After an EVO ICL: the first three months

Our ICL pages are written for the decision: how the lens is sized, what the endothelial cell count means, what the lens actually is and what it costs. This one starts the day after, because the questions that arrive in the weeks following an implantation are a different set and are poorly served by the laser-surgery recovery material that dominates the internet. Almost everything written in English about dryness, screen fatigue and fluctuating vision after refractive surgery is written about a cornea that has been cut. An ICL patient reading it is reading about someone else's operation.

The disturbance is real, and it is not nerve damage

The usual explanation for dry eye after refractive surgery is nerve transection. A LASIK flap or a SMILE lenticule extraction severs part of the sub-basal nerve plexus in the cornea; the surface loses some of its sensation, the reflex that drives blinking and tear production is blunted, and the tear film destabilises until the nerves regrow. That account is correct, it is the reason corneal sensation is worth understanding before a corneal procedure, and it is why an EVO ICL is sometimes the route offered to someone whose tear film is already marginal.

It also leads a great many patients to conclude that an ICL cannot produce dryness at all, and then to worry when it does. It can, and the mechanism is simply different. The lens is introduced through a clear corneal incision of about three millimetres; a viscoelastic gel protects the inside of the eye during implantation and is then irrigated out; and the eye is treated with several weeks of drops. Each of those touches the ocular surface without transecting a nerve plexus.

The most useful measurement we have found is a prospective cohort study from Taiyuan Central Hospital, published in BMC Ophthalmology in 2024 (Yao J, Feng J, Li W, Liu C, Li Y, et al., DOI 10.1186/s12886-024-03790-2, open access). It enrolled 67 myopic patients treated between January 2023 and January 2024 — 35 eyes receiving an ICL V4c and 32 undergoing SMILE — and examined every one of them before surgery and again at one week, one month and three months, using the same Oculus Keratograph 5M non-invasive analyser and the same operator throughout. The two groups were indistinguishable at baseline on every parameter measured. That combination is what makes the study worth reading in full rather than in summary: it is a direct, same-instrument, same-clock comparison between an intraocular and a corneal refractive procedure, which is rare, and it was done in China on the platforms and protocols a patient travelling to China would actually meet.

The recovery clock, parameter by parameter

The table below is the study's own findings, restated as timing rather than as significance tests. It is the clock, not a promise about any individual eye.

Source: Yao et al., BMC Ophthalmology 2024, DOI 10.1186/s12886-024-03790-2. ICL group n = 35 eyes, SMILE group n = 32 eyes, single centre, three-month follow-up. Read in full rather than from the abstract. These are group means over a small sample with a short horizon; they describe a typical course, not a schedule your eye is obliged to keep.

Two things in that table are worth pulling out because they are counter-intuitive. First, the instrument recovers before the patient does. In the ICL group the tear film measurements were back to baseline at one month while the symptom score was not — so feeling dry at six weeks, with an examination that looks normal, is a recognised combination and not evidence of a missed problem. Second, the meibomian glands move in opposite directions in the two groups between one and three months. Those glands supply the oily layer that stops tears evaporating; the corneal group's loss score was still climbing at three months while the ICL group's had turned back down. That is the evaporative route to dry eye rather than the aqueous one, and it is the part of post-refractive dryness that tends to outlast the rest.

The incision site is a question you are allowed to ask

The Taiyuan authors, explaining why their ICL group's break-up time dipped at one week at all, pointed at the incision. A clear corneal incision crosses corneal nerve fibres wherever it is placed, and the number and importance of the fibres it crosses depends on where that is. They put the argument plainly: a temporal incision disrupts more of the relevant fibres, a superior incision disrupts fewer, and they recommended a superior approach rather than a nasal or temporal one in order to minimise damage to the corneal sensory nerves and improve patient comfort. Their own 35 ICL patients were operated through a superior incision.

Two limits on that, stated rather than smuggled. The recommendation comes from a discussion section reasoning partly from the cataract-surgery literature it cites, not from a trial that randomised incision location in ICL surgery — we looked for one and did not find it. And incision placement is a surgical judgement made on your anatomy, your surgeon's technique and, if you are having a toric lens, on the axis that lens has to sit at. It is a legitimate thing to ask about at the consultation, in the same category as asking which lens size is planned and why. It is not a preference to be ordered, and a surgeon who explains that they use a different approach for a stated reason is answering the question properly.

Why near work can feel harder at 20/15

This is the question that arrives at about a month, usually phrased as a worry that the correction has overshot: my distance vision is better than it has ever been, so why do my eyes ache after an hour of reading? The premise is the problem. Distance acuity and near effort are different quantities, and correcting a short-sighted eye improves the first while slightly increasing the second.

The reason is where the correction sits. A myope wearing spectacles accommodates less than a normally-sighted person for the same near target — the minus lens in front of the eye is doing part of the near work for free, which is the same effect that lets a short-sighted person in their forties read comfortably by taking their glasses off. Move the correction from the spectacle plane to inside the eye, and that free assistance disappears. As García-Montero and colleagues put it in Int J Environ Res Public Health 2022;19(11):6716, a patient corrected by a phakic intraocular lens requires a much higher change in the crystalline lens's own refractive power for near than a spectacle-corrected one does. Nothing has been damaged. The eye is being asked to do work the spectacles used to do.

The measurements behave accordingly. That same study followed 38 patients through bilateral ICL implantation, split into a high myopia group and a low-to-moderate group, and measured push-up amplitude of accommodation, monocular accommodative facility, near point of convergence, near convergence amplitude, stereopsis, ocular deviation and the AC/A ratio before surgery and again at one week and one month. Amplitude of accommodation fell significantly in both groups at both visits, from preoperative means of 11.38 ± 2.82 dioptres in the high myopia group and 10.70 ± 2.16 in the low-to-moderate group. Every other binocular measurement was statistically unchanged. The authors' own conclusion, quoted rather than paraphrased because the caveat matters as much as the finding, was that ICL implantation did not modify baseline accommodative and binocular function except for that fall in amplitude, that the fall was not clinically relevant at the ages studied, and that future studies are needed to know its impact in pre-presbyopes and emerging presbyopes. They also name their own two limitations: a short follow-up, and a subjective measurement method.

So: a small, measurable reduction in accommodative amplitude, on top of a real increase in the accommodative demand of the same reading distance, in an eye whose tear film is transiently less stable, in a person who has just started using their eyes in a new way. That is a sufficient explanation for near-work ache at one month, and none of its components is being corrected to 20/15. If you are in your forties, the trade is a different and larger one, and it belongs to the presbyopia discussion rather than to this page.

Telling the two apart

Focusing effort and tear-film instability produce a similar ache, and the distinction is practical rather than academic: the first has no treatment beyond time and working habits, the second responds to lubrication and to blinking deliberately during screen work. Symptoms that ease briefly after a blink or a drop, and vision that fluctuates rather than being steadily soft, point towards the tear film. An ache that builds steadily with reading time regardless of lubrication, with no fluctuation at all, points away from it. Neither pattern is diagnostic and neither replaces the postoperative refraction, which is what actually settles it — a small residual refractive error is itself capable of producing exactly this complaint while leaving distance acuity excellent, and it is measured in two minutes at a follow-up visit.

The drops are part of the picture

The postoperative regimen after an intraocular procedure is heavier than after laser surgery, and patients are rarely warned how much is going onto the surface. The schedule recorded for the ICL patients in the Taiyuan cohort was: a combined antibiotic-steroid drop four times daily for three days, followed by a steroid tapered over about two weeks; a separate antibiotic four times daily for one week; a non-steroidal anti-inflammatory four times daily for two weeks; and artificial tears four times daily for a full month.

Add those up and the eye receives a large number of instillations during exactly the window in which the break-up time is shortest. Preservatives in multi-dose bottles are a recognised contributor to surface irritation in their own right, and the month of artificial tears is built into the schedule from day one rather than being offered as a rescue if symptoms appear — which tells you the protocol already assumes a transient surface effect. Discomfort that eases as the regimen tapers is consistent with the regimen being part of it.

What this evidence does not cover

Three stated negatives, because a recovery page that only carries the reassuring numbers is not telling you what they were measured on.

• Heavy screen users were excluded. The Taiyuan study's published exclusion criteria remove people with extensive computer usage, people who drive at night on a long-term basis, and people in long-term outdoor work or prolonged air-conditioned environments. Methodologically that is correct — those are environmental causes of dry eye that would contaminate a surgical measurement. Practically it means the clean one-month curve was measured on a population from which the heaviest screen users had been deliberately removed, and it does not describe a desk worker's experience of the same surgery. We have not found a series that follows heavy screen users specifically, and we are not going to estimate one.
• Three months is where the evidence stops. Both studies quoted here followed patients for three months or one month, not for years. Meibomian gland structure in particular was still moving at the last visit in both groups. Nothing on this page should be read as a statement about year two.
• These are small, single-centre samples. Thirty-five ICL eyes and thirty-two SMILE eyes at one hospital, with one surgeon's technique and one incision location; thirty-eight patients in the accommodation study, with a subjective measurement method the authors themselves flag. Group means from samples this size describe a typical course. They do not predict an individual eye, and they are not a standard against which your own recovery can be judged.

What is not part of ordinary recovery

Everything above describes a gradual settling. The rarest and most time-critical abrupt change, infection inside the eye, has its own page on what the published risk is and why its timing matters to a fly-in patient. Anything abrupt belongs to a different category and to the same day it happens — the operating surgeon, or the nearest ophthalmic emergency service if you have already travelled home.

• Vision that was clear and becomes blurred, rather than vision that has been slowly improving.
• Pain, as distinct from grittiness or a foreign-body sensation.
• A red eye with light sensitivity, particularly in combination.
• A sudden increase in glare or haloes around lights, rather than the gradual settling that is ordinary.
• New floaters in a shower, flashing lights, or a curtain or shadow across any part of the field. These matter independently of the surgery: a highly myopic eye carries a retinal risk that correcting the refraction does not remove, and it is the one thing on this list that has nothing to do with the lens.

Two things the scheduled visits are watching that you cannot feel are the pressure inside the eye and the vault — the gap between the implanted lens and your own natural lens sitting behind it. Both can move without producing any symptom you would notice, which is why those appointments are not optional and why they are the part of the plan most often abandoned by a patient who has flown home feeling well. If your itinerary ends before the schedule does, arrange the remaining reviews with an ophthalmologist where you live before you fly, and take your operative note and your lens details with you. How long to stay sets out which of those visits have to happen on the ground.

The two things you cannot feel, and why the early visits are where they are

Vault is the gap between the implanted lens and your own natural lens sitting behind it, and getting it into the right range is the central technical problem of ICL surgery. That page is about choosing the lens size before the operation. What happens to the vault after the operation is a separate fact, and it explains the shape of your follow-up schedule better than the schedule itself does.

The vault does not hold still. It settles downwards, and the settling is heavily front-loaded into the window this page covers. A study of the rate of vault change after ICL V4c implantation, published in Heliyon in 2024, measured it in three stages and found roughly −19.5 micrometres per month between one week and one month, roughly −19.9 micrometres per month between one and three months, and then only about −4.3 micrometres per month between three and six months. In other words the rate falls by a factor of four or five once the third month is passed. Over a much longer horizon the drift continues slowly: a series of 169 eyes followed in Ophthalmology and Therapy in 2022 recorded a mean vault of about 541 micrometres at one month, 520 at six months, 504 at one year and 471 at final follow-up.

Read those two together and the follow-up pattern stops looking arbitrary. Nearly all of the movement that will ever be measured happens while you are still in the first three months, which is exactly why the appointments are clustered there and thin out afterwards — and why an appointment skipped at week four is not equivalent to one skipped at month eight. The second unfeelable variable is the pressure inside the eye, which is being watched for two different reasons at once: the lens sits near the structures that drain fluid from the eye, and the steroid drops you are taking during the same weeks can themselves raise pressure in a minority of people. Neither vault nor pressure announces itself. You will not know.

Glare and haloes: the part that plateaus earlier than people expect

Haloes around lights at night are covered comparatively across all four procedures on our night-vision page, and that is the better place to start if you are still choosing. One ICL-specific finding belongs here instead, because it is about the clock rather than the choice. A prospective consecutive series of 22 patients — 44 eyes, mean spherical equivalent −10.52 dioptres — followed for five years after EVO ICL and published in BMC Ophthalmology in 2025 found that measured glare fell significantly at one week and again at one month, and then stabilised: no significant further change out to five years. At the final visit 72.7% of patients still reported glare, but only 31.8% found it bothersome and every one of those rated it mild, with 18.2% describing mild difficulty driving at night. The study's own regression found pupil dilation velocity, maximum pupil diameter and ICL size to be the main influences.

The useful reframing is that the magnitude of the optical effect settles early and the tolerance of it keeps improving — so somebody at month four who notices haloes but is no longer bothered by them is the expected outcome, not a partial one. It also means that if haloes are genuinely interfering at three months, waiting a further year for the effect itself to shrink is not what this data predicts; that is a conversation to have with the operating surgeon rather than a matter of patience.

Questions

Keep reading