vitreous hemorrhage that obscures the retina requires B-scan ultrasonography acutely and indirect re-examination within days as the hemorrhage clears, because the rate of underlying break is ~60% in that subgroup Hollands 2009. A third is missed delayed breaks — the patient with a clean first examination who does not return for the 4–6 week follow-up Coffee 2007, AAO PPP 2025.
Practicalities
The examination is dilated fundus exam with indirect ophthalmoscopy and scleral depression, performed by an ophthalmologist (general or retina sub-specialist). A community optometrist with a dilated indirect-ophthalmoscopy capability is an acceptable first stop only if they can refer onward same-day; many optometry practices appropriately decline and refer directly to retina. In urban systems most retina practices reserve same-day urgent slots for this presentation; out-of-hours, the emergency department triages and either calls in the on-call ophthalmologist or transfers to a tertiary centre. Patients should be warned: dilation lasts 4–6 hours, they will not be able to drive after, and bright light may be uncomfortable.
Payoff — what the urgent visit accomplishes
The realistic outcome distribution from Hollands 2009 and Flaxel 2020: of patients who present with acute symptoms, roughly 14% have a retinal tear and most of the remainder have an uncomplicated PVD. Tears identified at the visit are treated in office that day with focal laser or cryotherapy, preventing progression to detachment in over 90% of cases. Patients with a confirmed detachment are scheduled for surgical repair — anatomic reattachment is achieved in >85% of cases after a single procedure (>95% after second procedure), with the visual result driven by whether the macula was on or off at the time of surgery Williamson 2014, van Bussel 2014, Greven 2019. Macula-on repair preserves near-baseline central vision in most cases; macula-off repair often leaves residual reduction in central acuity, contrast sensitivity, and metamorphopsia even with successful reattachment.
The credibility range
Optimist case
The optimist case is the AAO Preferred Practice Pattern position. The three warning symptoms are highly specific to retinal pathology when sudden in onset and combined with subjective visual reduction; one dilated indirect examination by a retina specialist within hours has very high sensitivity for breaks; focal laser treatment of identified breaks is one of the highest-leverage prevention moves in medicine (single visit, low cost, prevents irreversible blindness in roughly 1 in 3 untreated tears). The reader who learns to recognise the symptoms and respond same-day captures essentially the full preventive benefit available. Even when symptoms turn out to reflect benign PVD only — the majority case — the reader has confirmed that and bought a known-safe baseline to detect future change against. The high LR for visual reduction (5.0) and vitreous hemorrhage (10) in Hollands 2009 means the rule "any new flashes, floaters, or curtain — present today" is not over-triage; it is matched to the true positive yield.
Skeptic case
The skeptic case has two parts. First, the recognition step has imperfect specificity in the lay population: floaters are extremely common at baseline, and "new" is a subjective judgement that produces some over-triage. Most acute presentations resolve as PVD without tear, so the per-visit yield is roughly 1 in 7 — not the urgent-action math the symptom severity implies. Second, the timing literature for surgical repair is observational and confounded: Mahmoudi 2016 and earlier reports (Wykoff 2010 et al) found no clear difference in outcomes between early and slightly-delayed repair of macula-on detachments, suggesting the "every hour matters" framing oversells the data. The Sothivannan 2022 macula-on effect of −0.02 logMAR is real but small in absolute terms. The skeptic position is not "ignore the symptoms" but "the urgent-presentation framing is rougher than its evidence."
Author's call
The action recommendation lands firmly on the optimist case. The asymmetry is decisive: the cost of a same-day visit for a false alarm is small and time-bounded; the cost of delay when the symptom reflects a macula-on detachment progressing to macula-off is permanent central vision loss in the affected eye, irreversibly. The skeptic point on macula-on surgical timing is taken — the article does not claim that the macula-on case requires repair within hours rather than within 1–2 days — but the timing argument depends on having an examination, and the examination depends on the patient presenting. evidence rates 4 (consistent guideline recommendation, RCEx-grade likelihood ratios, large observational outcome data on timing); controversy rates 1 (the symptom-to-presentation rule is universally agreed; minor disagreement persists on surgical timing windows for macula-on, which does not affect the reader's action).
Stakeholder and incentive map
- Academy of Ophthalmology and retina sub-specialty societies. Promote prompt presentation; the PPP is the canonical source Flaxel 2020, AAO PPP 2025. Incentive aligned with patient outcome and with sub-specialty volume.
- Optometrists. Often the first contact for flashes/floaters; competent dilated indirect examination is within scope, but practice culture varies. Most refer to ophthalmology for any suspected break; some attempt management. Reader-facing implication: same-day access matters more than provider type.
- Primary care. Frequently sees these complaints and underestimates urgency; chart-review studies repeatedly show under-referral and delayed dilated examination. The reader who self-routes to ophthalmology or ED bypasses this failure.
- Emergency departments. Standard care includes ophthalmology consultation. In rural systems, transfer to a centre with retina capability may be required.
- No commercial supplement or device industry exists around symptom recognition itself — this is one of the rare entries with no marketing layer to filter past.
Population variability
RRD incidence has a strong age gradient (peak 60–70s in the general population) and a second smaller peak in young high myopes in their 20s–30s Mitry 2010. Men are over-represented (roughly 1.5–2× incidence). High myopia is the dominant modifiable risk factor; ethnic differences exist (lower RRD rates in some African and Hispanic cohorts; higher prevalence of high myopia and consequently higher RRD rates in East Asian cohorts). Pseudophakia after cataract surgery raises lifetime risk roughly 3–5×, concentrated in the first 4 years post-op but extending for decades. Lattice degeneration, identified incidentally on fundus examination, is present in ~6–10% of the population and is associated with elevated RRD risk that may justify prophylactic laser in high-risk eyes (although routine prophylaxis remains debated; see Flaxel 2020). Family history elevates risk; specific genetic syndromes (Stickler, Marfan, Wagner, Ehlers-Danlos) carry markedly elevated risk and may warrant proactive surveillance regardless of symptoms.
Knowledge gaps
- The benefit of same-day vs next-day presentation for macula-on detachments is poorly quantified. Observational data favour early but the effect size is small and confounded by selection. An RCT is unlikely to be ethical given current consensus.
- Long-term outcomes of laser retinopexy for asymptomatic vs symptomatic breaks diverge, but populations are heterogeneous and the threshold for prophylactic treatment of asymptomatic breaks in low-risk eyes remains under-evidenced.
- Patient self-triage accuracy. The performance characteristics of public-health messaging on "flashes, floaters, curtain → present today" has not been formally measured; data are limited to clinician-level likelihood ratios.
- AI-assisted screening via smartphone fundus imaging or peripheral-vision self-tests is an active area but not yet validated for emergency triage.
Scope. The brief named "sudden flashes, new floaters, and curtain-like vision loss as red flags for retinal tear or detachment — their effects on the timing of urgent ophthalmologic evaluation and on visual outcomes." That is what the article covers end-to-end: symptom recognition, the same-day-exam rule, and the outcome curve as a function of time-to-presentation. No silent narrowing.
Separate-entry candidates. Several adjacent topics surfaced during research that warrant their own entries rather than being squeezed in here:
- Routine dilated eye exams — frequency, what they catch, recommendation cadence by age and risk factors. A
test-action entry under vision. - High myopia management — the long-term retinal consequences of axial elongation, peripheral retinal monitoring, lattice degeneration. A
knowortestentry under vision. - Retinal detachment repair — what scleral buckle, pneumatic retinopexy, and vitrectomy actually involve, recovery, return-to-activity. A
decide-action entry under medical. - Pediatric/young high-myope screening — the bimodal incidence curve has a young peak that current practice undertreats. Separate entry under screening.
Hard call on the timing literature. The article lands on the AAO Preferred Practice Pattern framing — present same-day, repair urgently. The skeptic literature (Mahmoudi 2016; older Wykoff data) argues that selectively urgent rather than emergent timing for macula-on detachment does not worsen outcomes meaningfully, and the Sothivannan 2022 macula-on benefit is small (−0.02 logMAR). The article does not litigate this because the disagreement is about surgical timing once a patient is in front of an ophthalmologist — a decision the reader does not make. The reader's decision is whether to present, and on that there is no dispute. Surfaced here so a reviewer sees the call was deliberate.
Rating difficulties. longevity was the hardest call. The dimension is defined as "disease prevention and mortality reduction over years" and the substance prevents irreversible vision loss rather than death. Scored 4 on the grounds that sight preservation is the canonical longevity case for an eye-care entry — bending the trajectory of monocular blindness, with the dominant condition (RRD) running 10–26 per 100,000 per year. focus at 2 and mood at 2 capture the downstream cognitive and emotional costs of unilateral vision loss without overclaiming; both deliberately conservative because most function rides on binocular vision and the contralateral eye usually compensates. beauty_* at 0 because the substance has no visible-appearance pathway. energy at 0 because the substance does not affect daily vitality independent of vision-related disability.
Future links. When the routine-dilated-exam entry exists, the audience section should link to it for the high-risk groups. When the retinal-detachment-repair entry exists, the stakes section should link to it for the reader who wants to know what the surgical pathway looks like in detail. Both currently handled by the out-of-scope pointer.
Voice. The action type is respond and the stakes are high; the editorial bias was toward concrete, calm, and unhedged rather than alarming. The friend test was applied especially strictly on the protocol callout — the steps a reader takes need to be readable while alarmed.
Retinal Warning Signs
One urgent ophthalmology visit; in-office laser retinopexy if a tear is found is typically covered by insurance and runs in the low hundreds of dollars out of pocket where it is not. Surgical repair for detachment is much more expensive but is a downstream consequence of the symptoms, not the recognition-response action itself.
One same-day visit, several hours including dilation and the inability to drive afterward. Not a recurring regimen, no ongoing willpower cost.
Sight-preservation is the longevity case here: an untreated rhegmatogenous detachment commonly progresses to legal blindness in the affected eye. RRD incidence is 10-26 per 100,000 per year and trending up with myopia and aging (Mitry 2010, van de Put 2013); macula-on repair preserves near-baseline acuity in most patients while macula-off repair leaves substantial residual deficits (van Bussel 2014, Williamson 2014). The fellow eye carries roughly 10% lifetime risk after a first detachment (Mitry 2011).
Anchored by the JAMA Rational Clinical Examination synthesis with pooled likelihood ratios (Hollands 2009), the AAO Preferred Practice Pattern (Flaxel 2020, Kim 2024), and converging meta-analyses on surgical timing (van Bussel 2014, Sothivannan 2022). Not a 5 because the recognition-to-action threshold has not itself been formally trialled.
Recognising and acting on flashes, new floaters, or a curtain shadow preserves vision in the affected eye within days: among acute presenters, 14% have a retinal tear treatable same-day with focal laser, and 33-46% of untreated tears progress to detachment (Hollands 2009). The functional difference between a treated tear and a delayed macula-off detachment is large and felt within the week (Greven 2019).
Preserved central vision is preserved deep-work and reading capacity in the affected eye. Macula-off detachment commonly leaves reduced acuity, contrast sensitivity, and metamorphopsia even after successful reattachment (Williamson 2014), all of which degrade reading and screen work. The score is modest because most function rides on binocular vision and the contralateral eye usually compensates.
Avoiding permanent unilateral vision loss avoids the substantial mood burden documented in vision-impairment cohorts (depression and anxiety rates roughly double in adults with significant visual impairment). The score is modest rather than high because the article's substance is the recognition-and-act step, not the chronic adaptation to vision loss it prevents.