LUMINEYES · SCIENTIFIC COMPENDIUM

Full manuscript · Chapter 2 of 6

Anterior-Segment Safety Staging (L-SAFE)

Research edition: The full manuscript is reproduced from the selected source document. Final reference, figure, DOI, and approval audit is pending.

Associated research DOI: https://doi.org/10.5281/zenodo.23046455. Exact manuscript–record relationship awaits verification.

Lumineyes Anterior-Segment Safety Staging (L-SAFE): Development of a Four-State Pre-Exposure Assessment Framework for Laser Iris Pigment Reduction

Running title: L-SAFE four-state pre-exposure framework

Abstract

Purpose. To develop an explicit, testable anterior-segment classification for documenting ocular response and determining when a subsequent staged laser iris pigment-reduction exposure should be withheld.

Methods. A conceptual methodological development process mapped targeted, non-systematically selected clinical reports and established ophthalmic assessment principles to four examination domains, a prespecified hierarchy of integrated safety states, and a checkpoint record. Framework rules were distinguished from evidence-based examination practices. No patient-level data were analysed and no empirical validation was performed.

Results. The proposed Lumineyes Anterior-Segment Safety Staging (L-SAFE) framework requires assessment of intraocular pressure, anterior-chamber response, corneal status, and angle/outflow status at an eye-specific checkpoint. Findings are integrated into S0 (READY), S1 (WATCH), S2 (PAUSE), or S3 (STOP). S1 and S2 withhold further exposure pending reassessment; S3 terminates the current exposure sequence. Resolving, stable, and progressive trajectory descriptors apply only to S1 and S2. Incomplete mandatory assessments cannot establish S0. Return to S0 after S3 does not reopen the terminated sequence.

Conclusion. The development output is an auditable classification and documentation architecture, not a validated clinical decision rule. Independent reproducibility, feasibility, and clinical evaluation are required before any safety-benefit claim.

Keywords: anterior segment; laser iris depigmentation; intraocular pressure; pigmentary glaucoma; safety classification; methodological framework

Introduction

Laser iris pigmentation reduction has been described in a clinical series reporting favorable outcomes, while separate case reports document serious complications [1-4]. Reported events include refractory pigmentary glaucoma and, in one report, paracentral acute middle maculopathy [2,3]. A subsequent report described marked pressure elevation and a temporally associated episode of herpetic keratitis with corneal scarring; temporal association alone does not establish the cause of the keratitis [4]. These different study designs cannot be combined into a reliable complication rate or used to establish safety across devices and exposure protocols.

When exposures are staged, initial treatment eligibility and the response to a preceding exposure are different questions. A record of pressure alone does not describe the anterior-chamber response, corneal condition, or angle findings. A compact structure that preserves those observations while stating the reason for withholding further exposure may support more interpretable records. Whether such a structure improves decisions or outcomes must be tested.

The Standardization of Uveitis Nomenclature (SUN) initiative illustrates the value of explicitly defined ophthalmic terminology [5,6]. Its formal development and validation work cannot be transferred to a new classification by analogy. L-SAFE is an author-proposed framework that integrates four established examination domains into a pre-exposure safety state. The proposal concerns documentation and exposure withholding; it does not establish the acceptability, efficacy, or overall safety of cosmetic laser iris pigmentation reduction.

The objective of this methodological paper is to specify an eye-level, response-guided safety-state architecture that can be independently assessed for reproducibility, completeness, and clinical utility. Its primary output is a proposed decision and recording specification rather than an estimate of complications or a demonstration that additional laser exposure is safe.

Methods

Design and source of inputs

This was a conceptual framework-development exercise. The starting problem was whether the observed response to one staged laser iris exposure should lead clinicians to withhold consideration of the next. The proposal was informed by longitudinal clinical reasoning and selected existing literature; it was not developed from a systematically analysed clinical cohort, a consensus panel, a formal Delphi process, or a derivation study. No participants were recruited and no patient-level research dataset was collected or analysed.

Evidence inputs were deliberately restricted to distinct clinical and methodological functions: reports describing laser iris pigment modification and complications [1-4]; established terminology for anterior-chamber inflammation [5,6]; intraocular-pressure, glaucoma, and pigment-dispersion assessment [7-9]; endothelial measurement considerations [10]; and independent evaluation methods [11-13]. Selection was targeted rather than systematic: no registered review protocol, reproducible database search, screening flow, pooled synthesis, or formal risk-of-bias analysis was performed. These sources contextualise the domains and validation plan but do not establish the validity of the newly defined categories.

Framework-development logic

The conceptual design followed four explicit steps. First, the unit of assessment was restricted to an individual eye at a documented checkpoint before consideration of another staged exposure. Second, four complementary anterior-segment domains were selected for mandatory recording, with clinical measurements retained rather than replaced by a composite score. Third, observations were mapped to mutually distinguished integrated states according to the most safety-relevant abnormality and the corresponding withholding consequence; no numerical IOP, endothelial, pigment, or waiting-interval cutoffs were derived. Fourth, longitudinal modifiers, recovery rules, missing-data handling, and a record template were specified to permit later testing.

For transparency, the development separates (a) examination findings that can be observed, (b) published terminology or measurement principles that support recording them, and (c) author-proposed classification and exposure-control rules. The last group is a design output, not an empirically demonstrated finding. No clinician-rating study, test–retest assessment, patient-outcome comparison, or performance statistic was performed.

Scope and interpretation

L-SAFE asks whether the current integrated anterior-segment response supplies a safety reason to withhold the next staged exposure. The unit of assessment is an individual eye at a documented checkpoint. Each eye is assessed separately, with findings interpreted against its own baseline and subsequent course. Initial candidacy, informed consent, regulatory compliance, device selection, and the overall decision to offer treatment remain outside the framework.

The four domains are mandatory for an integrated checkpoint assessment. They do not constitute a complete ophthalmic safety examination. Visual symptoms, visual acuity, optic nerve findings, retinal pathology, or other clinical concerns may independently preclude further treatment, including when the four-domain classification is S0. Assessment of an acute concern must not be deferred until a scheduled checkpoint.

L-SAFE specifies no universal numerical intraocular-pressure (IOP) or endothelial-cell-density (ECD) cutoffs, fixed waiting intervals, drug protocol, or laser dosing rule. This does not remove the need to measure, investigate, or manage clinically significant abnormalities. The timing and content of clinical care remain determined by the findings and applicable clinical standards.

The Lumineyes Iris Pigmentation Scale (LIPS), if used, provides optional descriptive baseline information to document the initial iris phenotype. It is neither a prerequisite for L-SAFE nor a predictor of pigment release, dose tolerance, treatment interval, or subsequent safety state within this proposal.

Results: proposed methodological framework

The output comprises four mandatory observation domains, one integrated four-state classification, restricted trajectory descriptors, a checkpoint documentation specification, and a proposed evaluation sequence. These are conceptual design outputs; the Results section contains no observed patient outcomes or measured classification accuracy.

Output 1: four assessment domains

Intraocular pressure. Record the actual measurement and interpret it alongside the individual baseline, preceding response, measurement conditions, and medication context. Glaucoma guidance and pressure-measurement literature support contextual interpretation rather than treating one number as a complete assessment [7,9]. A pressure measured during treatment should not be described as untreated recovery. L-SAFE does not require medication withdrawal to establish a state and does not translate an absolute value or a change from baseline into an automatic stage.

Anterior-chamber response. Document inflammatory cells and flare separately from pigment or other particulate material when clinically distinguishable. SUN terminology can describe inflammation [5]; its grades do not automatically determine L-SAFE state. If the nature of visible particles is uncertain, that uncertainty should be recorded. The proposal does not assume that every visible particle is an inflammatory cell or that residual particles are invariably benign.

Corneal status. Assess clarity, edema, and other clinically relevant corneal findings, including endothelial status where indicated. When ECD is measured serially, interpretation should account for image quality, sampling, analysis method, and comparability of examinations [10]. Corneal status remains mandatory even when a particular ancillary measurement is not indicated or available. Absence of an ECD measurement is not evidence that endothelial status is normal, and an apparent percentage change alone does not establish a L-SAFE state.

Angle and outflow status. Record gonioscopic and clinical findings, including pigment burden and concern about outflow dysfunction. Pigment dispersion literature provides a rationale for attending to the outflow pathway [8,9], without establishing a laser-specific staging threshold. Pigmentation and functional impairment are distinct observations. A reassuring IOP measurement does not replace angle assessment, particularly when medication or the preceding course affects its interpretation.

Output 2: integrated safety-state definitions

L-SAFE is not an additive score. The most safety-relevant abnormality governs the integrated interpretation; a reassuring domain cannot cancel a clinically significant abnormality elsewhere. Several mild findings do not automatically add up to S2, although their combined significance can justify a more concerning integrated state. The underlying findings and the reason for classification should remain visible in the record.

S0 (READY) indicates that no finding within the four mandatory domains constitutes a L-SAFE safety reason to withhold consideration of the next staged exposure. READY is a restricted state label, not a recommendation to perform treatment. Table 1 permits trace residual cells or pigment only when clinically interpreted as an expected resolving response without a significant pressure, corneal, inflammatory, or outflow abnormality. That boundary is a proposed feature requiring evaluation, not a validated tolerance threshold.

S1 (WATCH) describes a mild deviation from expected recovery without an abnormality requiring S2-level management or surveillance. S2 (PAUSE) denotes a clinically significant abnormality requiring active management or review and/or intensified surveillance. Both require withholding the next exposure. Their distinction rests on documented clinical significance rather than a universal numerical boundary. Variation in what clinicians consider expected recovery is an acknowledged source of uncertainty.

S3 (STOP) describes a severe, persistent, refractory, or clinically consequential abnormality and terminates the current staged exposure sequence. Persistence is interpreted with severity, function, and response to management; mere continuation of a mild finding across visits is not an automatic S3 criterion. Irreversibility is not required. Recovery from an abnormal safety state must not be equated with reversal of any established tissue damage.

Output 3: trajectory and recovery rules

Resolving (R), stable (S), and progressive (P) modifiers apply only to S1 and S2. They describe observed change relative to the most recent clinically comparable assessment after the same exposure. Resolving and progressive require a clinically meaningful directional change beyond expected measurement variability. Stable describes absence of meaningful directional change when comparison is possible; an isolated assessment cannot establish stability. If comparison is unavailable or an integrated direction cannot be supported, record trajectory as not determined and retain the domain-level explanation.

Improvement does not itself authorize another exposure: S1-R remains S1 and S2-R remains S2 unless a new integrated assessment supports a different state. A trajectory modifier describes the course, not its cause or guaranteed reversibility. No R/S/P modifier is attached to S0 or S3.

Recovery within the framework is the integrated return to S0, with no separate recovery decision layer. Following S2-to-S0 recovery, the prior S2 event remains clinically relevant. If continuation is independently considered appropriate, the proposal calls for more conservative exposure planning and closer surveillance without prescribing a method or magnitude of exposure reduction. Following S3-to-S0 recovery, the current sequence remains terminated; the recovery entry does not establish eligibility for renewed laser treatment. The preceding S3 event must remain explicit in the longitudinal record.

Output 4: checkpoint record and missingness

Table 2 specifies the mandatory checkpoint fields and records medication exposure and adherence as contextual information rather than independent biological safety domains. For each eye, document laterality, the date and timing of the checkpoint relative to the preceding exposure, the individual baseline used for comparison, and the clinical reasoning behind the integrated state. Prior S2 events and termination of an exposure sequence remain part of the longitudinal record even when later findings improve.

An unassessed mandatory domain must not be presumed normal or silently coded as S0. Where information is insufficient, document the incomplete assessment and its reason; it cannot establish S0 readiness. Incomplete documentation is not an additional biological state and should not be converted automatically into S1, S2, or S3. Lack of documentation must likewise remain distinguishable from a documented absence of abnormality in retrospective research.

Table 1 presents the proposed safety-state definitions with their exposure-control consequences. Table 2 lists the minimum checkpoint fields. Fig. 1 displays the relation between baseline observation, integrated state, withholding decisions, and longitudinal recovery.

Table 1 Integrated L-SAFE states and exposure-control consequences

L-SAFE stateIntegrated clinical interpretationExposure-control consequence
S0 - READYNo finding within the four mandatory L-SAFE domains constitutes a safety reason to withhold consideration of the next staged exposure. Trace residual cells or pigment may be compatible with S0 when interpreted as part of an expected resolving response and when no clinically significant pressure, corneal, inflammatory, or outflow abnormality is present.No L-SAFE safety contraindication to consideration of the next staged exposure is identified.
S1 - WATCHA mild deviation from expected recovery is present, without a clinically significant abnormality requiring S2-level management or surveillance.Withhold further exposure. Continue appropriate existing management, verify relevant medication/adherence context, and reassess. Return to S0 is required before another staged exposure may be considered.
S2 - PAUSEA clinically significant abnormality is present, requiring active management/review and/or intensified surveillance.Withhold further exposure. Manage and reassess until recovery to S0. If treatment continuation is subsequently considered, the preceding S2 response should inform more conservative subsequent exposure planning and closer surveillance. L-SAFE does not prescribe a specific method or magnitude of exposure reduction.
S3 - STOPA severe, persistent, refractory, or clinically consequential safety abnormality is present.Terminate the current staged exposure sequence. Further laser exposure is contraindicated while the eye remains in this state. Clinical management is directed toward resolution of the abnormality and restoration of ocular safety. Return to S0 documents ocular recovery but does not, by itself, establish eligibility for renewed laser treatment.

L-SAFE is not an additive score. All four mandatory safety domains are assessed at each checkpoint, but the most safety-relevant abnormality determines the integrated state. A reassuring finding in one domain cannot offset a clinically significant abnormality in another. Multiple lower-grade findings do not automatically produce a higher state; however, their combined clinical significance should be considered when determining the overall safety state.

The distinction between S1 and S2 is based on integrated clinical significance rather than a single numerical threshold. S1 represents a mild deviation requiring continued observation and reassessment, whereas S2 represents a clinically significant abnormality requiring active management/review and/or intensified surveillance. Persistence, progression, and functional dysfunction provide objective clinical anchors that may support classification toward S2 but are not automatic numerical cutoffs.

S0 is a safety classification rather than a recommendation to perform another exposure; other clinical considerations may still preclude treatment.

Table 2 Minimum pre-exposure checkpoint record

Checkpoint componentRequired documentation
Intraocular pressureActual IOP measurement and clinical interpretation relative to the individual baseline, preceding response, and documented medication context
Anterior-chamber responseClinical finding and interpretation; inflammatory cells/flare and pigment or other treatment-related particulate material documented separately when clinically distinguishable
Corneal statusClinical finding and interpretation, including clarity/edema and endothelial status when clinically relevant
Angle/outflow statusGonioscopic/clinical finding and interpretation, including pigment burden and evidence of functional outflow concern when present
Medication contextRelevant medication exposure during the observation interval documented separately from the four biological safety domains
AdherenceYes / Partial / No / Unknown
Integrated L-SAFE stateS0 / S1 / S2 / S3
TrajectoryFor S1 or S2 only: Resolving (R) / Stable (S) / Progressive (P) / Not determined
Prior S2 eventYes / No

Only S0 indicates absence of an L-SAFE safety contraindication to consideration of the next staged exposure. Any state above S0 precludes the next staged exposure.

Trajectory is assigned only to abnormal, potentially reversible S1 and S2 states and is determined relative to the most recent clinically comparable assessment following the same exposure. A clinically meaningful directional change beyond expected measurement variability is required. When no comparable prior assessment is available, trajectory is recorded as not determined. Progression describes observed deterioration and does not, by itself, establish its cause.

Illustration reproduced from source manuscript; see adjacent figure description

Fig. 1 L-SAFE pre-exposure safety architecture. Following baseline documentation and an interval observation, the four domains are integrated into one safety state. S1 and S2 require withholding and reassessment; S3 terminates the current staged sequence. Trajectory modifiers apply only to S1 and S2. An S0 assessment does not by itself authorize treatment or override prior S3 termination.

Discussion

Relation to existing ophthalmic evidence

The published evidence is heterogeneous in both design and reporting. A clinical series describes outcomes in treated eyes [1], whereas case reports describe serious presentations, including pigmentary glaucoma requiring surgery and concurrent retinal pathology [2, 3]. Another case report documents marked intraocular-pressure elevation and a temporally associated corneal infectious episode; the temporal association alone does not establish causation for keratitis [4]. Case reports identify plausible safety concerns but cannot quantify frequency; an uncontrolled series cannot exclude uncommon harms across different devices, operators, patient groups, or exposure schedules.

These findings motivate repeated assessment without specifying a universally safe treatment interval or acceptable pigment burden. A measured pressure at one visit cannot substitute for examination of the anterior chamber, cornea, or drainage angle. In turn, an apparently resolving clinical response does not prove restoration of baseline ocular reserve. These are assessment and inference problems, not claims of comparative efficacy or demonstrated risk reduction.

Four-domain assessment is particularly relevant when observations disagree. As a hypothetical documentation example, a pressure value within the individual's prior range does not erase clinically important angle pigment or corneal edema. Similarly, an improvement recorded while pressure-lowering medication is in use cannot be interpreted as untreated normalization. When the distinction between inflammatory cells and pigment is uncertain, the record should preserve that uncertainty instead of forcing a categorical biological interpretation. These examples illustrate reasoning within the proposal, not validated case classifications.

Published reports differ in treatment platform, follow-up, and study design [1-4]. Accordingly, case reports identify clinically important adverse-event possibilities without establishing population incidence, and an uncontrolled series cannot rule out infrequent harms. Existing uveitis and glaucoma measurement guidance [5-10] informs documentation; it is not indirect validation of L-SAFE states.

Design strengths and unresolved boundaries

The proposed contribution is a link between integrated examination findings and explicit exposure-control consequences. A vocabulary may make it easier to distinguish mild deviation from clinically significant abnormality, current improvement from recovery, and recovery from eligibility for renewed treatment. Those potential advantages remain hypotheses. Naming a framework does not establish consensus or show that its use reduces harm.

A particular tradeoff is the absence of universal IOP or ECD boundaries. Context-sensitive interpretation avoids presenting an unvalidated numerical rule as authoritative, but it can reduce agreement between observers. The framework therefore retains measurements and clinical reasoning rather than substituting a state label for source observations. The S1-S2 boundary and the allowance for trace residual findings in S0 are priorities for evaluation and may require revision.

The four-domain scope also creates a risk of false reassurance. Reported retinal and optic nerve consequences after laser iris treatment illustrate why an anterior-segment state cannot certify whole-eye safety [2,3]. Similarly, an improved value under medication does not demonstrate that the eye would remain stable without that medication. Historical adverse responses, treatment context, and concerns beyond the framework must remain available to the clinician.

The name Lumineyes identifies the proposed framework; it supplies no comparative evidence for a named provider, device, or technique. Findings from the cited laser reports should not be transferred to an unspecified platform as incidence estimates, nor should the framework be used to claim a superior safety profile. The proposed terminology should be assessed independently of commercial associations.

The framework is intentionally conservative in a narrow sense: an S0 label does not mandate re-exposure, while any abnormal state withholds it. The proposal does not assert that withholding by this classification reduces event rates; such a claim would require controlled evaluation and independent clinical outcomes. The most important unresolved design decisions are whether S0 can permit trace residual findings, whether clinicians reliably distinguish S1 from S2, and how to handle incomplete gonioscopy or non-comparable endothelial measurements.

Proposed independent validation sequence

The first evaluation stage should examine interpretability and reproducibility before clinical effectiveness. Independent raters could review complete, de-identified examination records or explicitly fictional vignettes, without knowing subsequent decisions or outcomes. Studies should pre-specify whether raters see historical medication exposure, serial photographs, and prior checkpoints. Outcomes should include exact four-state agreement, agreement across the crucial S0 versus S1-S3 withholding boundary, disagreement between S1 and S2, and uncertainty intervals. Repeated rating after a suitable interval could assess intrarater stability. GRRAS provides relevant reliability-reporting guidance [11]. Agreement in vignettes would establish neither clinical benefit nor safety.

A second stage should assess real-world feasibility: the proportion of checkpoints with all four domains, reasons for missing measurements, clarity of documentation, and frequency of unclassifiable encounters. Subsequent observational evaluation should preserve baseline and serial findings, exposure timing, medication context, paired-eye clustering, and repeat visits. Associations with independently adjudicated ocular outcomes could then be explored, separating the framework's classification performance from effects of downstream care. Decisions to intensify monitoring or to stop exposure cannot serve as sole validation outcomes because they already help define the proposed states. Observational findings should be reported transparently using applicable STROBE principles [12], without treating association as proof of preventive effectiveness.

Retrospective application can test how often existing records permit classification, but it cannot establish that earlier clinicians acted according to a classification devised later. Prospective work should first pre-specify state assignment, missing-data handling, blinding where feasible, and independent outcome definitions. As with evaluation of other complex clinical interventions, development, feasibility, implementation, and effectiveness should be considered distinct questions [13]. Any prospective evaluation must be ethically reviewed; no eye should be exposed to further laser solely to test whether withholding was necessary.

An additional prespecified falsification criterion would be substantial disagreement between independent raters about whether the same fully documented checkpoint is S0 or above S0. Persistent disagreement or a high proportion of unclassifiable encounters would argue for revising the rule boundaries, mandatory fields, or terminology before clinical implementation. The present paper supplies no measured estimate for those endpoints.

Limitations

L-SAFE is an unvalidated conceptual proposal informed by a selected, non-systematic literature base and general clinical experience. No patient-level derivation, interobserver study, external validation, or formal consensus process is reported. The four-domain selection and state boundaries may be incomplete or insufficiently reproducible. Terms such as clinically significant, expected recovery, and persistent require contextual judgment and may be interpreted differently across observers and settings.

The framework does not estimate complication incidence, predict an individual outcome, establish a safe exposure schedule, or demonstrate clinical benefit. A return to S0 does not establish restoration of pretreatment tissue reserve or absence of future risk. Generalization across laser platforms or to cosmetic iris implants, keratopigmentation, or other interventions has not been established. Clinical recommendations beyond the proposed withholding architecture require their own evidence.

Conclusion

The L-SAFE development output is a four-domain, four-state classification with explicit withholding, trajectory, recovery, and missing-assessment rules. It offers a reproducible specification for investigation, not validated evidence of patient safety, efficacy, or suitability for another laser exposure. Independent reliability and feasibility testing must precede conclusions about its clinical utility.

References

1. Grimaldos Ruiz P (2021) Photoablative cosmetic iridoplasty: effective, safe, and predictable—eye color change in 1176 eyes. Int Ophthalmol 41:1381–1393. https://doi.org/10.1007/s10792-021-01693-5

2. Ong AY, Ching-A-Sue G, Krilis M, Guirao-Navarro MC, Hornby S, Jutley G (2023) Refractory iatrogenic pigmentary glaucoma secondary to cosmetic laser treatment: a case report. Eur J Ophthalmol 33:NP1–NP4. https://doi.org/10.1177/11206721211050338

3. Liu J, Korban S, Moster MR, Rhéaume MA, Wang Q (2023) Bilateral severe iatrogenic pigmentary glaucoma following laser treatment for cosmetic iris color change. Am J Ophthalmol Case Rep 32:101927. https://doi.org/10.1016/j.ajoc.2023.101927

4. Jaber W, Nemet M, Waisbourd M (2026) Laser iris depigmentation resulting in markedly elevated intraocular pressure and herpes simplex keratitis with corneal scarring: case report. Case Rep Ophthalmol 17:312–317. https://doi.org/10.1159/000551226

5. Jabs DA, Nussenblatt RB, Rosenbaum JT; SUN Working Group (2005) Standardization of uveitis nomenclature for reporting clinical data: results of the First International Workshop. Am J Ophthalmol 140:509–516. https://doi.org/10.1016/j.ajo.2005.03.057

6. Jabs DA, McCluskey P, Palestine AG, Thorne JE; SUN Working Group (2022) The standardisation of uveitis nomenclature (SUN) project. Clin Exp Ophthalmol 50:991–1000. https://doi.org/10.1111/ceo.14175

7. Pazos M, Traverso CE, Viswanathan A; European Glaucoma Society (2025) European Glaucoma Society - terminology and guidelines for glaucoma, 6th edition. Br J Ophthalmol 109(Suppl 1):1–212. https://doi.org/10.1136/bjophthalmol-2025-egsguidelines

8. Niyadurupola N, Broadway DC (2008) Pigment dispersion syndrome and pigmentary glaucoma—a major review. Clin Exp Ophthalmol 36:868–882. https://doi.org/10.1111/j.1442-9071.2009.01920.x

9. Weinreb RN, Brandt JD, Garway-Heath D, Medeiros FA (eds) (2007) Intraocular pressure: World Glaucoma Association Consensus Series 4. Kugler Publications, Amsterdam

10. McCarey BE, Edelhauser HF, Lynn MJ (2008) Review of corneal endothelial specular microscopy for FDA clinical trials of refractive procedures, surgical devices, and new intraocular drugs and solutions. Cornea 27:1–16. https://doi.org/10.1097/ICO.0b013e31815892da

11. Kottner J, Audigé L, Brorson S, Donner A, Gajewski BJ, Hróbjartsson A et al (2011) Guidelines for Reporting Reliability and Agreement Studies (GRRAS) were proposed. J Clin Epidemiol 64:96–106. https://doi.org/10.1016/j.jclinepi.2010.03.002

12. von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP; STROBE Initiative (2007) The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. PLoS Med 4:e296. https://doi.org/10.1371/journal.pmed.0040296

13. Skivington K, Matthews L, Simpson SA, Craig P, Baird J, Blazeby JM et al (2021) A new framework for developing and evaluating complex interventions: update of Medical Research Council guidance. BMJ 374:n2061. https://doi.org/10.1136/bmj.n2061

Statements and Declarations

Funding: No funds, grants, or other support were received during the preparation of this manuscript.

Competing interests: Mustafa Mete is the developer of the Lumineyes method and owner of Mylumineyes and has a commercial interest related to the subject of the manuscript.

Author contributions: Mustafa Mete conceived and developed the L-SAFE framework, selected and interpreted the supporting literature, and takes responsibility for the final manuscript.

Ethics approval: Not applicable. No human participants, patient records, animal research, or patient-level datasets were included in this conceptual methodological development report.

Consent to participate and consent to publish: Not applicable; no participant data or identifiable material are reported.

Data availability: No original patient-level data or analytical dataset were generated or analysed. The proposed classification and checkpoint fields are fully specified in Tables 1 and 2.