Corneal abrasion is the most frequently reported perioperative ocular injury, and its consequences range from pain, photophobia and blurred vision to delayed recovery and, rarely, persistent visual morbidity. What makes the problem difficult to quantify is the extraordinary variability in reported incidence. A recent systematic review identified corneal injury rates ranging from 0.01% to 59% in surgical patients, depending on the population studied, the eye protection strategies employed and, notably, the diagnostic method used. Earlier work reported a corneal injury incidence of 39.5% in surgical patients protected with adhesive tape, compared with 15.8% in those given a hydrogel eye dressing. The protocol authors are careful to note that the hydrogel dressing differs from the polyethylene chamber now under evaluation, so those figures illustrate the variability of different ocular-protection strategies rather than serving as a confirmatory effect estimate for the chamber.

Standard perioperative eye care usually combines lubricating ophthalmic drops with eyelid closure using adhesive tape. Tape is inexpensive and familiar to every operating room team, but it carries practical weaknesses. It may loosen during patient repositioning, fail to maintain complete closure in participants with lagophthalmos, irritate the periorbital skin or require reapplication — each of which increases nursing workload and interrupts workflow. A moisture-retaining eye protection chamber, by contrast, is designed to reduce evaporation and mechanical exposure while maintaining eyelid closure. The new protocol, reported in accordance with the SPIRIT 2025 Statement, describes a single-centre, two-arm, parallel-group, assessor-blinded randomised pilot and feasibility trial with individual participant allocation in a 1:1 ratio.

The rationale for a feasibility study rests partly on a previous pilot randomised controlled trial in 80 critically ill adults, in which 40 participants received eye drops plus a polyethylene chamber and 40 received eye drops plus adhesive tape. Corneal injury was reported in six participants in the tape group and three in the chamber group. Because that study was conducted in critically ill patients — who differ from elective surgical patients in duration of sedation, organ dysfunction, mechanical ventilation and repeated eye-care exposure — those event counts are not treated as a definitive effect estimate for the surgical population. The protocol authors argue that these data support the need for perioperative feasibility and event-rate estimates before planning a definitive confirmatory trial.

The trial will be conducted in the operating rooms and immediate postoperative care areas of Hospital Universitário Antônio Pedro, a public hospital within the Brazilian Unified Health System managed by the Brazilian Hospital Services Company, located in Niterói, Rio de Janeiro State. Approximately 400 patients per year at the study site undergo general anaesthesia for procedures expected to last longer than 120 minutes. The trial targets 120 participants recruited over 12 months, with an average recruitment rate of 10 randomised participants per month. Participants must be adults aged 18 years or older scheduled for non-ocular surgery under general anaesthesia with an expected duration longer than 120 minutes, able to provide written informed consent or represented by a legally authorised representative when permitted. Exclusion criteria include corneal epithelial injury detected during the preoperative fluorescein examination, emergency polytrauma surgery when consent cannot be obtained, and impaired consciousness without an available legally authorised representative.

The sample size is deliberately pragmatic rather than powered for hypothesis testing. No formal sample-size calculation based on efficacy is used; an earlier calculation grounded in the hydrogel dressing study was withdrawn because that study evaluated a different device and assumed substantially higher event rates than those reported in the authors’ previous chamber pilot. The 120-participant target should not be interpreted as providing adequate statistical power to demonstrate or exclude superiority of the eye protection chamber. Randomisation uses a computer-based simple randomisation sequence generated before enrolment by a trial statistician who takes no part in screening, consent, intervention delivery, outcome assessment or the blinded clinical analysis. Each allocation is generated independently with a probability of 0.50 for each group, with no forced balance, block randomisation or stratification. Because simple randomisation does not force equal final group sizes, the realised numbers in the two groups may differ from 60 per arm. Allocations are concealed in sequentially numbered, opaque, sealed envelopes, opened in numerical order only after eligibility is confirmed, consent obtained and the baseline corneal examination completed. Any skipped, damaged or prematurely opened envelope will be documented as a protocol deviation, preserving an auditable allocation trail.

Blinding is engineered around the visible difference between the two devices. Participants, anaesthetised when the assigned device is applied, will not be informed of allocation before the postoperative corneal assessments. Nurses applying the chamber or tape cannot be blinded, but application, displacement, reapplication and protocol deviations will be documented. A second statistician with no access to the randomisation sequence or allocation key will conduct the prespecified analyses using masked group labels, with the code key withheld until those analyses are complete. Before the immediate postoperative assessment, a nurse who is not the outcome assessor will remove the eye-protection device and perform periorbital cleansing before the assessor enters the assessment area, with packaging, devices and intervention records kept out of the assessor’s view — a procedure intended to reduce the possibility of identifying group allocation.

Both groups receive identical baseline ocular lubrication with 0.5% sodium carmellose ophthalmic solution, so the only difference between the arms is the eyelid occlusion method. In the control group, after induction of general anaesthesia and loss of the blink reflex, the perioperative nurse instils the lubricating drops and closes both eyelids completely with microporous adhesive tape, which remains in place during surgery unless displacement, contamination, incomplete closure or a safety concern requires reapplication. In the intervention group, the nurse instils the identical drops and applies the flexible polyethylene eye protection chamber over the closed eyelids, which likewise remains in place unless displacement, loss of seal, contamination or a safety concern requires reapplication. Both devices are removed after surgery before the blinded corneal assessment.

The primary objective is to assess the feasibility of conducting a definitive randomised trial comparing the chamber with adhesive tape in this surgical population. Primary feasibility outcomes include recruitment rate, retention and completeness of the planned corneal assessments, intervention adherence, device displacement, slippage, reapplication or unplanned removal, and prespecified adverse events. Progression to a future definitive trial will be judged after the 12-month recruitment period using a traffic-light framework: green indicates the definitive trial may proceed using current procedures, amber indicates progression may be appropriate after targeted protocol modification, and red indicates progression should not occur without major redesign. Recruitment is judged across the whole trial, while retention, assessment completeness, adherence and device displacement are calculated separately for each arm, with the less favourable arm-specific category determining the domain classification. Any serious intervention-related adverse event in either arm classifies the safety domain as red. Threshold classification uses the exact unrounded proportion, with displayed percentages rounded to one decimal place.

The exploratory clinical outcome is any new corneal epithelial injury detected at the immediate postoperative assessment or at a 24-hour postoperative assessment that was absent at the preoperative baseline examination. Injury is assessed by fluorescein staining under cobalt-blue illumination using a slit lamp, with lesions classified morphologically as punctate epithelial erosions, corneal abrasion or macro-epithelial defect, or corneal ulceration. The participant is the unit of analysis: a participant with injury in one or both eyes contributes one participant-level event, and laterality is described as unilateral or bilateral. The primary outcome assessor is a perioperative nurse specialist with more than 10 years of operating room experience and doctoral-level training, who completed one month of specific training with an ophthalmologist. Separate evidence using fluorescein assessment by trained nurses found almost perfect inter-rater agreement, with kappa values from 0.84 to 0.93, supporting reliability after training without establishing diagnostic sensitivity or specificity against an ophthalmologist.

Statistically, corneal-injury event rates will be estimated for each group with exact Clopper-Pearson 95% confidence intervals. Exploratory between-group effects will be reported as relative risk with a 95% Koopman asymptotic score confidence interval and absolute risk difference with a 95% Newcombe hybrid-score confidence interval. Fisher’s exact test may be reported as an exploratory comparison, but no confirmatory superiority conclusion will be based on statistical significance. All randomised participants constitute the intention-to-treat population, missing outcomes will not be imputed as injury or no injury, and no separate per-protocol efficacy analysis is planned. Because few corneal-injury events are anticipated, no multivariable adjustment model is prespecified; analyses of participant- and procedure-related factors will be hypothesis-generating only, with candidate variables defined a priori on the basis of clinical relevance and prior evidence. Adverse-event surveillance will target periorbital erythema, skin reaction, eyelid indentation, conjunctival irritation, pain or visual symptoms, and any corneal lesion, with any unexpected safety concern prompting suspension of the relevant study procedures while the event is clinically evaluated and reported to the Research Ethics Committee. No independent Data Monitoring Committee has been established; oversight rests with the principal investigator and a multidisciplinary monitoring group comprising perioperative nursing, ophthalmology, statistics and injury-prevention research members.

The protocol’s strengths include randomised allocation, identical baseline ocular lubrication in both groups, blinded outcome assessment, objective evaluation of corneal injury using fluorescein staining and slit-lamp examination at baseline, immediately after surgery and 24 hours postoperatively, and prospective collection of feasibility measures. The addition of the 24-hour assessment is intended to reduce underascertainment of epithelial injuries that may not be readily detectable during the immediate postoperative examination. Limitations include the single-centre design, the inability to blind nurses delivering the visibly different interventions, and the limited sample size. No patient or public involvement occurred in the development of the protocol, and no participant has been enrolled to date; recruitment will begin only after acceptance of the protocol, ethics approval and registry update are complete. If feasibility criteria are met, the trial is expected to provide the recruitment, adherence, device-performance, safety and preliminary event-rate estimates needed to design, power and size a future definitive randomised trial — and, ultimately, to support the development of evidence-based perioperative nursing protocols for eye protection.

Subject of Research: Perioperative corneal injury prevention and eye protection devices under general anaesthesia

Article Title: Eye Protection Chamber for Preventing Corneal Injury in Surgical Patients: Protocol for a Randomised Pilot and Feasibility Trial

Article References: do Carmo, T. G., de Souza, R., dos Santos, A. B. J., Barbosa, L. F., de Marteloti, G. O., Palmares, M. E. M., Rocha, E. M., & do Prado, P. R. (2026). Eye Protection Chamber for Preventing Corneal Injury in Surgical Patients: Protocol for a Randomised Pilot and Feasibility Trial. Nursing Open, 13(10), Article e70964. https://doi.org/10.1002/nop2.70964

Image Credits: AI Generated

DOI: 10.1002/nop2.70964

Keywords: corneal abrasion, perioperative eye care, general anaesthesia, eye protection chamber, adhesive tape, randomised pilot trial, feasibility study, fluorescein staining, slit-lamp examination, perioperative nursing, SPIRIT 2025, Brazil