Teaching case. Amblyopic-eye acuity has held at 20/25 for six months, the fellow eye 20/20, patch 2 hours. The family asks: "Can we take it off for good?" Stopping abruptly after 6–8 hours a day is a separate risk, and it has been measured.
Recurrence after success. ATS2C (Holmes et al., J AAPOS 2004): 156 children under 8 years (145 completed the year) with successfully treated anisometropic or strabismic amblyopia, who had received ≥2 h of occlusion a day or ≥1 drop of atropine a week for at least 3 months and had improved by at least 3 logMAR levels. Treatment was stopped and they were followed for 52 weeks. Recurrence — a decrease of ≥2 logMAR levels confirmed on retesting, or resumption of treatment because of such a decrease. Recurrence in 35 of 145 (24%), 95% CI 17–32%; similar after patching (25%) and atropine (21%). Among those who had been patching 6–8 hours, recurrence was 42% (11/26) if treatment was stopped abruptly, versus 14% (3/22) if it was first reduced to 2 hours (OR 4.4; 95% CI 1.0–18.7). Most recurrences occur early, often within the first weeks to three months. The PPP carries this into practice: about one quarter of children have a recurrence within the first year without treatment; if acuity holds when intensity is reduced, treatment can be stopped, but with a follow-up plan.
In children aged 7–12 who responded to 2–6 h of occlusion, atropine and near work, recurrence in the first year in the related PEDIG follow-up was about 7% — lower than in the younger children of ATS2C. Age is again a modifier, not a guarantee.
After binocular treatment, a separate cohort (PMC10592044): among 100 children who improved on dichoptic therapy, the Kaplan–Meier risk of recurrence up to 24–30 months was 24% (95% CI 16–35); in those who received no additional treatment the 12-month risk was 28%, comparable to the 24% after patching/atropine in PEDIG. The durability of digital gains has not been shown to be better than that of the classical methods.
How to decide that "the maximum has been reached". Table 2 of the Amblyopia PPP is a panel consensus, not an RCT:
| Response | What to change |
|---|---|
| No improvement for 3 months | Maintain or increase patching/atropine, or switch modality |
| Severe skin irritation | Alternative |
| No improvement on occlusion | Alternative; reduce or stop if already treated adequately |
| Failure due to organic disease | Reduce / stop |
| Strabismus and/or diplopia has appeared | Stop temporarily, monitor the angle and vision |
| Fellow eye has lost ≥2 lines | Stop temporarily, re-evaluate the diagnosis; in reverse amblyopia — patch the formerly amblyopic eye |
| Acuity stably normal or near-normal for ≥4 months at two or more visits | Taper or stop |
PPP: once maximal acuity has been achieved, intensity is reduced and then treatment is stopped; if acuity holds on the reduced regimen — stop, but follow-up is mandatory.
The long-term result of ATS1: at 10 and 15 years the gain is generally maintained, residual amblyopia is common, and starting before age 5 is associated with a better proportion reaching 20/25. Treatment does not "erase the diagnosis from the biography". It reduces the deficit and holds it.
The assembled plan at the end of the course — a teaching plan, not an individual chart.
- Name the form by cause and do not miss organic disease.
- Full cycloplegic optics immediately; decide how long to wait for an optical response without turning the wait into a dogma.
- If a deficit remains: 2 h of patching or weekend atropine for moderate preschool amblyopia; 6 h for severe preschool amblyopia as a regimen that did not lose to full-time.
- Atropine is an equal, not a "back-up", with a different harm profile.
- Do not put digital dichoptic methods under one label. Desk-based dichoptic vs 2 h of patching (Wygnanski-Jaffe 2023; PPP update 2024, I+, Good, Discretionary) — basket 1 for this device and the 4–<9-year population. Luminopia vs glasses at 4–7 years — basket 2 as regards replacing the patch. The ATS18 iPad game did not prove non-inferiority. An advertising brochure for a "neuro-game instead of patching" does not make up basket 1.
- Levodopa, "neuro-courses", supplements — basket 3.
- Taper high doses, do not cut them off; monitor for the first year, knowing about the 24% recurrence rate in younger children.
- Do not promise 1.0. Promise an honest follow-up timeframe and readiness to return to treatment if a line is lost.
What the course still does not cover: the platform's virtual patient, a national Ministry of Health of the Republic of Kazakhstan protocol on amblyopia (not found), cataract surgery as an operative skill, adults as a PPP target group. These are not holes in Lesson 4.5. This is the edge beyond which a different programme begins.