Eye Surgery Risks: The Honest Complication Numbers

Every vision correction consult ends with a risk paragraph, and most patients leave unable to repeat a single number from it. This article is the missing reference: the published complication rates for LASIK, PRK, SMILE, ICL, and lens surgery — what's common and temporary, what's uncommon and manageable, what's rare and serious — with the context that makes the numbers usable instead of scary. Two framing rules before the figures: rates come from published literature and vary by study, platform, and screening rigor (ranges, not gospel); and the single biggest variable behind every number is patient selection — the rates below describe appropriately screened eyes, which is exactly why screening is the safety system.

The Laser Procedures: What the Numbers Say

The common and temporary tier. Dry-eye symptoms lead every honest post-LASIK list: study ranges commonly report symptoms in a substantial minority of patients in the early months — figures around 20–40% at the early marks are typical — declining steadily, with persistent significant dryness beyond a year in the low single digits. Night-vision phenomena (halos, glare, starbursts) follow the same arc: common early, fading with neural adaptation and corneal healing for most, persisting in a bothersome way for a low-single-digit minority — with higher corrections and larger pupils weighting the odds. Neither is a surprise complication; both belong in expectations, and pre-existing dry eye is precisely why screening defers or pre-treats some candidates.

The uncommon and manageable tier. Enhancement (retreatment) rates land in the ~1–5% range depending on prescription magnitude and platform era — a scheduling event, not a crisis, though abroad patients should note it's a second trip. LASIK's flap adds a category all its own: slipped or wrinkled flaps (striae), interface inflammation (with its own name, DLK), and epithelial ingrowth under the flap — each in the fraction-of-a-percent to low-single-digit range in modern series, each treatable, and each the reason day-one and week-one checks exist. PRK trades flap risks for surface risks: haze (largely retired by modern medication protocols and mitomycin use at higher corrections) and a slower infection-vulnerability window while the surface heals. SMILE's profile: no flap events by design, a small rate of intraoperative lenticule difficulties (surgeon-experience-dependent), and interface issues at rates comparable to or below LASIK's in the meta-analyses — with the comparative literature detailed in our LASIK vs SMILE evidence piece.

The rare and serious tier. Infection after laser vision correction is genuinely rare — published rates in the roughly 1-in-several-thousand territory, lower for LASIK than surface procedures, and treatable when caught by the follow-up schedule doing its job. The most feared laser complication, corneal ectasia — progressive post-surgical warpage — sits at published rates around a few hundredths of a percent in the tomography-screening era, down dramatically from early decades precisely because modern imaging catches the at-risk corneas beforehand: ectasia is overwhelmingly a screening-failure story, which is why this site repeats the tomography rule like a liturgy. Vision loss meaningful enough to matter (losing two or more lines of best-corrected vision) sits well under 1% across modern series.

The Lens Procedures: A Different Risk Shape

ICL in the EVO generation publishes reassuring numbers: clinically significant cataract induction and pressure complications each in the fraction-of-a-percent range in recent series (down from the early-generation rates that built the procedure's outdated reputation), vault-related lens exchanges in the low single digits, and endophthalmitis — the serious intraocular infection — vanishingly rare but never zero, which is the honest sentence for all intraocular surgery. RLE and cataract surgery carry the best-quantified profile in medicine: endophthalmitis around 1-in-a-few-thousand with modern antibiotic protocols, retinal detachment in the fraction-of-a-percent range overall — but meaningfully elevated in young, highly myopic eyes, the specific reason our RLE guide gates that procedure by age and prescription — capsule complications in the low single digits, and posterior capsule opacification ("secondary cataract") common enough over years to be considered routine, treated with a brief laser polish. The lens procedures' distinguishing feature: their serious risks are intraocular and front-loaded into a monitored early window, which is why their post-op check schedules are denser and less skippable than laser surgery's.

A note on where these figures come from, because sourcing is half of trust: the modern refractive literature includes large registry studies, multi-year FDA-monitored trials, and the patient-reported-outcome projects of the past decade that specifically asked about satisfaction and symptoms rather than just chart measurements. When a clinic quotes you a number, the fair follow-up is "from whose data?" — their own audited series is a great answer, the published literature is a good one, and a shrug is a third answer of its own kind.

Reading Risk Like an Adult

Three tools make these numbers usable. Compare to the alternative, not to zero: decades in contact lenses carry their own published infection risk — microbial keratitis rates that, compounded over a wearing lifetime, bear honest comparison against one-time surgical figures; glasses-only life has zero procedural risk and its own daily costs. The relevant question is never "is surgery risk-free" but "against what." Distinguish expectations from complications: early dryness and halos are the predictable price of the biology, priced into the satisfaction rates around 95–98%; counting them as complications double-books the ledger. Weight the variables you control: screening rigor (tomography, tear-film workup, honest history), surgeon volume, platform generation, and follow-up compliance move your personal odds more than any choice between the mainstream procedures — the between-procedure differences are real but second-order against the quality-of-care differences. Which is the entire logic of this site's vetting obsession, at home or abroad: the safety checklist operationalizes it, and the complete guide puts the procedures' risk profiles side by side in the decision context where they belong.

The Bottom Line

Modern vision correction's honest risk picture: common temporary nuisances, uncommon manageable events, rare serious complications sitting well under one percent — in screened eyes, on current platforms, with follow-up schedules kept. The numbers are good because the screening era made them good; they stay good for you by insisting on the screening, the named platforms, the experienced surgeon, and the check schedule — the four variables that were always the real risk decision.