This randomized controlled trial compares two techniques for eyelid retraction during intravitreal injection (IVI) of anti-VEGF agents: the standard wire eyelid speculum (Group A) versus cotton-tipped applicator retraction (Group B) in patients with neovascular AMD, diabetic macular edema, or retinal vein occlusion.

The study evaluates four outcomes: (1) patient pain perception measured by a 10-cm visual analogue scale immediately after injection; (2) procedure duration from retraction device placement to removal; (3) patient satisfaction assessed by a 5-item Likert scale; and (4) safety including rates of subconjunctival hemorrhage, corneal abrasion, endophthalmitis, and intraocular pressure elevation.

A novel syringe cap technique using the Terumo 31G insulin syringe plastic cap as an injection-site marker (3.5 mm for pseudophakic eyes, 5.0 mm for phakic eyes from the limbus) is employed in both groups, replacing the traditional caliper.

Randomization is stratified by diagnosis and prior injection history using permuted block randomization (block sizes 4 and 6). The target sample size is 120 patients (60 per group) at Walailak University Hospital, Nakhon Si Thammarat, Thailand.

Background:

Intravitreal injection is one of the most frequently performed ophthalmic procedures worldwide. Eyelid specula are standard for keeping the eye open during injection but may cause discomfort, corneal exposure, and patient anxiety. Cotton-tipped applicator retraction is an alternative technique that may reduce pain and improve patient experience, but comparative randomized data on procedure time, satisfaction, and safety are lacking.

Study Design:

Single-center, prospective, parallel-group randomized controlled trial. Patients are randomly assigned 1:1 to speculum (Group A) or cotton-tipped applicator retraction (Group B). Stratified permuted block randomization is performed using Study Randomizer software with six strata: diagnosis (nAMD / DME / RVO) × prior injection history (injection-naïve / previously injected).

Interventions:

All patients receive standard preparation: topical tetracaine 0.5% (two instillations), 5% povidone-iodine ophthalmic drops (≥5 min contact), and 10% povidone-iodine periocular skin preparation. Anti-VEGF agents used are bevacizumab, aflibercept, or faricimab (per clinical indication), injected via Terumo 31G insulin syringe at the superotemporal quadrant. Post-injection levofloxacin 0.5% drops are instilled immediately after needle withdrawal.

Outcome Measures:

Primary: Pain VAS (0-10 cm) immediately post-procedure; procedure duration (seconds) from retraction device placement to removal. Secondary: Patient satisfaction (5-item Likert scale, total 5-25); safety outcomes (subconjunctival hemorrhage, corneal abrasion, endophthalmitis, IOP elevation) at Day 0, Day 1 (telephone), and Week 4.

Blinding:

Patients cannot be blinded to retraction method. Pain VAS is assessed by an independent assessor blinded to group assignment. Procedure timing is recorded by a circulating nurse blinded to the randomization list.

Sample Size:

Based on Chen et al. (2025), assuming pooled SD of 1.71 for pain VAS (Cohen's d = 0.57), with Bonferroni correction (α = 0.025, power 80%), 52 patients per group are required for pain and 46 for procedure duration. With 15% dropout allowance, the final target is 60 patients per group (120 total).

Inclusion Criteria:

1. Age 18 years or older
2. Diagnosed with neovascular age-related macular degeneration (nAMD), diabetic macular edema (DME), or retinal vein occlusion (RVO)
3. Scheduled for intravitreal anti-VEGF injection (bevacizumab, aflibercept, or faricimab)
4. Able to provide written informed consent

Exclusion Criteria:

1. Active ocular infection or inflammation
2. Known allergy to povidone-iodine, tetracaine, or levofloxacin
3. Inability to cooperate with the injection procedure
4. Concurrent participation in another interventional clinical trial
5. Anatomical eyelid abnormality precluding use of either retraction technique

The outcomes assessor (nurse measuring VAS pain score and procedure time) is blinded to group allocation. Blinding of participants and the care provider (injecting surgeon) is not feasible due to the nature of the intervention.